identifier
stringlengths 11
32
| pdf_url
stringlengths 17
4.62k
⌀ | lang
stringclasses 120
values | error
stringclasses 1
value | title
stringlengths 2
500
⌀ | source_name
stringlengths 1
435
⌀ | publication_year
float64 1.9k
2.02k
| license
stringclasses 3
values | word_count
int64 0
1.64M
| text
stringlengths 1
9.75M
|
|---|---|---|---|---|---|---|---|---|---|
https://openalex.org/W2152968182
|
https://journals.plos.org/plosone/article/file?id=10.1371/journal.pone.0005725&type=printable
|
English
| null |
Using Biotic Interaction Networks for Prediction in Biodiversity and Emerging Diseases
|
PloS one
| 2,009
|
cc-by
| 9,709
|
Introduction A fundamental underlying goal of biology is to model the
distribution of biota and identify their interactions, thus permitting
both an understanding of current distributions and the possibility
of predicting future ones [1]. Such models have important
applications, such as in biodiversity [2] and emerging diseases. Networks
offer
an
important
tool
for
understanding
and
visualizing biotic interactions and have been used in a variety of
contexts [4,5,6]. They are constructed by linking nodes of the
network, usually species that have a known interaction, such as in
trophic webs. However, as it is not feasible to exhaustively track
the large numbers of ecological interactions, the question arises:
can biotic interaction networks be constructed other than by direct
observation, using other available data? g
p
Collection
data
offer
an
important
proxy
for
modeling
distributions. Here, we show how such data can be used to infer
potential inter-species interactions, construct an associated net-
work and, further, show how that network can be used to construct
prediction models. Collection data are already widely used in
biodiversity informatics [10,11], and have been principally used
for constructing species distributions from abiotic niche variables
only. The data are taxonomic in nature and georeferenced, the set
of point collections of a species in a geographical region giving a
sampling for the distribution of the species in that region. Of
course, there is an important question of sample bias in the data
[12,14] (see also the Materials and Methods section), though its
extensive use and utility, even in areas where data are scarce [13],
is testament to the fact that it can yield important information if
treated carefully. Additionally, in the case of urgent problems of
great social impact, such as that of emerging diseases, it is
important to try to leverage the data that actually exist, at least
until better, more bespoke, data become available. There is evidence that the evolutionary dynamics of inter-
species interactions create rich geographic mosaics [7]. Moreover,
phylogenetic research has shown that species are conservative
when it comes to the taxa with which they interact, both spatially
and temporally. As an example relevant to this paper, blood
sucking insects have evolved phenotypic traits to optimize host-
seeking and feeding [8]. Co-distributions of host and parasite will
then reflect the strong biotic relation that exists between them. Introduction Similarly, as a reflection of the potential confrontation of species,
co-occurrence could also engender an interaction in the absence of Abstract Networks offer a powerful tool for understanding and visualizing inter-species ecological and evolutionary interactions. Previously considered examples, such as trophic networks, are just representations of experimentally observed direct
interactions. However, species interactions are so rich and complex it is not feasible to directly observe more than a small
fraction. In this paper, using data mining techniques, we show how potential interactions can be inferred from geographic
data, rather than by direct observation. An important application area for this methodology is that of emerging diseases,
where, often, little is known about inter-species interactions, such as between vectors and reservoirs. Here, we show how using
geographic data, biotic interaction networks that model statistical dependencies between species distributions can be used to
infer and understand inter-species interactions. Furthermore, we show how such networks can be used to build prediction
models. For example, for predicting the most important reservoirs of a disease, or the degree of disease risk associated with a
geographical area. We illustrate the general methodology by considering an important emerging disease - Leishmaniasis. This
data mining methodology allows for the use of geographic data to construct inferential biotic interaction networks which can
then be used to build prediction models with a wide range of applications in ecology, biodiversity and emerging diseases. Citation: Stephens CR, Heau JG, Gonza´lez C, Ibarra-Cerden˜a CN, Sa´nchez-Cordero V, et al. (2009) Using Biotic Interaction Networks for Prediction in Biodiversity
and Emerging Diseases. PLoS ONE 4(5): e5725. doi:10.1371/journal.pone.0005725 Editor: Laurent Re´nia, BMSI-A*STAR, Singapore Received March 19, 2008; Accepted November 20, 2008; Published May 28, 2009 Copyright: 2009 Stephens et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits
unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Funding: Funds and support received from the UNAM Macroproyectos: Tecnologias para la Universidad de la Informacion y de la Computacion and SIBA and
from CONACyT. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing Interests: The authors have declared that no competing interests exist. * E-mail: stephens@nucleares.unam.mx a pre-existing one [9]. We are thus led to consider distributional
data for constructing inter-species interaction networks. Using Biotic Interaction Networks for Prediction in
Biodiversity and Emerging Diseases Christopher R. Stephens1,2*, Joaquı´n Gime´nez Heau1,3, Camila Gonza´lez1,3, Carlos N. Ibarra-Cerden˜ a1,3,
Victor Sa´nchez-Cordero1,3, Constantino Gonza´lez-Salazar1,3 1 C3 - Centro de Ciencias de la Complejidad, Universidad Nacional Auto´noma de Me´xico, Ciudad de Me´xico, Me´xico, 2 Instituto de Ciencias Nucleares, Universidad
Nacional Auto´noma de Me´xico, Ciudad de Me´xico, Me´xico, 3 Instituto de Biologı´a, Universidad Nacional Auto´noma de Me´xico, Ciudad de Me´xico, Me´xico PLoS ONE | www.plosone.org Results A very useful and transparent one,
that can be deduced using only the properties of the network, is the
naive Bayes approximation [22] (see the Materials and Methods
section), wherein a score function, S (Bi | I9), that is a monotonic
function of P (Bi | I9), can be constructed. The score consists of a
sum of contributions from each niche variable, both biotic and
abiotic, from which it is possible to observe which are the most
important niche variables. As we have no underlying theory with which to construct
P(Bi(xa)|I(xa)) we will use a data mining approach to estimate it,
using point collection data as a proxy for the actual distribution of
taxa. Point collection data here represent the set of georeferenced
localities (latitude, longitude and date) of museum voucher
specimens. It is important to remember that the distribution of
taxa is a direct result of the past and present interactions of all
relevant causative factors - climactic, phylogenetic, co-evolution-
ary, ecological etc. Hence, part of the task of any analysis is to
determine, out of the myriad of factors that contribute to I, which
ones
are the
most predictive
in determining
a particular
distribution. An immediate problem is that, as every spatial cell
is unique, for each xa one has a statistical sample of size one and
hence P(Bi(xa)|I(xa)) = 0 or 1. To overcome this, one first constructs the relationship between
Bi and I, such as P(Bi| I), via a sampling of all spatial cells in order
to obtain the relationship between a given distribution measure
and the associated niche variables. With this in hand, a ‘‘profile’’
of any given spatial cell xa can be constructed in terms of the biotic
and abiotic niche variables and the relationship between Bi and I
used to determine P(Bi(xa)|I(xa)) (see Materials and Methods
section). As an example of the general methodology we consider an
important emerging disease - Leishmaniasis - a vector borne
disease widely distributed in tropical regions that is estimated to
affect 12 million people in 88 countries. Since Leishmaniasis is a
zoonotic tropical disease, sylvan reservoirs are crucial to the
maintenance of the parasite in ecological communities and,
further, are intimately associated with human transmission [18]. Results Dividing up a geographic region into spatial cells, xa, we take as
our underlying variable of interest, Bi(xa), a measure of the PLoS ONE | www.plosone.org May 2009 | Volume 4 | Issue 5 | e5725 May 2009 | Volume 4 | Issue 5 | e5725 1 Biotic Interaction Networks where the binomial distribution associated with P(Bi) can be
approximated by a normal distribution then values of |e(Bi
|Bj)|.2 would indicate an inconsistency between the data and the
null hypothesis to the 95% confidence level. distribution of the ith taxon in the spatial cell xa. The specific form
of Bi is determined by the available data - relative or absolute
abundance, presence/absence or presence only. A fundamental
object of interest is P(Bi(xa)|I(xa)), the probability that the
distribution measure Bi(xa) takes a certain value in the spatial cell
xa conditioned on, I(xa), which is composed of, in principle, all
biotic and abiotic variables that affect species distributions, and
which constitute the biotic and abiotic profiles of the correspond-
ing niche [10]. An example of interest would be that Bi(xa)
represents presence of the ith species in the spatial cell xa. For any pair of taxa, Bi and Bj, taken as nodes of a network, a
link between them, whose ‘‘strength’’ is given by e(Bi |Bj), or P (Bi
|Bj), can be graphed. The resulting interaction network then offers
a visualization of the inferred statistical dependencies between
different taxa. Note that, contrary to networks that are common in
the literature, that represent known interactions, such as between
predator and prey in a trophic web [15], this network represents
statistical associations from which inferences about real causal
interactions can be made and then tested. Higher order statistical
associations, such as P (Bi |Bj Bk), can also be examined. Such an
interaction could be represented by three nodes, with links from Bj
and Bk to Bi, and would represent the degree of statistical
dependence of taxon Bi on the co-occurrence of the taxa Bj and
Bk. From the network, for a given node Bi, a ranked list of values of
e(Bi |Bj), or P (Bi |Bj), can be taken as a model for predicting the
most important potential biotic interactions of the species Bi. To
determine P (Bi |I9) when I9 is high dimensional, a statistical model
must be used to approximate it. Results Reservoirs of Leishmania can be classified as primary and
incidental, according to their importance in the long-term
transmission of the parasite, being considered incidental if they
are dead ends that do not transmit to vectors [19]. Although direct
experiment could determine to which type a given reservoir
belongs,
when
there
are
many
potential
reservoirs
other
alternatives, such as that presented here, are more feasible. As P(Bi| I) involves counting the number of spatial cells where
there is a co-occurrence of the ith species with a particular
configuration of the niche variables I, if I is of high dimension then
the number of cells where there are co-occurrences will be small or
zero. We thus restrict attention for the moment to the case where I
is a single variable, Ij, so that P BijIj
~NBi&Ij
NIj, where NBi&Ij
is the number of cells with a co-occurrence of the distribution
variable Bi and the niche variable Ij, and NIj is the number of cells
with niche variable Ij. In the case where Ij is also a taxon
distribution, and we consider presence, then P (Bi |Bj) measures
the probability of presence of taxon Bi given the presence of taxon
Bj and is thus a measure of the statistical association between Bi
and Bj. As P (Bi |Bj) does not take into account statistical
confidence however, we consider rather We used collection data points for 427 terrestrial mammal
species occurring in Mexico as potential or confirmed reservoirs
and 11 species of Lutzomyia as confirmed or potential vectors for
Leishmania. The description of the data set can be found in the
Materials and Methods section. Lutzomyia is a genus of ‘‘sand flies’’
that in the New World is responsible for the transmission of the
Leishmania parasite. Only females suck blood for egg production. In
Mexico there is little information about which vectors are involved
in transmission of the parasite in different geographic regions. The
only confirmed vector is Lutzomyia olmeca olmeca [20]. However,
several species have been found with the parasite - Lutzomyia olmeca
olmeca, Lutzomyia cruciata and Lutzomyia ovallesi [21]. With respect to
transmission of the visceral form of the disease the principle vector
Lutzomyia longipalpis has been collected in Mexico but has not been
reported with the parasite. For the secondary vector Lutzomyia
evansi, there exists only one collection in the state of Chiapas which
was without infection [22]. Biotic Interaction Networks Moreover, if there are many connections to mammals that are not
connected to other vectors, then all else being equal, it would be
evolutionarily suboptimal for the vector not to exploit them. L. cruciata and L. longipalpis, in particular are associated with large
numbers of mammals that have no statistical relation with other
vectors. On the other hand, L. olmeca, L. ovalesi, Lutzomyia shannoni
and Lutzomyia panamanensis are all within a highly connected part of
the network that corresponds geographically to the peninsula of
Yucatan, where many mammals are associated with several
different vectors. In such circumstances, a vector may adopt a
strategy of specializing to a smaller group of species in order to
avoid competition. Interestingly, four of the eight infected rodent
reservoirs - Peromyscus yucatanicus, Ototylomys phyllotis, Reithrodontomys
gracilis and Heteromys gaumeri, all restricted to the peninsula of
Yucatan, have very high vertex degrees, a fact that associates them
with higher risk, as potentially many different vector species can
exchange parasites with them. reservoirs. It is important, therefore, to be able to predict which
currently unidentified mammals are most likely to be important as
actual or potential reservoirs for the disease. As a measure of
statistical association we consider P(vi |mj) and e(vi |mj), where vi
represents the ith vector and mj the jth potential reservoir. There
are 4697 potential vector-reservoir pairs. In Figure 1 we show the
241 most important positive associations (highest values of e)
between Lutzomyia as vectors and mammals as suspected and
confirmed reservoirs for Leishmania. The vector species are marked
as red nodes, while the confirmed reservoirs are marked as green. The darker the link, the stronger is the associated statistical
dependence between the associated Lutzomyia and mammal. The connectivity of the network is related to the geographical
distribution of the different species and has consequences for the
way in which a parasite could propagate across the network from
one geographical region to another. The separated subnetwork
corresponds to L. anthophora, a species indigenous only to the north
of Mexico and the United States. For Lutzomyia nodes, the vertex
degree dictates with how many mammals a given vector shares
important positive statistical associations, while, for mammal
nodes, the vertex degree tells us how many vectors are potentially
exploiting the mammal. A high vertex degree for a given vector
shows that it could potentially exploit many different mammals. Results In Mexico, there are only eight
mammal species found infected with Leishmania mexicana parasites,
responsible for the cutaneous form of the disease, identified in the
state of Campeche in southern Mexico [23,24,25]; a very small
number when compared to the total number of potential e BijBj
~ NBj P BijBj
{P Bi
ð
Þ
NBjP Bi
ð
Þ 1{P Bi
ð
Þ
ð
Þ
1=2 which also measures the degree of confidence one can have in the
statistical association between Bi and Bj relative to the null
hypothesis, P(Bi), that the distribution of Bi is independent of Bj
and distributed with this probability over the region of interest. Essentially, this is a one-sided binomial test where the null
hypothesis is that the distribution of Bi is random over the sample
space; in this case the cells of the region of interest. It can, of
course, be useful to consider other null hypotheses. For instance,
one could use as null hypothesis P(Bi| A) where A represents a set
of abiotic factors, or the result of a niche-model such as GARP or
MaxEnt [15,16]. Values of |e(Bi |Bj)| greater than a certain
threshold (see Materials and Methods section) measure the degree
to which the data is consistent with the null hypothesis. In the case PLoS ONE | www.plosone.org PLoS O May 2009 | Volume 4 | Issue 5 | e5725 PLoS ONE | www.plosone.org 2 Biotic Interaction Networks Biotic Interaction Networks Biotic Interaction Networks In Figure 2 we see the results for the grid partition of Mexico we
used earlier. The redder/whiter the area the higher/lower the
predicted probability for finding Lutzomyia based only on co-
occurrence with mammal species, the mid-range being associated
with the probability P(Bi) associated with the null hypothesis. Also
shown is the georeferenced set of point collections of Lutzomyia. As
can be seen, the agreement is good, though there are one or two
outliers. Finally, on the map we also see those geographical regions
where cases of Leishmaniasis have been reported. The shaded
regions correspond to ‘‘municipios’’ (municipalities) where Leish-
maniasis cases have been reported in the last 40 years. Note that the
area of different municipios can vary greatly. In regions where
there is no cross-hatching there are no cases that have been
reported to the Secretaria de Salu´d Pu´blica (Governmental Public
Health Agency) in Mexico. This does not necessarily imply that
there are none, as there is no obligatory reporting of cases of
Leishmaniasis in Mexico. In this sense reported cases are the
equivalent of presence data, while no reported cases does not
imply ‘‘absence’’. A noteworthy feature of the map is that there are
no areas with reported cases where the model does not predict a
higher than random probability for presence of Lutzomyia. In
interpreting the apparent overprediction several comments are in
order: First of all, as mentioned, the quality of reporting data of
cases of Leishmaniasis varies significantly from state to state in
Mexico. Secondly, the map is of degree of risk due to biotic factors
only; the output being a score that measures the probability of
Lutzomyia being present in a given spatial cell. In that sense, it is a
map associated with only one type risk factor, all be it an
important and necessary one for the presence of the disease in the
human population which, obviously, depends on many other
factors. By including such factors, for example, abiotic or socio-
demographic variables, more complex risk models can be simply
created using our methodology. p
Such a ranked list provides a general model for predicting the
most important likely reservoirs for any given disease. Note that, of
the eight infected reservoirs of Leishmania in Mexico, six of them,
including the four confirmed, appear in the top 7% of ranked
predictions of most important potential reservoirs. Biotic Interaction Networks Biotic Interaction Networks Turning first to the prediction of potential reservoirs: with e(vi |mj)
in hand, for a given vector vi, we can construct a ranked list, from
maximum to minimum value, of e(vi |mj), over all pairs (vi,mj), i.e.,
a ranking of the links of a given node according to their strength. Those mammals with the highest values of e are predicted to
correspond to the most important potential reservoirs for that
vector. In Table 1 we show the results for the highest 150 values of
e(v|mj), where to obtain the list we have grouped together the
different Lutzomyia species into one group v to form a list of 427
values of e(v|mj) as a function of j. The highest ranked mammals
have the highest degree of statistical correlation with Lutzomyia,
with the implication that these mammals are the most important
potential reservoirs for Leishmania. By grouping together the
different Lutzomyia species we are considering association between
a given mammal species and the different species of the Lutzomyia
genus present in Mexico, rather than with individual species, thus
increasing the sample size and allowing for more robust statistics. A secondary logic for this is also that the biomass of parasite that
can pass from vector to mammal in a given spatial cell depends on
the number of different vector species that are present in that cell. Thus, a mammal with a high probability of co-occurrence with
more than one Lutzomyia will, all else being equal, present a higher
degree of risk of having the Leishmania parasite transmitted to them
than one that has a high degree of occurrence with only one
species. being associated with directed links from vector to reservoir in
the network. The advantage of this approximation is that the
contribution of each biotic niche variable, Bi, is independent of the
rest, so that, in the case where abiotic variables are also explicitly
included, the relative importance of both biotic and abiotic factors
can be studied. As one would expect in the present case, biotic
variables play a more important role than abiotic ones, due to the
direct dependence of a vector on its associated reservoirs. With
S(B) in hand, the biotic niche profile of any geographical area can
be determined using a ranked list of niche characteristics and
allows one to see at a glance which species are playing an
important role. Biotic Interaction Networks If we take as
null hypothesis that the confirmed reservoirs are distributed
randomly in the ranked list, then the probability that they appear
with their actual rankings is less than 1028, thus showing that the
model’s results are statistically significant and that the model
predicts very well, especially given the relative lack of information
on which it is based, in that at no point was information on
confirmed reservoirs used to ‘‘train’’ the model. Of course, one
could argue that, all else being equal, there should be a higher
degree of co-occurrence between Lutzomyia species and those
mammals that are most widespread, as these will have had a
higher probability of having being tested as potential reservoirs. Of
course, if this were true, it would greatly reduce the predictive
power of the model. We tested this hypothesis on a subset of 360
mammal species where distribution data was readily available. The positions of the confirmed reservoir species ranked according
to their area of distribution were: 25, 152, 154, 200, 224, 230, 249,
255 and 257; while when ranked according to our prediction
model the positions were 4, 6, 8, 22, 27, 28, 40, 88 and 130. As a
simple statistical comparison one can compare the mean rank
from both methods using an independent two sample t-test. The
test statistic value is 5.4 corresponding to a p value of less than
0.001 clearly indicating that the predictive power of our model
cannot be explained by assuming that those species with larger
distributions are more likely to be confirmed reservoirs. Biotic Interaction Networks Besides offering substantial insight into the ecological interac-
tions between potential vectors and reservoirs of a disease, the
interaction network can also be used to obtain predictive models. Here we consider two such models - one for directly predicting the
most important potential disease reservoirs and another for
predicting a measure of disease risk for a given geographic area. Figure 1. Interaction network between potential and confirmed vectors and reservoirs for Leishmania in Mexico. Mammal species
confirmed as reservoirs for Leishmania mexicana, responsible for the cutaneous form of the disease are marked with a double circle. One species,
Didelphis marsupialis is the known sylvatic reservoir for the visceral form. doi:10.1371/journal.pone.0005725.g001 Figure 1. Interaction network between potential and confirmed vectors and reservoirs for Leishmania in Mexico. Mammal species
confirmed as reservoirs for Leishmania mexicana, responsible for the cutaneous form of the disease are marked with a double circle. One species,
Didelphis marsupialis is the known sylvatic reservoir for the visceral form. doi:10.1371/journal.pone.0005725.g001 May 2009 | Volume 4 | Issue 5 | e5725 PLoS ONE | www.plosone.org 3 Discussion The main contribution of this paper is to show how biotic
interaction networks may be constructed inferentially using a data
mining approach applied, in this case, to point collection data,
rather than by direct observation, and to show that these networks
can be used, not only to understand and visualize potential inter-
species interactions, but also to formulate prediction models. The
important area of emerging diseases was used as a test bed to show
the utility of the approach. The main logic of this methodology is
that current distributions of biota, as proxied by point collection
data for the example given here, adequately reflect all causal
influences, both biotic and abiotic. The task, for a given set of
input variables, is to discriminate which ones are of greater
influence for a particular distribution. In this paper we used only
biotic variables. A statistical dependence between two species
infers, but does not prove, a direct biotic causal relationship. Thus,
for a pair of nodes the strength of the link between them measures The third step we will take is to construct a predictive model to
quantify disease ‘‘risk’’ in any given geographic cell. Here we take
as risk measure the probability that disease vectors are present,
while the prediction itself is based only on biotic factors, i.e., the
presence of potential mammal reservoirs. Explicitly, a score
function, S (Bi | B), for predicting class membership is constructed,
where Bi is associated with the ith vector species and B represents
the presence of mammal species, B1, …, BN, and related to the
posterior classifier probabilities, P (Bi| B), using the naive Bayes
approximation, P BjBi
ð
Þ~ P
N
j~1 P Bj
Bi
, the factors P(Bj |Bi) PLoS ONE | www.plosone.org May 2009 | Volume 4 | Issue 5 | e5725 4 May 2009 | Volume 4 | Issue 5 | e5725 Biotic Interaction Networks Table 1. Ranked list of potential mammal reservoirs for Leishmania in Mexico. Mammals
Epsilon
Conf. Mammals
Epsilon Conf. Mammals
Epsilon Conf. May 2009 | Volume 4 | Issue 5 | e5725 Discussion Biotic risk map for Leishmania using the mapped score function. doi:10.1371/journal.pone.0005725.g002 Figure 2. Biotic risk map for Leishmania using the mapped score function. doi:10.1371/journal.pone.0005725.g002 Figure 2. Biotic risk map for Leishmania using the mapped score function. doi:10.1371/journal.pone.0005725.g002 emphasized, based on the relative importance of the potential
reservoir in terms of what biomass of parasite can potentially be
harbored in a given spatial cell rather than what mammals are
physiologically capable of being reservoirs. To test this, the
following scenario may be envisaged: consider the known
distribution of a given mammal from the list; select spatial cells
at random from this distribution; in each cell capture the chosen
mammal species and test for the presence of Leishmania. The
appropriate metric is the proportion of spatial cells in which
specimens were found with the parasite or, alternatively, if
sufficient statistics may be obtained, e(cells with specimens with
parasite | total cells with specimens). This would be repeated for
different mammal species. The hypothesis is that a highly ranked
species will yield higher values for these two metrics than a low
ranked one. To facilitate testing the hypothesis, the most
appropriate species would be those chosen from different points
in the ranked list that are common in a given geographical region
and easy to capture. Of course, many mammals simply do not
have any geographical overlap with the vectors. Strictly speaking
one should consider these mammals too and test for presence of
the parasite. Common sense would dictate that for those species
far away from the known distribution of the vectors there is
effectively zero probability of finding the parasite thus obviating
the need to explicitly check these areas. Work is currently being
planned to undertake these tests. the degree to which two species tend to co-occur. If they co-occur
in a statistically significant way we are prompted to identify as a
plausible explanation a vector-reservoir interaction. In the case of Lutzomyia and mammals this understanding comes
from the natural potential direct causal relationship there: that the
Lutzomyia feed on the corresponding mammal. The properties of
the corresponding biotic network show to what extent a given
vector is exploiting its potential food sources, evolutionary
dynamics giving a logic as to why this usage should be optimal. Discussion 1
Eira barbara
10.1683
51
Molossus sinaloae
5.8518
101
Balantiopteryx plicata
3.8590
2
Rhogeessa aeneus
9.3649
52
Artibeus lituratus
5.8422
102
Peromyscus leucopus
3.7994
3
Artibeus intermedius
9.1628
53
Mormoops megalophylla
5.8374
103
Sturnina ludovici
3.7888
4
Reithrodontomys gracilis
8.8921
Yes
54
Habromys lepturus
5.7848
104
Enchisthenes hartii
3.6929
5
Carollia sowelli
8.8303
55
Myotis keaysi
5.6148
105
Vampyrodes caraccioli
3.6929
6
Heteromys gaumeri
8.8000
Yes
56
Chiroderma villosum
5.5562
106
Eptesicus furinalis
3.6453
7
Peromyscus mexicanus
8.7859
57
Tamandua mexicana
5.4845
107
Liomys pictus
3.6107
8
Heteromys desmarestianus
8.7164
Yes
58
Tylomys nudicaudus
5.4510
108
Glossophaga commissarisi
3.4861
9
Molossus rufus
8.6277
59
Saccopteryx bilineata
5.2984
109
Lonchorhina aurita
3.4781
10
Glossophaga soricina
8.5713
60
Macrotus mexicanus
5.2472
110
Phyllostomus discolor
3.4781
11
Carollia perspicillata
8.5030
61
Sciurus aureogaster
5.2267
111
Peromyscus gymnotis
3.4516
12
Orthogeomys hispidus
8.3468
62
Baiomys musculus
5.2092
112
Anoura geoffroyi
3.4201
13
Pteronotus parnellii
8.1632
63
Rhogeessa tumida
5.1950
113
Platyrrhinus helleri
3.3586
14
Desmodus rotundus
8.1519
64
Sciurus deppei
5.1414
114
Eumops bonariensis
3.3398
15
Dasyprocta mexicana
8.1128
65
Dermanura watsoni
5.1338
115
Sciurus variegatoides
3.3398
16
Sturnira lilium
8.0290
66
Otonyctomys hatti
5.1338
116
Uroderma bilobatum
3.3373
17
Dermanura phaeotis
8.0055
67
Orthogeomys grandis
5.0556
117
Lasiurus intermedius
3.2197
18
Dasyprocta punctata
7.9678
68
Alouatta palliata
5.0457
118
Lasiurus ega
3.1739
19
Oryzomys couesi
7.7253
69
Choeroniscus godmani
5.0457
119
Peromyscus megalops
3.1410
20
Potos flavus
7.7246
70
Peropteryx macrotis
5.0457
120
Eumops glaucinus
3.0564
21
Conepatus semistriatus
7.6879
71
Pteronotus personatus
5.0266
121
Urocyon cinereoargenteus 2.9697
22
Ototylomys phyllotis
7.5587
Yes
72
Lontra longicaudis
4.9330
122
Procyon lotor
2.9502
23
Ateles geoffroyi
7.4787
73
Reithrodontomys mexicanus 4.9120
123
Hylonycteris underwoodi
2.9343
24
Cryptotis magna
7.4207
74
Oryzomys rostratus
4.8681
124
Rhynchonycteris naso
2.8580
25
Cuniculus paca
7.3220
75
Mimon cozumelae
4.8327
125
Eptesicus brasiliensis
2.8106
26
Lampronycteris brachyotis
7.2852
76
Pteronotus davyi
4.7943
126
Myotis albescens
2.8106
27
Sigmodon hispidus
7.2805
Yes
77
Herpailurus yagouaroundi
4.7100
127
Lophostoma evotis
2.8106
28
Peromyscus yucatanicus
7.2486
Yes
78
Glossophaga leachii
4.6849
128
Tapirus bairdii
2.8106
29
Oryzomys chapmani
7.1242
79
Rhogeessa gracilis
4.6317
129
Vampyrum spectrum
2.8106
30
Didelphis virginiana
7.1150
80
Sylvilagus brasiliensis
4.6317
130
Marmosa mexicana
2.7731
Yes
31
Peromyscus melanocarpus
7.0260
81
Hodomys alleni
4.5155
131
Peromyscus furvus
2.7731
32
Microtus umbrosus
6.9630
82
Leopardus wiedii
4.4420
132
Myotis velifera
2.5757
33
Thyroptera tricolor
6.9630
83
Peromyscus simulatus
4.4195
133
Spilogale putorius
2.5411
34
Nasua narica
6.8953
84
Sigmodon alleni
4.3707
134
Microtus mexicanus
2.5268
35
Megadontomys cryophilus
6.6830
85
Bassariscus sumichrasti
4.3110
135
Dasypus novemcinctus
2.4725
36
Oryzomys alfaroi
6.6816
86
Oryzomys fulvescens
4.3110
136
Myotis nigricans
2.4704
37
Sorex veraepacis
6.6797
87
Diphylla ecaudata
4.3013
137
Lophostoma brasiliense
2.4407
38
Carollia subrufa
6.6316
88
Oryzomys melanotis
4.2907
Yes
138
Diclidurus albus
2.4407
39
Peromyscus aztecus
6.6173
89
Micronycteris microtis
4.2338
139
Sciurus niger
2.4407
40
Didelphis marsupialis
6.4390
Yes
90
Mazama americana
4.2274
140
Leptonycteris curasoae
2.4268
41
Sciurus yucatanensis
6.3865
91
Microtus oaxacensis
4.2061
141
Nyctomys sumichrasti
2.4026
42
Philander opossum
6.2546
92
Rheomys thomasi
4.2061
142
Sigmodon mascotensis
2.3815
43
Habromys ixtlani
6.1120
93
Oryzomys saturatior
4.2061
143
Alouatta pigra
2.3374
44
Microtus waterhousii
6.1120
94
Myotis elegans
4.2024
144
Peromyscus melanophrys
2.2204
45
Pteronotus rubiginosus
6.1120
95
Oligoryzomys fulvescens
4.1984
145
Dermanura tolteca
2.1920
46
Reithrodontomys microdon
6.0967
96
Natalus stramineus
4.0626
146
Trachops cirrhosus
2.1663
47
Coendou mexicanus
6.0268
97
Balantiopteryx io
4.0522
147
Bauerus dubiaquercus
2.1612
48
Centurio senex
6.0076
98
Nyctinomops laticaudatus
4.0522
148
Spilogale pygmaea
2.1612
49
Artibeus jamaicensis
5.9786
99
Tlacuatzin canescens
4.0119
149
Leptonycteris nivalis
2.1402
50
Glossophaga morenoi
5.8847
100
Odocoileus virginianus
3.9265
150
Sylvilagus floridanus
2.1002
doi:10.1371/journal.pone.0005725.t001 May 2009 | Volume 4 | Issue 5 | e5725 5 Biotic Interaction Networks Figure 2. Discussion From the network, the corresponding list of predicted reservoirs
for a given Lutzomyia is not based on the physiological possibility
that a given mammal is a reservoir but, rather, on the fact that a
mammal with a high fraction of co-occurrences is more likely to be
an important food resource for Lutzomyia than one with a small
fraction and, therefore, that there is greater transmission of the
parasite from one to the other. Moreover, as e(vi |mj) increases as
the range of the mammal mj grows, then this measure also predicts
the degree of importance of the reservoir, a reservoir of small
range being of less potential impact, all else being equal, than one
of ample range. As mentioned, the utility of the model is clearly in
evidence, given that all known reservoirs in Mexico are highly
ranked in the complete list of 427 possible candidates. To create spatial prediction models we used a model that
utilized only information that came from the biotic interaction
network. The associated score is a measure of the probability that
Lutzomyia are present, which we can take as a proxy for the
probability that the disease is present. To relate this to the number
of cases in a more sophisticated model would require the inclusion
of socio-economic and socio-demographic variables among others. PLoS ONE | www.plosone.org Materials and Methods p y
[
]
With respect to the data set for Mexican mammals, this data has
been collected over a period of more than 100 years with a
consequently large number of collectors [24,25]. Hence, although
the data has not been collected systematically, it has probably led
to an adequate sampling. Additionally, mammals are the best
known and collected group in Mexico. In the case of Lutzomyia the
coverage is less but still represents the best available. With the
registered cases of Leishmaniasis, unfortunately, there is no
compulsory reporting of these in Mexico. So one can infer where
the disease is present but not where it is absent. In problems of
great social impact, such as that of emerging diseases, it is
important to try to leverage the data that actually exists, at least
until better more bespoke data becomes available. Parasite
detection studies in potential reservoirs have been carried out
principally in the state of Campeche. Van Wynsberghe et al [21]
analyzed the evolution of the infection using parasitological
methods in 29 naturally infected rodents. The mammals belonged
to four species: Sigmodon hispidus (2), Oryzomys melanotis (12),
Ototylomys phyllotis (9) and Peromyscus yucatanicus (6). In a second
study [22], infection by Leishmania mexicana was detected in eight
mammal species using two methods – in culture and PCR. The
Leishmania parasite was confirmed by both methods in six species:
O. phyllotis, Heteromys gaumeri, O. melanotis, P. yucatanicus, S. hispidus,
and Heteromys desmarestianus. In the other two species it was
confirmed using only via one of the methods: in culture for
Marmosa mexicana and by PCR for Reithrodontomys gracilis. p
y
g
p
g
The class, we will take to be a taxon distribution, Bi, while the
feature vector set is taken to be a subset of niche variables I’(I. In
this case, I9, represents a niche profile with both biotic and abiotic
components which constitute the biotic and abiotic profiles of the
niche. For a given taxon, B’(B, and niche variables, I’(I, our
chief object of study is the probability P(Bi | I’9) = NBiAND I9/NI9,
where NBiAND I9 is the number of spatial cells where there is a co-
occurrence of the taxon Bi and the niche variables I9, and NI9 is
the number of cells where the niche variables take their stated
values. Materials and Methods (
b)
(
(
b)
(
b)
(
b)
(
b))
A full model would consist of determining P (Bi(xa, tb)) = F (I (xr,
ts)), relating the distribution of a subset of biota at one place and
time to all biotic and abiotic factors considered at different places
and times. Of course, there are no underlying fundamental
principles on which to build the function F. We therefore adopt a
non-parametric ‘‘data mining’’ approach, modeling the distribu-
tion directly using available data, rather than constructing an a
priori parametric model. An advantage of this approach is that the
observed distribution is a direct result of the past and present
interactions of all relevant causative factors - climactic, phyloge-
netic, co-evolutionary, ecological etc. Nothing is omitted. Howev-
er, an observation of P (Bi(xa, tb)) in itself does not provide a
predictive model. To create such a model we consider the problem
as a classification task, relating a class, such as the class of cells with
presence of a given species, to a feature vector I using the
conditional probabilities P (Bi| I). Converting the problem to one
of classification is very natural from the point of view of presence
or presence/absence. In the case of abundance a coarse graining
of the abundance data in a given spatial-temporal cell is required. This can be achieved in many ways, depending on how many
classes are posited and the criterion by which a given abundance
fits in a given category. For example, one might classify abundance
into three categories – Low, Normal and High – where Low is any
abundance at least one standard deviation below the average and
High is any abundance at least one standard deviation above the
average. One can then
naturally
consider the conditional
probability that a High abundance of species Bi is found given a
High abundance of species Bj. Of course, in order to do this, one
requires abundance data in the first place. As this is less common
than presence or presence/absence data, and simply not available
in the context of emerging diseases such as Leishmaniasis, we will
here focus on the latter. For the same reason, in the following, we
will also restrict attention to the spatial dependence of the
distributions and ignore the temporal aspect, as the data simply is
not capable of reliably describing temporal changes. Materials and Methods The data set consisted of point collection data associated with
one Class, Mammalia, and one genus - Lutzomyia. The mammal
data set consisted of 37,297 point collections from georeferenced
localities for 427 terrestrial mammals occurring in Mexico. The The results of this paper clearly lead us in the direction of
making corresponding hypotheses that can be verified by further
empirical research. Our ranked list of potential reservoirs is, as May 2009 | Volume 4 | Issue 5 | e5725 6 Biotic Interaction Networks data were obtained from museum voucher specimens from
national and international museum collections, public electronic
databases (MaNIS; www.manis.gob.mx, and CONABIO; www. conabio.gob.mx) and published records [27,28]. For Lutzomyia,
there were 270 point collections, taken from published literature
and from national entomological collections (Instituto de Diag-
no´stico y Referencia Epidemiolo´gica (InDRE, Mexico City), the
Coleccio´n Entomolo´gica Regional Universidad Auto´noma de
Yucata´n (UADY, Me´rida) and the Laboratorio de Medicina
Tropical at the Universidad Nacional Auto´noma de Me´xico
(UNAM, Mexico City), associated with 11 species. For both data
sets, each locality was georeferenced to the nearest 0.01 degrees of
latitude and longitude using 1:250,000 topographic maps (INEGI;
www.inegi.gob.mx, Instituto de Geografı´a, Universidad Nacional
Auto´noma de Me´xico; www.igeograf.unam.mx). Point collection
data was, of course, not collected in order to provide an unbiased
sampling of underlying species abundance and therefore must be
considered carefully to understand potential statistical biases that
might be present. The utility and limitations of point collection
data have been amply discussed in [12,14]. geographic region xa in the time interval tb. As Bi (xa, tb) is a
stochastic variable, the distribution of any taxon Bi(xa, tb) is
described by a probability distribution, P(Bi(xa, tb)), whose
evolution, in principle, depends on both biotic factors, Bj(xr, ts),
associated with other species, and abiotic factors, A(xr, ts) = (A1(xr,
ts), …, ANA(xr, ts)), such as temperature, precipitation etc., where
we consider cells xr, ts that may be different to a given cell xa, tb to
indicate that, in principle at least, there may be statistical
associations between a given spatio-temporal cell and others. The full ecological niche at xa and tb can be described by a vector
I(xa, tb) = (A1 (xa, tb),…, ANA (xa, tb); B1(xa, tb),…, BNB (xa, tb)). PLoS ONE | www.plosone.org Biotic Interaction Networks The numerator of equation
(1) then, is the difference between the actual number of co-
occurrences of Bi and Ik relative to the expected number if the
distribution of point collections were obtained from a binomial
with sampling probability P(Bi). As we are talking about a
stochastic
sampling
the
numerator
must
be
measured
in
appropriate ‘‘units’’. As the underlying null hypothesis is that of
a binomial distribution, it is natural to measure the numerator in
standard deviations of this distribution and that forms the
denominator of equation (1). In general, the null hypothesis will
always be associated with a binomial distribution as in each cell we
are carrying out a Bernoulli trial (‘‘coin flip’’). However, the
sampling probability can certainly change. For instance, one could
take as null hypothesis a binomial distribution with sampling
probability P(Bi|M = 1) = NBj/NM = 1, where M here is a binary
variable associated with the fact that a niche-variable model, such
as GARP or MaxEnt, says whether the species Bi is present or
absent. NM is then the number of cells where the niche model says
there is presence. Taking P(Bi|Bk,M) relative to the null hypothesis
P(Bi|M) tells us how the presence of species Bj is associated with
the presence of Bk in the context of cells where a niche model has
indicated the presence/absence of Bk. In other words, how Bk
affects the distribution of Bi in those places where the niche model
says Bk is present/absent. Probabilities P (Bi |I9), where I9 is of high dimension, can be
constructed using different classification models, such as neural
networks, discriminant analysis etc. A particularly transparent,
simple and effective approximation is the Naive Bayes approxi-
mation [26] with P BijI
ð
Þ~ P IjBi
ð
ÞP Bi
ð
Þ
P Ið Þ
~
P
N
k~1 P IkjBi
ð
ÞP Bi
ð
Þ
P Ið Þ P BijI
ð
Þ~ P IjBi
ð
ÞP Bi
ð
Þ
P Ið Þ
~
P
N
k~1 P IkjBi
ð
ÞP Bi
ð
Þ
P Ið Þ where, in the first equality, Bayes rule has been used, and in the
second it has been assumed that the niche variables Ik are
independent. The product here is over the N niche variables under
consideration as conditioning factors for Bi. In the case of the
relationship between Lutzomyia and mammals, N represents the
number of mammal species. Biotic Interaction Networks A score function that can be used as a
proxy for P (Bi |I9) is S BijI’
ð
Þ~
X
N
k~1
S BijIk
ð
Þ~
X
N
k~1
ln
P IkjBi
ð
Þ
P IkjBi
! where Bi is the complement of the set Bi. For example, if Bi is the
set of cells with presence of taxon Bi then Bi represents the set of
cells without presence. S(Bi | I9) is a measure of the probability to
find the distribution variable Bi when the niche profile is I9. It can
be applied to a spatial cell xa by determining the niche profile of
the cell, I9(xa). As an example, for two biotic niche variables, B2
and B3, that take values 1 (corresponding to the fact that there is a
point collection associated with that cell) and 0 (there is no point
collection associated with the cell), the four possible biotic niche
profiles of any cell are (B2, B3) = (0,0); (0,1), (1,0) and (1,1). The
score contributions of each biotic variable are S(Bi|B2) and
S(Bi|B3), calculated using the above formula. Hence, S(Bi |
I9) = S(Bi | B2, B3) = S(Bi|B2)+S(Bi|B3). Thus, for any given spatial
cell xa one can assign a niche profile, i.e. values of B2 and B3, from
whence it is possible to assign a corresponding score. If there is no
statistical association between Bi and B2 or B3 then the
corresponding score contributions are zero. An overall zero score
then signifies that the probability to find Bi js the same as would be
found if Bi were distributed randomly. If the score is positive then
there is a higher than random probability to find Bi present and on
the contrary if the score is negative. The quantitative values of e(Bi |Bk) can be interpreted in the
standard sense of hypothesis testing by considering the associated
p-value as the probability that |e(Bi |Bk)| is at least as large as the
observed one and then comparing this p-value with a required
significance level. In the case where NBj§5{10 then a normal
approximation for the binomial distribution should be a decent
approximation and in this case e(Bi |Bk) = 2 would represent the
standard 95% confidence interval. In the case where a normal
approximation is not accurate then other approximations to the
cumulative probability distribution of the binomial must be used. Materials and Methods The niche profile I9(xa) associated with a spatial cell xa then
determines the probability of the distribution variable, Bi(xa), in
that cell, and one now has a predictive model. Note that, although
we concentrate on biotic variables in the present paper, in the
current approach, all niche variables can be treated on a
democratic footing. The problem of calculating P (Bi | I9) directly
is that both NBi AND I9 and NI9 are likely to be zero when the
number of taxa or niche variables considered simultaneously is
large, as there will tend to be no co-occurrences of so many As collection data is fundamentally tied to a taxonomic
classification, it is natural to describe the biota in terms of taxa
and consider the spatio-temporal distribution of a species for
example. For a data set that covers a spatial area A and time
interval T one may divide the space and interval into spatio-
temporal cells, (xa, tb) which form a mesh that partitions both the
geographic region and time interval. The labels xa and tb simply
indicate the particular spatio-temporal cell we are considering. A
point collection associated with this cell is such that it corresponds
to a latitude and longitude within the spatial cell xa and to a
collection date in the temporal cell tb. We can consider the
distribution of the set of species, B(xa, tb) = (B1(xa, tb), … BNB(xa,
tb)), where Bi(xa, tb) is a measure of the distribution of the ith taxon
in a spatial cell xa, in the time interval tb. A natural realization of
Bi (xa, tb) would be the abundance of the taxon i in the spatial cell
xa, in the time interval tb as measured by its frequency or relative
frequency. A less discriminating realization for Bi (xa, tb) would be
a function that indicates only presence or presence/absence in the PLoS ONE | www.ploson May 2009 | Volume 4 | Issue 5 | e5725 PLoS ONE | www.plosone.org 7 Biotic Interaction Networks Biotic Interaction Networks variables. This can be ameliorated by considering a reduced
number of both class and feature variables. For instance, P (Bi | Ik)
is determined by the number of co-occurrences of the taxon Bi and
the niche variable Ik and, in principle, allows us to find the most
important statistical associations between the niche variables and
the taxa distributions. However, P(Bi | Ik) being a probability does
not account for sample size. For example, if P(Bi | Ik) = 1 this may
be as a result of there being a coincidence of Bi and Ik in one
spatial cell or 1,000. Obviously, the latter is more statistically
significant. To remedy this we consider the following test statistic From either P(Bi |Bk) or e(Bi |Bk), an inferential interaction
network between taxa can be constructed where the nodes are the
taxa and the links represent the degree of statistical dependence of
one on the other. The links must represent the degree of
interaction as otherwise one has a uniform fully connected
network. This can be done, for instance, by only showing the
principle interactions above a certain threshold of e or P, or by
having the link width or size depend on their values. Note that
such an interaction network, being based on point collection data,
is inferential with respect to real biotic interactions between the
taxa. This is distinct to other networks where network links are
determined observationally. P (Bi |Bk) and e(Bi |Bk) are measures
of pair-wise dependencies between taxa. They can be generalized
to take into account higher order interactions. For instance, e(Bi
|Bk Bm) measures the statistical interaction between the joint
presence of taxa Bm and Bk and that of taxon Bi. e BijIk
ð
Þ~
NIj P BijIk
ð
Þ{P Bi
ð
Þ
ð
Þ
NIjP Bi
ð
Þ 1{P Bi
ð
Þ
ð
Þ
1=2
ð1Þ ð1Þ which measures the statistical dependence of Bi on Ik relative to
the null hypothesis that the distribution of Bi is independent of Ik
and randomly distributed over the grid, i.e., P Bi
ð
Þ~NBi=N,
where NBij is the number of grid cells with point collections of
species Bi and N is the total number of cells in the grid. The
sampling distribution of the null hypothesis is a binomial
distribution where, in this case, every cell is given a probability
P(Bi) of having a point collection of Bi. PLoS ONE | www.plosone.org Biotic Interaction Networks The quantitative values of e(Bi |Bk) can be interpreted in the
standard sense of hypothesis testing by considering the associated
p-value as the probability that |e(Bi |Bk)| is at least as large as the
observed one and then comparing this p-value with a required
significance level. In the case where NBj§5{10 then a normal
approximation for the binomial distribution should be a decent
approximation and in this case e(Bi |Bk) = 2 would represent the
standard 95% confidence interval. In the case where a normal
approximation is not accurate then other approximations to the
cumulative probability distribution of the binomial must be used. In the case where Ik = Bk, another taxon, then P(Bi |Bk) and e(Bi
|Bk) are measures of the statistical association between the two
taxa, e(Bi |Bk) having the added advantage of having built into it
the degree of statistical confidence that one may have about the
association. Note that such a statistical association does not
necessarily prove that there is a direct ‘‘causal’’ interaction
between the two taxa. Rather, it allows for a statistical inference
that may be validated subsequently. I9 In the case where Ik = Bk, another taxon, then P(Bi |Bk) and e(Bi
|Bk) are measures of the statistical association between the two
taxa, e(Bi |Bk) having the added advantage of having built into it
the degree of statistical confidence that one may have about the
association. Note that such a statistical association does not
necessarily prove that there is a direct ‘‘causal’’ interaction
between the two taxa. Rather, it allows for a statistical inference
that may be validated subsequently. The geographical region of interest for the data of the present
study is Mexico. Within this specified region there is an important
question of how to choose an appropriate mesh size. The right
degree of coarse graining is essentially governed by the size of the
data set available relative to the data necessary to construct a given PLoS ONE | www.plosone.org May 2009 | Volume 4 | Issue 5 | e5725 8 Biotic Interaction Networks Table 2. Relative rank by score of known reservoirs for
Leishmania in Mexico as a function of grid size. Biotic Interaction Networks Species
5 km
10 km
25 km
50 km
100 km
Didelphis marsupialis
52
31
40
17
22
Heteromys gaumeri
1
13
6
47
38
Sigmodon hispidus
17
19
27
50
90
Ototylomys phyllotis
2
5
22
60
40
Oryzomys melanotis
90
54
88
72
51
Peromyscus yucatanicus
3
10
28
84
62
Average Rank
27.50
22.00
35.17
55.00
50.50
z-score
212.54
225.93
215.48
216.69
216.91
doi:10.1371/journal.pone.0005725.t002 Table 2. Relative rank by score of known reservoirs for
Leishmania in Mexico as a function of grid size. probability function. For instance, to calculate P(Bi, Bk), where Bi
represents presence of species i in grid cell xa : If the mesh size is
too small then the probability of a co-occurrence of species i and k
is very small. On the other hand, if the mesh size is too big then, as
well as a lack of statistical significance, discrimination will also be
lost. A reasonable estimate of the appropriate cell size can be
determined by assuming that the N collections are distributed
randomly in an area A. An appropriate cell size is then A1/2/N,
which corresponds to having, on average, one collection per cell. Given that we are emphasizing here pairwise associations between
species, the appropriate value of N is the average number of
collections for any species. A more sophisticated methodology is to
consider the number of co-occurrences as a function of cell size
and look for the maximum of this function. This can be done for a
particular pair of species, or one may consider an average over
different pairs. For our study we used 3,337 square cells of linear
size 25 km which corresponds to an average number of point
collections of about 20. doi:10.1371/journal.pone.0005725.t002 values of epsilon will change as a function of cell size, principally
due to the effect of reducing the number of co-occurrences as one
passes to large cell sizes or to very small cell sizes. However,
relative values of epsilon will remain quite stable. Checks were made with other cell sizes of 5 km, 10 km, 50 km
and 100 km to assure the robustness of our conclusions. In Table 2,
for the ranked list of potential reservoirs we see how the average
position in the ranked list changes as a function of cell size. References 1. Lomolino M, Brown JH, Riddle B (2005) Biogeography. Sunderland, MA, USA:
Sinauer. 18. Wolfe, et al. (2007) Origins of major human infectious diseases. Nature 447:
279–283. 2. Volkov I, Banavar JR, Hubbell SP, Maritan A (2007) Patterns of relative species
abundance in rainforests and coral reefs. Nature 450: 45–49. 19. Chaves LF, Herna´ndez MJ, Dobson A, Pascual M (2007) Sources and sinks:
revisiting the criteria for identifying reservoirs for American cutaneous
Leishmaniasis. Trends Parasit 23: 311–316. 3. Peterson AT, Sanche´z-Cordero V, Beard CB, Ramsey JM (2002) Ecologic niche
modeling and potential reservoirs for Chagas disease in Mexico. Emerg Inf Dis
8: 662–667. 20. Biagi F, de Biagi AM, Beltra´n y F (1965) Phlebotomus flaviscutellatus, transmisor
natural de Leishmania mexicana. Prensa Me´dica Me´xico 30: 276, 272. 4. Montoya JM, Pimm SL, Sole´ RV (2006) Ecological networks and their fragility. 4. Montoya JM, Pimm SL, Sole´ RV (2006) Ecological networks and their fragility. Nature 442: 259 264 21. Rebollar-Te´llez E, Ramı´rez-Fraire A, Andrade-Narvaez y FJ (1996) A two years
study on vectors of cutaneous leishmaniasis. Evidence for sylvatic transmission in
the state of Campeche, Mexico. Memorias Instituto Oswaldo Cruz 91(5):
555–560. 5. Strogatz SH (2001) Exploring complex networks. Nature 410: 268–276. g
p
g
p
6. McCann K (2007) Protecting biostructure. Nature 446: 29. 6. McCann K (2007) Protecting biostructure. Nature 446: 29. 7. Thompson JN (1999) The evolution of species interactions. Science 284:
2116–2118. 22. Iba´n˜ez-Bernal S, Rodrı´guez-Domı´nguez G, Gomez-Herna´ndez CH, Ricardez-
Esquinca JR (2004) First record of Lutzomyia evansi in Mexico. Mem Inst
Oswaldo Cruz 99(2): 127–129. 8. Balashov YS (1984) Interaction between blood-sucking arthropods and its hosts,
and its influence on vector potential. Ann Rev Ent 29: 137–156. . Balashov YS (1984) Interaction between blood-sucking arthropods a and its influence on vector potential. Ann Rev Ent 29: 137–156. 23. Van Wynsberghe NR, Canto-Lara SB, Damia´n-Centeno AG, Itza´-Ortiz MF,
Andrade-Narva´ez FJ (2000) Retention of Leishmania mexicana in Naturally
Infected Rodents from the State of Campeche, Mexico. Mem Inst Oswaldo
Cruz 95(5): 595–600. 9. Lossos J (1996) Phylogenetic Perspectives on Community Ecology. Ecology 77:
1344–1354. 10. Sobero´n J, Peterson AT (2004) Biodiversity informatics: Managing and applying
primary biodiversity data. Phil Trans Roy Soc B 359: 689–698. 24. Canto-Lara SB, Van Wynsberghe NR, Vargas-Gonza´lez A, Ojeda-Farfan FF,
Andrade-Narva´ez FJ (1999) Use of monoclonal antibodies for the identification
of Leishmania spp. Biotic Interaction Networks This
shows that the relative ranking is quite insensitive to the cell size, as
the z-scores of the average rank of six of the known reservoirs
relative to the expected average rank if the distribution were
random are highly statistically significant. In other words, the
predictions as to which species are most likely to be reservoirs are
robust to large changes in the cell size. In general, the absolute Author Contributions Analyzed the data: CRS JG CG CI VSC CGS. Wrote the paper: CRS. Designed the models: CRS. References isolated from humans and wild rodents in the State of
Campeche, Mexico. Mem Inst Oswaldo Cruz 94(3): 305–9. 11. Elith J, et al. (2006) Novel methods improve prediction of species distributions
from occurrence data. Ecography 29: 129–151. from occurrence data. Ecography 29: 129–151 12. Ponder WF, Carter GA, Flemons P, Chapman RR (2001) Evaluation of
Museum Collection Data in Biodiversity Assessment. Cons Biol 15: 648–657. 25. Erika Ivett Sosa Bibiano (2004) Diagno´stico e identificacio´n del subge´nero de
Leishmania en mamı´feros silvestres mediante la te´cnica de reaccio´n en cadena de
la polimerasa (RCP). Me´rida, Yucata´n, Me´xico: Universidad Auto´noma de
Yucata´n, Facultad de Medicina, Unidad de Posgrado e Investigacio´n. En opcio´n
al tı´tulo de Maestro en Ciencias. 13. Lim BK, ATownsendPeterson, Engstrom MD (2004) Robustness of ecological
niche modeling algorithms for mammals in Guyana. J Biodiv Cons 11:
1237–1246. 14. Graham CH, Ferrier S, Huettman F, Moritz C, A Townsend Peterson (2004)
New developments in museum-based informatics and applications in biodiversity
analysis. Trends Ecol Evol 19: 497–503. 26. Hand D, Mannila H, Smyth P (2001) Principles of Data Mining. MA, USA:
MIT Press. y
15. Stockwell DRB, Peters D (1999) The GARP modeling system: problems and
solutions to automated spatial prediction. Int J Geogr Inform Sci 13: 143–158. 27. Hall ER (1981) The Mammals of North America, Vol. 1 and 2. NY: Ronald
Press. 16. Peterson AT, Papes M, Eaton M (2007) Transferability and model evaluation in
ecological niche modeling: a comparison of GARP and Maxent. Ecography 30:
550–560. 28. Guevara-Chumacero L, Lo´pez-Wilchis R, Sa´nchez-Cordero V (2001) 105 an˜os
de investigacio´n mastozoolo´gica en Me´xico (1890–1995): una revisio´n de sus
enfoques y tendencias. Acta Zool Mex, N S 83: 35–72. 17. Krebs CJ, et al. (1995) Impact of Food and Predation on the Snowshoe Hare
Cycle. Science 269: 1112–1115. PLoS ONE | www.plosone.org May 2009 | Volume 4 | Issue 5 | e5725 9
|
https://openalex.org/W4361885713
|
https://figshare.com/articles/journal_contribution/Supplemental_Figure_4_from_Mitochondrial_Inhibition_Augments_the_Efficacy_of_Imatinib_by_Resetting_the_Metabolic_Phenotype_of_Gastrointestinal_Stromal_Tumor/22464080/1/files/39915398.pdf
|
English
| null |
Supplemental Figure 3 from Mitochondrial Inhibition Augments the Efficacy of Imatinib by Resetting the Metabolic Phenotype of Gastrointestinal Stromal Tumor
| null | 2,023
|
cc-by
| 563
|
Supplemental Figure 4. BUN (mg/dL)
CREA (mg/dL)
BUN/CREA ratio
P (mg/dL)
Ca (mg/dL)
GLU (mg/dL)
CO2 (mEq/L)
Na (mEq/L)
K (mEq/L)
Cl (mEq/L)
0
1
3
50
100
150
200
300
400
500
Basic Metabolic Panel
Control
VLX
Imatinib
Combination
ALP (U/L)
ALT (U/L)
AST (U/L)
T BILI (mg/dL)
TOTAL PROTEIN (g/dL)
ALB (g/dL)
0
1
2
50
100
150
200
250
Hepatic Panel
Control
VLX
Imatinib
Combination
CD45
Macrophages
Monocytes
Neutrophils
CD3
CD4
CD8
NK
0
1
1.5
10
20
30
80
90
100
Percent of CD45
Immune Cell Frequencies
Control
VLX
Imatinib
Combination
BUN (mg/dL)
CREA (mg/dL)
BUN/CREA ratio
P (mg/dL)
Ca (mg/dL)
GLU (mg/dL)
CO2 (mEq/L)
Na (mEq/L)
K (mEq/L)
Cl (mEq/L)
0
10
20
30
100
150
200
250
300
350
Basic Metabolic Panel
Control
VLX
Imatinib
Combination
Supplemental Figure 4. A
B
upplemental Figure 4. Serum metabolic chemistries (top, middle) and splenic immune cell frequencies
erformed by flow cytometry (bottom) in (A) KitV558del/+ and (B) NOD.Cg-Prkdcscid Il2rgtm1Wjl/SzJ (NSG) mice
=3-5 mice per group. KitV558del/+ mouse
NSG mouse BUN (mg/dL)
CREA (mg/dL)
BUN/CREA ratio
P (mg/dL)
Ca (mg/dL)
GLU (mg/dL)
CO2 (mEq/L)
Na (mEq/L)
K (mEq/L)
Cl (mEq/L)
0
1
3
50
100
150
200
300
400
500
Basic Metabolic Panel
Control
VLX
Imatinib
Combination
ALP (U/L)
ALT (U/L)
AST (U/L)
T BILI (mg/dL)
TOTAL PROTEIN (g/dL)
ALB (g/dL)
0
1
2
50
100
150
200
250
Hepatic Panel
Control
VLX
Imatinib
Combination
CD45
Macrophages
Monocytes
Neutrophils
CD3
CD4
CD8
NK
0
1
1.5
10
20
30
80
90
100
Percent of CD45
Immune Cell Frequencies
Control
VLX
Imatinib
Combination
BUN (mg/dL)
CREA (mg/dL)
BUN/CREA ratio
P (mg/dL)
Ca (mg/dL)
GLU (mg/dL)
CO2 (mEq/L)
Na (mEq/L)
K (mEq/L)
Cl (mEq/L)
0
10
20
30
100
150
200
250
300
350
Basic Metabolic Panel
Control
VLX
Imatinib
Combination
Supplemental Figure 4.
A
B
upplemental Figure 4. Serum metabolic chemistries (top, middle) and splenic immune cell frequencie
erformed by flow cytometry (bottom) in (A) KitV558del/+ and (B) NOD.Cg-Prkdcscid Il2rgtm1Wjl/SzJ (NSG) m
=3-5 mice per group.
KitV558del/+ mouse
NSG mouse BUN (mg/dL)
CREA (mg/dL)
BUN/CREA ratio
P (mg/dL)
Ca (mg/dL)
GLU (mg/dL)
CO2 (mEq/L)
Na (mEq/L)
K (mEq/L)
Cl (mEq/L)
0
1
3
50
100
150
200
300
400
500
Basic Metabolic Panel
Control
VLX
Imatinib
Combination
A
KitV558del/+ mouse BUN (mg/dL)
CREA (mg/dL)
BUN/CREA ratio
P (mg/dL)
Ca (mg/dL)
GLU (mg/dL)
CO2 (mEq/L)
Na (mEq/L)
K (mEq/L)
Cl (mEq/L)
0
1
3
50
100
150
200
300
400
500
Basic Metabolic Panel
Control
VLX
Imatinib
Combination
BUN (mg/dL)
CREA (mg/dL)
BUN/CREA ratio
P (mg/dL)
Ca (mg/dL)
GLU (mg/dL)
CO2 (mEq/L)
Na (mEq/L)
K (mEq/L)
Cl (mEq/L)
0
10
20
30
100
150
200
250
300
350
Basic Metabolic Panel
Control
VLX
Imatinib
Combination
A
B
KitV558del/+ mouse
NSG mouse B A Control
VLX
Imatinib
Combination ALP (U/L)
ALT (U/L)
AST (U/L)
T BILI (mg/dL)
TOTAL PROTEIN (g/dL)
ALB (g/dL)
0
1
2
50
100
150
200
250
Hepatic Panel
Control
VLX
Imatinib
Combination Hepatic Panel Control
VLX
Imatinib
Combination CD45
Macrophages
Monocytes
Neutrophils
CD3
CD4
CD8
NK
0
1
1.5
10
20
30
80
90
100
Percent of CD45
Immune Cell Frequencies
Control
VLX
Imatinib
Combination Control
VLX
Imatinib
Combination Supplemental Figure 4. Serum metabolic chemistries (top, middle) and splenic immune cell frequencies
performed by flow cytometry (bottom) in (A) KitV558del/+ and (B) NOD.Cg-Prkdcscid Il2rgtm1Wjl/SzJ (NSG) mice. n=3-5 mice per group.
|
https://openalex.org/W2158306844
|
https://zenodo.org/record/3264349/files/166.pdf
|
English
| null |
Authorization Using the Publish-Subscribe Model
| null | 2,008
|
cc-by
| 7,840
|
Authorization Using the Publish-Subscribe Model
Qiang Wei, Matei Ripeanu, and Konstantin Beznosov
Dept. of Electrical and Computer Engineering, University of British Columbia
Vancouver, BC, Canada
{qiangw, matei, beznosov}@ece.ubc.ca client
subject
application server
authorization
server
policy
decision point
(PDP)
resource
policy
enforcement point
(PEP)
application
request
application
response
authorization
request
authorization
response
Figure 1. The request-response model. Abstract Traditional authorization mechanisms based on the
request-response model are generally supported by point-
to-point communication between applications and autho-
rization servers. As distributed applications increase in
size and complexity, an authorization architecture based
on point-to-point communication becomes fragile and dif-
ficult to manage. This paper presents the use of the
publish-subscribe (pub-sub) model for delivering autho-
rization requests and responses between the applications
and the authorization servers. Our analysis suggests that
using the pub-sub architecture improves authorization sys-
tem availability and reduces system administration over-
head. We evaluate our design using a prototype implemen-
tation, which confirms the improvement in availability. Al-
though the response time is also increased, this impact can
be reduced by bypassing the pub-sub channel when return-
ing authorizations or by caching coupled with local infer-
ence of authorization decisions based on previously cached
authorizations. Figure 1. The request-response model. the same time, large-scale commodity computing is becom-
ing a reality, with eBay having over 12,000 servers and
15,000 application server instances [23], and Google esti-
mated to have “more than 450,000 servers spread in at least
25 locations around the world” [17]. When a distributed
system evolves to this scale, a tightly-coupled point-to-point
architecture suffers from two drawbacks: fragility and high
management overhead. Fragility occurs in a large-scale, point-to-point architec-
ture when authorization service is unavailable. In that case,
all applications depending on that authorization server may
not work properly, if at all. The authorization server may be
unavailable due to a failure (transient, intermittent, or per-
manent) of the network, of the software located in the crit-
ical path (e.g., OS), of the hardware, or even due to a mis-
configuration of the supporting infrastructure. A conven-
tional approach to improving the availability of a distributed
infrastructure is failure masking through redundancy (ei-
ther information, time, or hardware). However, redundancy
and other general purpose fault-tolerance techniques for dis-
tributed systems scale poorly and become technically and
economically infeasible when the number of entities in the
system reaches thousands [14, 25]. 1
Introduction Modern access control solutions are based on the
request-response model [15, 11, 18, 21, 19, 10], as illus-
trated in Figure 1. In this model, a policy enforcement point
(PEP) intercepts application requests, obtains access con-
trol decisions (a.k.a. authorizations) from a policy deci-
sion point (PDP), and enforces those decisions. The sep-
aration into PEP and PDP in the request-response model
enables using PDPs in the form of authorization servers,
thereby reusing the authorization logic and enforcing con-
sistent policies across multiple PEPs. Additionally, management of large-scale authorization
systems with a point-to-point architecture becomes increas-
ingly costly. Due to the coupling between PEPs and PDPs,
even relatively simple changes made on a PDP might be
difficult to implement. Implementation may involve several The request-response model, however, commonly leads
to a point-to-point communication architecture, where PEPs
obtain decisions from PDPs through synchronous remote
procedure calls (RPC). This point-to-point architecture, in
turn, results in tight coupling between PEPs and PDPs. At COM+ and EJB). A PEP can also be a part of the corre-
sponding application resource, e.g., implemented via static
or dynamic “weaving” using aspect oriented software devel-
opment techniques [16]. The PDP is usually implemented
in the form of authorization servers. It can be designed
specifically for the application or use third party compo-
nents. The PDP stores policy information which is usually
specified by the security administrator. different application development teams sitting down to co-
ordinate and make sure that the necessary changes do not
break anything else. For instance, a failed PDP needs to be
brought back up and possibly relocated. Each PEP depend-
ing on that PDP needs to update its information to reflect
the PDP’s relocation. This paper presents an authorization system that uses
a publish-subscribe (pub-sub) architecture to replace the
existing point-to-point architecture. Our goals are to re-
duce system administration overhead, thus improving sys-
tem manageability, and to increase the system availabil-
ity, thus improving overall system robustness. Unlike in
a point-to-point architecture, where PEPs are configured to
send their requests to specific PDPs, a pub-sub architecture
enables PEPs to send their requests without knowing which
PDP will receive them. Similarly, the PDPs show interest in
requests without knowing which PEPs generate them. By
using the pub-sub architecture, the coupling between PEPs
and PDPs is removed; as a result, system availability is im-
proved and system administration is simplified. 1
Introduction We distinguish between an application request, which is
generated by the subject and is dependent on the application
logic, and an authorization request, which is generated by
the PEP and is independent from the application logic. This
decoupling, for instance, is performed by the context han-
dler in the XACML-compliant PEP [28]. The context han-
dler generates an XACML request context, which is sent to
the PDP for processing. We define the authorization request
as a tuple (s, o, a), where s is the subject, o is the object, a
is the access right and the authorization response as a tuple
(r, d), where r is the request and d is the decision. This paper makes the following contributions. First, we
present the design of an authorization system based on the
pub-sub architecture and examine the requirements for the
supporting event notification service (ENS). Second, we de-
velop a prototype using an open-source ENS and evaluate it
experimentally. Our evaluation suggests that the pub-sub
architecture improves the availability of the authorization
infrastructure. Although the response time is increased, this
overhead can be reduced by returning authorizations by-
passing the ENS, or by caching coupled with local infer-
ence of authorization decisions based on cached authoriza-
tions [8]. 2.2
The Publish-Subscribe Model A pub-sub model is a common communication model
in large-scale enterprise applications, enabling loosely cou-
pled interaction between entities whose location and behav-
iors may vary throughout the lifetime of the system [12]. Generally, entities that send messages “publish” them as
events, while entities that wish to receive certain events
“subscribe” to those events. Often, an entity may become
both a publisher and subscriber, sending and receiving mes-
sages within the system. The rest of this paper is organized as follows. Section 2
presents background information on access control archi-
tectures and the pub-sub model. Section 3 describes our
system design and analysis. Section 4 evaluates our proto-
type implementation. Section 5 discusses related work. We
conclude in Section 6. Pub-sub systems usually provide one of two types
of subscription schemes:
topic-based subscription and
content-based subscription. In a topic-based scheme, a
message belongs to one of a fixed set of topics. A subscrip-
tion targets a topic, and the subscriber receives all events
that are associated with that topic. A content-based scheme
is not constrained to the notion that each message must be-
long to a particular topic. Instead, the message delivery de-
cisions are based on a query or predicate issued by a sub-
scriber. The advantage of a content-based scheme is its flex-
ibility: it provides the subscriber with the ability to specify
just the information he/she needs without having to learn a
set of topic names and their content before subscribing. 2
Background This section provides background on the request-
response authorization model and on the pub-sub model. 3.1
System Architecture and Data Flow PEP to wait for all of them. Therefore, once the PEP re-
ceives a response to a request, it immediately unsubscribes
to the response for that request (step 10). The architecture of the pub-sub authorization system is
illustrated in Figure 2(a). The system consists of multiple
PEPs, PDPs and a logically centralized event notification
service (ENS). The ENS mediates the communications be-
tween PEPs and PDPs, thereby fully decoupling PEPs and
PDPs. The ENS is responsible for: (1) receiving events
from publishers, (2) receiving subscriptions from event sub-
scribers, and (3) matching each event to subscriptions and
routing the event, in the form of notifications, to the inter-
ested subscribers. Note that we assume that policies are consistent across
PDPs, so that each PEP can enforce the decision made by
any PDP in the system. When they are not consistent (due
to policy changes), our solution does not make the system
worse than the point-to-point system. To handle the incon-
sistency issue, one possible solution is for each response to
include a time-stamp which represents the time that the PDP
receives a policy change. When the PEP receives inconsis-
tent responses, it always uses the one with the most recent
time-stamp. PEPs and PDPs can be both subscribers and publishers. A PEP subscribes to the responses published by the PDP. This can be achieved simply by passing the request as the
parameter of the subscribe() operation of the ENS for each
incoming request. A PDP subscription shows the ability of
the PDP to resolve certain requests. A PDP can precisely
subscribe to all the request tuples (s, o, a) it can resolve,
or subscribe to those requests that contain the objects for
which it is responsible, in order to reduce the number of
subscriptions it creates. Figure 2. Publish-subscribe architecture for delivering authorization requests and responses. Figure 2. Publish-subscribe architecture for delivering authorization requests and responses. 3
System Design and Analysis In the request-response model, PEP intercepts applica-
tion requests from subjects and enforces decisions from the
PDP. Subjects are the processes that make application re-
quests on behalf of users. A PEP can be a security inter-
ceptor (as in CORBA Security, ASP.NET, and most Web
servers), or can be a part of the component container (as in This section presents the design and analysis of an au-
thorization system based on a pub-sub architecture. This
replaces the point-to-point communication used between
PEPs and PDPs in existing authorization systems. PEP
PEP
PEP
PDP
PDP
event
notification
service
(ENS)
notification
notification
notification
notification
notification
subscribe,
publish
subscribe,
publish
subscribe,
publish
subscribe,
publish
subscribe,
publish
(a) System architecture. PEP
PDP
event
notification
service
(ENS)
1. subscribe to
all its requests
2. generate a
request
3. subscribe to
the response
for the request
4. publish the
request
5. notify the
matching of
a request
6. compute the
response
7. publish the
response
8. notify the
matching of a
response
9. enforce the response
10. unsubscribe
to the received
response
(b) Typical data flow. Figure 2. Publish-subscribe architecture for delivering authorization requests and responses. PEP
PEP
PEP
PDP
PDP
event
notification
service
(ENS)
notification
notification
notification
notification
notification
subscribe,
publish
subscribe,
publish
subscribe,
publish
subscribe,
publish
subscribe,
publish
(a) System architecture. PEP
PDP
event
notification
service
(ENS)
1. subscribe to
all its requests
2. generate a
request
3. subscribe to
the response
for the request
4. publish the
request
5. notify the
matching of
a request
6. compute the
response
7. publish the
response
8. notify the
matching of a
response
9. enforce the response
10. unsubscribe
to the received
response
(b) Typical data flow. event
notification
service
(ENS) event
notification
service
(ENS) (a) System architecture. (b) Typical data flow. (b) Typical data flow. 3.4.1
Bypassing the ENS when returning responses The first technique requires a modification to the data flow. Instead of returning responses through the ENS, the PDP re-
turns them directly to the PEP, as shown in Figure 3(a). To
preserve the decoupling property of the pub-sub architec-
ture, each request contains the address of the PEP that pub-
lishes it. Each PDP then can make a remote call to return
the response without knowing the PEP address in advance. • Performance. In the pub-sub architecture, the inser-
tion of an additional software component (the ENS)
between PEPs and PDPs, imposes run-time overhead,
which may in turn degrade application performance. In particular, the ENS needs to spend additional time to
find the potential subscribers for each message and to
route the message to them. Therefore, the ENS should
be efficient in finding a match even with a large num-
ber of subscriptions. Otherwise, it may become the
performance bottleneck. When the time used by the
ENS is far lower than the network latency or the com-
putational cost at the PDP, its impact on the overall
performance is negligible. Additionally, Section 3.4
presents two performance-oriented optimizations that
reduce ENS overhead and improve performance. • Performance. In the pub-sub architecture, the inser-
tion of an additional software component (the ENS)
between PEPs and PDPs, imposes run-time overhead,
which may in turn degrade application performance. In particular, the ENS needs to spend additional time to
find the potential subscribers for each message and to
route the message to them. Therefore, the ENS should
be efficient in finding a match even with a large num-
ber of subscriptions. Otherwise, it may become the
performance bottleneck. When the time used by the
ENS is far lower than the network latency or the com-
putational cost at the PDP, its impact on the overall
performance is negligible. Additionally, Section 3.4
presents two performance-oriented optimizations that
reduce ENS overhead and improve performance. We expect that returning responses without the ENS can
improve the system performance for the following two rea-
sons: 1) the number of calls involved in resolving a request
is reduced; and 2) the load of the ENS is also reduced. Our
evaluation results in Section 4.4 confirm that this approach
improves the performance. 3.2
Benefits Using the pub-sub model to build authorization systems
is expected to provide the following benefits: • Increased availability. In the point-to-point architec-
ture, a PEP is generally configured to send a request
to only one PDP. In existing enterprise systems, how-
ever, the same data often resides in multiple locations,
on multiple machines, and within a variety of appli-
cations. Consequently, multiple PDPs may have to
be setup to resolve access requests for the same data. Our solution exploits this situation: using the pub-sub
architecture enables a request to reach all PDPs that
are able to resolve it. Even though some of the PDPs
may fail, the chances that at least one will provide a
response on time are higher. In other words, our de-
sign allows the collection of PDPs to appear as a single
large reliable PDP matrix. Figure 2(b) shows a typical data flow. Each PDP first
subscribes to all the requests it can resolve when it boots
(step 1). After a PEP intercepts an application request, it
generates an authorization request (step 2). It first sub-
scribes to the response to the request (step 3) and then im-
mediately publishes the request (step 4). If some PDP can
resolve the request, the ENS will find a match and notify
that PDP of the request (step 5). After the PDP computes
a response (step 6), it immediately publishes the response
(step 7). The ENS then notifies the PEP of the response
(step 8) and the PEP finally enforces the authorization deci-
sion (step 9). Although multiple PDPs may resolve the same
request and publish the response, it is not necessary for the • Reduced administration overhead. We expect that
decoupling PEPs and PDPs will reduce the human • Security. Using a pub-sub architecture introduces new
threats to the system. We focus on two attacks. First,
performance attacks: a malicious participant can at-
tempt to lower the system performance. For exam-
ple, a malicious publisher can send a large number of
junk events to the ENS to increase its load, and a ma-
licious subscriber can register a large number of junk
subscriptions in the ENS to increase its time in finding
an interested subscriber. Second, correctness attacks:
a malicious participant can subscribe to any ongoing
request and publish false responses. If these false re-
sponses are used by the PEP, the system’s correctness
is compromised. • Security. 3.4
Performance-Oriented Optimizations • Robustness. The ENS itself should be robust to
achieve the benefits of improving the overall system
availability. Otherwise, it becomes the single point of
failure. Therefore, the ENS should only be logically
centralized, but have a robust implementation. The ENS introduces additional overhead to process sub-
scriptions and deliver events. This section proposes two
approaches to reduce the impact of ENS overhead on the
overall system performance. Exiting pub-sub systems usually provide robust ENS
implementations, where the ENS is implemented not
only by a single server but also by a set of federated
servers to reduce the single point of failure problem. For instance, in Siena [7] and Elvin [22], two wide-
area content-based pub-sub networks, the underlying
pub-sub infrastructure can be implemented as a collec-
tion of network servers communicating with each other
in a peer-to-peer fashion. The failure of a server will
not impact the system performance significantly. 3.2
Benefits Using a pub-sub architecture introduces new
threats to the system. We focus on two attacks. First,
performance attacks: a malicious participant can at-
tempt to lower the system performance. For exam-
ple, a malicious publisher can send a large number of
junk events to the ENS to increase its load, and a ma-
licious subscriber can register a large number of junk
subscriptions in the ENS to increase its time in finding
an interested subscriber. Second, correctness attacks:
a malicious participant can subscribe to any ongoing
request and publish false responses. If these false re-
sponses are used by the PEP, the system’s correctness
is compromised. costs of operating and administering authorization in-
frastructures. Consider the previous example of a
failed PDP. After it is brought back and possibly re-
located, only the ENS might need to be reconfigured
rather than the case of all the PEPs depending on the
failed PDP as in a system based on point-to-point com-
munication. • Improved software development process. Using an
event-driven, standards-based pub-sub channel to pro-
vide a comprehensive communication framework be-
tween PEPs and PDPs will improve the software de-
velopment cycle. The integration of PEPs and PDPs is
faster and less expensive using a pub-sub channel than
using a point-to-point architecture. To prevent these attacks, the ENS needs to provide an
authentication and authorization mechanism to allow
only authorized parties to subscribe or publish. Most
existing commercial pub-sub systems provide such a
mechanism. If the ENS does not have this feature, the
PEP will need to verify each response to ensure its cor-
rectness. Detailed analysis of the security model and
solutions will be the subject of our future work. 3.4.1
Bypassing the ENS when returning responses Note that this approach prevents
cooperative caching at the ENS, in which the ENS caches
responses made by PDPs and uses them to resolve the re-
quests from PEPs. 3.3
Requirements for the ENS The ENS is the core component of our architecture. For
the pub-sub system to provide the aforementioned benefits,
the ENS should meet the following requirements: 4.1
Prototype Implementation (SAAM) [8]. SAAM adds a secondary decision point (SDP)
to the request-response model. The SDP is usually collo-
cated with the PEP and caches the returned authorizations
from the PDP. When the PDP is overloaded or unavail-
able, the SDP can resolve authorization requests not only by
reusing previous cached authorizations (a.k.a. precise recy-
cling) but also by inferring new authorizations from cached
authorizations (a.k.a. approximate recycling). Therefore,
the SDP provides an alternative decision making source. We implemented a prototype of the proposed pub-sub
authorization system on top of Siena [7], a content-based
ENS developed in Java. We chose Siena for two reasons:
first, it provides the necessary functionality to study the fea-
sibility and performance; second, it allows us to customize
the code for further development and investigation. In Siena, events are published as a set of attribute and
value pairs. Attribute names are simply strings, and values
are from a predefined set of primitive types, for which a
fixed set of operators is defined. The following sample code
shows how a PEP publishes a request to Siena. Figure 3(b) shows the pub-sub authorization architecture
after adding the SDP. The SDP first tries to resolve the re-
quest from the PEP. If the SDP fails, it then publishes the
request to the ENS and subscribes to the response. The
returned response is added to the cache maintained by the
SDP. As the cache size increases, more requests can be re-
solved by the local SDP, thus reducing the time to obtain
responses and the load of the ENS. Notification e = new Notification();
e.putAttribute("subject", "Sean");
e.putAttribute("object", "/etc/passwd");
e.putAttribute("access", "read");
siena.publish(e); Notification e = new Notification();
e.putAttribute("subject", "Sean");
e.putAttribute("object", "/etc/passwd");
e.putAttribute("access", "read");
siena.publish(e); One challenge of introducing an SDP is maintaining
cache consistency. If the SDP is unaware of the policy up-
date at the PDP, it may compute false responses. The pub-
sub architecture provides an efficient way to propagating
policy update messages: the SDP can subscribe to the pol-
icy update messages published by the PDP; after removing
the affected cache, the SDP can publish a result message to
the ENS indicating whether or not the update is successful. The subscriber subscribes to events by specifying fil-
ters using the subscription language. 4.1
Prototype Implementation The filters define
constraints, usually in the form of name-value pairs of at-
tributes and basic comparison operators (=, <, ≤, >, ≥),
which identify valid events. The following sample code
shows how a PDP subscribes to the request for an object
in Siena. Figure 3. Performance-oriented optimization techniques. Figure 3. Performance-oriented optimization techniques. 3.4.2
Integration with authorization recycling Another technique for improving system performance
is by integrating the pub-sub authorization system with
the Secondary and Approximate Authorization Model PEP
PDP
ENS
1. subscribe to
all its requests
2. generate a
request
3. publish the
request
4. notify the
matching of
a request
5. compute the
response
7. enforce the response
6. return the
response to
PEP directly
(a) Technique I: Changed data flow to improve performance: the PDP
returns responses to the PEP directly, bypassing the ENS. PEP
PEP
PEP
PDP
PDP
ENS
SDP
SDP
SDP
(b) Technique II: A secondary decision point (SDP) is added to the
pub-sub system. Figure 3. Performance-oriented optimization techniques. PEP
PDP
ENS
1. subscribe to
all its requests
2. generate a
request
3. publish the
request
4. notify the
matching of
a request
5. compute the
response
7. enforce the response
6. return the
response to
PEP directly
(a) Technique I: Changed data flow to improve performance: the PDP
returns responses to the PEP directly, bypassing the ENS. PEP
PEP
PEP
PDP
PDP
ENS
SDP
SDP
SDP
(b) Technique II: A secondary decision point (SDP) is added to the
pub-sub system. PEP
PEP
PEP
PDP
PDP
ENS
SDP
SDP
SDP
(b) Technique II: A secondary decision point (SDP) is added to the
pub-sub system. (a) Technique I: Changed data flow to improve performance: the PDP
returns responses to the PEP directly, bypassing the ENS. (b) Technique II: A secondary decision point (SDP) is added to the
pub-sub system. 4.3
Evaluating Availability We expect that using the pub-sub architecture leads to
increased availability in the presence of PDP failures as a
request can reach multiple PDPs. Our evaluation aimed to
quantify this effect. We used the percentage decrease of
failed requests perceived by the PEP as a metric to mea-
sure the increase in availability. In each experiment, we first
measured the number of failed requests in the point-to-point
implementation, and then repeated the same experiment and
measured the failed requests in the pub-sub implementation
with various configurations. Finally we calculated the per-
centage difference in the number of failed requests of two
implementations. An increase in this difference indicates
an improvement of availability as more requests have been
successfully resolved. Second, we studied the impact of the overlap rate be-
tween PDPs on the percentage decrease of failed requests. In the experiment, we varied the overlap rate from 0% to
100% while fixing the number of PDPs at 4 (Figure 4(b)). As expected, the larger the overlap rate, the larger the per-
centage decrease of failed requests. When the overlap rate
was 100%, the percentage decrease of failed requests was
also 100%, which means all requests have been resolved. In addition, the results indicate that an increase of overlap
rate provided diminishing returns. For instance, the first
10% overlap rate increase (from 0% to 10%) led to a 23%
improvement, while the last 25% increase (from 75% to
100%) led to only 2% improvement. This suggests that the
benefit of the pub-sub approach is most remarkable with a
small resource overlap between PDPs. We studied the influence of the following two factors on
the percentage decrease of failed requests: (a) the number of
PDPs and (b) the overlap rate between the resource spaces
of two PDPs, defined as the ratio of the objects owned by
both PDPs to the objects owned only by the studied PDP. The overlap rate served as a measure of similarity between
the resources of two PDPs. 4
Experimental Evaluation Filter f = new Filter();
f.addConstraint("object", "/etc/passwd");
siena.subscribe(f); Filter f = new Filter();
f.addConstraint("object", "/etc/passwd");
siena.subscribe(f); The previous section presents the design of the autho-
rization system based on a pub-sub architecture. This sec-
tion presents an experimental evaluation of our design. Our
evaluation sought to understand the availability gain of us-
ing the pub-sub model and studied the impact of the ENS
on system performance in terms of response time. Matching in Siena is accomplished by a Binary Decision
Diagram. The routing paths for notifications are set at the PDPs. Each PDP then booted, subscribing to all the objects
for which it is responsible. The PEP then started sending
5,000 randomly generated requests sequentially. Each re-
quest was resolved with the flow described in Section 3 (see
Figure 2(b)). time of subscription. In a distributed Siena deployment, a
new subscription is stored and forwarded from the originat-
ing server to all the servers in the network. This forms a
tree that connects the subscriber with servers. Notifications
are then routed towards the subscriber following the reverse
path of the tree. To simulate the failure of PDPs, each PDP switched be-
tween two modes: work mode and failure mode. In the
work mode, the PDP computed and published a response af-
ter receiving a request. In the failure mode, the PDP simply
ignored the request notification from Siena and did noth-
ing. Therefore, a time-out request perceived by the PEP
indicated the failure of the PDP. We used the number of
elapsed requests as a measurement of time. We ensured that
all the PDPs received all requests at the same pace so that
their time clock was always synchronized. We used time-to-
failure (TTF) to represent the time between two consequent
failures and time-to-repair (TTR) to represent the time a
failure lasted. We used the exponential distribution to sim-
ulate both times. The mean TTF (MTTF) was 100 requests
and the mean TTR (MTTR) was 10 requests. We controlled
the maximum TTF to be 500 requests and the maximum
TTR to be 50 requests. As a basis for comparison, we also implemented a sim-
ple authorization system using a point-to-point architecture. The PDP was implemented as an JAVA Remote Method In-
vocation (RMI) service, and the PEP obtained authoriza-
tions from the PDP over RMI calls. 4.2
Evaluation Platform The experimental pub-sub authorization system con-
sisted of one PEP, one or more PDPs, and a centralized
Siena ENS server. We assumed that all these components
trust each other. The PEP was responsible for randomly
generating requests and performing system measurement. The PDP resolved requests using a role-based access con-
trol (RBAC) policy. The Siena ENS mediated the commu-
nication between PEPS and PDPs. All these components
ran in separate cluster nodes, each equipped with four Intel
Xeon 2.33 GHz processors and 4 GB of memory, running
Fedora Linux 2.6.24.3. First, we studied the impact of the number of PDPs on
the percentage decrease of failed requests. In the exper-
iment, we varied the number of PDPs from 1 to 6 while
fixing the overlap rate between PDPs at 10%. Figure 4(a)
shows the results. As expected, the larger the number of the
PDPs, the larger the percentage decrease of failed requests,
which implies that the availability is improved. The reason
is that the possibility to receive a response from other PDPs
increased even when the studied PDP failed. Figure 4. Availability results. Figure 4. Availability results. as a metric of performance. The response time is measured
as the time elapsed after the PEP generates a request until it
receives the response for that request. In these experiments,
we did not simulate the failure of PDPs. The PEP recorded
the response time for each request. After every 500 requests
the PEP calculated the mean response time and used it as an
indicator of the response time for that period. 0
1
2
3
4
5
2
3
4
5
6
7
8
9
10
response time (ms)
number of requests (x500)
pub-sub
pub-sub-RMI
point-to-point
Figure 5. Response time comparison. Pub-
Sub-RMI is the architecture in which the PDP
returns responses directly to the PEP, by-
passing the ENS. 0
1
2
3
4
5
2
3
4
5
6
7
8
9
10
response time (ms)
number of requests (x500)
pub-sub
pub-sub-RMI
point-to-point We have implemented two variations of the pub-sub ar-
chitecture proposed in Section 3.4 to improve the system
performance. In the first architecture, the PEP included its
address in the request and the PDP returned response di-
rectly to that address over a Java RMI call (Figure 3(a)). We refer this approach as pub-sub-RMI. In the second ar-
chitecture, we integrated pub-sub with SAAM by adding an
SDP implementation [26] which is responsible for recycling
previous authorizations and providing secondary authoriza-
tions to the PEP (Figure 3(b)). Figure 5. Response time comparison. Pub-
Sub-RMI is the architecture in which the PDP
returns responses directly to the PEP, by-
passing the ENS. Figure 5 compares the response time of the original
pub-sub architecture (top curve), pub-sub-RMI architec-
ture (middle curve) and point-to-point architecture (bottom
curve). As expected, the pub-sub architecture led to longer
response time, i.e., by 1.6ms on average, compared with the
point-to-point architecture. In addition, using pub-sub-RMI
helps to reduce the pub-sub response time, i.e., by 0.8ms. subscribe” call (step 10 in Figure 2(b)) by the PEP, and thus
the number of subscriptions kept increasing with the num-
ber of requests. Figure 6(b) shows that, without the “un-
subscribe” call, the response time increased almost linearly
with the number of requests, while with the “unsubscribe”
call, the response time remains the same. This suggests that
it is important for PEPs to remove stale subscriptions. 4.4
Evaluating Performance In evaluating performance, our goal was to understand
how the use of the ENS impacts the system performance
and how our two proposed performance improvement tech-
niques help to mitigate this impact. We used response time In the experiment, a RBAC policy was first generated
for each PDP. Each policy contained 100 subjects, 1000 ob-
jects, and 3 access rights. We controlled the object space of
each PDP to simulate certain overlap rates between different 0
20
40
60
80
100
0
20
40
60
80
100
percentage decrease of failed requests (%)
overlap rate (%)
(b) Percentage decrease of failed requests as a function of overlap rate
when there are 4 PDPs. 0
20
40
60
80
100
1
2
3
4
5
6
percentage decrease of failed requests (%)
number of PDPs
(a) Percentage decrease of failed requests as a function of the number
of PDPs when the overlap rate is 10%. 0
20
40
60
80
100
0
20
40
60
80
100
percentage decrease of failed requests (%)
overlap rate (%)
(b) Percentage decrease of failed requests as a function of overlap rate
when there are 4 PDPs. Figure 4. Availability results. 0
20
40
60
80
100
0
20
40
60
80
100
percentage decrease of failed requests (%)
overlap rate (%)
(b) Percentage decrease of failed requests as a function of overlap rate
when there are 4 PDPs. 0
20
40
60
80
100
1
2
3
4
5
6
percentage decrease of failed requests (%)
number of PDPs 0
20
40
60
80
100
0
20
40
60
80
100
percentage decrease of failed requests (%)
overlap rate (%) (b) Percentage decrease of failed requests as a function of overlap rate
when there are 4 PDPs. (a) Percentage decrease of failed requests as a function of the number
of PDPs when the overlap rate is 10%. 4.5
Discussion to 4. Each PDP subscribed to 1,000 identical objects (with
100% overlap rate). The result shows that an increase of
the number of PDPs almost has no impact on the response
time. A possible reason is that Siena optimized the match-
ing algorithm for the situation when multiple subscribers
have common interests (e.g., by storing only one copy of
the subscription). Evaluating the case when PDPs post di-
verse subscriptions is subject to our future work. Our evaluation confirms the analysis in Section 3.2: the
use of pub-sub model leads to higher availability, and this
improvement increases with the number of PDPs and their
resource overlap rate. In particular, when the overlap rate
is small, even a slight increase in the overlap rate leads to a
significant improvement in availability. Therefore, even in a
large-scale system without a large resource overlap between
applications, our approach can be still beneficial. We also ran an experiment to study the extent to which
the integration of pub-sub with SAAM reduces the response
time. The PEP sent 10,000 randomly generated requests in
total, which accounted for one third of the total possible re-
quests. We evaluated both precise recycling, where the SDP
only used previously cached responses, and approximate re-
cycling, where the SDP inferred new responses from cached
responses. Figure 7(a) shows the results. As expected, both
precise recycling and approximate recycling helps to reduce
the response time, while approximate recycling achieve bet-
ter results. This is because many requests can be resolved
by the collocated SDP. When the cache size increases, the
decrease in response time is more significant. Additionally, our results confirm that the system per-
formance is impacted: comparing the pub-sub architecture
with the point-to-point architecture, we observed an 1.6ms
increase in response time. However, the importance of this
overhead must be judged in conjunction with the cost of
making decisions at the PDP, which can be one order of
magnitude larger. We have also demonstrated that it is im-
portant to reduce the number of subscriptions to the ENS,
either by using alternative subscription mechanisms at the
PDP or having “unsubscribe” call to remove stale subscrip-
tions by the PEP. Our results also indicate that providing
high availability by using multiple PDPs with large overlap
rate does not impact the response time. Finally, our results
confirms that our two proposed approaches to improving
performance are effective. Figure 4. Availability results. We also studied the extent to which the system perfor-
mance depends on the number of subscriptions in the ENS. First, we varied the number of subscriptions posted by the
PDP: the PDP subscribed to either 100 or 1,000 objects. Figure 6(a) shows that the response time almost doubled
with 1,000 object subscriptions. This result suggests that it
is important to keep the number of PDP subscriptions small. The results of availability evaluation suggest show that
multiple PDPs with a large overlap rate provide high avail-
ability. Next, we studied whether this happens at the cost of
system performance. We varied the number of PDPs from 1 Second, we varied the number of subscriptions posted by
the PEP. We ran an experiment where we disabled the “un- 0
1
2
3
4
5
2
3
4
5
6
7
8
9
10
response time (ms)
number of requests (x500)
1,000 objects
100 objects
(a) As the subscriptions posted by the PDP vary. 0
10
20
30
40
50
60
2
3
4
5
6
7
8
9
10
response time (ms)
number of requests (x500)
with unsubscribe
without unsubscribe
(b) As the subscriptions posted by the PEP vary. Figure 6. The impact of the number of subscriptions on response time. 0
1
2
3
4
5
2
3
4
5
6
7
8
9
10
response time (ms)
number of requests (x500)
1,000 objects
100 objects 0
10
20
30
40
50
60
2
3
4
5
6
7
8
9
10
response time (ms)
number of requests (x500)
with unsubscribe
without unsubscribe response time (ms) (a) As the subscriptions posted by the PDP vary. (b) As the subscriptions posted by the PEP vary. (a) As the subscriptions posted by the PDP vary. Figure 6. The impact of the number of subscriptions on response time. 4.5
Discussion Specifically, by integrating pub-
sub and SAAM, the response time is significantly reduced
when the PDP’s authorization logic is complex. In the above experiment, we used a simple authorization
logic: the PDP resolves any request in less than 1ms. In
reality, however, the authorization logic might be complex
and takes longer time. Therefore, in the next experiment,
we manually added a 20ms delay to each response returned
by the PDP to simulate a complex authorization logic. Fig-
ure 7(b) shows that using SAAM helped to reduce the re-
sponse time more significantly in this case. In particular,
the response time decreased from 22ms to 6ms when ap-
proximate recycling is used. The reason is that more and
more requests can be resolved by the local SDP, instead of
being resolved by the slow PDP. Acknowledgments Initial ideas of using pub-sub for access control have
benefited significantly from the presentation and discussion
of [3] at the New Security Paradigms Workshop (NSPW)
’05. The authors are grateful to the anonymous reviewers
for their helpful comments, and Kirstie Hawkey for improv-
ing the readability of this paper. Members of the Labo-
ratory for Education and Research in Secure Systems En-
gineering (LERSSE) gave valuable feedback on the earlier
drafts of this paper. Research on pub-sub by the first and
third authors have been partially supported by the Cana-
dian NSERC Strategic Partnership Program, grant STPGP
322192-05. Figure 7. Response time results when pub-sub is integrated with SAAM. over systems based on topics by introducing a subscription
scheme based on the actual content of the considered events
(Gryphon [1], Siena [6], Elvin [22].) overall system design and define the requirements for the
core event notification service: robustness, performance
and security. Our evaluation demonstrates that our design
achieved better availability, and the impact of integrating
the pub-sub channel on performance is small and can be
further reduced by returning authorizations bypassing the
pub-sub channel or by integration with SAAM. Due to their flexibility and scalability, pub-sub systems
have been used as the basic communication and integration
infrastructure for many application domains, such as Inter-
net games [4], mobile agents [20], user and software moni-
toring [13], mobile systems [9], among others. The contri-
bution of this paper lies in applying the pub-sub model to
access control systems. For future research, we will further refine the security
analysis of our pub-sub authorization model and develop
countermeasures to the identified attacks. Second, we will
further explore how PEPs can handle the case when poli-
cies become inconsistent among PDPs. Third, we will de-
velop application-specific load balancing techniques, since
an ENS node in a distributed pub-sub authorization system
may suffer from uneven load distribution due to different
population densities and interests of PDPs. To improve the availability of access control systems,
caching authorizations has been employed in a number of
commercial systems [15, 11, 18], as well as several aca-
demic distributed access control systems [2, 5]. However,
these and other approaches only compute precise authoriza-
tions and therefore are only effective for resolving repeated
requests. Beznosov [3] introduces the concept of recycling
approximate authorizations, and later Crampton et al. [8]
formally define SAAM and introduce the concept of SDP. The SDP can resolve new requests by extending the space of
supported responses to approximate ones. In other words,
SAAM provides a richer alternative source for authorization
responses than the existing approaches do, thus improve the
availability of access control systems. To further improve
the performance and availability of access control systems,
Wei et al. [27] explore the cooperation between multiple
SDPs and combine their cooperation with the inference. 5
Related Work Pub-sub systems have been an active research area. Early
pub-sub systems use a subscription scheme based on the no-
tion of topics, and have been implemented by several indus-
trial solutions (e.g., [24]). Content-based systems improve 0
0.5
1
1.5
2
2.5
3
2
4
6
8
10
12
14
16
18
20
response time (ms)
number of requests (x500)
no recycling
precise recycling
approximate recycling
(a) PDP With simple authorization logic ( authorization compute time
at the PDP < 1 ms ). 0
5
10
15
20
25
2
4
6
8
10
12
14
16
18
20
response time (ms)
number of requests (x500)
no recycling
precise recycling
approximate recycling
(b) PDP With Complex authorization logic ( time = 20ms ). Figure 7. Response time results when pub-sub is integrated with SAAM. 0
5
10
15
20
25
2
4
6
8
10
12
14
16
18
20
response time (ms)
number of requests (x500)
no recycling
precise recycling
approximate recycling
(b) PDP With Complex authorization logic ( time = 20ms ). 0
0.5
1
1.5
2
2.5
3
2
4
6
8
10
12
14
16
18
20
response time (ms)
number of requests (x500)
no recycling
precise recycling
approximate recycling 0
5
10
15
20
25
2
4
6
8
10
12
14
16
18
20
response time (ms)
number of requests (x500)
no recycling
precise recycling
approximate recycling response time (ms) (a) PDP With simple authorization logic ( authorization compute time
at the PDP < 1 ms ). (b) PDP With Complex authorization logic ( time = 20ms ). References [16] G. Kiczales, J. Lamping, A. Mendhekar, C. Maeda,
C. Lopes, J.-M. Loingtier, and J. Irwin. Aspect-oriented
programming. In Proceedings of the 1997 11th European
Conference on Object-Oriented Programming, ECOOP, Jun
9-13 1997, volume 1241 of Lecture Notes in Computer Sci-
ence, page 220, Jyvaskyla, Finl, 1997. [1] G. Banavar, T. Chandra, B. Mukherjee, J. Nagarajarao,
R. Strom, and D. Sturman. Efficient multicast protocol for
content-based publish-subscribe systems. In Proceedings of
the 19th International Conference on Distributed Comput-
ing Systems (ICDCS’99)., pages 262–272, 1999. [17] J. Markoff and S. Hansell. Google’s not-so-very-secret
weapon, 2006. [2] L. Bauer, S. Garriss, and M. K. Reiter. Distributed proving
in access-control systems. In Proceedings of the 2005 IEEE
Symposium on Security and Privacy (S&P’05), pages 81–
95, Oakland, CA, 2005. IEEE Computer Society. [18] Netegrity. Siteminder concepts guide. Technical report,
Netegrity, 2000. [19] OMG. Common object services specification, security ser-
vice specification v1.8, 2002. [3] K. Beznosov. Flooding and recycling authorizations. In
Proceedings of the New Security Paradigms Workshop
(NSPW’05), pages 67–72, Lake Arrowhead, CA, USA, 20-
23 September 2005. ACM Press. [20] A. Padovitz, S. Loke, and A. Zaslavsky. Using the publish-
subscribe communication genre for mobile agents. In Pro-
ceedings of the First German Conference on Multiagent Sys-
tem Technologies, Sep 2003. [4] A. R. Bharambe, S. Rao, and S. Seshan. Mercury: a scalable
publish-subscribe system for internet games. In NetGames
’02: Proceedings of the 1st workshop on Network and sys-
tem support for games, pages 3–9, 2002. [21] Securant. Unified access management: A model for inte-
grated web security. Technical report, Securant Technolo-
gies, June 25 1999. [22] B. Segall, D. Arnold, J. Boot, M. Henderson, and T. Phelps. Content Based Routing with Elvin4. In Proceedings of
AUUG2K, June 2000. [5] K. Borders, X. Zhao, and A. Prakash. CPOL: high-
performance policy evaluation. In Proceedings of the 12th
ACM conference on Computer and Communications Secu-
rity (CCS’05), pages 147–157, New York, NY, USA, 2005. ACM Press. [23] P. Strong. How Ebay scales with networks and the chal-
lenges. In the 16th ACM/IEEE International Symposium
on High-Performance Distributed Computing (HPDC’07),
Monterey, CA, USA, 2007. ACM Press. Invited talk. [6] A. Carzaniga, D. S. Rosenblum, and A. L. Wolf. Achiev-
ing scalability and expressiveness in an internet-scale event
notification service. In PODC’00: Proceedings of the nine-
teenth annual ACM symposium on Principles of distributed
computing, pages 219–227, 2000. y
[24] I. TIBCO. 6
Summary and Future Work This paper presents an authorization architecture that
uses the pub-sub model to improve the robustness and man-
ageability of the access control system. We introduce the References TIB/Rendezvous White Paper. Palo Alto, Cali-
fornia, 1999. [25] W. Vogels. How wrong can you be? Getting lost on the road
to massive scalability. In the 5th International Middleware
Conference, Toronto, Canada, October 20 2004. ACM Press. Keynote address. [7] A. Carzaniga, D. S. Rosenblum, and A. L. Wolf. Design and
evaluation of a wide-area event notification service. ACM
Trans. Comput. Syst., 19(3):332–383, 2001. [26] Q. Wei, J. Crampton, K. Beznosov, and M. Ripeanu. Au-
thorization recycling in RBAC systems. In Proceedings of
the thirteenth ACM Symposium on Access Control Models
and Technologies (SACMAT), pages 63–72, Estes Park, Col-
orado, USA, June 11–13 2008. ACM. [8] J. Crampton, W. Leung, and K. Beznosov. Secondary
and approximate authorizations model and its application to
Bell-LaPadula policies. In Proceedings of the 11th ACM
Symposium on Access Control Models and Technologies
(SACMAT’06), pages 111–120, Lake Tahoe, CA, USA, June
7–9 2006. ACM Press. [27] Q. Wei, M. Ripeanu, and K. Beznosov. Cooperative
secondary and approximate authorization recycling. In
Proceedings of the 16th ACM/IEEE International Sympo-
sium on High-Performance Distributed Computing (HPDC),
pages 65–74, Monterey Bay, CA, June 27-29 2007. ACM
Press. [9] G. Cugola and H.-A. Jacobsen. Using publish/subscribe
middleware for mobile systems. SIGMOBILE Mob. Com-
put. Commun. Rev., 6(4):25–33, 2002. [10] L. G. DeMichiel, L. ¨U. Yalc¸inalp, and S. Krishnan. Enter-
prise JavaBeans Specification, Version 2.0. Sun Microsys-
tems, 2001. [28] XACML-TC. OASIS eXtensible Access Control Markup
Language (XACML) version 2.0. OASIS Standard, 1 Febru-
ary 2005. [11] Entrust. GetAccess design and administration guide. Tech-
nical report, Entrust, September 20 1999. [12] P. T. Eugster, P. A. Felber, R. Guerraoui, and A.-M. Kermar-
rec. The many faces of publish/subscribe. ACM Comput. Surv., 35(2):114–131, 2003. [13] D. M. Hilbert and D. F. Redmiles. An approach to large-
scale collection of application usage data over the internet. In ICSE ’98: Proceedings of the 20th international confer-
ence on Software engineering, pages 136–145, 1998. [14] Z. Kalbarczyk, R. K. Lyer, and L. Wang. Application fault
tolerance with Armor middleware. IEEE Internet Comput-
ing, 9(2):28–38, 2005. [15] G. Karjoth. Access control with IBM Tivoli Access Man-
ager. ACM Transactions on Information and Systems Secu-
rity, 6(2):232–57, 2003.
|
https://openalex.org/W4362697711
|
https://sprinpub.com/sjahss/article/download/sjahss-2-4-3-27-42/251
|
English
| null |
Language Learning Strategies and Grammatical Competence of the English-Major Students in Zamboanga del Norte HEIs
|
Sprin Journal of Arts, Humanities and Social Sciences
| 2,023
|
cc-by
| 8,946
|
Electronic reference (Cite this article): Electronic reference (Cite this article):
Rodriguez, E. (2023). Language Learning Strategies and Grammatical Competence of the English-Major
Students in Zamboanga del Norte HEIs. Sprin Journal of Arts, Humanities and Social Sciences, 2(04),
27–42. https://doi.org/10.55559/sjahss.v2i04.99 Copyright Notice:
© 2023 Author(s). This is an Open Access article distributed under the terms of the Creative Commons
Attribution 4.0 International License (CC-BY 4.0: https://creativecommons.org/licenses/by/4.0/), allowing
third parties to copy and redistribute the material in any medium or format and to remix, transform, and
build upon the material for any purpose, even commercially, provided the original work is properly cited
and states its license. Copyright Notice:
© 2023 Author(s). This is an Open Access article distributed under the terms of the Creative Commons
Attribution 4.0 International License (CC-BY 4.0: https://creativecommons.org/licenses/by/4.0/), allowing
third parties to copy and redistribute the material in any medium or format and to remix, transform, and
build upon the material for any purpose, even commercially, provided the original work is properly cited
and states its license. Keywords: language learning strategies, grammatical competence, metacognitive strategies,
word formation, syntax ABSTRACT Language learning requires strategies that increase our competence in communication. This
study identified the language learning strategies and the grammatical competency level of the Bachelor
of Secondary Education (BSED) English major students at Jose Rizal Memorial State University
(JRMSU), School Year 2017-2018. This research is mixed methods since it determined the language
learning strategies mostly used by the students and identified the grammatical competency level of the
respondents. The researcher triangulated the data from the survey using the focus group discussion
through semi-structured interviews. It further examined whether there is a significant relationship
between language learning strategies and the grammatical competence level of the BSED-English. The
respondents of this study were the 138-3rd year BSED-English major students of JRMSU chosen
through simple random sampling by the lottery technique. The students took the Strategy Inventory of
Language Learning (SILL) survey questionnaire, developed by Oxford (1990), and the grammatical
competence test. Twelve students from the group participated in the focus group discussion. The data
were checked, tallied, and analyzed utilizing the weighted mean, ranking, and percentage. The
researcher employed Spearman's Rho test to find the relationship between language learning strategies
and grammatical competency levels. She then transcribed and coded the data from the focus group
discussion. The results of this study showed that language learning strategies usually used by the
BSED-English major students were metacognitive, social, cognitive, affective, and compensation
strategies. In addition, it showed that the overall grammatical competence level of the students is
competent, which revealed their higher competence level in word formation and syntax but less
competence in spelling, pronunciation, and vocabulary. The results showed a significant relationship
between language learning strategy use and grammatical competence level, indicating that the more
language learning strategies are mostly used by the students, the higher their grammatical competence
level. Keywords: language learning strategies, grammatical competence, metacognitive strategies,
word formation, syntax rds: language learning strategies, grammatical competence, metacognitive strategies, Sprin Journal of Arts, Humanities and Social Sciences
Abbreviated Key Title: Spr. J. Arts. Humanit. Soc. Sci. ISSN: 2583-2387 (Online)
Journal homepage: https://sprinpub.com/sjahss
Vol.02(4). Apr 2023, pp, 27-42
Language Learning Strategies and Grammatical Competence of
the English-Major Students in Zamboanga del Norte HEIs
Elmira C. Rodriguez
Associate Professor 2, Jose Rizal Memorial State University, Philippines
DOI: 10.55559/sjahss.v2i04.99
Received: 24.03.2023 | Accepted: 03.04.2023 | Published: 07.04.2023
Electronic reference (Cite this article):
Rodriguez, E. (2023). Language Learning Strategies and Grammatical Competence of the English-Major
Students in Zamboanga del Norte HEIs. Sprin Journal of Arts, Humanities and Social Sciences, 2(04),
27–42. https://doi.org/10.55559/sjahss.v2i04.99
Copyright Notice:
© 2023 Author(s). This is an Open Access article distributed under the terms of the Creative Commons
Attribution 4.0 International License (CC-BY 4.0: https://creativecommons.org/licenses/by/4.0/), allowing
third parties to copy and redistribute the material in any medium or format and to remix, transform, and
build upon the material for any purpose, even commercially, provided the original work is properly cited
and states its license. Sprin Journal of Arts, Humanities and Social Sciences
Abbreviated Key Title: Spr. J. Arts. Humanit. Soc. Sci. ISSN: 2583-2387 (Online)
Journal homepage: https://sprinpub.com/sjahss
Vol.02(4). Apr 2023, pp, 27-42
Language Learning Strategies and Grammatical Competence of
the English-Major Students in Zamboanga del Norte HEIs
Elmira C. Rodriguez
Associate Professor 2, Jose Rizal Memorial State University, Philippines
DOI: 10.55559/sjahss.v2i04.99
Received: 24.03.2023 | Accepted: 03.04.2023 | Published: 07.04.2023
Electronic reference (Cite this article):
Rodriguez, E. (2023). Language Learning Strategies and Grammatical Competence of the English-Major
Students in Zamboanga del Norte HEIs. Sprin Journal of Arts, Humanities and Social Sciences, 2(04),
27–42. https://doi.org/10.55559/sjahss.v2i04.99
Copyright Notice:
© 2023 Author(s). This is an Open Access article distributed under the terms of the Creative Commons
Attribution 4.0 International License (CC-BY 4.0: https://creativecommons.org/licenses/by/4.0/), allowing
third parties to copy and redistribute the material in any medium or format and to remix, transform, and
build upon the material for any purpose, even commercially, provided the original work is properly cited
and states its license. Sprin Journal of Arts, Humanities and Social Sciences
Abbreviated Key Title: Spr. J. Arts. Humanit. Soc. Sci. ISSN: 2583-2387 (Online)
Journal homepage: https://sprinpub.com/sjahss
Vol.02(4). Apr 2023, pp, 27-42
Language Learning Strategies and Grammatical Competence of
the English-Major Students in Zamboanga del Norte HEIs
Elmira C. Rodriguez
Associate Professor 2, Jose Rizal Memorial State University, Philippines
DOI: 10.55559/sjahss.v2i04.99
Received: 24.03.2023 | Accepted: 03.04.2023 | Published: 07.04.2023
Electronic reference (Cite this article):
Rodriguez, E. (2023). Language Learning Strategies and Grammatical Competence of the English-Major
Students in Zamboanga del Norte HEIs. Sprin Journal of Arts, Humanities and Social Sciences, 2(04),
27–42. https://doi.org/10.55559/sjahss.v2i04.99
Copyright Notice:
© 2023 Author(s). This is an Open Access article distributed under the terms of the Creative Commons
Attribution 4.0 International License (CC-BY 4.0: https://creativecommons.org/licenses/by/4.0/), allowing
third parties to copy and redistribute the material in any medium or format and to remix, transform, and
build upon the material for any purpose, even commercially, provided the original work is properly cited
and states its license. ces
ISSN: 2583-2387 (Online)
Vol.02(4). Apr 2023, pp, 27-42 Associate Professor 2, Jose Rizal Memorial State University, Philippines DOI: 10.55559/sjahss.v2i04.99 1.0 Introduction Learning is a continuing process gained through the interaction of study, instruction,
and experience (Francis & Rivera, 2007). Learning strategies are ways taken by students to
improve their learning. Strategies are especially significant for language education since they
aid in active, self-directed involvement, which is extremely important for developing
communicative competence. According to Oxford (1990), suitable language learning
strategies improve proficiency and self-confidence. One of the most critical issues for educating English Language Learners (ELLs) is
their dearth of academic language skills needed for school achievement (Scarcella, 2003;
Bailey & Butler, 2007). A lack of academic language proficiency affects the ELL's capability
to comprehend and analyze texts in middle and high school, restricts their ability to write and
express themselves effectively, and can hamper their gain of academic content in all academic
fields. Moreover, in light of the role of vocabulary and grammar in educational content zones,
ELLs face difficulties in acquiring content area knowledge: their academic language and,
therefore, achievement lags behind that of their native English-speaking peers (National
Center for Education Statistics, 2005). Thus, it is relevant to give special attention to the learning strategies of each student
that can work together with - or in conflict with - a given instructional methodology. Oxford
(2003) hypothesized that if there is a balance between the student's strategy preferences and
the fusion of instructional methods and materials, students are likely to perform well, feel
confident, and experience low anxiety. Conversely, if clashes occur between these factors, the
students often perform poorly, feel diffident, and experience significant anxiety. Sometimes,
such disagreements lead to a severe breakdown in teacher-student interaction. These conflicts
may also lead to the depressed student's outright rejection of the teaching methodology, the
teacher, and the subject matter. Equally important as the language learning strategies is grammatical competence as
part of communicative competence. According to Diaz-Rico and Weed (2010),
communicative competence is a function of the language awareness of a language consumer,
which helps the consumer to know when, where, and how to use language correctly. Grammatical competence is one of the four areas of the communicative competence theory by
Canale and Swain (1980). Grammatical competence focuses on the command of the language code, including the rules
of the shape of terms and sentences, names, spellings, and pronunciations (Gao, 2001). Keywords: language learning strategies, grammatical competence, metacognitive strategies,
word formation, syntax Sprin Journal of Arts, Humanities and Social Sciences | Published by Sprin Publisher, India | https://sprinpub.com 27 Elmira C. Rodriguez, Spr. J. Arts. Humanit. Soc. Sci., Vol.02 (4). Apr 2023, pp, 27-42 1.0 Introduction Sprin Journal of Arts, Humanities and Social Sciences | Published by Sprin Publisher | https://sprinpub.com
28 1.0 Introduction The
goal is to gain awareness of and the capacity to use grammatically appropriate and consistent
modes of speech (Diaz-Rico & Cannabis, 2010; Gao, 2001). Grammatical abilities encourage
consistency and fluency in second-language development (Gao, 2001) and rise in value as the
learner progresses (Diaz-Rico & Weed, 2010). Grammar skills grow in value as the learner
progresses in skills (Diaz-Rico & Weed, 2010). As students travel through the stages of language proficiency, grammatical
competence becomes more critical since grammar is the glue that binds the English language
together (Gaard, n.d.). That is why it is necessary to determine the language learning
strategies the BSEd English major students use to examine their English language
deficiencies, especially in grammar. In the previous study conducted by the researcher and her
co-author on evaluating the weakness of the graduates of the JRMSU-Tampilisan Campus
concerning the DepEd requirements in hiring teachers, Liboon and Rodriguez (2017) found 28 Elmira C. Rodriguez, Spr. J. Arts. Humanit. Soc. Sci., Vol.02 (4). Apr 2023, pp, 27-42 that the graduates are most deficient in terms of teaching experience, followed by the English
Proficiency Test (EPT) (Liboon & Rodriguez, 2017) as conducted with the Bachelor of
Elementary Education (BEED) graduates LET passers who applied for Teacher I position in
DepEd Zanorte. On the other hand, the Bachelor of Secondary Education major in English (BSEd-
English) is a four-year degree program that will prepare students for teaching English subjects
in high school offered by the JRMSU system. Based on the researcher's observation, the
college students in the institution performed poorly in writing composition, evident in their
grammatical errors committed. Another problem the school faces is the low Licensure Examination for Teachers
(LET) rating. For the past few years, the JRMSU-TC LET passing rate was always below the
National Passing Rate set by the Philippine Regulation Commission (PRC). The College of
Education in JRMSU-Tampilisan Campus, during September 2016 LET, registered only 17
out of 92 takers, or 18.48%, while BSED recorded only six from the 102 takers, or 5.88%
(The State Collegian, 2017). With the above premises, this study assessed the language learning strategies mostly
used by the JRMSU BSED-English major students and their grammatical competence level. 2.0 Objectives of the Study This study aimed to identify the language learning strategies mostly used by the
JRMSU BSED-English major students and their grammatical competency level. Specifically, the study sought to answer the following questions: Specifically, the study sought to answer the following questions: 1. What are the language learning strategies mostly used by JRMSU English major st hat are the language learning strategies mostly used by JRMSU English major students? 2. What is the grammatical competence level of the JRMSU BSEd English major students in
terms of Pronunciation, Spelling, Word Formation, Syntax, and Vocabulary? 3. Is there a significant relationship between the Language Learning Strategies and the
Grammatical Competency Level of the BSEd English major students in JRMSU? terms of Pronunciation, Spelling, Word Formation, Syntax, and Vocabulary? 3. Is there a significant relationship between the Language Learning Strategies and the
Grammatical Competency Level of the BSEd English major students in JRMSU? 1.0 Introduction Specifically, the study aimed to 1) identify the language learning strategies mostly used by the
JRMSU major in English students in terms of Memory strategies, Cognitive strategies,
Compensation strategies, Metacognitive strategies, Affective strategies, and Social strategies;
2) determine the grammatical competency level of the JRMSU BSEd English major students
in terms of Pronunciation, Spelling, Word Formation, Syntax, and Vocabulary; 3) determine
whether there is a significant relationship between the Language Learning strategies and the
Grammatical Competency Level of the BSEd, English major students in JRMSU. 3.0 Theoretical/Conceptual Framework This study is anchored on the Theory of Oxford (1990) on Language Learning
Strategies as well as on Grammatical Competence as one of the areas of Canale and Swain’s
(1980) Communicative Competence. Oxford (1990) developed a comprehensive classification system of learning strategies
using two major groups: direct and indirect strategies. Each category was broken down into
subcategories reflecting the specific strategies that would fit under the labels. For example,
direct strategies, which are directly related to learning/producing the target language, are
subdivided into Memory strategies (retrieving and storing new information), Cognitive rin Journal of Arts, Humanities and Social Sciences | Published by Sprin Publisher | https://sprinpub.com 29 Elmira C. Rodriguez, Spr. J. Arts. Humanit. Soc. Sci., Vol.02 (4). Apr 2023, pp, 27-42 strategies (operating new input), and Compensation strategies (overcoming missing
knowledge of a target language). Indirect strategies are those that enable direct strategies to occur and/or increase their
successful application. Indirect strategies include Metacognitive strategies for managing the
cognitive process, Affective strategies for controlling emotions in language learning, and
Social strategies for interacting with others. In general, these strategies help students (a) to
become more autonomous, (b) to diagnose their learning strengths and weaknesses, and (c) to
self-direct their learning process (Oxford, 1990). Learning strategies, therefore, help learners
become efficient in learning and using a language. In Canale and Swain’s (1980, 1981) model, grammatical competence is mainly
defined in terms of Chomsky’s linguistic competence, which is why some theoreticians (e.g.,
Savignon, 1983), whose theoretical and/or empirical work on communicative competence was
largely based on the model of Canale and Swain, use the term linguistic competence for
grammatical competence. According to Canale and Swain, grammatical competence is
concerned with mastery of the linguistic code (verbal or non-verbal), which includes
vocabulary knowledge and knowledge of morphological, syntactic, semantic, phonetic, and
orthographic rules. This competence enables the speaker to use the knowledge and skills
needed for understanding and expressing the literal meaning of utterances. The independent variable of this study consists of the six major groups of second
language learning strategies identified by Oxford (1990). These are memory strategies,
cognitive strategies, compensation strategies, metacognitive strategies, affective strategies,
and social strategies, which are hypothesized to have a significant association with the
grammatical competency level in any of the sub-variables, such as pronunciation, spelling,
word formation, grammar, and vocabulary. 3.0 Theoretical/Conceptual Framework The types mentioned above of language learning strategies are assumed to have a
significant association with the grammatical competence level based on what was revealed in
the studies previously discussed in this paper. Sprin Journal of Arts, Humanities and Social Sciences | Published by Sprin Publisher | https://sprinpub.com
30 4.0 Methodology The current study employed a mixed-methods design, including quantitative and
qualitative research methods. According to Creswell (2003), this is "an inquiry strategy that is
focused on converging or triangulating different quantitative and qualitative data sources" (p. 210). The design combined both approaches, offering a much more accurate and informative
picture of what was discussed in the study. The researcher utilized something similar to what
Creswell (2003) called a sequential explanatory model, a mixed-method design in which
quantitative data collection was undertaken before qualitative data collection. With the
priority being placed on the quantitative data (a questionnaire was given to the whole sample),
the qualitative data (interviews conducted with a subsample) were used to explain and
elucidate the quantitative data, thus deepening the understanding and interpretation of the
results. This study employed three instruments to gather the data needed. These were the
Strategy Inventory for Language Learning (SILL), the Grammatical Competency Test, and
the Focus Group Discussion. The first instrument used in this research is a survey
questionnaire Oxford developed in 1990 called Strategy Inventory for Language Learning
(SILL). It is a self-report, paper, and pencil survey. The SILL was initially designed to assess Elmira C. Rodriguez, Spr. J. Arts. Humanit. Soc. Sci., Vol.02 (4). Apr 2023, pp, 27-42 the frequency of use of language learning strategies by students at the Defense Language
Institute in California. Two versions of the SILL were available in Oxford's (1990) language
learning strategy book for language teachers. The first is used with foreign language learners
whose native language is English, containing 80 items. English learners use the second test as
a second or international language. It consists of 50 items. This study used the latter variant. Oxford and Burry-Stock (1995) asserted that the results of the studies regarding the reliability
of the ESL/EFL SILL have shown that it is a highly reliable instrument. Concerning the
content validity of the inventory, Oxford and Burry-Stock (1995) stated that the instrument's
content validity was determined by professional judgment and was found to be very high. The
SILL (Version 7.0) consists of six subsections. Each section represents one of the six
categories of LLS, which the learners do not know when taking the inventory. 4.0 Methodology The 50
statements in the list follow the general format 'I do such and such.' Students respond on a 5-
point Likert scale ranging from 1, 'Almost never true of me,' to 5, 'Almost always true of me.' The current study required the participants to circle the number corresponding to the
survey questionnaire's answer. Upon completing all the answers, the values assigned to each
item in each section have been added and then divided into the number of items in each
section. The same procedures were repeated for each section, and values between 1 and 5
were obtained. These values showed the profile of a learner; in other words, the strategy
groups employed by the learner and their frequency. The second instrument was used to
determine the level of grammatical competence of the BSED-English major students. The
Grammatical Competence Test was adapted from different online resources. The competency
test is made up of multiple sections, each comprised of 20 items. This test included
pronunciation, spelling, word formation, syntax, and vocabulary. The third instrument is the
question guide for the Focus Group Discussion in the form of a semi-structured interview. It
listed seven questions regarding the language learning strategies used by the students. The
first instrument, the Strategy Inventory for Language Learning (SILL) developed by Oxford
(1990), was considered valid since it is a standardized survey questionnaire. However, it was
administered to the 3rd year BSED-English major students who were not participants in the
study to test its reliability. Nevertheless, the SILL survey was highly reliable since its
Cronbach’s Alpha was .95. Meanwhile, the panel of experts corrected and validated the grammatical competency
test before administering it to the respondents. The validators qualified the items to be
included and those to be rejected or revised. It was then pilot tested with the 3rd year English
major students at Saint Joseph College of Sindangan. The pilot test result was later subjected
to item analysis. The test did not include items that were found to be very easy and most
difficult. Some statements were revised, and the majority were accepted. The study was conducted at the four different campuses of Jose Rizal Memorial State
University, Zamboanga Del Norte, particularly in the College of Education of each campus:
Dapitan, Dipolog, Tampilisan, and Siocon. Sprin Journal of Arts, Humanities and Social Sciences | Published by Sprin Publisher | https://sprinpub.com 4.0 Methodology The study's respondents were the Bachelor of
Secondary Education (BSED) major in English students from the College of Education of
each campus. Meanwhile, the FGD participants were taken from the BSED-English major of
Tampilisan Campus only since it was financially unfeasible to collect data from the four
schools and beyond the time allotted by CHED to complete the study. The respondents of this study were the 138-3rd year BSEd English major students of
Jose Rizal Memorial State University during the 2nd Semester, School Year 2017-2018. 31 31 Elmira C. Rodriguez, Spr. J. Arts. Humanit. Soc. Sci., Vol.02 (4). Apr 2023, pp, 27-42 These students were enrolled in January 2018 since the institution is an August starter due to
the ASEAN 2015 Integration. These 3rd-year BSED-English students of JRMSU had passed
the major English courses such as Language and Curriculum for Secondary, Afro-Asian
Literature, Literary Criticism, American Literature, Teaching of Listening and Reading, and
Mythology and Folklore. In addition, the participants of this study had also taken their
Structure of English course. Thus, it was relevant to determine their grammatical competence
in pronunciation, spelling, word formation, syntax, and vocabulary. Furthermore, this study
was conducted at the 3rd year college level since the data collection started. The four
mentioned schools did not have 1st- and 2nd-year BSED-English enrollees due to the K to 12
Transition Period. There is a 138 or 64% sample size from the 214-total number of 3rd-year BSED-
English students enrolled in the 2nd Semester of School Year 2017-2018. The 64% sample
size was determined using a sample size calculator with a 95% confidence level. After
determining the 64% sample in each campus, the respondents for this study were selected
using probability sampling, particularly simple random sampling using the lottery technique,
to have an equal chance to participate in the survey and test. For the data collection, first, the researcher sought permission from the President of
the JRMSU System, different Campus Administrators, and the various Deans of the College
of Education of the four JRMSU campuses. After the consent was granted, the researcher
informed the Campus Administrator, the Registrar, the Dean of the College of Education, and
the BSED Program Chair of each campus regarding the study. Then, the schedule was
arranged for when to administer the survey questionnaire and the grammatical competence
test. When the negotiation was finalized, the researcher conducted the study. 4.0 Methodology The respondents
were given 20-30 minutes to answer the SILL survey, which was immediately retrieved so
that they could proceed to the grammatical competency test. After that, they were allotted one
and a half hours to finish the grammatical competency test, which has five parts, namely:
pronunciation test, spelling test, word formation test, syntactic test, and vocabulary test. After
conducting the SILL and the test on grammatical competence in the four campuses, the Focus
Group Discussion (FGD) followed. Due to budgetary constraints, the 12 FGD participants
from JRMSU-Tampilisan BSED-English students were chosen. However, the 12 participants
were still selected using the lottery technique, and they were from the 25 students of JRMSU-
Tampilisan who took the SILL. The FGD was in the form of a semi-structured interview. The
researcher acted as an interviewer, and she requested two facilitators to help: one as secretary
and the other was assigned to the audio recording of the session. The survey and test results
were checked, tallied, and interpreted. In contrast, the results of the FGD were transcribed and
coded according to the language learning strategies indicated in the SILL instrument. In finding the Language Learning Strategies mostly used by the JRMSU English
major students in terms of memory, cognitive, compensation, metacognitive, social, and
affective strategies, the weighted mean and ranking (Zhou, 2010) were utilized with a
qualitative description within the established limits as follows: Weight
Range of Values
Description
5
4.50 – 5.00
Almost always used (AU)
4
3.50 – 4.49
Usually used (UU)
3
2.50 – 3.49
Sometimes used (SU)
4
1.50 – 2.49
Generally not used (GnU)
1
1.00 – 1.49
Almost never used (NU) In identifying the level of grammatical competence of the BSED-English major
students, frequency counts and percentages were employed. The grammatical competency test Sprin Journal of Arts, Humanities and Social Sciences | Published by Sprin Publisher | https://sprinpub.com
32 32 Elmira C. Rodriguez, Spr. J. Arts. Humanit. Soc. Sci., Vol.02 (4). Apr 2023, pp, 27-42 Elmira C. Rodriguez, Spr. J. Arts. Humanit. Soc. Sci., Vol.02 (4). Apr 2023, pp, 27-42 scoring followed the 50% passing (raw score x 50 / total no. of items + 50 = rating or grade)
based on the JRMSU Grading System (JRMSU Code, 2017). The following rating scale was
adopted to identify the grammatical competence level of the BSED-English students. adopted to identify the grammatical competence level of the BSED-English students. 4.0 Methodology Rating Scale
Description
Code (JRMSU Grading System)
99.00 - 100.00
Highly Competent (HC)
(Excellent)
91.00 - 98.99
Much Competent
(MC)
(Very Good)
80.00 - 90.99
Competent
(C)
(Good)
75.00 - 79.99
Less Competent
(LC)
(Fair)
74.00 and below
Not Competent
(NC)
(Failed) In determining the significant relationship between the Language Learning Strategies In determining the significant relationship between the Language Learning Strategies
and the Grammatical Competency level of the BSEd English Major Students, Spearman's
Rho, with the aid of Stata, was employed since this study tested the non-parametric
correlation of variables. In determining the significant relationship between the Language Learning Strategies
and the Grammatical Competency level of the BSEd English Major Students, Spearman's
Rho, with the aid of Stata, was employed since this study tested the non-parametric
correlation of variables. Open coding was employed to analyze the focus group discussion result. First, the
researcher read through the data several times and then created tentative labels for chunks of
data summarizing the results. Next, the researcher recorded examples of participants' words
and established properties of each code, whether the sub-strategies they were using fell on
memory, cognitive, compensation, metacognitive, affective, and social strategies. 5.0 Results and Discussion The language learning strategies used by the BSED-English major students of Jose
Rizal Memorial State University, Zamboanga del Norte, are shown in Table 1. The result
indicates that the BSED-English major students of JRMSU usually used sub-strategies that
fall under the metacognitive strategies, such as centering, arranging, planning, and evaluating
their learning. In addition, they also utilized social strategies such as asking questions,
cooperating, and empathizing with others. Likewise, the students employed mental or
cognitive strategies such as practicing, receiving, sending messages, analyzing and reasoning,
and creating structures for input and output. Meanwhile, these students also used sub-strategies such as managing emotions by
lowering their anxiety, encouraging themselves, and taking emotional temperature, which fell
under the affective strategies. They also utilized strategies to make up for missing knowledge,
such as guessing intelligently and overcoming limitations in speaking and writing, which are
the compensation strategies. Table 1 Sprin Journal of Arts, Humanities and Social Sciences | Published by Sprin Publisher | https://sprinpub.com
33
Table 1
Language Learning Strategies Mostly Used by the JRMSU BSED-English Major Students
Language Learning
Strategies
Weighted
Mean
Rank
Description
Memory Strategies
Cognitive Strategies
Compensation Strategies
Metacognitive Strategies
Affective Strategies
Social Strategies
3.45
3.80
3.61
4.07
3.65
3.91
6th
3rd
5th
1st
4th
2nd
Sometimes Used
Usually Used
Usually Used
Usually Used
Usually Used
Usually Used
Overall Mean
3.75
Usually Used ge Learning Strategies Mostly Used by the JRMSU BSED-English Major Students Language Learning Strategies Mostly Used by the JRMSU BSED-English Major Elmira C. Rodriguez, Spr. J. Arts. Humanit. Soc. Sci., Vol.02 (4). Apr 2023, pp, 27-42 Legend: Range of Values
Description
4.50 - 5.0
Almost always used
3.50 - 4.49
Usually used
2.50 - 3.49
Sometimes used
1.50 - 2.49
Generally not used
1.00 - 1.49
Almost never used Legend: Range of Values
Description
4.50 - 5.0
Almost always used
3.50 - 4.49
Usually used
2.50 - 3.49
Sometimes used
1.50 - 2.49
Generally not used
1.00 - 1.49
Almost never used In general, the respondents of this study "usually used" the language learning
strategies, which implies that the students pay attention, consciously search for practice
opportunities, plan for language tasks, self-evaluate their progress and monitor their errors. They likewise ask questions, communicate with others using the English language, and
become culturally aware of other people's cultures. The students also reason, analyze, and
summarize - all reflective of deep processing. They also reduce anxiety, encourage
themselves, and do self-reward. They, too, guess meanings from the context of listening and
reading and use synonyms and gestures to convey a meaning when the precise expression is
not known. The JRMSU BSED-English major students employed the abovementioned
strategies in learning the second language. These findings are supported by the focus group discussion data, whereby the students
asserted that they sometimes do memorization. For example, when they find new words, they
use them in a sentence or write in their vocabulary notebook for them to look for their
meanings in a dictionary. The respondents also scanned before skimming and looking for the
text's main idea. Likewise, they do silent reading to comprehend the passage before reading it
loudly to practice pronunciation. Meanwhile, when they become speechless in a conversation,
they try to use sign language, gestures, and facial expressions. Sprin Journal of Arts, Humanities and Social Sciences | Published by Sprin Publisher | https://sprinpub.com Table 1 They also use words
synonymous with what they have to say with word association. The BSED-English major students affirmed that they usually watch English movies
with subtitles to help them understand fully the lines of the characters, which results in their
language learning, studying the grammatical rules of the English language and trying to
identify errors in a written text, read books with a dictionary at hand and check the meaning
of the problematic words encountered if using context clues could not help them and practice
speaking the English language in front of the mirror. Besides, they read books, English
novels, and stories written in English and try to use a second language whenever they have
conversations with friends and peers. In using English in communications, the respondents
overcome anxiety by taking deep breaths and trying to relax and be calm, confident, and
optimistic about accepting criticism. They also search online and watch English movies to
learn more about American culture and the target language. The above finding corroborates the study of Hong-Nam and Leavell (2006), who
found that the students preferred to use metacognitive strategies mostly. In contrast, they
showed the least use of affective and memory strategies. They have executive control over the
learning process since the metacognitive strategies are mostly behaviors used for centering,
arranging, planning, and evaluating their learning. These are "beyond the cognitive" strategies
(Akbarov & Arslan, 2010, p. 16). Likewise, Al-Qahtani’s (2013) study showed that her
respondents utilized all language learning strategies, and cognitive strategies were the most
frequently used. Yaimin's (2006) research also found that the pupils employed cognitive
strategies in learning English with the highest total average of 3.50. The other strategies 34 Elmira C. Rodriguez, Spr. J. Arts. Humanit. Soc. Sci., Vol.02 (4). Apr 2023, pp, 27-42 employed were the metacognitive strategy, affective strategy, memory strategy, social
strategy, and compensation strategy. Both studies indicate the students' skills, which involve
manipulation and transformation of language in the same direct way, such as note-taking,
functional practice in naturalistic settings, formal training with structures and sounds, as well
as through reasoning analysis. Figure 1. Sprin Journal of Arts, Humanities and Social Sciences | Published by Sprin Publisher | https://sprinpub.com Table 1 Apr 2023, pp, 27-42 in the spelling test, meaning they could not recognize the correct spelling of the words when
given options. On the whole, the grammatical competency level of the BSED-English major of
JRMSU is "competent," which shows that the respondents are competent in word formation
(morphology), grammar (syntax), vocabulary (lexis), pronunciation (phonology), and spelling
ability. The finding of this study is supported by the study of Lasala (2014), which found that
the level of communication in the oral and writing skills of senior secondary students is
acceptable. Still, they differ in numerical values since their grammatical competence average
in oral skills is 3.10. In contrast, they obtained an average rate of 2.91 in writing skills. Meanwhile, Cortez's (2016) analysis of intervention materials in producing Grade 7 grammar
skills contradicts the findings of this study. Lasala's (2014) research, like Cortez's (2016),
found that the level of competence of Grade 7 graduates in English grammar is only fairly
competent. He concluded that the students had not developed adequate and necessary skills to
master. Meanwhile, the study of Cortez (2016) on the intervention materials in developing
grammatical competence of Grade 7 students opposes the findings of the present study and
Lasala’s (2014) research, as Cortez (2016) found that the level of competence of Grade 7
students in English grammar is only fairly competent. He concluded that the students had not
developed adequate and needed skills to master the different grammar structures. Table 2
Test on the Relationship between the Language Learning Strategies and Grammatical
Competency Level of the JRMSU BSED-English Major Students
Variable
Mean
SD
Spearman’s
Rho
Coefficient
P-value
@
0.05 level of
significance
Interpretation
Language
Learning
Strategies
3.74
0.6090
0.9967
0.0000
Highly
Significant
Grammatical
Competency
Level
81.09
7.6439 Table 2 presents the correlation test between the language learning strategies and the
grammatical competency level of the BSED-English major students. The results reveal
Spearman's Rho coefficient of 0.9967 and a P-value of 0.0000, which is lower than the α0.05
level of significance that indicates a strong positive relationship between the language
learning strategies mostly used by the students and their grammatical competency level. It
means that the more they used a specific language learning strategy, the higher the level of
their grammatical competence and vice versa. Therefore, the correlation indicates that the
students who usually use language-learning strategies could be competent in grammatical
competency tests. Table 1 The grammatical competence level of the JRMSU BSED-English major students
Legend:
Rating Scale
Description
99.00 - 100.00
Highly Competent ( HC )
91.00 - 98.99
Much Competent ( MC )
80.00 - 90.99
Competent ( C )
75.00 - 79.99
Less Competent ( LC )
74.00 and below
Not Competent ( NC )
78.73
75.43
88.46
82.07
79.37
80.81
65.00
70.00
75.00
80.00
85.00
90.00
Students' Rating
Grammatical Competency Level
Pronunciation
Spelling
Word Formation
Syntax
Vocabulary
Average Rating Grammatical Competency Level Figure 1. The grammatical competence level of the JRMSU BSED-English major students
Legend:
Rating Scale
Description Figure 1. The grammatical competence level of the JRMSU BSED-English major students Figure 1. The grammatical competence level of the JRMSU BSED-English major stud Figure 1. The grammatical competence level of the JRMSU BSED-English major students
Legend:
Rating Scale
Description
99.00 - 100.00
Highly Competent ( HC )
91.00 - 98.99
Much Competent ( MC )
80.00 - 90.99
Competent ( C )
75.00 - 79.99
Less Competent ( LC )
74.00 and below
Not Competent ( NC ) Rating Scale
Description
99.00 - 100.00
Highly Competent ( HC )
91.00 - 98.99
Much Competent ( MC )
80.00 - 90.99
Competent ( C )
75.00 - 79.99
Less Competent ( LC )
74.00 and below
Not Competent ( NC ) Figure 1 shows the grammatical competency level of the JRMSU BSED-English
major students. The respondents here are seen to be "competent" in word formation and
grammar. It means that the students are competent in creating new words based on other
words or morphemes, also called derivational morphology. Likewise, they are also skilled in
encoding meanings into words in English, which include the structure of words, phrases,
clauses, and sentences up to the construction of the whole text. Meanwhile, the respondents were "less competent" in vocabulary, pronunciation, and
spelling. These results mean that the students were less competent in a vocabulary test
intended to look into their range (or stock) of words. The respondents were also less confident
in pronunciation practice, which examined how the students tried to pick the correct sound
output of the given words based on the phoneme. Likewise, the students were less competent Sprin Journal of Arts, Humanities and Social Sciences | Published by Sprin Publisher | https://sprinpub.com 35 35 Elmira C. Rodriguez, Spr. J. Arts. Humanit. Soc. Sci., Vol.02 (4). Sprin Journal of Arts, Humanities and Social Sciences | Published by Sprin Publisher | https://sprinpub.com
36 Sprin Journal of Arts, Humanities and Social Sciences | Published by Sprin Publisher | https://sprinpub.com Acknowledgement g
The author would like to thank everyone who contributed to this article. Table 1 In this case, Ha, which states that there is a significant relationship between
language learning strategies and grammatical competency level, is accepted. 36 Elmira C. Rodriguez, Spr. J. Arts. Humanit. Soc. Sci., Vol.02 (4). Apr 2023, pp, 27-42 This finding aligns with Cohen's (1998) study, which manifested the connection
between grammar learning strategies and grammatical competence. He argues that strategies
likely contribute to more grammatically accurate speech, and he claims that determining what
grammatical features are needed is one of the steps learners follow when talking. Learning
strategies can facilitate learning grammatical items by helping learners explicitly notice them,
structure them into working, and automatize them through practice to be available for
spontaneous use. Also, the study of O'Malley and Chamot (1990) indicates that more
successful second language learners use language learning strategies more frequently and
appropriately than less successful ones. 6.0 Conclusion and Recommendation The JRMSU BSED-English major students usually used metacognitive strategies in
learning the second language. Therefore, they are learners who can be considered to be
independent learners since they utilize metacognition or thinking about thinking. In addition,
the students also used social, cognitive, affective, and compensation strategies in engaging in
learning the second language. With these, it can be inferred that these students employed
various strategies in learning the second language. It is further concluded that the overall
grammatical competency level of the students is competent. Nevertheless, the students are competent in word formation and syntax if considered
by component. In contrast, they are less competent in vocabulary, pronunciation, and spelling. Furthermore, the student's language learning strategy use significantly relates to their
grammatical competency level. Hence, training in language learning strategies should always
be part of every language classroom. English teachers may be encouraged to utilize teaching strategies matching the
language mentioned above learning strategies of the students, such as providing them an
avenue for listening to the second language. Language teachers of JRMSU may still offer
extra effort in their teaching. However, it is needed to raise their competency to a "much
competent" or "highly competent," more specifically in spelling, pronunciation, and
vocabulary. English teachers may expose language learners to different language learning
strategies. So be it in the form of language learning strategy training for them to learn English
more effectively. Language teachers may do away with teaching the language through
memory sub-strategies except for learning vocabulary, definitions, and literary texts. Conflict of interests The authors declare no conflict of interest in the publication of this article. Sprin Journal of Arts, Humanities and Social Sciences | Published by Sprin Publisher | https://sprinpub.com 37 37 Elmira C. Rodriguez, Spr. J. Arts. Humanit. Soc. Sci., Vol.02 (4). Apr 2023, pp, 27-42 References Abraham, R. G., & Vann, R. J. (1987). Strategies of two language learners: A case study. In
A. L. Wenden & J. Rubin (Eds.), Learner Strategies in Language Learning. Englewood
Cliffs, NJ: Prentice-Hall. Abraham, R. G., & Vann, R. J. (1987). Strategies of two language learners: A case study. In
A. L. Wenden & J. Rubin (Eds.), Learner Strategies in Language Learning. Englewood
Cliffs, NJ: Prentice-Hall. Akbarov, A. & Arslan, M. (2010). SLA implications to language learning strategies and
pedagogy. In 2nd International Symposium on Sustainable Development. A Conference
Paper, 14-20. https://www.researchgate.net/publication/274080524/download Al-Jarf, R. (2008). Phonological and orthographic problems in EFL college spelling. First
Regional Conference on English Language Teaching and Literature (ELTL1) at Islamic
Azad University, Iran. Al-Qahtani, M. F. (2013). Relationship between English Language, Learning Strategies,
Attitudes, Motivation, and Students’ Academic Achievement. Department of Health
Information Management and Technology, University of Dammam, Saudi Arabia. http://web.a.ebscohost.com/ehost/pdfviewer/pdfviewer?sid=523e0914-667f-4f84-990b-
3ff6e79eea55%40sessionmgr4006&vid=0&hid=4201 Arellano, M. D. C. (2017). Memory learning strategies in English as a Foreign Language in
Vocational Studies. In Tendencias Pedagogicas N ֯29, 229-248. Bachelor
of
Secondary
Education
major
in
English
[BSED-Eng.]
(n.d.). http://www.courses.com.ph/bsed-bachelor-of-secondary-education-major-english-
philippines/ Bailey, A. L., & Butler, F. A. (2007). A conceptual framework of academic English language
for broad application to education. In A. Bailey (Ed.), The Language Demands of
School: Putting Academic English to the Test. New Haven, CT: Yale University Press. School: Putting Academic English to the Test. New Haven, CT: Yale University Press. Bialystok, E. (1981). The role of conscious strategies in second language proficiency. Modern
Language Journal, pp. 65, 24–35. doi:10.1111/j.1540-4781.tb00949.x Bialystok, E. (1981). The role of conscious strategies in second language proficiency. Modern
Language Journal, pp. 65, 24–35. doi:10.1111/j.1540-4781.tb00949.x Brown, D. (1987). Principles of Language Learning and Teaching. New Jersey: Prentice-
Hall. Brown, C. L. (2007). Strategies for making social studies texts more comprehensible for
English language learners. The Social Studies 98 (5), 185–188. tandfonline.com Cabrejas-Penuelas, A. B. (2012). The compensation strategies of an advanced and less
advanced language learner: A case study. In Theory and Practice in Language Studies,
2, 1341-1354. doi:10.4304/tpls.2.7.1341-1354. Canale, M., & Swain, M. (1980). Conflict of interests Theoretical bases of communicative approaches to second
language teaching and testing. Applied Linguistics, 1, 1-47. Canale, M., & Swain, M. (1981). A Theoretical Framework for Communicative Competence. In Palmer, A., Groot, P., & Trosper, G. (Eds.), The contract validation of a test of
communicative competence, 31-36. Canale, M., & Swain, M. (1981). A Theoretical Framework for Communicative Competence. In Palmer, A., Groot, P., & Trosper, G. (Eds.), The contract validation of a test of
communicative competence, 31-36. Celce-Murcia, M. (1988). On directness in communicative language teaching. In TESOL
Quarterly, 32(1), 116-119. doi:10.2307/3587905 Chamot, A. U., Barnhardt, S., El-Dinary, P., & Robbins, J. (1996). Methods for teaching-
learning strategies in the foreign language classroom. In R. L. Oxford (Ed.) Language
learning strategies around the world: Cross-cultural perspectives (175–188), Honolulu,
HI: University of Hawai'i Press. Creswell, J. W. (2003). Research design: Qualitative, quantitative, and mixed methods
approaches (2nd ed.). Thousand Oaks, Sage. Sprin Journal of Arts, Humanities and Social Sciences | Published by Sprin Publisher | https://sprinpub.com Sprin Journal of Arts, Humanities and Social Sciences | Published by Sprin Publisher | https://sprinpub.com 38 Crystal, D. (2003). English as a global language, 2nd ed., UK, University Press, Cambridge. Cohen, A. D. (1996). Strategies in Learning and Using a Second Language. Essex, U.K.:
Longman. Crystal, D. (2003). English as a global language, 2nd ed., UK, University Press, Cambridge. Cohen, A. D. (1996). Strategies in Learning and Using a Second Language. Essex, U.K.:
Longman. C h
A D (1998) S
i
i L
i
d U i
S
d L
E
U K Crystal, D. (2003). English as a global language, 2nd ed., UK, University Press, Cambridge. Cohen, A. D. (1996). Strategies in Learning and Using a Second Language. Essex, U.K.:
Longman. Crystal, D. (2003). English as a global language, 2nd ed., UK, University Press, Cambridge. Cohen, A. D. (1996). Strategies in Learning and Using a Second Language. Essex, U.K.:
Longman. Cohen, A. D. (1998). Strategies in Learning and Using a Second Language. Essex, U.K
L Cohen, A. D. (1998). Strategies in Learning and Using a Second Language. Essex, U.K.:
Longman. Cohen, A. D. (1998). Strategies in Learning and Using a Second Language. Essex, U.K.:
Longman. (
)
g
g
g
g
g
Longman. Cornwall,
T. (2010). Benefits
of
Teaching
Grammar:
Teacher-2
teacher. http://www.speechwork.co.th Cortez, R. A. (2016). Supplementary intervention material in developing grammatical
competence of Grade 7 students. Master’s Thesis, Graduate School. Conflict of interests Ilocos Sur
Polytechnic State College. Diaz-Rico, L. T. & Weed, K. Z. (2010). The cross-cultural, language, and academic
development handbook: A complete K-12 reference guide (4th ed.) Boston: Allyn &
Bacon. Debata, P. (2013). The importance of grammar in English Language Teaching – A
reassessment. In Language in India, 13(5), 482–486. El-Dakhs, D., & Mitchell, A. (2011). Spelling errors among EFL high school graduates. In
The 4th Annual KSAALT Conference, a paper presented in Al Khobar, Prince
Mohammed Bin Fahad University. Ellis, R. (2006). Current issues in the teaching of grammar: An SLA perspective. TESOL
Quarterly, 40(1), 83–107. Francis, D. J. & Rivera, M. O. (2007). Principles underlying English language proficiency
tests and academic accountability for ELLs. In Abedi, J. (Ed.) English Language
Proficiency Assessment in the Nation: Current Status and Future Practice. School of
Education, University of California, Davis. Gaard, E. D. (n.d.). Critique/Classroom Implications on Grammatical Competence. Retrieved
from
https://slaencyclopediaf10.wikispaces.com/Grammatical+Competence+%28Michael+C
anale+%26+Merrill+Swain%29 Gao, C. Z. (2001). Second Language learning and the teaching of grammar. Education, pp. 2,
326–336. Grammar Test. Retrieved from http://esl.fis.edu/grammar/q7m/1.htm Green, J., & Oxford, R. L. (1995). A closer look at learning strategies, L2 proficiency, and
gender. TESOL Quarterly, 29, 261-297. Hinkel, E., & Fotos, S. (2002). New Perspectives on Grammar Teaching in Second Language
Classrooms. Mahwah, N.J.L: Erlbaum Associates. Hong-Nam, K. & Leavell, A. G. (2006). Language learning strategy use of ESL students in an
intensive
English
learning
context. In
Science
Direct,
34(3),
399–415. https://doi.org/10.1016/j.system.2006.02.002 Khamees, K. S. (2016). An evaluative study of memorization as a strategy for learning
English. In
International
Journal
of
English
Linguistics,
6(4),
248-259. DOI:10.5539/ijel.v6n4p248 Krashen, S. (1981). Second Language Acquisition and Second Language Learning. Oxford:
Pergamon Press. Krashen, S. (1981). Second Language Acquisition and Second Language Learning. Oxford:
Pergamon Press. Krashen, S. (1982). Principles and Practice in Second Language Acquisition. Oxford:
Pergamon Press. Krashen, S. (1982). Principles and Practice in Second Language Acquisition. Oxford:
Pergamon Press. rin Journal of Arts, Humanities and Social Sciences | Published by Sprin Publisher | https://sprinpub.com
39 39 39 Elmira C. Rodriguez, Spr. J. Arts. Humanit. Soc. Sci., Vol.02 (4). Apr 2023, pp, 27-42 Krashen, S. & Terrell, T. D. (1983). The Natural Approach: Language acquisition in the
classroom. London: Prentice-Hall Europe. Krashen, S. & Terrell, T. D. (1983). The Natural Approach: Language acquisition in the
classroom. London: Prentice-Hall Europe. Kumaravadivelu, B. (1991). Language-learning tasks: Teacher intention and learner
interpretation. ELT Journal, 45(2), 98-107. Conflict of interests https://doi.org/10.1093/elt/45.2.98 Lasala, C. B. (2014). Communicative competence of senior secondary students: language
instructional pocket. In Procedia – Social and Behavioral Sciences, 134, 226-237. doi:10.1016/j.sbspro.2014.04.243 Liboon, Jr., A., & Rodriguez, E. (2017). Teaching Deficiencies of Teacher-Education
Graduates of JRMSU-TC Based on the DepEd Criteria for Hiring Teachers. JPAIR
Institutional Research, 10(1), 14-28. https://doi.org/10.7719/irj.v10i1.529 Lightbown, P., & Spada, N. (2006). How Languages are Learned (3rd ed.). Oxford: Oxford
University Press. Liu (2003). EFL proficiency, gender, and language learning strategy use among a group of
Chinese
technological
institute
English
majors. ARECLS
E-Journal,
1. http://www.ecls.ncl.ac.uk Lyster, R. (1996). Question forms, conditionals, and second-person pronouns used by
adolescent native speakers across two levels of formality in written and spoken French. The Modern Language Journal, pp. 80, 165–182. Margolis, D. P. (2001). Compensation Strategies by Korean Students. In The PAC Journal,
1(1), 163-174. McLaughlin, B., Rossman, T., & McLeod, B. (1983). Second language learning: An
information-processing perspective. Language Learning, pp. 33, 135–158. Millrood, R. (2014). Cognitive models of grammatical competence of students. Procedia –
Social and Behavioral Sciences, pp. 154, 259–261. Mohamed, A. R., Ismail, S. A. M. M., & Eng, L. S. (n.d.). Now everyone can measure
grammar ability through the use of a grammar assessment system. In International
Journal of Teaching and Education, 2(3), 127–136. Muncie, J. (2002). Finding a place for grammar in EFL composition classes. ELT Journal,
56(2), 180–186. Oxford University Press. National Center for Education Statistics. (2005). National Assessment of Educational
Progress, 2005, reading assessments. Washington, DC: U.S. Department of Education,
Institute of Education Sciences. In Abedi, J. (Ed.) English Language Proficiency
Assessment in the Nation: Current Status and Future Practice. School of Education,
University of California, Davis. Nassaji, H., & Fotos, S. (2011). Teaching Grammar in Second Language Classrooms:
Integrating Form-Focused Instruction in Communicative Context. New York:
Routledge Press. Nunan, D. (1991). Communicative tasks and the language curriculum. TESOL Quarterly
25(3), 279–295. Nunan, D. (1998). Teaching grammar in context. ELT Journal, 52(2), 101-109. https://doi.org/10.1093/elt/52.2.101 Olshtain, E. (1987). The acquisition of new word-formation processes in second language
acquisition. Studies
in
Second
Language
Acquisition,
9(2),
221–231. doi
10.1017/S0272263100000486 Sprin Journal of Arts, Humanities and Social Sciences | Published by Sprin Publisher | https://sprinpub.com 40 Elmira C. Rodriguez, Spr. J. Arts. Humanit. Soc. Sci., Vol.02 (4). Apr 2023, pp, 27-42 O’Malley, M., & Chamot, A. U. (1990). Learning strategies in second language acquisition. Cambridge: Cambridge University Press. Oxford, R. L. (1989). Conflict of interests The role of styles and strategies in second language learning. Retrieved
from http://www.ericdigests.org/pre-9214/styles.htm Oxford, R. (1990). Language Learning Strategies: What Every Teacher Should Know. Boston: Heinle & Heinle. Oxford, R. (1990). Strategy Inventory for Language Learning (SILL). Retrieved from
http://homework.wtuc.edu.tw/sill.php Oxford, R. (2003). Language learning styles and strategies: An overview. Retrieved from
http://hyxy.nankai.edu.cn/jingpinke/buchongyuedu/learning%20strategies%20by%20O
xford.pdf Oxford, R. & Crookall, D. (1989). Language learning strategies, the communicative approach,
and their classroom implications. Foreign Language Annals, 22(1), 29–30. Oxford, R. L., & Burry-Stock, J. A. (1995). Assessing the use of language learning strategies
worldwide with the ESL/EFL version of the strategy inventory for language learning
(SILL). System, 23 (1), 1–23. Pontarolo, L. (2013). The Role of Grammar in EFL Instruction: A Study on Secondary School
Students and Teachers. Dissertation, Italy: University of Padova. Pronunciation Test. https://www.proprofs.com/quiz-school/story.php?title=pronunciation-test Pronunciation
Test. http://highered.mheducation.com/sites/0072982772/student_view0/part6/ten_reading_se
lections-999/multiple_choice_quiz_4.html Ramirez, D.C. (2012). Higher English Proficiency Level among Secondary Students. https//:www.academia.edu.com Richards, J., & Rodgers, T. (2001). Approaches and Methods in Language Teaching (2nd ed.). Cambridge: Cambridge University Press. Rubin, J. (1975). What the "good language learner" can teach us. TESOL Quarterly, pp. 41–
51. Jstor.org Salahshour, F., Sharifi, M., & Salahshour, N. (2012). The relationship between language
learning strategy use, language proficiency level, and learner gender. In Procedia –
Social and Behavioral Sciences, pp. 70, 634–643. doi:10.1016/j.sbspro.2013.01.103 Scarcella, R. (2003). Academic English: A Conceptual Framework. UC Linguistic Minority
Research Institute Technical Report 2003-1. In Abedi, J. (Ed.) English Language
Proficiency Assessment in the Nation: Current Status and Future Practice. School of
Education, University of California, Davis. Skehan, P. (2009). Modeling second language performance: Integrating complexity, accuracy,
fluency,
and
lexis. Applied
Linguistics,
30(4),
510–532. https://doi.org/10.1093/applin/amp047 Spelling Test. http://www.businesswriting.com/cgi-bin/commonmisspelled.cgi ing Test. http://www.businesswriting.com/cgi-bin/commonmisspelled.cgi Su, M. M. (2005). A study of EFL technological and vocational college students’ language
learning strategies and their self-perceived English proficiency. Electronic Journal of
Foreign Language Teaching, 2, 44-56. Sprin Journal of Arts, Humanities and Social Sciences | Published by Sprin Publisher | https://sprinpub.com 41 Elmira C. Rodriguez, Spr. J. Arts. Humanit. Soc. Sci., Vol.02 (4). Apr 2023, pp, 27-42 Elmira C. Rodriguez, Spr. J. Arts. Humanit. Soc. Sci., Vol.02 (4). Apr 2023, pp, 27-42 Swain, M. (1985). Communicative competence: Some roles of comprehensible input and
comprehensible output in its development. In S. Gass & C. Madden (eds.), Input in
Second Language Acquisition. Rowley, MA.: Newbury House. Szyszka, M. (n.d.). Foreign language anxiety and self-perceived English pronunciation
competence. In Studies in Second Language Learning and Teaching, 1(2), 283–300. The State Collegian (2017). The Official Student Publication of Jose Rizal Memorial State
University – Tampilisan Campus. The State Collegian (2017). The Official Student Publication of Jose Rizal Memorial State
University – Tampilisan Campus. Toit, S. D. & Kotze, G. (2009). Metacognitive strategies in teaching and learning
mathematics. Pythagoras, 70, 57-67. Toit, S. D. & Kotze, G. (2009). Metacognitive strategies in teaching and learning
mathematics. Pythagoras, 70, 57-67. Vicenta, V. G. (2002). Grammar learning through strategy training: A classroom study on
learning conditionals through metacognitive and cognitive strategy training. Universitat
de Valencia Servei de Publications. Vicenta, V. G. (2002). Grammar learning through strategy training: A classroom study on
learning conditionals through metacognitive and cognitive strategy training. Universitat
de Valencia Servei de Publications. Vocabulary
Test. http://www.english-grammar.at/worksheets/general-vocabulary/general-
vocabulary-false-friends-1.pdf Vocabulary
Test. http://www.english-grammar.at/worksheets/general-vocabulary/general-
vocabulary-false-friends-1.pdf Wenden A. L. & Rubin, J. (1987). Learner strategies in language learning. Englewood Cliffs,
NJ: Prentice-Hall. Wenden A. L. & Rubin, J. (1987). Learner strategies in language learning. Englewood Cliffs,
NJ: Prentice-Hall. Wharton, G. (2008). Language learning strategy use of Bilingual Foreign Language Learners
in Singapore. In Language Learning, 50(2). https://doi.org/10.1111/0023-8333.00117. Wharton, G. (2008). Language learning strategy use of Bilingual Foreign Language Learners
in Singapore. In Language Learning, 50(2). https://doi.org/10.1111/0023-8333.00117. White, C. J. (1993). A comparative study of metacognitive, cognitive, social, and affective
strategy use in foreign language learning. Dissertation. Massey University. White, C. J. (1993). A comparative study of metacognitive, cognitive, social, and affective
strategy use in foreign language learning. Dissertation. Massey University. Wong, C. Y. & Barrea-Marlys, M. (2012). The role of grammar in communicative language
teaching: An exploration of second language teachers’ perceptions and classroom
practices. ing Test. http://www.businesswriting.com/cgi-bin/commonmisspelled.cgi Electronic Journal of Foreign Teaching, 9(1), 61–75. Wong, C. Y. & Barrea-Marlys, M. (2012). The role of grammar in communicative language
teaching: An exploration of second language teachers’ perceptions and classroom
practices. Electronic Journal of Foreign Teaching, 9(1), 61–75. Word
Formation
Test. http://www.english-grammar.at/online_exercises/word-
formation/wf050-sentences.htm Word
Formation
Test. http://www.english-grammar.at/online_exercises/word-
formation/wf050-sentences.htm Word
Formation
Test. http://www.english-grammar.at/online_exercises/word-
formation/wf037- sentences.htm Word
Formation
Test. http://www.english-grammar.at/online_exercises/word-
formation/wf037- sentences.htm Yaimin (2006). A Study of Language Learning Strategies of Elementary School Pupils at SDN
1 Bengkulu. Thesis. Department of Languages and Arts. Universitas Bengkulu. Yaimin (2006). A Study of Language Learning Strategies of Elementary School Pupils at SDN
1 Bengkulu. Thesis. Department of Languages and Arts. Universitas Bengkulu. Yang, F. C., & Wu, W. C. V. (2015). Using Mixed-Modality Learning Strategies via e-
Learning for Second Language Vocabulary Acquisition. Educational Technology &
Society,
18
(3),
309–322. http://web a ebscohost com/ehost/pdfviewer/pdfviewer?sid=2c73f2fd-2e1d-40b1-a64d- Yang, F. C., & Wu, W. C. V. (2015). Using Mixed-Modality Learning Strategies via e-
Learning for Second Language Vocabulary Acquisition. Educational Technology &
Society,
18
(3),
309–322. http://web.a.ebscohost.com/ehost/pdfviewer/pdfviewer?sid=2c73f2fd-2e1d-40b1-a64d-
16394d2c5c97%40sessionmgr4007&vid=0&hid=4201 Zhang, J. (2009). Necessity of grammar teaching. International Education Studies, 2(2), 184–
187. Zhang, J. (2009). Necessity of grammar teaching. International Education Studies, 2(2), 184–
187. Zhou, Y. (2010). English language learning strategy used by Chinese senior high school
students. In English Language Teaching, 3(4), 152–158. www.ccsenet.org/elt Zhou, Y. (2010). English language learning strategy used by Chinese senior high school
students. In English Language Teaching, 3(4), 152–158. www.ccsenet.org/elt Sprin Journal of Arts, Humanities and Social Sciences | Published by Sprin Publisher | https://sprinpub.com 42 42
|
https://openalex.org/W2418187829
|
https://zenodo.org/records/1995762/files/article.pdf
|
English
| null |
PUERPERAL GENERAL PERITONITIS
|
Annals of surgery
| 1,907
|
public-domain
| 7,165
|
OF NEW YORK. The problems of puerperal infection are still unsolved and,
since the introduction of aseptic methods into obstetric prac¬
tice, no great advance lias been made in the prevention and
treatment of the condition. While women continue to lose their
lives in such large numbers from this condition, little help has
been offered by investigation. This disease annually accounts
for thousands of deaths in this country, and is more worthy
of study than many more abstruse hut less important problems
investigated by special committees and under special grants. The course of puerperal infection shows hut little differ¬
ence from that of severe infection in other parts of the body. The great severity of the condition is due to several factors. The pregnant woman is more susceptible to infection, in gen¬
eral, than the non-pregnant, and is prone to disturbances of
metabolism which lessen resistance and decrease elimination. The traumatism of labor causes a local disturbance of circula¬
tion and the raw surface of the uterus bares a huge surface
for the entrance and growth of micro-organisms. The ana¬
tomical relations and blood supply of the organs have been
temporarily altered and an easy portal of infection is presented
by the genital parts of the recently pregnant woman. Puerperal infection shows varied manifestations and, in
itself, is not a sufficient diagnosis of a condition whose varied
phases require different treatment. The clinical evidences of
puerperal infection arc those of a rapidly advancing genital
infection. The endometrium, in the majority of cases, is the
point of entrance of the micro-organisms and their products. Wounds of the perineum and vagina may become infected
but, as a rule, the infection here remains local. If the endome- ellice McDonald. 204 triiun is the seat of the infection, the micro-organism spreads
into the veins and lymphatics and may gain entrance to the
general circulation. This is the common course of strep¬
tococcus puerperal infection. In this form of infection, the
essential lesion is a peri-uterine lymphangitis. This lymphan¬
gitis follows a similar course to the common and well known
streptococcus infection of the arm, save that it is modified by
the more generous lymphatic and blood supply and an adjacent
serous membrane—the peritoneum. OF NEW YORK. This lymphangitis is more commonly associated with
thrombosis than with the presence of actual pus in the lym¬
phatics, which is a residt only in severe cases, and accounts for
Trendelenberg’s1 findings of thrombosis of the uterine and
spermatic veins in 21 out of 43 fatal cases, and lymphatic infec¬
tion in only four. Grossman,2 in a study of 51 post-mortem
examinations in puerperal infection, found peri-uterine
thrombo phlebitis alone in fourteen instances, and associated
with lymphangitis in 13 other cases. However, the frequency
of peritonitis, which must be from direct lymphatic extension,
shows that the condition is more common than the post-mortem
records show. The thrombosis of the peri-uterine veins is a
matter of secondary consideration and a comparatively unim¬
portant complication of the disease. The essential lesion is a
spreading lymphangitis following the course of lymphangitis
elsewhere in the body. Cuff 8 has recorded a case in which laparotomy was done,
of a severe puerperal infection with rigors and very high
temperature. There was, on vaginal examination, a mass on
the right side of the uterus which proved to be the broad liga¬
ment, “ the thickness of three fingers stretching from the uterus
wall internally to the pelvic wall externally.” The ovarian vein
was tied and the patient recovered. While assisting Noble of
Philadelphia, I saw two such cases of puerperal infection in
which the broad ligament was thickened, respectively, to one
and to two and half fingers in breadth. Both cases got well
after closure of the abdomen without further surgical treatment. There was no abdominal pus in either Cuff’s case or the two
referred to. This condition is, however, a preliminary state PUERPERAL GENERAL PERITONITIS. 205 before peritoneal infection. However, the condition may sub¬
side before peritoneal infection results, and become spontan¬
eously cured. This form of peritonitis is usually streptococcic and is
marked by severe clinical manifestations, with little evidence of
effort at repair on the part of the peritoneum. The local seat
of the disease is about the broad ligaments and the pus is
found generally distributed within the abdominal cavity, with
little or no evidence of repair in the distant parts. Intestinal
paralysis soon follows and the streptococcic form of the dis¬
ease is almost always fatal. OF NEW YORK. Sargent4 believes that no recov¬
eries follow this type of peritonitis, but I have seen two cases
of general peritoneal infection with streptococcus recover after
a simple laparotomy and drainage. While this type of peritonitis is a most common one in the
puerperium, there are other forms in which the infecting
organism does not pass through the genital canal, but is the
result of the lighting up of a previous focus of infection. Into
this class fall all cases of infection from bruising of tumors,
rupture of pus tubes, or purulent collections about the adjacent
pelvic organs. Any variety of a pelvic tumor may be bruised
or have its blood supply cut off by torsion at the pedicle and
become inflamed and gangrenous, thus setting up a peritonitis. The important point to recognize is that, in this form, the lym¬
phatics are not involved, and that the treatment is that of a
non-puerperal peritonitis from a similar cause. However, it
frequently happens, particularly in infection from pus tubes,
that there may be associated two infections, one of which is of
the nature of a lymphangitis, and the other is the old infected
focus lighted up. Thus, as occurred in one of my cases, an old
purulent gonorrhoeal salpingitis may become reinfected with
streptococci by lymphatic extension and form pus collections. This may rupture, causing a general peritonitis. In this way it
will be seen that it is difficult to make hard and fast subdivisions
between lymphatic puerperal peritonitis and peritonitis from
previous infection. While the streptococcus is the most common cause of
lymphatic peritonitis, other organisms also have a share in its ellice McDonald. 206 production. The streptococcus is the most frequent cause of
puerperal infection, and in a previous paper 0 it was estimated
to be the causative organism in 40 per cent, of all cases. In more recent communications on this subject, Lloyd,"
iti 159 cases, says streptococci were found in 33 cases, staphy¬
lococci in 30 cases, pneumococci in 17 cases, gonococci in 21,
colon bacilli in 22 cases and bacillus capsulatus ;erogenes in
2 cases. While these statistics are based upon a study of the
infection in the puerperium, it has also been shown that the
streptococcus is a frequent cause of infection in premature
labor and abortion. * Some of these cases have been referred to before in other com¬
munications. A number of them were studied in the Bender Laboratory
and some are from the records of that place. Others were studied at the
N. Y. Lying-in Hospital and Kensington Hospital for Women. OF NEW YORK. The frequency of the incidence of the various infecting
organisms, is shown by tile following table of eleven cases, in
which they occur in about the usual ratio: TABLE OF CASES*
Case. I-ocal Lesions. Character
of Peritonitis. Complications. Organisms,
' 1 Acute hxmorrhagic
endometritis. Acute suppurative. Acute suppurative en¬
docarditis; septic
pneumonia. Staphylococcus
aureus. 11 Abortion; acute sup¬
purative endome¬
tritis. Acute purulent. Acute suppurative
salpingitis. Pneumococcus. III
Premature labor. Acute purulent. Bilateral gonorrhoeal
salpingitis. B. coli; streptococ¬
cus; gonococcus. IV
Abortion; gangren¬
ous endometritis. Acute purulent, with
petechial haemor¬
rhages. Acute yellow atrophy
of liver. Streptococcus; 11. cap¬
sulatus aerogenes. V
Acute exudative en¬
dometritis. Acute plastic. Salpingitis; hydrosal¬
pinx. Streptococcus. VI
General plastic. Acute purulent. Pyosalpinx. Streptococcus. Purulent endome-
Iritis,
Streptococcus; 13. coll. VIII
Acute endometritis. Acute purulent. Abscess of cul-de-sac. Streptococcus. IX
1 Acute purulent endo¬
metritis. Acute purulent endo¬
metritis. Fibro-purulent. Acute purulent. Pericarditis. Staphylococcus
aureus. Streptococcus. XI
Acute purulent, with
multiple local collec-
tions. Bilateral purulent sal- Gonococcus. pingitis. * Some of these cases have been referred to before in other com¬
munications. A number of them were studied in the Bender Laboratory
d
f
th
d
f th t
l
Oth
t di d
t th * Some of these cases have been referred to before in other com¬
munications. A number of them were studied in the Bender Laboratory
and some are from the records of that place. Others were studied at the
N. Y. Lying-in Hospital and Kensington Hospital for Women. PUERPERAL GENERAL PERITONITIS. 207 These cases show that the streptococcus was the infecting
organism alone in 3 cases, and associated with other bacilli in 3
cases. Thus, it will be seen that the streptococcus occupies a
greater share as a cause of puerperal peritonitis than it does
of puerperal infection in general. Also, Sargent4 states that,
in a study of 258 cases of peritoneal lesions, the staphylococcus
albus was found in 108 cases; and he states that this and the
colon bacillus form the most important and frequent causative
organism concerned in peritonitis, outside of pregnancy. In
the cases in which streptococcus was concerned, there were very
acute symptoms and marked clinical disturbances. The course
was fulminating, and was associated with marked anatomical
disturbances in other parts of the body. OF NEW YORK. Case III, a case of premature labor at the sixth month,
showed few symptoms until the seventh day, when there
was a high temperature of 104°, with abdominal tenderness and
rigidity. The uterus seemed well involuted and there was
profuse purulent discharge. The physical examination showed
a moderate sized fixed mass on the right side of the uterus. No swelling or mass in the cul-de-sac. The uterine discharge
contained the gonococcus and streptococcus. The leukocyte
count at this time was 29,000, with 90.5 per cent, polynuclears. On the ninth day after delivery, the patient was taken with
sudden severe abdominal pains and marked tenderness and
rigidity. The pain was peristaltic in character, intermittent
and intense. The tongue was dry and there was slight vomit¬
ing. The physical examination showed fulness in the cul-de-sac,
with marked tenderness and a disappearance of the mass on
the right side; leukocyte count was 11,000, with 90 per cent,
polynuclears. Laparotomy showed general distribution of
sero-purulent fluid with marked congestion and redness of the
viscera. This is most marked in the pelvic region. The right
tube had formed part of the wail of an abscess beside the uterus
and pus exuded from the fimbriated end and from the cut
surface of the broad ligament. Death occurred nine hours
after rupture of the abscess. The gonococcus, streptococcus and
colon bacillus were recovered from the abdominal fluid. ELLICE MCDONALD. 208 Case IV was one of criminal abortion at the seventh
month, and was seen five days after labor. The pulse was
140 and temperature 102°. There was slight jaundice. The
leukocyte count was 19,000, and the polynuclears 86 per cent. Vomiting was continuous until stupor intervened and increased
to coma. The patient died on the fifth day after the operation. The autopsy showed acute yellow atrophy of the liver, with
general peritonitis. In the fluid was found the streptococcus
and B. capsulatus aarogenes. Case V was one in which streptococcic infection was asso¬
ciated with an old pyosalpinx; this caused an acute peritonitis,
more marked in the lower part of the abdominal cavity, but
which invaded the upper part as well. There was little pus. The hydrosalpinx on the left side was uninfected. Case VI was a similar case to Case V, save that the peri-
tonitial infection was not as acute, and there had been some
attempt at repair by adhesions. OF NEW YORK. Case VII showed general peritonitis with a large amount
of pus containing colon bacilli and streptococci. This followed
upon an acute purulent endometritis. Death occurred on the
tenth day. Case X was a similar case, being marked by severe vomit¬
ing, a high temperature of 105° and intense pain and rigidity. Death occurred on the ninth day. While the cases of streptococcus infection showed severe
symptoms, the cases of infection with staphylococcus aureus
also showed acute clinical manifestations with marked anatomi¬
cal lesions. In this series the staphylococcus aureus was pres¬
ent in two cases. In the study of puerperal infection, the staphylococcus
aureus has been seldom isolated. Fullerton and Bonney7 have
found this organism in one out of 54 cases. Lloyd,0 in a study
of 159 cases, found staphylococcus in 30. The type of infec¬
tion with this organism seems to be one of great severity, with
a tendency toward pyemia and the production of metastatic
abscesses. Such was the result in Case I where, after an apparently PUERPERAL GENERAL PERITONITIS. 209 normal pregnancy, the patient was delivered by a midwife. She was seen on the eighth day of the puerperium, profoundly
infected. There was marked tenderness and rigidity. Tem¬
perature was 104.30. Death occurred on the next day. No
operative measures were attempted. Post-mortem examina¬
tion showed a large quantity of peritoneal pus with a dull
injected peritoneum covered with a small amount of fibrin. The pericardial cavity also contained pus. There were areas
of septic pneumonia. Bacteriological examination of the
organs and pus showed staphylococcus pyogenes aureus. Case IX was somewhat similar but ran a more prolonged
course. The patient entered the hospital on the eighteenth day
and succumbed to the profound infection on the same day. Autopsy showed a pericarditis, empyema and general peritonitis
with little evidence of repair. The broad ligaments were
thickened and firm and there was macroscopic evidence of
lymphangitis, pus exuding from the cut surface of the broad
ligament. In these two cases, the tendency of staphylococcus
infection toward extension to the neighboring cavities and
metastatic abscesses is well shown. Gonorrhoeal infection is one of the most common varieties
of infection in the puerperium, and is the least frequently dis¬
covered. It usually causes little or no rise of temperature,
but may cause high fever, serious morbid disturbances and
death. OF NEW YORK. Stone and myself8 found the gonococcus in the lochia
of puerperal women in 17 out of 171 cases. This organism
was associated with peritonitis in only one case of that series. This is Case III, in which there was a bilateral purulent salpin¬
gitis from gonorrheeal infection and, following upon this,
lymphatic peritonitis from the streptococcus, rupture of an
abscess and general peritoneal infection. Case XI was one of gonorrheeal infection of the uterus
with extension to the tubes and peritoneum. In this case,
there was a lesion of gonorrheeal infection before pregnancy,
and the disease followed its usual course of rapid extension
in the puerperium. The temperature went to ioi° and the
third day of the puerperium, the pulse was 96. The gonococ- ELLICE MCDONALD. 210 cus was isolated from the vaginal discharge. Pain, rigidity
and a temperature of 103.6° appeared on the fifth day. On
the sixth day, indefinite masses in the abdomen and pelvis were
made out on abdominal and vaginal examination. Pulse at
this time was 120. Pulse and temperature kept high until the
seventeenth day when death occurred. Operation was refused. Autopsy showed bilateral purulent salpingitis and a general
peritonitis of some duration. Collections of pus were found
encapsulated by intestines and adhesions in various parts of the
abdomen. The gonococcus was isolated in pure culture. Mann0 has reported a somewhat similar case, in which the
symptoms came on the tenth day. Temperature was 107°,
pulse 120. Death occurred and the post-mortem examination
showed a peritonitis from pure gonococcus infection. Gonorrhoeal infection usually extends by direct continuity
of mucous membrane, but may, in the soft condition of the
genitalia, penetrate the uterine muscle and extend into the
broad ligament. Salpingitis is a frequent complication of this
infection in the puerperium, as is shown by the study of Stone
and myself,8 in which we found clinical evidence of extension
of the infection beyond the uterus, in 7 out of 17 cases. This
extension of the disease may continue, as in Cases III and XI,
to cause a peritonitis which is, however, usually localized in
the cul-de-sac. The late complications of gonorrhoeal infec¬
tion are more to be feared than is the earlier uncommon gen¬
eral peritonitis. Salpingitis and pelvic peritonitis are the most
common results. OF NEW YORK. This is the cause of “ one child sterility.” Pneumococcus infection is one of the more uncommon
forms of puerperal peritoneal infection, and one of compara¬
tively slight severity, save in isolated cases, as in Case II. This
condition resulted in a woman four months pregnant, follow¬
ing an induced abortion. She was seen five days after the
induction and was curetted for retained secundines. The pneu¬
mococcus was isolated from smears and cultures taken at the
time of curettage. The temperature was 102 and continued
high until death three days afterwards from peritonitis. g
y
p
Post-mortem examination showed all the peritoneal sur- PUERPERAL GENERAL PERITONITIS. 211 faces to be covered try a sticky, greenish-yellow purulent exu¬
date which was thin between the adherent intestines, but thick
between surfaces held apart by collections of fluid. All the
dependent parts of the abdomen contained yellowish turbid
fibrino-purulent exudate. There was an acute suppurative
salpingitis and endometritis. The pneumococcus was recov¬
ered from the peritoneal exudate. While pneumococcus puerperal infection is uncommon, a
number of cases have recently been reported. Weichelbaum 14
as well as Bar and Tissicr," have reported cases; and Cohn
one similar to this. Fullerton and Bonney1 found six cases of
pneumococcic infection in 54 cases of puerperal fever; and
Lloyd4 found the pneumococcus in 17 out of 159 cases of
puerperal infection. The pneumococcus is being more fre¬
quently isolated in such conditions, as the bacteriological
methods of study improve. That the condition is by no means
rare as a cause of infection of the peritoneum, is shown by
Annand’s and Bowen’s12 collection of 91 cases of pneumococ¬
cus peritonitis in children. They found that in half the cases
the pus was encysted and that the peritoneal infection was
usually secondary to some remote pneumococcal lesion. The
exhaustive discussion by Jensen13 of this form of peritonitis,
gives a thorough idea of the subject. He reports several very
interesting cases and gives a list of 143 references. The character of this form of peritoneal infection is
usually that of Case II, reported here. The infection is char¬
acterized by a plastic exudate, very rich in fibrin, which causes
adhesions and encapsulation of the exudate; quite rarely, in the
severer forms, the entire mass of the intestines adhere together
and are surrounded by pus. OF NEW YORK. In typical cases, the clinical pic¬
ture is quite striking; the onset is that of an acute peritonitis,
followed, very soon, by a chronic stage with-mild symptoms
and indefinite masses on abdominal palpation. The diag¬
nosis is never certain without bacteriological examination,
although one might suspect this infection from the thin, odor¬
less, greenish-yellow pus and the abundant fibrinous adhesions. The prognosis is, as a rule, favorable; but recovery without ellice McDonald. 212 operation is rare. Simple evacuation of pus and drainage of
the collections are all that is usually required. There is, in this series of cases, no instance of peritoneal
infection resulting from direct traumatism or necrosis of a
tumor from pressure or torsion of the pedicle. A fibroid or
ovarian cyst may take on rapid growth during pregnancy, and
from torsion of the pedicle or sudden loss of nourishment from
lessening of the blood supply after labor may become necrotic
and infected. Lepage and Mouchotte14 have collected a num¬
ber of such cases. Similar causes lead to infection from ova¬
rian cysts; and Getter15 has reported 21 cases where, in spite
of normal labors, infection of the cyst has occurred and led
to fatal peritonitis. The infection is commonly due to the
colon bacillus and is usually widespread and severe. Lawrence
has reported ten such cases sent to the hospital as peritonitis
following a puerperal genital infection. Patton 17 has collected
321 cases of ovarian cysts in pregnancy. In 95 cases treated
expectantly until labor, torsion of the pedicle occurred 29
times—4 times during labor and 25 times during the puerpe-
rium. Rupture happened 13 times—3 before and 10 after
or during labor. There were 25 deaths in the 95 cases, only
4 of which occurred in patients who had operations after labor,
and 21 in those who were treated wholly expectantly. Gen¬
eral peritonitis occurred in 7 of the 95 cases. There were 184
cases treated by operation, with a mortality of 8 (4.3 per
cent.). Infection of ovarian cysts is especially likely to hap¬
pen in the early puerperium. The teratoid ovarian tumors are
particularly liable to be aroused from quiescence to rapid
growth during pregnancy. This is well shown by a review
of 35 cases by Neuhauser.18 Appendicitis is another lesion which may cause an exten¬
sive and fatal form of peritonitis in pregnancy and the puer¬
perium. OF NEW YORK. The lessened resistance to infection of the pregnant woman
has added to it the local abdominal disturbances of the pres¬
ence of the mass of the gravid uterus and the increased vascu¬
larity of the pelvic viscera. Futh,10 in his recent papers, up- PUERPERAL GENERAL PERITONITIS. 213 holds YValdeyer’s statement that the caecum and appendix are
pushed up during pregnancy. This displacement begins about
the fourth month, when the uterus rises out of the true pelvis. It then opens up the broad ligaments and the ovarian vessels
are enormously increased in size. This elevation of the ctecum
is of clinical significance as, in addition to predisposition to
disease by reason of the alteration of position, the focus of
inflammation is thereby placed in a more dangerous position—
i.e., higher in the abdomen, where adhesions are more easily
torn and where inflammatory processes spread with greater
ease. The caecum returns to its proper place after labor; but,
if the appendix is adherent to the uterus or adnexa, it is
dragged into the true pelvis by the involution of the uterus. This may cause rupture of an abscess and increase the extent
of the inflammation to the general peritoneum. Seven casts of the abdominal cavity of women, dying
during pregnancy, are described by Filth 20 to uphold this view. Five cases of appendicitis in pregnancy are also reported, and
he states that appendicitis is much more dangerous after the
fourth month, on account of the size of the uterus influencing
the position of the caecum. He has divided Boije’s series into
two groups. In the first, under four months, there were 10
cases, with three deaths. In the second, from the fourth to
the ninth month, there were 32 cases with 19 deaths—a much
higher mortality. This seems to be clear proof of the effect
of the continuance of pregnancy upon the mortality of this
condition. Hlawccek,21 in 1897, collected 13 cases of perito¬
nitis from this cause with 11 deaths; but under early operative
treatment, the prognosis is brighter. The diagnosis of generalized peritonitis in the puerperium
is by 110 means easy. The extension of an infection in a
woman already severely infected, shows few additional symp¬
toms and small increase of pathognomonic signs. The diag¬
nosis of puerperal genital infection is in itself not always easy,
and the presence of pathogenic micro-organisms in the lochia is
not proof of infection. OF NEW YORK. Bumm and Sigwart22 found that, by
very careful examination, the streptococcus was isolated in the ellice McDonald. 214 vaginal discharge of 38 per cent, of women in the later months
of pregnancy. Leo,23 in an examination of the lochia in 38
normal women, in the puerperiuni, found the streptococcus in
the vagina in 50 per cent, and in the uterine lochia in 17.6 per
cent. In the later days of the puerperiuni, streptococcus is
more commonly found, as was shown by a study of Schenk and
Schieb,24 who found this organism four times more frequently
late in the puerperiuni than at the beginning. They found that
streptococci existed in the lochia of one-third of all normal
women. Also the discovery of a certain organism in the
vagina, although an indication is not pathognomonic of the
cause of infection. Sargent4 reports a case of gonococcus in¬
fection found in the vagina while the peritoneum was infected
with the pneumococcus. Stone and myself25 have shown, in a
study of the gonococcus, that this organism may exist in the
uterine lochia without causing temperature. However, the
presence of certain bacteria in the uterine lochia, as shown by
smears and cultures, is some indication of the cause of infection. The time at which the extension of the infection to the
peritoneum occurs seems to be very variable. Lymphatic peri¬
tonitis from streptococcus usually appears from the third to
the tenth day of the puerperiuni. It is seldom earlier but often
later. The time of onset of symptoms of peritonitis from
previous lesions also varies within a wide limit, but is usually
later in appearance than the lymphatic form. Pain is usually a prominent symptom, and occurred in all
my cases. The pain of peritonitis is, I believe, fairly charac¬
teristic, and is a great aid in the diagnosis of the condition. It depends upon two conditions: First, it is now recognized
that most of the pain in peritonitis is due to an accompanying
lymphangitis. This, in part, causes the crampy pains of peri¬
tonitis—the lymph vessels of the intestines press upon the
sensory nerves as the lymphangitis extends. The pain in the
lymphatic peritonitis is not usually localized, as in appendicitis,
but is sometimes referred, like that of appendicitis, to the epi¬
gastrium. This is supposed to be due to the infection in the
peri-lymphatic tissue and the lymph glands around the aorta. OF NEW YORK. PUERPERAL GENERAL PERITONITIS. 215 There seems to be but little pain from lymphangitis of the
broad ligament. The second source of pain is from the exterior of the gut
itself, and is due to the presence of an irritant causing an
inflammation of the peritoneum. The movement of the intes¬
tine causes severe crampy pains. That this pain is due some¬
what to the irritant, and not directly to the inflammation, is
shown by a case of abdominal hremorrhage following C;esarian
hysterectomy upon which I did a laparotomy to control the
bleeding. After clean hysterectomy there was a sudden flow
of blood into the peritoneal cavity. The patient, who had
been resting quietly, immediately complained of intense pain
in the epigastrium and in the diaphragmatic region, generally. There was difficulty in breathing and nausea. The pain was
intermittent and intense. Immediate laparotomy showed that
the unclotted blod was generally distributed throughout the
abdominal cavity. The character of the pain was similar to
that of Case III, in which there was a sudden rupture of a
pus collection, causing intense intermittent pain, crampy or
peristaltic in character. In addition to the direct irritation
and inflammation of the peritoneum, the pain is further caused
by the rubbing of the inflamed intestinal covering against the
parietal peritoneum. The visceral serosa has comparatively
few sensory nerves while the parietal peritoneum is exception¬
ally well supplied. The pain in the diaphragmatic region, in
the two cases cited, was probably due to irritation of the parie¬
tal serosa in the area. It is also recognized, however, that
peritonitis may exist in the center of the belly, beneath the
colon and above the pelvis, amongst the coils of the small intes¬
tine, for some lime and become widespread without causing
marked pain. That acute abdominal symptoms may be caused by a
lymphangitis alone, is shown by a report of Rowjand,20 of two
cases of operation upon supposed perforation in typhoid fever. Masses of enlarged lymphatic glands were found in the mesen¬
tery of the gut without any evidence of perforation or peri¬
tonitis. The pain was intermittent and peristaltic. There was ellice McDonald. 2l6 localization of the tenderness with little or no rigidity. OF NEW YORK. Arm¬
strong27 has reported a similar case in typhoid fever; and
McCrae,28 in his study of the pain in typhoid fever, cites two
cases in which the explanation for the acute symptoms was the
enlarged mesenteric glands. In peritonitis, while the intestines are quiet, pain is not
usually a marked feature; but, during peristalsis, it is usually
present. For this reason, it is often useful for diagnostic pur¬
poses to set up peristalsis and elicit the pain by giving a purga¬
tive enema. This usually also gives a clue to the location of
the point of greatest inflammation. Rigidity is usually an early and trustworthy sign. It is
present in all cases of peritonitis which were seen at their
inception. It was present in the 7 cases of localized gonor¬
rhoeal peritonitis, before referred to. Rigidity, however, de¬
pends, in some measure, upon the suddenness of the onset of the
peritonitis and sometimes does not last long. Intestinal disten¬
sion usually overcomes it and causes it to disappear. To the
educated hand of the surgeon, it is the most reliable early
symptom. Tenderness on palpation is not usually a marked symptom
unless there has been marked effort at repair and the formation
of much exudate and many adhesions. It may sometimes be
produced vaginally by movement of the uterus. It can also
be elicited after peristalsis has been set up. Vomiting is a fairly constant symptom and occurs at two
different periods in the disease, First, at the onset, there is
usually vomiting, regurgitative in character, and later there
is more persistent bile-stained vomiting, often frecal or hsenior-:
rhagic in character. The temperature was rather variable in character. In
the infection with the more virulent organisms, it went very
high and was intermittent in character; but in the last days of
the disease, this intermission was not present. In the less
virulent infection by gonococcus and pneumococcus, the tem¬
perature while high, did not rise above 103°. The pulse in
these two instances was also lower than in the more virulent PUERPERAL GENERAL PERITONITIS. 217 form. There was, however, nothing characteristic about the
less virulent peritonitis. In Case III, after rupture of the
abscess, the pulse suddenly rose from 90 to 120 and became
of high tension and thready. The pulse is usually a better indi¬
cation of the condition of the patient than is the amount of
fever. OF NEW YORK. The blood changes of general peritonitis are interesting
and instructive. There is usually a diminution in the red cells. This is more marked in puerperal peritonitis, and, in puerperal
infections generally, than in infection in the 11011-pregnant. The leukocyte count is increased as it is in all septic conditions. This depends upon many conditions, as the patient’s resistance
and the virulence of the infection. It may be said that, as a
general rule, in streptococcus infection, the leukocyte count is
less in the pregnant than in the non-pregnant. A sudden fali
in the leukocyte count (as in Case III, from 29,000 to 11,000)
is suggestive of an overpowering of the system by toxins of
the infective organisms. The polymorphonuclear leukocytes
are usually increased in percentage. Iodophilia is another useful sign and gives reliable evi¬
dence in all septic conditions. Alter staining with weak solu¬
tion of iodine (Ehrlich’s method), the blood, in cases of septic
infection, usually shows a reaction in the cytoplasm of the
leukocytes. This iodophilia usually occurs in the polymor¬
phonuclear neutrophile cells and sometimes in the lymphocytes. It docs not bear a definite relation to the leukocytosis; but de¬
pends upon the amount of toxemia, not upon the leukocyte
range. Thus, the blood of a profoundly septic person may
show intense iodophilia with a fall in the leukocyte count;
while a high leukocytosis, without iodophilia, is not incompat¬
ible with an infection exciting a toxaemia sufficient to stimulate
the cells to overproduction, but not of a character to affect
them structurally. This fact makes the reaction especially
valuable in cases of puerperal peritonitis, where there is often
sudden overpowering of the system by toxins. An interesting
review of the work of Cabot, Locke,2B Dunham,30 Dunn 81 and
Keen 32 upon this subject is given by Da Costa,33 with a report ellice McDonald. 218 of ioo cases, including 30 cases of sepsis in which the iodo-
philia was present. It must be remembered, however, that the diagnosis of
peritonitis can never depend upon the blood changes which only
give confirmative evidence of infective processes. No hard
and fast rule in regard to the degree of leukocytosis can be
laid down, in spite of efforts of some of the camp followers of
scientific investigation to have us diagnose our cases of sepsis
in their private laboratories. OF NEW YORK. The bacteriology of the blood and loebia is of importance
in forming an exact idea of the extent and character of the
infection. Smears of the uterine lochia often give immediate
and useful information in regard to the possible cause of the
peritonitis. Careful physical examination should be made upon all
cases of suspected puerperal peritonitis, as it is the exception
and not the rule for all patients with severe puerperal peri¬
tonitis to be without complications and infection in other
organs. Care should be taken to recognize lung and heart
complications, and search should be made for metastatic collec¬
tions. Pyelitis and pyonephrosis are also not uncommon re¬
sults of infection in pregnancy. The use of the Pravaz or Hol¬
lenbeck 31 needle has been suggested for the detection of free
pus in the abdominal cavity, but should be used with great care,
as there may be adhesions which would cause the bowel to be
punctured. The treatment of uncomplicated puerperal peritonitis is
that of peritonitis in the non-pregnant. Immediate laparotomy
and evacuation of the purulent matter with removal of the
focus of infection, if possible, is recognized by all surgeons to
be the proper procedure. While this is true in all cases in
which peritoneal infection has resulted from previous foci as
pus tubes, abscesses or necrosis of tumors, there is more dif¬
ference of opinion as regards the value of operation and the
procedure in lymphatic peritonitis. Trendelenberg has advised, in this condition, resection or
ligature of the veins of the broad ligament. It is decidedly PUERPERAL GENERAL PERITONITIS. 219 questionable whether this operation has any great value, as
cases are noted in which recovery has taken place after simple
laparotomy. The thrombosis of the veins is merely an incident
of the infection, and offers no excuse for the laceration of the
broad ligament and the division of the lymphatic channels, to
allow the escape of micro-organisms. Hysterectomy, an opera¬
tion which has been practically abandoned unless there is a local
lesion, has more raison d’etre, for it removed a greater part of
the infected surface and allowed of better drainage. In order to thoroughly discuss operative measures in the
lymphatic form of general puerperal peritonitis, a proper
knowledge of the processes of infection with the most common
causative organism, the streptococcus, is necessary. OF NEW YORK. Strepto¬
coccus infection differs from most other infections in that the
blood serum does not acquire streptococcidal properties, but
the destruction of the cocci and relief from the infection is
brought about by the leukocytes. There is a relatively small
amount of toxin developed and no bactericidal properties in tbc
serum, and it is reasonable to suppose that phagocytosis is an
important factor in recovery. I11 addition to a leukocytosis,
there must be an increase in the opsonin, i.c., an increase in the
power which prepares the leukocytes to engulf the bacteria. Any substance, then, which aids the body defenses to resist the
infection is of use. Anti-streptococcic sera have been tried for many years
with poor results; but recent reports show that properly pre¬
pared polyvalent serum has given more satisfaction. Bumm 35
reviews his results in 32 cases of severe infection, and in four
particularly striking cases of high fever, but without phagocy¬
tosis. Hyperlcukocytosis occurred 12 hours after the injection
of the serum and was intense, the leukocytes incorporating the
steptococci with great avidity. Ran 30 has reported similar
good results, and a reduction of his mortality from 60 to 36
per cent, in streptococcus infection. Escherich 37 has also re¬
ported beneficial effects in scarlet fever with the Moser poly¬
valent serum. Anti-streptococcus serum is, however, more of
a prophylactic measure against peritonitis and it is doubtful ELLtcii McDonald. 220 whether it has much effect after the infection has passed the
confines of the uterus. It should, however, be used. Other substances, however, are used to enhance the resist¬
ance of the peritoneum and to increase the activity of the
leukocytes. Mickulicz 38 used intra-peritoneal injections' of
nucleic acid and performed operations as soon as the leukocytes
began to increase. In 45 cases, the increase varied from 9 per
cent, to 452 per cent. Seven cases died, but none from peri¬
tonitis, although two recovered in spite of this complication. Diez 38 also used nucleic acid and advocates the injection of a
2 per cent, solution as a preliminary to operation. Hanner40 reports 51 cases in which 50 c.c. of the same
acid was injected in the form of sodium nucleate 13 hours be¬
fore operation. Marked hyper-leukocytosis was present in
every case. The reaction was always striking, the temperature
rising one to two degrees C., with a chill in some instances. OF NEW YORK. I-Ie also, as did Gray,41 used quantities of dead organisms as
vaccine, but the effect was doubtful. It would, therefore, seem that the injection of nucleic acid
is a useful adjunct to early operation in generalized peritoneal
infection, particularly in infection from the streptococcus. Operation should remove, if possible, any focus of infec¬
tion with as little traumatism and handling of the intestines as
possible. Lavage should not be performed, as the risk of
spreading the infection is too great. The treatment should be
after the method of Murphy and consist of making a small
opening with the introduction of drainage tubes. Vaginal
drainage is often useful. Evacuation of the intestinal contents
in cases with paralyzed intestinal walls is sometimes of use. A high rectal tube may answer when the sigmoid is involved. If paralysis be higher than this, enterostomy is necessary. Peritonitis characterized by abundant sero-purulent exu¬
date and unaccompanied by signs of deep inflammation is a
comparatively benign affection provided that operation is
timely; but that form characterized by little or no exudate
and with the gut wall red, dry, distended and paralyzed, gives
an almost hopeless prognosis. PUERPERAL GENERAL PERITONITIS. 221 The prognosis, as a whole, in puerperal peritonitis is bad
unless early operation is done. However, results are improv¬
ing, as is shown by 121 cases collected by Jeannin.49 These
histories extend back to the early days of abdominal surgery,
yet there were 60 recoveries and 61 deaths. This showing will
be improved with more modern methods of treatment and
earlier operation. In cases of puerperal general peritonitis
treated expectantly, the result is almost invariably fatal, while
in those cases treated surgically there are frequent successes,
hence every case of puerperal general peritonitis should be oper¬
ated upon as soon as diagnosed. In the after-prognosis of such severe puerperal infections,
it should be kept in mind that the late appearance of metastatic
abscesses is not an uncommon condition. REFERENCES. I Trciulclcnberg. Mount f. Geb. ti. Gyn., 1902, October. * Cuff. Jour, of Obs. and Gyn. of B. E., 1906, ix, 5. * Sargent. Jour, of Obs. and Gyn. of B. E., 1906, ix, 3. ‘McDonald. Aitier. Med., 1906, Feb. u. ‘McDonald. Aitier. Med., 1906, Feb. u. ' Lloyd. Intercolonial Med. Jour, of Australasia, 1906, Oct. ' Lloyd. Intercolonial Med. Jour, of Australasia, 1906, Oct. y
' Fullerton and Bonney. Trans. London Obstct. Soc., 1904. ' Fullerton and Bonney. Trans. London Obstct. Soc., 1904. y
* Stone and McDonald. Surgery, Gyn. and Obst. 1906, Feb. * Stone and McDonald. Surgery, Gyn. and Obst. 1906, Feb. * Mann. Amcr. Gymccology, 1903, ii, 6. 10 Wcichelbautn. Wcin klin. Woch., 1888, x. 28 10 Wcichelbautn. Wcin klin. Woch., 1888, x. 28. " Bar and Tissicr. L'obstctrique, 1896, 97. q
"Atmand and Bowen. Lancet, 1906, June. "Atmand and Bowen. Lancet, 1906, June. “Jensen. Arcbiv f. klin. Cltir., 1903, lxx, p. 91. “Jensen. Arcbiv f. klin. Cltir., 1903, lxx, p. 91. " Lepage and Moucbottc. Annalcs de Gyn. d’Obstet., 1906, Feb. " Getter. Loc. cit. “ Lawrence. Quoted by Kynock, Jour, of Obs. and Gyn. of B. E.,
1906, June. “ Lawrence. Quoted by Kynock, Jour, of Obs. and Gyn. of B. E.,
1906, June. " Patton. Surgery Gyn. and Obs., 1906, iii, 3. " Patton. Surgery Gyn. and Obs., 1906, iii, 3. “ Ncwhauser. Archiv f. Gyn., 1906, lxxix, 3 “Futli. Archiv f. Gyn. Ixxvi, 3, p. 507. “Futli. Archiv f. Gyn. Ixxvi, 3, p. 507. p 5
II Hlawecek. Monat f. Geb. u. Gyn., 1897, vi, .4. II Hlawecek. Monat f. Geb. u. Gyn., 1897, vi, .4. y
“ Bumm and Sigwart. Hegar’s Beitrage zur Geb. u. Gyn., 1906, viii, 3. “ Bumm and Sigwart. Hegar’s Beitrage zur Geb. u. Gyn., 1906, viii “Leo. Munch, med. Woch., 1904, li, 2131. “Leo. Munch, med. Woch., 1904, li, 2131. "Schenk and Schieb. Munch med. Woch., 1904, li, 2129. "Schenk and Schieb. Munch med. Woch., 1904, li, 2129. “ Stone and McDonald. Surgery Gyn. and Obs., 1906, Feb. “ Stone and McDonald. Surgery Gyn. and Obs., 1906, Feb. "Rowland. Jour. A. M. A., 1906, Feb. 17. "Rowland. Jour. A. M. A., 1906, Feb. 17. ellice McDonald. 222 " Armstrong. Annals of Surgery, 1902, November. “ McCrac, J. N. Y. Med. Jour., 1904, May 4. ” Cabot and Locke. Quoted by Da Costa. " Dunham. Loc cit. REFERENCES. 11 Dunn. Loc. cit. ” Keen. Loc cit. “ Da Costa. Therapeutic Gaz., 1906, Sept. sl Hollenbeck. Jour. A. M. A., 1906, Oct. 20. 55 Bumm. Berlin klin. Woch., 1905, xli, 44. x Rau. Med. Press, 1905, Nov. 8. " Eschcrich. Weill klin. Woch., 1903, 23. “Mickulicz. Lancet, 1904, July 2. ”Dicz. Gaz. di Ospedali, 1905, xxvii, 57. "Hanner. Centr. f. Gyn., 1905, xxx, 24. "Gray. Lancet, 1906, April 21. “Jeannin. L’Obstctrique, 1906, May. " Armstrong. Annals of Surgery, 1902, November. “ McCrac, J. N. Y. Med. Jour., 1904, May 4. ” Cabot and Locke. Quoted by Da Costa. " Dunham. Loc cit. 11 Dunn. Loc. cit. 11 Dunn. Loc. cit. ” Keen. Loc cit. ” Keen. Loc cit. “ Da Costa. Therapeutic Gaz., 1906, Sept. sl Hollenbeck. Jour. A. M. A., 1906, Oct. 20 " Eschcrich. Weill klin. Woch., 1903, 23. ”Dicz. Gaz. di Ospedali, 1905, xxvii, 57. "Hanner. Centr. f. Gyn., 1905, xxx, 24. "Gray. Lancet, 1906, April 21. “Jeannin. L’Obstctrique, 1906, May.
|
https://openalex.org/W4387010442
|
https://ieeexplore.ieee.org/ielx7/6287639/6514899/10262319.pdf
|
English
| null |
Optimized Artificial Intelligence Model for DDoS Detection in SDN Environment
|
IEEE access
| 2,023
|
cc-by
| 14,756
|
This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2023.3319214 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2023.3319214 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2023.3319214 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2023.3319214 4 4 Date of publication xxxx 00, 0000, date of current version xxxx 00, 0000. Date of publication xxxx 00, 0000, date of current version xxxx 00, 0000. Date of publication xxxx 00, 0000, date of current version xxxx 00, 0000. Digital Object Identifier 10.1109/ACCESS.2023.0322000 Digital Object Identifier 10.1109/ACCESS.2023.0322000 YOUSIF AL-DUNAINAWI 1, BILAL R. AL-KASEEM 2, (Member, IEEE), AND HAMED S.
AL-RAWESHIDY 3, (Senior, IEEE) 1Department of Cybersecurity Engineering, College of Engineering and Information Technology, AlShaab University, Baghdad, Iraq
2Department of Communication Engineering, College of Engineering and Information Technology, AlShaab University, Baghdad, Iraq
3Department of Electronic and Electrical Engineering, College of Engineering, Design and Physical Sciences, Brunel University London, London, United
Kingdom Corresponding author: Hamed S. Al-Raweshidy (e-mail: hamed.al-raweshidy@brunel.ac.uk). ABSTRACT
Distributed denial of service (DDoS) attacks continue to be a major security concern,
threatening the availability and reliability of network services. Software-defined networking (SDN) has
emerged as a promising solution to address this issue, enabling centralized network control and management. However, conventional SDN-based DDoS mitigation techniques often struggle to detect and mitigate
sophisticated attacks due to their limited ability to analyze complex traffic patterns. This paper proposes an
innovative and optimized approach that effectively combines mininet, Ryu controller, and one dimensional-
convolutional neural network (1D-CNN) to detect and mitigate DDoS attacks in SDN environments. The
proposed approach involves training the 1D-CNN model with labeled network traffic data to effectively
identify abnormal patterns associated with DDoS attacks. Furthermore, seven hyperparameters of the trained
1D-CNN model were tuned using non-dominated sorting genetic algorithm II (NSGA-II) to achieve the
best accuracy with minimum training time. Once the optimized 1D-CNN model detects an attack, the
Ryu controller dynamically adapts the network policies and employs appropriate mitigation techniques
to protect the network infrastructure. To evaluate the effectiveness of the optimized 1D-CNN model,
extensive experiments were conducted using a simulated SDN environment with a realistic DDoS attack
dataset. The experimental results demonstrate that the developed approach achieves significantly improved
detection accuracy of 99.99% compared to other machine learning (ML) models. The NSGA-II enhances
the optimized model accuracy with an improvement rate of 9.5%, 8%, 5.4%, and 2.6% when it is compared
to logistic regression (LR), random forest (RF), support vector machine (SVM), and k-nearest neighbor
(KNN) optimized models respectively. This research paves the way for future developments in leveraging
deep learning (DL) driven techniques and SDN architectures to address evolving cybersecurity challenges. INDEX TERMS Artificial Intelligence; Distributed Denial of Service; Hyperparameters Tuning; Mininet;
NSGA-II; Optimized Model; Ryu Controller; Software Defined Networking. INDEX TERMS Artificial Intelligence; Distributed Denial of Service; Hyperparameters Tuning; Mininet;
NSGA-II; Optimized Model; Ryu Controller; Software Defined Networking. This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/ I. INTRODUCTION the network, making high-level decisions and orchestrating
the configuration of network devices. The data plane, also
known as the forwarding plane, is responsible for the actual
forwarding of network traffic, including data packets, from
source to destination based on the instructions received from
the control plane. It is the operational layer of the network
where the physical network devices (such as switches and
routers) perform the essential task of transmitting packets. The data plane devices in an SDN architecture follow instruc-
tions provided by the centralized controller [1]. S S
OFTWARE-defined networking (SDN) is a networking
approach that separates the control plane from the data
plane in a network. The control plane is one of the core
components responsible for managing and controlling the
behavior of the network. It involves making decisions about
how network traffic should be handled, how data should
flow through the network, and how network resources should
be allocated. The control plane essentially defines the rules
and policies that govern the operation of the network. The
functions of the control plane are shifted from individual net-
working devices to a centralized software application known
as the SDN controller. The controller acts as a brain for In traditional or legacy networks, network devices such as
switches and routers have control planes and make indepen-
dent forwarding decisions based on their routing tables. This 1 VOLUME 11, 2023 VOLUME 11, 2023 Yousif Al-Dunainawi et al.: Optimized Artificial Intelligence Model for DDoS Detection in SDN Environment
This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2023.3319214 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2023.3319214 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2023.3319214 Yousif Al-Dunainawi et al.: Optimized Artificial Intelligence Model for DDoS Detection in SDN Environment DDoS is a cyber attack involving overwhelming a network
or server with traffic from multiple sources. A DDoS attack
aims to disrupt or bring down the targeted network or server. I. INTRODUCTION DDoS attacks are typically carried out using a botnet, which
is a network of compromised computers that are controlled
by a single attacker. DDoS attacks can be very difficult to
defend against and are a common threat to online services
and businesses [5].i feature can lead to inefficient use of network resources and
difficulty in managing network configurations [2]. In con-
trast, SDN centralizes the control plane and enables network
administrators to manage network traffic and configurations
from a central location using software applications. These
applications allow more efficient use of network resources,
easier network management, and greater flexibility in adapt-
ing to changing network needs. Fig. 1 illustrates the legacy
network and SDN architectures. Some of the key differences
between SDN and legacy networks include [3] [4]: DDoS attacks significantly threaten network security, and
SDN environments are not immune to them. In fact, SDN
environments are particularly vulnerable to DDoS attacks
because of their centralized control and programmability. A
DDoS attack in an SDN environment can occur in several
ways. Attackers may flood the network with a high volume of
traffic to overwhelm the network and cause it to crash. Alter-
natively, they may exploit vulnerabilities in SDN controllers
or switches to gain control of the network and launch DDoS
attacks against other targets. The programmability of SDN
also allows attackers to modify the network configuration to
evade detection or launch more sophisticated attacks. For ex-
ample, attackers may modify the flow rules in SDN switches
to redirect traffic to a victim or bypass DDoS mitigation
mechanisms. To prevent DDoS attacks in SDN environments,
it is important to implement effective security measures [6].i 1) Centralized control: In SDN, the control plane is central-
ized and managed by a software controller, whereas, in
legacy networks, each network device has its own control
plane. 2) Programmability: SDN networks are programmable,
meaning network administrators can write applications to
manage network traffic and configurations. In contrast,
legacy networks are not as flexible and require manual
configuration. 3) Virtualization: SDN enables network virtualization, which
allows multiple virtual networks to run on a single phys-
ical network infrastructure. Network virtualization can
improve network efficiency and reduce costs. 4) Open standards: SDN is based on open standards, meaning
network devices from different vendors can work together
seamlessly. I. INTRODUCTION The open standard is in contrast to legacy
networks, which often require proprietary technologies
that are specific to each vendor. Artificial Intelligence (AI) is a branch of computer science
that focuses on creating intelligent machines that can learn
and make decisions like humans. AI is used in various ap-
plications, from image recognition to robotics. Some of the
popular AI technologies include machine learning (ML), deep
learning (DL), and natural language processing (NLP) [7] [8]. AI, SDN, and DDoS are all related to the field of computer
networking and security, and integrated them can be carried
out in many ways [9] [10]: FIGURE 1. The legacy network and SDN architectures • AI for DDoS detection and mitigation: AI can be used to de-
tect and mitigate DDoS attacks by utilizing ML algorithms
that can analyze network traffic patterns and identify ab-
normal traffic behavior that could indicate a DDoS attack. Once a DDoS attack is detected, an SDN controller can be
used to reconfigure the network to block or mitigate the
attack dynamically. • AI for DDoS detection and mitigation: AI can be used to de-
tect and mitigate DDoS attacks by utilizing ML algorithms
that can analyze network traffic patterns and identify ab-
normal traffic behavior that could indicate a DDoS attack. Once a DDoS attack is detected, an SDN controller can be
used to reconfigure the network to block or mitigate the
attack dynamically. • SDN for AI-enabled network management: SDN can be
used to provide a flexible and programmable network in-
frastructure that enables AI to manage network resources. With SDN, network administrators can dynamically allo-
cate network resources, prioritize traffic, and control net-
work policies. AI can be used to optimize network per-
formance, predict network failures, and automate network
management tasks based on current traffic status. FIGURE 1. The legacy network and SDN architectures SDN is a rapidly evolving technology that is transforming
how networks are managed and operated. Recent trends in
SDN technology have focused on enhancing network per-
formance, improving security, and increasing the scalability
of networks. SDN is used to enhance network security by
providing greater visibility into network traffic and enabling
dynamic security policies that include using SDN to detect
and mitigate distributed denial of service (DDoS) attacks and
other types of cyber threats [4]. This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/ I. INTRODUCTION • AI-enabled SDN for DDoS defense: AI can be used to
enhance the effectiveness of SDN-based DDoS defense
mechanisms. For example, AI algorithms can be used to
automatically configure SDN-based DDoS mitigation poli-
cies based on real-time network traffic patterns. The auto-
matic configuration can help to detect and mitigate DDoS
attacks more quickly and efficiently without affecting the
network performance. 2 VOLUME 11, 2023 2 This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/ nawi et al : Optimized Artificial Intelligence Model for DDoS Detection in SDN Environment
This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2023.3319214 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2023.3319214 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2023.3319214 4 Yousif Al-Dunainawi et al.: Optimized Artificial Intelligence Model for DDoS Detection in SDN Environment was adopted to tune the SVM parameters to find their opti-
mum values. The authors compared their developed approach
to a single SVM classifier, and the results showed that their
classifier achieved a high detection accuracy of 98.90%. The main contributions of this paper navigate the intricate
intersection of DDoS attack mitigation and detection mecha-
nisms within the context of SDN using AI techniques and can
be summarized as follows: Zhijun et al. in [13] introduced a DDoS detection approach
using factorization machines (FM) for low-rate DDoS attacks
that targeted the data pane in SDN architecture. The obtained
experimental results from their developed approach showed
a moderate detection accuracy of 95.80% due to the fine-
grained detection for low-rate DDoS attacks. 1) Creating SDN simulated traffic dataset by utilizing
mininet emulator. The created dataset launches DDoS
flooding attacks for multiple protocols (i.e., user data-
gram protocol (UDP), Internet control message protocol
(ICMP), and transmission control protocol (TCP)). II. RELATED WORKSi II. RELATED WORKSi Swami et al. in [16] presented a detailed analysis of var-
ious classifiers to detect the ingress DDoS attack traffic
in an SDN environment. TCP-SYN flooding attacks were
employed to study the effectiveness of the developed ML
classifiers, such as decision tree (DT), multilayer perceptron
(MLP), AdaBoost (AB), RF, and logistic regression (LR). The authors utilized a cross-validation technique to validate
the introduced classification models, and the obtained results
from the conducted experiments showed that their introduced
approach achieved high performance. The field of utilizing AI in the SDN environment has wit-
nessed significant advancements in recent years, with numer-
ous scholars and researchers investigating various aspects of
the subject matter. This section provides a comprehensive
review of the existing literature, highlighting the key studies,
theories, and methodologies that have shaped the understand-
ing of protecting the SDN environment from DDoS attacks. By examining the works of prominent researchers, this review
aims to identify the gaps and limitations in the current knowl-
edge, paving the way for the present study’s contribution. Sangodoyin et al. in [17] developed multiple ML mod-
els to detect and classify DDoS attacks against SDN. Their
introduced SDN model had been emulated using mininet,
while three DDoS flooding attacks were launched to study
the effectiveness of their ML models. Four popular classi-
fiers were employed by the authors (i.e., classification and
regression tree (CART), KNN, Gaussian naïve Bayes (GNB),
and quadratic discriminant analysis (QDA)), and the obtained
results showed that the CART model outperformed other
models with prediction accuracy of 98%. Swami et al. in [11] introduced a voting-based intrusion de-
tection framework that could detect DDoS attacks and secure
the SDN environment. The authors employed three ensemble
classifiers called Voting-CKM, Voting-RKM, and Voting-
CMN. While NSL-KDD, CICIDS2017, and UNSW-NB15
were utilized for training and testing their developed models. The authors argued that their ensemble models had better
performance when compared to other existing models with
a classification accuracy of 99.68%, 97.77%, and 89.29% for
Voting-CKM, Voting-RKM, and Voting-CMN, respectively. Maheshwari et al. in [18] introduced an optimized
weighted voting ensemble model that employed a hybrid
metaheuristic optimization algorithm (BHO) to find the op-
timal set of the model’s weights. Their introduced ensemble
classifier was based on a couple of each of the following
classifiers: SVM, RF, and gradient-boosted machines (GBM). Sahoo et al. This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/ I. INTRODUCTION 2) Developing a DL model based on one dimensional-
convolutional neural network (1D-CNN) that employs a
stacked convolution technique for improving the model’s
ability to extract complex features from the input data,
handle variations, and prevent overfitting. Ahuja et al. in [14] utilized ML techniques to classify
the incoming traffic in an SDN environment into normal
and DDoS attack traffic. The authors developed their hybrid
model by integrating two classification algorithms: SVM and
random forest (RF). The obtained results from their experi-
mental scenario showed that the support vector classifier with
random forest (SVC-RF) model achieved a high classifica-
tion accuracy of 98.8% when a realistic SDN dataset was
employed in training and testing their model. 3) Optimizing the 1D-CNN model’s hyperparameters using
non-dominated sorting genetic algorithm II (NSGA-II) for
achieving the best accuracy with minimum training time. The rest of the paper is organized as follows: Section II
reviews the recent state-of-the-art research that targeted the
integration of AI to detect or mitigate DDoS attacks in the
SDN environment. Section III summarizes the DDoS opera-
tion and its effects on network resources. While Section IV
and Section V provide a clear explanation of the mininet em-
ulator and the employed ryu controller. Section VI introduces
the proposed approach steps in detail and highlights the main
contributions of this work. On the other hand, Section VII
discusses the obtained results from the developed approach. Finally, Section VIII concludes the presented work. Musumeci et al. in [15] developed an automated DDoS
attacks detection (DAD) approach for SDN environment that
aimed at detecting DDoS attacks in OpenFlow switch with
the marginal intervention of SDN controllers. The authors
introduced two versions of the DAD approach: standalone
and correlated. These approaches were able to detect SYN
flood attacks locally on the OpenFlow switch level to protect
the SDN controller from DDoS attacks. The numerical results
for the DAD model showed a classification accuracy of 98%
for all tested ML algorithms (k-nearest neighbors (KNN), RF,
SVM, and artificial neural network (ANN)). III. SDN ENVIRONMENT UNDER DDOS ATTACK The informa-
tion entropy detected the spoofed switch ports, while the CNN
model was used to differentiate between suspicious traffic
and normal traffic. The authors argued that their proposed
approach exhibited high detection accuracy of 98.98% when
detecting DDoS attacks. FIGURE 2. The impact of DDoS attack in SDN environment The main motivation behind the conducted research was
that DDoS attacks are becoming increasingly sophisticated,
making them challenging to detect using traditional methods. DL techniques have shown promise in capturing and learning
intricate patterns in data, making them suitable for detecting
the complex and evolving nature of DDoS attacks. The po-
tential research gaps for developing an AI model to detect
DDoS attacks, relative to existing works, could be: (i) none
of the papers cited above considered the imbalance of classes
in the employed dataset, and (ii) some of the papers cited
above did not utilize SDN emulated traffic dataset. However,
publicly accessible datasets created for legacy networks were
not applicable in the SDN context. This work utilizes 1D-
CNN that aims to leverage the strengths of DL, specifically in
capturing complex patterns, automatic feature extraction, and
adaptability. By harnessing these capabilities, the developed
1D-CNN model will enhance the accuracy, efficiency, and ef-
fectiveness of DDoS detection systems, ultimately improving
the resilience and security of network infrastructures. FIGURE 2. The impact of DDoS attack in SDN environment Fig. 2 illustrates how the attacker strikes the SDN environ-
ment from many potential points of entry and the impacts of
a DDoS attack in the SDN environment are [23] [24]: • Impact on OpenFlow switch: DDoS attacks generate a mas-
sive volume of malicious traffic that floods the network in-
frastructure. OpenFlow switches, responsible for forward-
ing and controlling network flows in the SDN environment,
may become overwhelmed by excessive traffic. These large
number of malicious flows may flood the switch, resulting
in the exhaustion of flow table entries. Once the flow table
reaches its capacity, the switch may either stop accepting
new flows or evict existing flows, affecting the network’s
ability to process traffic effectively. Consequently, this can
lead to network congestion, causing delays, packet loss, and
overall performance degradation. To the best of our knowledge, this paper underscores the
significant contributions that have been made to the field of
SDN security. III. SDN ENVIRONMENT UNDER DDOS ATTACK The author argued that the employed six classifiers were
differentiated by their hyperparameter values and eliminated
false negatives through the introduced dynamic fitness func-
tion. The obtained results from their developed ensemble
model showed that their model had high classification ac-
curacy of 99.35% and 99.41% for CAIDA-2007 and CIC-
DDoS-2019 datasets, respectively. DDoS attack is one of the most regular and devastating
attacks against SDN. This kind of attack has an impact on
programmable networks performance and its behavior. It
stops or degrades network services by depleting the available
resources, and hence, valid hosts cannot communicate with
the SDN controller or deliver packets across the network [22]. In an SDN environment, a DDoS attack is accomplished
by generating several new flows that overwhelm the SDN
controller, the OpenFlow switches, and the secure channel,
causing the network to go down for valid hosts. It is important
to note that while some attack vectors are typical of traditional
networks, programmable networks (e.g., SDN) have their
own distinct set of dangers. For example, an attacker may
take advantage of low-volume traffic flows rather than high-
volume traffic flows to create a large number of ingress mes-
sages, which, in turn, would overwhelm the ingress switch
and the controller. To be more explicit, attackers will produce
numerous new flows that have faked IP addresses but will
send them from several sources. A table-miss has occurred
because these faked addresses do not match any of the current
flow rules in the flow table of the OpenFlow switch. Alanazi et al. in [19] proposed an efficient DDoS detec-
tion approach for SDN environment using DL. The authors
employed the CIC-IDS2017 dataset to train their classifier
using long-short-term memory (LSTM), gated recurrent unity
(GRU), and CNN to enhance the traffic classification. The
authors argued that their proposed approach achieved a detec-
tion accuracy of 99.77% while a few features were adopted. Mbasuva et al. in [20] proposed a DDoS detection ap-
proach for SDN environment that utilized recurrent neural
network (RNN), deep neural network (DNN), and CNN. The
authors employed the CIC-IDS2017 dataset to train their
developed ensemble model, and they argued that their exper-
imental results outperformed other ensemble classifiers (i.e.,
ensemble voting, ensemble RNN, and ensemble CNN) by
achieving a high detection accuracy of 99.05%. Liu et al. in [21] introduced a DDoS attack detection ap-
proach comprising information entropy and DL. This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/ II. RELATED WORKSi in [12] utilized a support vector machine
(SVM) as the main classifier for DDoS detection and miti-
gation in the SDN environment. The classifier was followed
by kernel principal component analysis (KPCA) to perform
feature selection. At the same time, a genetic algorithm (GA) 3 VOLUME 11, 2023 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2023.3319214 Yousif Al-Dunainawi et al.: Optimized Artificial Intelligence Model for DDoS Detection in SDN Environment
een accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2023.3319214 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2023.3319214 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2023.3319214 Yousif Al-Dunainawi et al.: Optimized Artificial Intelligence Model for DDoS Detection in SDN Environment Yousif Al-Dunainawi et al.: Optimized Artificial Intelligence Model for DDoS Detection in SDN Environment Yousif Al-Dunainawi et al.: Optimized Artificial Intelligence Model for DDoS Detection in SDN Environment Yousif Al-Dunainawi et al.: Optimized Artificial Intelligence Model for DDoS Detection in SDN Environment V. RYU CONTROLLER An SDN controller is a software program that manages and
directs network traffic flow in an SDN environment. The
SDN controller manages the network and communicates with
the switches that forward the traffic. It provides a single
point of control for the network, allowing network admin-
istrators to manage the network more easily and efficiently. The controller communicates with the switches using the
OpenFlow protocol which is a standardized protocol used
in SDN environments. The OpenFlow protocol allows the
controller to program the switches to forward traffic in a
specific way based on the network policies and rules defined
by the administrator [28]. Various networks can be created with mininet using virtual
hosts, switches, controllers, and links that behave just like
a real network but without the expense or limitations of
physical hardware. In addition, it provides an environment to
run experiments, test different network topologies, and evalu-
ate network performance under different conditions. Mininet
supports various network protocols, including OpenFlow, and
can be integrated with other network simulation tools and em-
ulators. The architecture of mininet consists of the following
components [25] [26] [27]: Ryu is an open-source SDN controller that provides a
programmable feature to network infrastructure for cloud
computing and data center environments. Ryu is written in
Python and supports the OpenFlow protocol, which is widely
used in SDN environments. The Ryu controller architecture
is designed to provide a flexible and modular framework
for building SDN applications. It is based on a component-
based architecture, where each component is responsible for
a specific function or feature. The key components of the Ryu
controller architecture are [29] [30]: • Hosts in Mininet represent end devices such as computers
or servers in a network. Each host is implemented as a
Linux network namespace, providing a fully functional net-
work stack. Hosts can run applications, generate traffic, and
communicate with other hosts within the mininet network. • Switches in mininet emulate OpenFlow switches, which
are central to SDN. Each switch is implemented as a vir-
tual Ethernet bridge and supports OpenFlow protocol for
communication with the SDN controller. Switches in the
mininet handle the forwarding of network traffic based on
the flow table entries provided by the SDN controller. • Ryu application framework provides a programming
framework for building SDN applications. It is based on
the Python programming language and provides a simple
and easy-to-use API for developing SDN applications. III. SDN ENVIRONMENT UNDER DDOS ATTACK Through meticulous investigation and rigorous
analysis, this paper shed new light on DDoS detection in the
SDN environment, providing valuable insights and advancing
our understanding of the benefits of optimizing the hyperpa-
rameters of the 1D-CNN model. The findings of this paper
pave the way for future investigations in this area. VOLUME 11, 2023 4 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2023.3319214 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2023.3319214 Yousif Al-Dunainawi et al.: Optimized Artificial Intelligence Model for DDoS Detection in SDN Environment • Impact on the secure channel between SDN planes: The se-
cure channel, which is responsible for transmitting control
messages and instructions between the SDN planes, can be
compromised or disrupted during a DDoS attack, leading
to various consequences like (i) communication disruption
between control and data planes, and (ii) increasing latency
in the secure channel by consuming network resources that
lead to increased packet loss and queuing delays. protocol to install flow table entries, handle network events,
and control the behavior of the network. • Links in mininet represent the virtual network connections
between switches and hosts. They emulate network links
and provide connectivity between different components
within the mininet network. Links can have configurable
characteristics such as bandwidth, delay, and packet loss,
allowing users to simulate various network conditions. • Impact on SDN controller: DDoS attacks can overwhelm
the resources of the SDN controller, which is the central
intelligence that manages and controls the network infras-
tructure, including CPU, memory, and network bandwidth.i • Custom network topologies in mininet can be created by
users throughout the arrangement of switches, hosts, and
links. Topologies range from simple linear or tree-like
structures to more complex and realistic configurations. Mininet provides a flexible application programming in-
terface (API) for defining topologies programmatically or
using predefined topologies. • Impact on SDN controller: DDoS attacks can overwhelm
the resources of the SDN controller, which is the central
intelligence that manages and controls the network infras-
tructure, including CPU, memory, and network bandwidth. Massive malicious traffic can consume these resources,
resulting in performance degradation. III. SDN ENVIRONMENT UNDER DDOS ATTACK As a result, the con-
troller may struggle to process control messages, make
intelligent decisions, and effectively manage the network,
leading to delays and unresponsiveness. Hence, critical
network services and applications relying on the controller
may experience downtime or degraded performance. • Command line interface (CLI) and API in mininet allow
users to manage and control the network interactively. Users can use the CLI to execute commands on hosts,
view network configurations, and test network behavior. Additionally, mininet offers a Python API that enables
programmatic control and automation of network setup and
testing. IV. MININET EMULATOR Mininet is an open-source network emulator that allows de-
velopers and academic scholars to create a virtual network on
their own computers or in the cloud. It enables developers
to build, test, and debug complex network topologies and
protocols without the need for physical hardware. Mininet
is a powerful tool for network emulation and testing, with a
wide range of features and capabilities that make it useful for
research, education, and network development [25]. This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/ Yousif Al-Dunainawi et al.: Optimized Artificial Intelligence Model for DDoS Detection in SDN Environment of the employed network. The mininet provides a vir-
tual network with multiple hosts, switches, and con-
trollers. The employed topology comprises a single
remote Ryu controller connected to eight OpenFlow
switches that are connected to three hosts. • The OpenFlow library is responsible for communicating
with OpenFlow switches and other network devices. It
provides a high-level abstraction of the OpenFlow protocol,
allowing developers to write SDN applications without
worrying about the protocol’s low-level details. of the employed network. The mininet provides a vir-
tual network with multiple hosts, switches, and con-
trollers. The employed topology comprises a single
remote Ryu controller connected to eight OpenFlow
switches that are connected to three hosts.ii Step 2: • Network application manager is responsible for managing
the lifecycle of SDN applications. It provides an interface
for registering and unregistering applications with the con-
troller and managing application dependencies. Step 2: Traffic Generation: To simulate normal traffic, vari-
ous communication patterns have been executed be-
tween hosts that mimic real-world communication
patterns, and this can be achieved by having hosts
send requests to each other using common protocols
like web servers might receive HTTP requests from
clients, and clients might fetch data from servers. On the other hand, to simulate DDoS traffic, a high
volume of traffic from multiple hosts should be gen-
erated using various tools like “iperf” and “hping.” A
custom script was used to simulate attack traffic for
ICMP, UDP, and TCP protocols. • Representational state transfer (REST) API server provides
a RESTful interface for interacting with the Ryu controller. It allows developers to interact with the controller using
hypertext transfer protocol (HTTP) requests and provides
access to the controller’s features and functionality.i • Database is responsible for storing the configuration and
state information of the controller. It provides a persistent
storage mechanism for the controller, allowing it to recover
from failures and restarts. p
Step 3: Monitoring and Logging: A standalone application
runs inside the Ryu controller that monitors the
global topology construction and then logs different
OpenFlow switch flows statistics on a regular basis. The OpenFlow switch flow details and requests to the
SDN controller were written into a comma-separated
values (CSV) file. The collected dataset was anno-
tated automatically using a simple code that set the
label column of the dataset to “0” during the gen-
eration of benign traffic and set the label column of
the dataset to “1” during the generation of malicious
traffic. Ryu has gained popularity in the SDN community due to
its simplicity, extensibility, and Python-based programming
model. Researchers and developers widely use Ryu controller
to build applications and network management solutions. VI. PROPOSED APPROACH This section presents an innovative and optimized approach
that leverages the integration of the mininet network emulator,
the Ryu controller, and the 1D-CNN model to tackle the
challenge of detecting and mitigating DDoS attacks in SDN
environments. The developed approach aims to enhance the
network’s ability to identify and respond to malicious traffic
patterns in a timely and automated manner. By combining the
flexibility and programmability of SDN, the realistic network
emulation capabilities of mininet, the centralized control pro-
vided by the Ryu controller, and the DL capabilities of the 1D-
CNN model, this work envisions a comprehensive solution
that addresses the limitations of traditional DDoS mitigation
techniques. FIGURE 3. Dataset collection steps of the simulated network This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/ V. RYU CONTROLLER • Controllers in mininet represent the SDN controller that
is responsible for managing the network and providing
instructions to the switches. Mininet supports various SDN
controllers such as OpenDaylight, Ryu, and POX. The con-
troller communicates with switches using the OpenFlow • Event manager is responsible for receiving and handling
network events, such as switch connection/disconnection
events, packet-in events, and flow modification events. It
uses an event-driven programming model to handle events
in real-time. 5 VOLUME 11, 2023 Yousif Al-Dunainawi et al.: Optimized Artificial Intelligence Model for DDoS Detection in SDN Environment
This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2023.3319214 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2023.3319214 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2023.3319214 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2023.3319214 Yousif Al-Dunainawi et al.: Optimized Artificial Intelligence Model for DDoS Detection in SDN Environment A. DATASET COLLECTION AND PREPROCESSING This work utilizes a simulated approach to construct the
whole dataset specifically targeted to capture the DDoS at-
tacks in an SDN-based environment. There are many avail-
able datasets, but they lack diversity in attack types and are
unrealistic. Therefore, training DL models will be difficult to
capture real-world scenarios’ behavior effectively. To address
this issue, this work tried to generate various benign and
malicious network traffic for three different protocols using
various tools. Simulating benign traffic and DDoS attacks in
mininet involve creating a virtual network environment that
mimics real-world communication patterns between hosts
and the characteristics of a real DDoS attacks. Fig. 3 depicts
the simulated network architecture and how the dataset was
collected. The dataset collection steps were identical except
the tools that utilized during network traffic generation: FIGURE 3. Dataset collection steps of the simulated network Fig. 4 depicts the details of the proposed approach compris-
ing three main phases: dataset collection, AI model training,
and optimized model benchmarking. Step 1: Create Topology: Define the network topology using
mininet’s Python API. This Python program creates
hosts, switches, and links to simulate the architecture Step 1: Create Topology: Define the network topology using
mininet’s Python API. This Python program creates
hosts, switches, and links to simulate the architecture 6 VOLUME 11, 2023 This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/ This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2023.3319214 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2023.3319214 4 ce Model for DDoS Detection in SDN Environment Yousif Al-Dunainawi et al.: Optimized Artificial Intelligence Model for DDoS Detection in SDN Environment l.: Optimized Artificial Intelligence Model for DDoS Detection in SDN Environment
d approach phases to develop the optimized DDoS detection model GURE 4. The proposed approach phases to develop the optimized DDoS detection model
LUME 11, 2023
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/ FIGURE 4. The proposed approach phases to develop the optimized DDoS detection model This work is licensed under a Creative Commons Attribution 4.0 License. A. DATASET COLLECTION AND PREPROCESSING The sample count for benign and malicious traffic in the collected dataset (c) SMOTE to whole dataset (a) Original dataset (b) SMOTE to individual protocol (b) SMOTE to individual protocol (b) SMOTE to individual protocol (c) SMOTE to whole dataset (a) Original dataset FIGURE 6. The sample count for benign and malicious traffic in the collected dataset Fig. 5(a) and Fig. 6(a) show the detailed statistics of the col-
lected dataset in this work, with classes 0 and 1 corresponding
to benign traffic and malicious traffic, respectively. Table 1
shows the features of the collected dataset that were employed
to train the developed model. phase: (i) removing redundant entries, unchanged features,
infinite values, and null; (ii) encoding categorical variables
into a format that can be processed by the 1D-CNN model,
preventing biases, and enhancing the model’s ability to learn
relevant patterns and features from the data; and (iii) normal-
izing the numeric features between 0 and 1 is a key prepro-
cessing step that enhances the training efficiency, stability,
and generalization capability of 1D-CNN models, leading to
better overall performance. TABLE 1. The features of the employed dataset
No. Feature Name
Data Type
1
Switch ID
Numerical (integer)
2
Source IP
Categorical (object)
3
Destination IP
Categorical (object)
4
Packet Count
Numerical (integer)
5
Byte Count
Numerical (integer)
6
Duration
Numerical (integer)
7
Packet In
Numerical (integer)
8
Packet / Flow
Numerical (integer)
9
Byte / Flow
Numerical (integer)
10
Packet Rate
Numerical (integer)
11
Protocol
Categorical (object)
12
Port Number
Numerical (integer)
13
Transmitted Bytes
Numerical (integer)
14
Received Bytes
Numerical (integer)
15
Label
Numerical (integer) TABLE 1. The features of the employed dataset The main observation from Fig. 5(a) and Fig. 6(a) is that the
classes of the employed dataset are unbalanced. Therefore,
the synthetic minority over-sampling technique (SMOTE)
is utilized to improve the balance in dataset classes. This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/ A. DATASET COLLECTION AND PREPROCESSING For more information, see https://creativecommons.org/licenses/by/4.0/ Y
if Al D
i
i
t
l
O ti
i
d A tifi i l I t lli
M d l f
DD S D t
ti
i
SDN E
i
t
This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2023.3319214 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2023.3319214 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2023.3319214 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2023.3319214 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2023.3319214 Yousif Al-Dunainawi et al.: Optimized Artificial Intelligence Model for DDoS Detection in SDN Environment (a) Original dataset
(b) SMOTE to individual protocol
(c) SMOTE to whole dataset
FIGURE 5. The percentage of benign and malicious traffic in the collected dataset
(a) Original dataset
(b) SMOTE to individual protocol
(c) SMOTE to whole dataset
FIGURE 6. The sample count for benign and malicious traffic in the collected dataset (c) SMOTE to whole dataset (a) Original dataset
(b) SMOTE to individual protocol
(c) SMOTE to whole dataset
FIGURE 5
The percentage of benign and malicious traffic in the collected dataset (a) Original dataset
(b) SMOTE to individual protocol
FIGURE 5. The percentage of benign and malicious traffic in the collected dataset (a) Original dataset (c) SMOTE to whole dataset (c) SMOTE to whole dataset (a) Original dataset (b) SMOTE to individual protocol (a) Original dataset FIGURE 5. The percentage of benign and malicious traffic in the collected dataset (a) Original dataset
(b) SMOTE to individual protocol
FIGURE 6. The sample count for benign and malicious traffic in the collected dataset (a) Original dataset
(b) SMOTE to individual protocol
FIGURE 6. A. DATASET COLLECTION AND PREPROCESSING Otherwise, the point P is neglected; 5) Synthesis the new sample using Pnew = ηP + (1 −η)Oi
when dis is less than its corresponding di and 0 < η < 1. Otherwise, the point P is neglected; patterns that indicate DDoS attacks. In an SDN environment,
there is no need to reshape or transform the data into a one-
dimensional format to reduce the input dimensionality and
simplify the data preprocessing steps, making the detection
process more efficient. In this work, two approaches have been implemented when
applying SMOTE to the original dataset: (i) applying SMOTE
on a protocol basis and (ii) applying SMOTE on a label basis. It is worth noting from Fig. 5(b) and Fig. 5(c) that both
approaches above balanced the total number of classes in
the dataset. However, applying SMOTE on a protocol basis
achieves better balancing in the total number of class labels,
as shown in Fig. 6(b) because each protocol will have equal
value for benign and malicious traffic labels. On the other
hand, applying SMOTE to the class label will not provide a
balance at the protocol level as shown in Fig. 6(c) because
the variance in protocol samples still exists as SMOTE did
not pay any attention to the type of protocol when synthesis
the new samples of the training dataset. Two techniques were adopted in the developed model to
improve its accuracy: stacked convolutional neural network
(SCNN) [32] and early stopping (ES) [33]. SCNN involves
stacking multiple convolutional layers in a 1D-CNN model
that can improve model accuracy in several ways, including
increased model capacity, non-linear transformations, and
feature reusability. At the same time, ES involves monitoring
the model’s performance on a validation set and stopping the
training early when the performance starts to deteriorate. The
effects of ES on accuracy include preventing overfitting, find-
ing optimal generalization point, and efficient model training. g p
g
pi
g
Tuning the hyperparameters of a 1D-CNN is an important
step in optimizing its performance and achieving the best
possible results. Hyperparameters are configuration settings
that define the behavior of a model and affect its performance. In the context of a 1D-CNN, the key hyperparameters include
the number of employed filters and size of the employed ker-
nel, the learning rate, batch size, activation functions, dropout
rates, and the number of dense layers. A. DATASET COLLECTION AND PREPROCESSING Start by defining a
search space for each hyperparameter, representing the range
of values explored during the tuning process. The NSGA-II
approach was utilized in tuning seven hyperparameters of the
developed 1D-CNN model: The number of employed filters
in both convolutional and stacked convolutional layers, the
size of employed filters in both convolutional and stacked
convolutional layer, the learning rate, the dropout rate, and
the batch size. The NSGA-II is a multi-objective optimization
algorithm that uses a GA to optimize a set of objectives
while ensuring that the solutions found are non-dominated,
meaning no solution is better than another in all objectives si-
multaneously. NSGA-II is a favored choice for multi-criteria
optimization due to its capacity to identify a diverse range
of non-dominated solutions that capture trade-offs between
conflicting objectives, enabling decision-makers to navigate
complex decision landscapes and select from a variety of
Pareto-optimal solutions tailored to their preferences and re-
quirements. NSGA-II algorithm works by performing a series
of steps in each generation [34]: In summary, this paper adopted the utilization of the
SMOTE approach on a protocol basis when generating the
final dataset due to the following reasons: • Increasing the representation of the minority class by gen-
erating synthetic samples because SMOTE balances the
class distribution, allowing AI models to learn from both
classes more effectively; • Reducing bias towards the majority class and promoting
fairer representation for both classes as a class imbalance
can lead to biased models towards the majority class. This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/ A. DATASET COLLECTION AND PREPROCESSING The
employed SMOTE algorithm consists of the following main
steps [31]: 1) Split the dataset samples into minority class samples and
majority class samples; 2) Find the center points of the minority sample distribu-
tion {O1, O2, ..., On} for n regions that observed using k-
clustering method, where Oi = 1
m
Pm
j=1 Dij and Dij is the
jth point in ith region; The collected dataset should be preprocessed before the
training phase of the optimized model begins because the raw
data may have NaN/infinity values, errors, duplicate values,
and missing values that certainly affect the accuracy of the
AI model when classifying the ingress traffic. In this work,
the below actions were performed during the preprocessing 3) Calculate the closest distance di of all the majority sample
points to the points Oi where i = 1 to i = n; 4) Calculate the distance dis for each minority class sample
P to its corresponding center point; P to its corresponding center point; 8 VOLUME 11, 2023 This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/ nawi et al.: Optimized Artificial Intelligence Model for DDoS Detection in SDN Environment
This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2023.3319214 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2023.3319214 Dunainawi et al : Optimized Artificial Intelligence Model for DDoS Detection in SDN Environment
This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2023.3319214 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2023.3319214 4 Yousif Al-Dunainawi et al.: Optimized Artificial Intelligence Model for DDoS Detection in SDN Environment Yousif Al-Dunainawi et al.: Optimized Artificial Intelligence Model for DDoS Detection in SDN Environment 5) Synthesis the new sample using Pnew = ηP + (1 −η)Oi
when dis is less than its corresponding di and 0 < η < 1. B. OPTIMIZED AI MODEL TRAINING Citation information: DOI 10.1109/ACCESS.2023.3319214 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2023.3319214 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2023.3319214 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2023.3319214 Yousif Al-Dunainawi et al.: Optimized Artificial Intelligence Model for DDoS Detection in SDN Environment Yousif Al-Dunainawi et al.: Optimized Artificial Intelligence Model for DDoS Detection in SDN Environment 6) Crossover: Create new candidate solutions by recombin-
ing genetic information from the selected parents. 6) Crossover: Create new candidate solutions by recombin-
ing genetic information from the selected parents. 6) Crossover: Create new candidate solutions by recombin-
ing genetic information from the selected parents. where F(x) represents the fitness value of a candidate solu-
tion with hyperparameters x. By adjusting the weights w1 and
w2, the right trade-off between maximizing model accuracy
and minimizing training time can be achieved. It’s like finding
the sweet spot that meets our specific requirements. 7) Mutation: Introduce small changes to the genetic informa-
tion of the new candidate solutions to add diversity to the
population. 8) Replacement: Replace the current population with the
newly created candidate solutions. 8) Replacement: Replace the current population with the
newly created candidate solutions. The optimization process utilizing NSGA-II involves cre-
ating a population of potential solutions with different sets of
hyperparameters presented in Table 2. They were then evalu-
ating these solutions based on their fitness values. NSGA-II
employs evolutionary operations such as selection, crossover,
and mutation to evolve the population and uncover a diverse
range of non-dominated solutions. These solutions represent
different trade-offs between the objective functions. The tournament selection was employed in the utilized
NSGA-II. A random subset of individuals was selected in a
tournament, and the one with the best fitness value (based on
non-dominated sorting and crowding distance) was chosen
as a parent. Tournament selection helped maintain diversity
in the population and selected solutions with different trade-
offs. At the same time, simulated binary crossover (SBX)
was employed in the utilized NSGA-II. C. ML AND DL BENCHMARKING This benchmarking section provides a comprehensive eval-
uation and comparison of various ML techniques in the
context of DDoS detection in the SDN environment. By
conducting an extensive analysis, this paper aims to assess
the performance and effectiveness of these techniques in
relation to the existing body of work in the literature. The
benchmarking process involves examining diverse state-of-
the-art algorithms, including LR, RF, SVM, and KNN, by
comparing their performance metrics, such as accuracy, sen-
sitivity, specificity, precision, and F1-Score. Furthermore,
by contrasting our findings with prior research, this paper
aims to contribute to the ongoing efforts in advancing the
field and identifying the most promising directions for future
developments in DDoS detection and mitigation in the SDN
environment.i Multi-objective optimization problems, such as optimizing
the hyperparameters of the ID-CNN model using NSGA-
II, pose interesting challenges for researchers and academic
scholars. These problems involve juggling multiple objectives
that often compete with each other. This paper’s proposed
approach strives to maximize model accuracy while min-
imizing training time. However, finding the right balance
between these objectives is not a simple task. It requires
careful exploration of trade-offs and identifying the best set
of hyperparameters that can deliver the desired performance. Four classifiers have been adopted to study the effective-
ness of the optimized model developed in this work. These
classifiers have different strengths and are suitable for dif-
ferent data types and problem domains. The employed ML
classifiers are: • LR is a classification algorithm used to predict the prob-
ability of a binary or categorical outcome. It models the
relationship between a set of input features and the proba-
bility of a particular class using a logistic function. LR is
a linear model that applies a sigmoid function to the linear
combination of input features to estimate the probability of
the positive class [35]. Table 2 summarizes the decision variables of the intro-
duced optimization problem and their search space with the
obtained optimum value from the employed NSGA-II that
systematically explores the search space by evaluating dif-
ferent combinations of values for the decision variables and
adjusting them iteratively to find the best solution. The first objective function is the model accuracy (MA). It aims to push the 1D-CNN model to perform at its best by
training it with various hyperparameters and measuring the
resulting validation accuracy. B. OPTIMIZED AI MODEL TRAINING SBX emulated binary
crossover in the continuous domain by mixing the genetic
information of two parents based on a probability distribution. It created offspring that explore the search space around
the parents, encouraging diversity and exploration with a
crossover rate of 0.8. Finally, the polynomial mutation was
adopted to perturb a decision variable by adding a small
random value while ensuring the new value stayed within
the variable’s defined range. The polynomial mutation with a
mutation rate of 0.1 was important for maintaining diversity
and escaping local optima. This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/ F(x) = w1 × MA(x) + w2 × TT(x).
(1) B. OPTIMIZED AI MODEL TRAINING The dataset will be divided into 70% as a training dataset, and
30% will be divided evenly between validation and testing
datasets. TensorFlow is an open-source ML framework de-
veloped by Google. It provides a comprehensive ecosystem of
tools, libraries, and resources for building and deploying var-
ious ML models. The 1D-CNN is a neural network architec-
ture commonly used for analyzing sequential or time-series
data. It is particularly effective in capturing local patterns and
dependencies within the data. When using TensorFlow for
1D-CNN, developers can leverage the TensorFlow library’s
extensive capabilities to build, train, and deploy models effi-
ciently. TensorFlow provides a high-level API called Keras,
which simplifies the process of building neural networks,
including 1D-CNNs. Keras provides a user-friendly interface
for defining the model architecture, specifying the layers, and
configuring hyperparameters. Once the 1D-CNN model is
trained, TensorFlow offers tools for model evaluation, visu-
alization, and deployment. Developers can export the trained
model in a format compatible with different platforms and
frameworks, allowing seamless integration into various ap-
plications. 1) Initialization: Create an initial population of candidate
solutions. 2) Evaluation: Evaluate each candidate’s solution using the
objectives of the problem. 3) Non-dominated sorting: sort the population into fronts
based on the non-dominated relationships between solu-
tions. 4) Crowding distance assignment: Assign a crowding dis-
tance value to each solution in each front, representing the
density of solutions around that solution. 1D-CNNs can capture local patterns and dependencies
in the temporal dimension that are particularly relevant for
DDoS detection, as certain attack patterns may manifest over
time in the network traffic. By applying convolutional opera-
tions in the 1D-CNN, the model can extract relevant features
from the temporal data, allowing it to identify suspicious 5) Selection: Select parents for the next generation using a
combination of the non-dominated sorting and crowding
distance assignment. VOLUME 11, 2023 9 Yousif Al-Dunainawi et al.: Optimized Artificial Intelligence Model for DDoS Detection in SDN Environment
This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2023.3319214 Yousif Al Dunainawi et al : Optimized Artificial Intelligence Model for DDoS Detection in SDN Environment
This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/ C. ML AND DL BENCHMARKING It does not build
a model but stores all the training instances as reference
points. KNN determines the class of a new instance by
comparing it to the k closest training instances based on
Euclidean distance [38]. The system performance that utilized the optimized model
has been evaluated using the dataset created in Section VI-A. The confusion matrix (CM) is a table that is commonly
used to evaluate the performance of a classification model. It provides a summary of the predictions made by the model
on a set of data points, comparing them to the actual labels
of those data points. Receiver operating characteristic (ROC)
is a graphical representation and evaluation metric used in
ML to assess the performance of binary classification models. It plots the false positive rate (1-specificity) against the true
positive rate (sensitivity) at various classification thresholds. ROC provides a way to observe how well the classifier
separates the positive and negative classes across different
thresholds. Selecting the NSGA-II technique for hyperparameter tun-
ing of 1D-CNN models can offer several compelling advan-
tages that make it a suitable choice for the work presented in
this paper: (i) NSGA-II excels in multi-objective optimization
scenarios, enabling the exploration of trade-offs between con-
flicting objectives to find a set of Pareto-optimal solutions, (ii)
NSGA-II explores a wide range of solutions across the Pareto
front providing various options that cater to different priori-
ties, (iii) NSGA-II’s navigates complex and high-dimensional
hyperparameter spaces makes it well-suited for optimizing
hyperparameters in ML and DL scenarios, (iv) NSGA-II em-
ploys techniques like crowding distance to maintain diversity
among solutions, striking a balance between exploration of
the search space and exploitation of promising regions, and
(v) NSGA-II’s relatively straightforward implementation and
intuitive parameter settings make it accessible to researchers
from different backgrounds. VII. RESULTS AND DISCUSSIONS C. ML AND DL BENCHMARKING Hyperparameter
Search Space
Optimized Value
1
No. of Filters (Convolutional Layer)
16, 32, 64, 128, 256
64
2
Kernel Size (Convolutional Layer)
8, 16, 32, 64, 128
8
3
No. of Filters (Stack Convolutional Layer)
8, 16, 32, 64, 128
128
4
Kernel Size (Stack Convolutional Layer)
4, 8, 16, 32, 64
32
5
Dropout Rate
0.1, 0.2, 0.3, 0.4, 0.5
0.5
6
Learning Rate
0.0001, 0.001, 0.01, 0.1
0.001
7
Batch Size
100, 200, 300, 400, 500
200 TABLE 2. The tuned hyperparameters (decision variable x)
No. Hyperparameter
Search Space
Optimized Value
1
No. of Filters (Convolutional Layer)
16, 32, 64, 128, 256
64
2
Kernel Size (Convolutional Layer)
8, 16, 32, 64, 128
8
3
No. of Filters (Stack Convolutional Layer)
8, 16, 32, 64, 128
128
4
Kernel Size (Stack Convolutional Layer)
4, 8, 16, 32, 64
32
5
Dropout Rate
0.1, 0.2, 0.3, 0.4, 0.5
0.5
6
Learning Rate
0.0001, 0.001, 0.01, 0.1
0.001
7
Batch Size
100, 200, 300, 400, 500
200 TABLE 2. The tuned hyperparameters (decision variable x) TABLE 2. The tuned hyperparameters (decision variable x) • KNN is a simple and intuitive classification algorithm that
classifies new data points based on the majority vote of their
k-nearest neighbors in the feature space. It does not build
a model but stores all the training instances as reference
points. KNN determines the class of a new instance by
comparing it to the k closest training instances based on
Euclidean distance [38]. multi-objective function that aimed at simultaneously opti-
mizing multiple objectives or criteria, considering the trade-
offs between them: the model’s accuracy and the model’s
computation time (training time). The optimization process
explores the hyperparameters search space to find diverse
solutions that balance multiple decision variables. Once the
experimental run finishes, the optimized model will have a
set of hyperparameters with the optimized value that provides
the best accuracy with minimum training time. It is clear that
the NSGA-II maintains diversity in the population due to the
adoption of the crowding distance technique as a selection
criterion. As a result, the NSGA-II will select solutions that
are not too similar to each other in the objective space. • KNN is a simple and intuitive classification algorithm that
classifies new data points based on the majority vote of their
k-nearest neighbors in the feature space. C. ML AND DL BENCHMARKING On the other hand, the second
objective function is the training time (TT). It focuses on
reducing the time it takes to train the 1D-CNN model. This
time was measured by determining the time it takes to build
the model using the chosen hyperparameters. • RF is an ensemble learning method that combines multiple
decision trees to make predictions. Each decision tree is
built on a random subset of the training data and a random
subset of the input features. RF improves the accuracy
and reduces overfitting by aggregating the predictions of
multiple decision trees. It can handle both classification and
regression tasks [36]. • SVM is a powerful classification algorithm that finds the
best hyperplane to separate classes in a high-dimensional
feature space. It works by mapping the input data into a
higher-dimensional space and finding the hyperplane that
maximizes the margin between the closest data points of
different classes. It can handle linear and nonlinear classi-
fication tasks using different kernel functions [37]. A fitness function (F) is formulated as a weighted sum to
merge the above objective functions. It calculates the fitness
value of a potential solution, represented by a set of hyperpa-
rameters, using the below equation: F(x) = w1 × MA(x) + w2 × TT(x). (1) (1) 10 VOLUME 11, 2023 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2023.3319214 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2023.3319214 f Al-Dunainawi et al.: Optimized Artificial Intelligence Model for DDoS Detection in SDN Environment
This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2023.3319214 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2023.3319214 4 Yousif Al-Dunainawi et al.: Optimized Artificial Intelligence Model for DDoS Detection in SDN Environment Yousif Al-Dunainawi et al.: Optimized Artificial Intelligence Model for DDoS Detection in SDN Environment TABLE 2. The tuned hyperparameters (decision variable x)
No. VII. RESULTS AND DISCUSSIONS The experimental scenarios of the developed approach were
carried out using an HP All-in-One 24-df1xxx machine that
equipped with Intel® Core i5-1135G7 processor and 8GB
of random access memory (RAM). The HP machine runs
Windows 10 as an operating system while the mininet soft-
ware runs under a Linux virtual machine. The proposed ap-
proach tried to develop an optimized 1D-CNN model for SDN
traffic classification to prevent DDoS attacks. Fig. 9 depicts the visual representation of how the multi-
objective function value evolves throughout the optimization
process. Over 100 trials, the multi-objective function curve
exhibits a steady improvement with minimum fluctuation
around the initial value (reference point). It is clearly noted
that the optimum values for the optimized 1D-CNN model’s
hyperparameters were reached in trial 33, and the multi-
objective function value did not improve after that point. ii
p
The optimized 1D-CNN model developed in this work
utilized NSGA-II to tune its hyperparameters. The tuned
hyperparameters are the number of employed filters in both
convolution and stacked convolution layers, the size of the
employed kernel in both convolution and stacked convolution
layers, batch size, learning rate, and dropout. Fig. 7 shows the
importance of hyperparameters to the optimized model accu-
racy. In summary, hyperparameters play a crucial role in de-
termining the final accuracy of the trained model. Optimizing
these hyperparameters through careful tuning and experimen-
tation can lead to improved performance. Understanding the
relationship between specific hyperparameters and accuracy
and their impact on model complexity and generalization is
essential for achieving the best possible model performance. Fig. 10 illustrates the learning curve in terms of the model’s
accuracy for the optimized 1D-CNN model. The training
curve shows the optimized model’s performance on the train-
ing data as the amount of training data increases. While
the validation curve represents the optimized model’s perfor-
mance on a separate validation dataset, which contains sam-
ples not seen during training. The purpose of the validation set
is to evaluate the model’s generalization ability. It is worth Fig. 8 shows the relationship between the model’s hyper-
parameters and fitness function. The NSGA-II employed a VOLUME 11, 2023 11 VOLUME 11, 2023 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. VII. RESULTS AND DISCUSSIONS This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2023.3319214 Yousif Al-Dunainawi et al.: Optimized Artificial Intelligence Model for DDoS Detection in SDN Environment classify the incoming traffic correctly. While various decision
trees were adopted by the RF classifier in order to classify
the class labels. Correct decisions made by other trees can
balance out an incorrect decision resulting from one decision
tree. The final class label can be obtained from the maxi-
mum votes provided by each decision tree. Therefore, the
RF classifier provides better accuracy than the LR classifier. On the other hand, the SVM classifier provided moderate
accuracy because it utilized support vectors to properly find
a decision plane that had the greatest degree of separation
between the two classes. The dataset features are strongly
correlated; therefore, it may not always be able to get a perfect
decision boundary without overfitting. Finally, the KNN clas-
sifier provided better accuracy than the above-mentioned ML
classifiers as it employed a straightforward approach to classi-
fying the ingress traffic using just elementary math. It is worth
noting from Fig. 12 and Table 3 that the optimized 1D-CNN
model outperformed all the optimized LR, RF, SVM and
KNN models because the dataset has complex patterns and
automatic feature extraction is important. The improvements
in model accuracy were achieved by the NSGA-II approach
that was employed to tune the hyperparameters of the 1D-
CNN model and reach the optimum values for them. In addi-
tion, employing stacked convolution layers in the optimized
1D-CNN model can improve the model’s ability to extract
complex features from the input data, handle variations, and
prevent overfitting during model training. noting from Fig. 10 that the training and validation curves
increased together as more data was added. This indicates that
the model learns from the data and generalizes well to unseen
samples. The narrow generalization gap indicates that the
optimized model has a good ability to generalize its learning
from the training data to unseen samples. In other words,
the optimized model captures the underlying patterns of the
data rather than memorizing specific instances. In addition,
the narrow generalization gap indicates that the model has
reached its optimal capacity with the available training data. This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/ VII. RESULTS AND DISCUSSIONS Citation information: DOI 10.1109/ACCESS.2023.3319214 Yousif Al-Dunainawi et al.: Optimized Artificial Intelligence Model for DDoS Detection in SDN Environment
This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2023.3319214 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2023.3319214 Yousif Al-Dunainawi et al.: Optimized Artificial Intelligence Model for DDoS Detection in SDN Environment Yousif Al-Dunainawi et al.: Optimized Artificial Intelligence Model for DDoS Detection in SDN Environment FIGURE 7. The importance of hyperparameters to the optimized model accuracy
FIGURE 8. The relationship between the hyperparameters and the optimized model accuracy FIGURE 7. The importance of hyperparameters to the optimized model accuracy
FIGURE 8. The relationship between the hyperparameters and the optimized model accuracy
FIGURE 9. The objective function value vs optimization trials
12
VOLUME 11, 202
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/ FIGURE 7. The importance of hyperparameters to the optimized model accuracy FIGURE 7. The importance of hyperparameters to the optimized model accuracy The importance of hyperparameters to the optimized model accuracy p
yp
p
p
y
FIGURE 8. The relationship between the hyperparameters and the optimized model accuracy FIGURE 8. The relationship between the hyperparameters and the optimized model accuracy FIGURE 9. The objective function value vs optimization trials 12 VOLUME 11, 2023 This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/ This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2023.3319214 awi et al.: Optimized Artificial Intelligence Model for DDoS Detection in SDN Environment
This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2023.3319214 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2023.3319214 This article has been accepted for publication in IEEE Access. VII. RESULTS AND DISCUSSIONS In order to evaluate
the effectiveness of the proposed approach, a comparison is
made between the model developed in this work (optimized
1D-CNN) and the previous works that have already been
done in the field of DDOS attack detection using a simulated
dataset. As seen from Table 4, the optimized 1D-CNN model
gets the maximum accuracy, which is 99.99%. The developed
DL-based approach capitalizes on the inherent capacity of
CNN to discern intricate patterns and anomalies that might
elude conventional methods. This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/ VII. RESULTS AND DISCUSSIONS Citation information: DOI 10.1109/ACCESS.2023.3319214 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2023.3319214 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2023.3319214 Yousif Al-Dunainawi et al.: Optimized Artificial Intelligence Model for DDoS Detection in SDN Environment Yousif Al-Dunainawi et al.: Optimized Artificial Intelligence Model for DDoS Detection in SDN Environment (d) K-NN model
(e) Optimized model (a) LR model
(b) RF model
(c) SVM model
(d) K-NN model
(e) Optimized model
FIGURE 12. The CM of the testing dataset for the benchmark optimized models (a) LR model
(b) RF model
(c) SVM model
(d) K-NN model
(e) Optimized model
FIGURE 12. The CM of the testing dataset for the benchmark optimized models FIGURE 12. The CM of the testing dataset for the benchmark optimized models FIGURE 12. The CM of the testing dataset for the benchmark optimized models FIGURE 12. The CM of the testing dataset for the benchmark optimized models TABLE 3. The detailed performance evaluation metrics for different classifiers Performance Metric
LR Model
RF Model
SVM Model
KNN Model
Optimized Model
Accuracy (%)
90.87
92.22
94.69
97.37
99.99
Sensitivity
0.9468
0.9219
0.9544
0.9784
0.9999
Specificity
0.8673
0.9225
0.9385
0.9686
0.9999
Precision
0.8860
0.9336
0.9455
0.9714
0.9999
False Positive Rate
0.1327
0.0775
0.0615
0.0314
0.0001
False Negative Rate
0.0532
0.781
0.0456
0.0216
0.0001
F1-Score
0.9154
0.9277
0.9499
0.9749
0.9999 TABLE 4. The comparisons with state-of-the-art works
No. Author Name
Year
Testing
Accuracy
1
Sahoo et al. in [12]
2020
98.90%
2
Ahuja et al. in [14]
2021
98.80%
3
Musumeci et al. in [15]
2021
98.00%
4
Alanazi et al. in [19]
2022
99.77%
5
Liu et al. in [21]
2022
98.98%
6
Optimized Model
2023
99.99% TABLE 4. The comparisons with state-of-the-art works challenges of DDoS attacks and redefines the boundaries of
efficacy in SDN security. A distinct differentiation emerges
by juxtaposing the introduced method against the state-of-
the-art solutions prevalent in the literature. VII. RESULTS AND DISCUSSIONS Adding more data may not significantly improve its perfor-
mance, suggesting that the model has learned the essential
patterns and features necessary for accurate predictions. FIGURE 10. The accuracy of the optimized 1D-CNN model FIGURE 11. The loss of the optimized 1D-CNN model FIGURE 10. The accuracy of the optimized 1D-CNN model Fig. 11 illustrates the learning curve regarding the model’s
loss for the optimized 1D-CNN model. The narrow gener-
alization gap in the loss training curve indicates that the
optimized model is able to minimize its loss not only on the
training data but also on the validation data, indicating its
ability to capture the underlying patterns and make accurate
predictions. This implies that the model is not overfitting or
underfitting but is achieving a reliable level of performance on
unseen data. In summary, a loss training curve with a narrow
generalization gap is a positive sign, reflecting the model’s
ability to be deployed in real-world scenarios. In order to have a fair comparison, the benchmarking ML
models (e.g., LR, RF, SVM, and KNN optimized models)
have their hyperparameters tuned using NSGA-II. Fig. 12
visually represents the CM comparisons, highlighting the
contrasts between the approach developed in this paper and
other optimized ML models. Concurrently, Table 3 encap-
sulates a comprehensive summary of intricate comparisons
among diverse optimized classifiers. Despite the LR classifier
being widely employed in the literature of ML, however, it did
not provide adequate accuracy as shown in Table 3 because
the LR model failed to capture the exact relationship between
dataset features and dataset labels. Hence, LR could not FIGURE 11. The loss of the optimized 1D-CNN model Fig. 13 depicted the ROC for the developed model versus
other ML models. It is clear that the ROC of the optimized 1D-
CNN model is closer to the top-left corner, indicating a better
classifier with higher specificity and lower sensitivity when
compared to optimized LR, RF, SVM, and KNN models. In the landscape of DDoS detection and mitigation within
SDN environments, the proposed approach in this paper
stands as an innovative stride, leveraging the power of DL
techniques. The introduced approach addresses the evolving 13 VOLUME 11, 2023 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. VIII. CONCLUSIONS The developed 1D-CNN model was trained
using NSGA-II to tune its hyperparameters for achieving the
best accuracy with minimum training time. [9] N. Ravi and S. M. Shalinie, ‘‘Learning-Driven Detection and Mitigation
of DDoS Attack in IoT via SDN-Cloud Architecture,’’ IEEE Internet of
Things Journal, vol. 7, no. 4, pp. 3559–3570, 2020. [10] J. Jiang, C. Lin, G. Han, A. M. Abu-Mahfouz, S. B. H. Shah, and
M. Martínez-García, ‘‘How ai-enabled sdn technologies improve the se-
curity and functionality of industrial iot network: Architectures, enabling
technologies, and opportunities,’’ Digital Communications and Networks,
2022. Various ML algorithms were adopted to compare the de-
veloped approach among them, and other works existed in
the literature to study the effectiveness of the optimized 1D-
CNN model. The obtained results showed that the optimized
1D-CNN model outperformed the other ML models by 9.5%,
8%, 5.4%, and 2.6% when compared to LR, RF, SVM, and
KNN models, respectively. [11] R. Swami, M. Dave, and V. Ranga, ‘‘Voting-Based Intrusion Detection
Framework for Securing Software-Defined Networks,’’ Concurrency and
Computation: Practice and Experience, vol. 32, no. 24, Jul. 2020. [12] K. S. Sahoo, B. K. Tripathy, K. Naik, S. Ramasubbareddy, B. Balusamy,
M. Khari, and D. Burgos, ‘‘An Evolutionary SVM Model for DDOS
Attack Detection in Software Defined Networks,’’ IEEE Access, vol. 8, pp. 132 502–132 513, 2020. The developed methods of detecting and mitigating DDoS
attacks in an SDN environment allow researchers and net-
work security experts to test new detection and mitigation
strategies without risking actual network infrastructure. This
is particularly beneficial for fine-tuning algorithms, exploring
different attack scenarios, and assessing the effectiveness
of novel techniques. However, the simulated environments
might not capture the complexities of real-world networks. The performance of detection and mitigation strategies in
a controlled simulation might not directly translate to real-
world effectiveness due to variations in network dynamics,
traffic patterns, and the adaptability of attackers. [13] W. Zhijun, X. Qing, W. Jingjie, Y. Meng, and L. Liang, ‘‘Low-Rate DDoS
Attack Detection Based on Factorization Machine in Software Defined
Network,’’ IEEE Access, vol. 8, pp. 17 404–17 418, 2020. [14] N. Ahuja, G. Singal, D. Mukhopadhyay, and N. Kumar, ‘‘Automated
DDOS Attack Detection in Software Defined Networking,’’ Journal of
Network and Computer Applications, vol. 187, p. 103108, 2021. [15] F. Musumeci, A. C. Fidanci, F. Paolucci, F. Cugini, and M. VIII. CONCLUSIONS This paper presented a comprehensive literature review cov-
ering the state-of-the-art solutions to address DDoS attacks
in the SDN environment. Simple network administration fa-
cilities could be achieved when utilizing SDN technology
with more flexible tools as compared to legacy network tech-
nology. Various security concerns still exist when deploying
SDN controller, and these issues need to be solved to im-
prove network performance and increase security. This paper
aimed to develop an optimized 1D-CNN model to classify the
incoming traffic in an SDN environment and prevent DDoS
attacks. Ryu controller was employed to monitor the network
traffic generated in the mininet emulator. Extensive simula-
tion scenarios were run through the utilization of OpenFlow
switches and hosts. Once the simulation scenarios finished,
the raw dataset was saved in a CSV file for further processing
before training the developed model. FIGURE 13. The ROC for the optimized model versus other ML models The developed approach enhanced the dataset through
SMOTE in order to eliminate the effect of unbalanced classes
in the collected dataset. The SMOTE was carried out at the
protocol level rather than the label level to ensure that each
type of traffic from different protocols had equal labels (be-
nign and malicious). Then, the dataset was divided into 70%
training, and 30% was divided evenly between validation and FIGURE 13. The ROC for the optimized model versus other ML models 14 VOLUME 11, 2023 This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/ This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2023.3319214 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2023.3319214 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2023.3319214 Yousif Al-Dunainawi et al.: Optimized Artificial Intelligence Model for DDoS Detection in SDN Environment Yousif Al-Dunainawi et al.: Optimized Artificial Intelligence Model for DDoS Detection in SDN Environme testing datasets. REFERENCES [1] W. Rafique, A. S. Hafid, and S. Cherkaoui, ‘‘Complementing IoT Services
Using Software-Defined Information Centric Networks: A Comprehensive
Survey,’’ IEEE Internet of Things Journal, vol. 9, no. 23, pp. 23 545–
23 569, 2022. [23] M. Dabbagh, B. Hamdaoui, M. Guizani, and A. Rayes, ‘‘Software-Defined
Networking Security: Pros and Cons,’’ IEEE Communications Magazine,
vol. 53, no. 6, pp. 73–79, 2015. [2] Y. Maleh, M. Shojafar, A. Darwish, and A. Haqiq, Eds., Cybersecurity and
Privacy in Cyber-Physical Systems. CRC Press, may 2019.i [24] I. Ahmad, S. Namal, M. Ylianttila, and A. Gurtov, ‘‘Security in Software
Defined Networks: A Survey,’’ IEEE Communications Surveys & Tutori-
als, vol. 17, no. 4, pp. 2317–2346, 2015. [3] Y. Al Mtawa, A. Haque, and H. Lutfiyya, ‘‘Migrating From Legacy to
Software Defined Networks: A Network Reliability Perspective,’’ IEEE
Transactions on Reliability, vol. 70, no. 4, pp. 1525–1541, 2021. [25] Mininet Project, ‘‘Mininet: An Instant Virtual Network on Your Laptop
(or Other PC),’’ 2012, [Accessed on June 2023]. [Online]. Available:
http://mininet.org/ [4] T. Alharbi and M. Portmann, ‘‘The (In)Security of Virtualization in Soft-
ware Defined Networks,’’ IEEE Access, vol. 7, pp. 66 584–66 594, 2019. [26] F. Keti and S. Askar, ‘‘Emulation of Software Defined Networks Using
Mininet in Different Simulation Environments,’’ in 2015 6th International
Conference on Intelligent Systems, Modelling and Simulation, 2015, pp. 205–210. [5] R. Jmal, W. Ghabri, R. Guesmi, B. M. Alshammari, A. S. Alshammari,
and H. Alsaif, ‘‘Distributed Blockchain-SDN Secure IoT System Based on
ANN to Mitigate DDoS Attacks,’’ Applied Sciences, vol. 13, no. 8, 2023.i [6] L. F. Eliyan and R. Di Pietro, ‘‘DoS and DDoS Attacks in Software Defined
Networks: A Survey of Existing Solutions and Research Challenges,’’
Future Generation Computer Systems, vol. 122, pp. 149–171, 2021. [27] J. Yan and D. Jin, ‘‘VT-Mininet,’’ in Proceedings of the 1st ACM SIG-
COMM Symposium on Software Defined Networking Research. ACM,
Jun. 2015. [7] O. Campesato, Artificial Intelligence, Machine Learning, and Deep Learn-
ing. Mercury Learning and Information, 2020. [28] L. Zhu, M. M. Karim, K. Sharif, F. Li, X. Du, and M. Guizani,
‘‘SDN Controllers: Benchmarking & Performance Evaluation,’’ CoRR, vol. 1902.04491, 2019. [8] R. Olusegun, T. Oladunni, H. Audu, Y. Houkpati, and S. Bengesi, ‘‘Text
Mining and Emotion Classification on Monkeypox Twitter Dataset: A
Deep Learning-Natural Language Processing (NLP) Approach,’’ IEEE
Access, vol. 11, pp. 49 882–49 894, 2023. VIII. CONCLUSIONS Torna-
tore, ‘‘Machine-Learning-Enabled DDoS Attacks Detection in P4 Pro-
grammable Networks,’’ Journal of Network and Systems Management,
vol. 30, no. 1, Nov. 2021. [16] R. Swami, M. Dave, and V. Ranga, ‘‘Detection and Analysis of TCP-
SYN DDoS Attack in Software-Defined Networking,’’ Wireless Personal
Communications, vol. 118, no. 4, pp. 2295–2317, Feb. 2021. [17] A. O. Sangodoyin, M. O. Akinsolu, P. Pillai, and V. Grout, ‘‘Detection and
Classification of DDoS Flooding Attacks on Software-Defined Networks:
A Case Study for the Application of Machine Learning,’’ IEEE Access,
vol. 9, pp. 122 495–122 508, 2021. The employed dataset in this work provides controlled
environments where the characteristics of attacks can be pre-
cisely defined. They are useful for benchmarking and com-
paring the performance of various detection and mitigation
methods. They can also help in understanding how different
attacks affect network behavior. However, the utilized simu-
lated datasets aim to replicate real-world attack patterns; they
might not fully capture the diversity and sophistication of
actual DDoS attacks. Real-world DDoS attacks can evolve
rapidly, employing new techniques that the simulated dataset
might not cover. Additionally, the scale of real attacks can
vary widely, from small-scale attacks to large and complex
botnet-driven assaults, which might not be adequately repre-
sented in the simulated data. [18] A. Maheshwari, B. Mehraj, M. S. Khan, and M. S. Idrisi, ‘‘An Optimized
Weighted Voting Based Ensemble Model for DDoS Attack Detection and
Mitigation in SDN Environment,’’ Microprocessors and Microsystems,
vol. 89, p. 104412, 2022. [19] F. Alanazi, K. Jambi, F. Eassa, M. Khemakhem, A. Basuhail, and K. Al-
subhi, ‘‘Ensemble Deep Learning Models for Mitigating DDoS Attack in
Software-Defined Network,’’ Intelligent Automation & Soft Computing,
vol. 33, no. 2, pp. 923–938, 2022. [20] U. Mbasuva and G.-A. L. Zodi, ‘‘Designing Ensemble Deep Learning
Intrusion Detection System for DDoS attacks in Software Defined Net-
works,’’ in 2022 16th International Conference on Ubiquitous Information
Management and Communication (IMCOM), 2022, pp. 1–8. [21] Y. Liu, T. Zhi, M. Shen, L. Wang, Y. Li, and M. Wan, ‘‘Software-Defined
DDoS Detection with Information Entropy Analysis and Optimized Deep
Learning,’’ Future Generation Computer Systems, vol. 129, pp. 99–114,
2022. [22] P. M. Ombase, N. P. Kulkarni, S. T. Bagade, and A. V. Mhaisgawali,
‘‘Survey on DoS Attack Challenges in Software Defined Networking,’’
International Journal of Computer Applications, vol. 173, no. 2, pp. 19–25,
2017. REFERENCES [29] Ryu SDN Framework Community, ‘‘Ryu Controller,’’ 2012, [Accessed on
June 2023]. [Online]. Available: https://ryu-sdn.org/ 15 VOLUME 11, 2023 VOLUME 11, 2023 This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/ This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2023.3319214 Yousif Al-Dunainawi et al.: Optimized Artificial Intelligence Model for DDoS Detection in SDN Environment
This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2023.3319214 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2023.3319214 Yousif Al-Dunainawi et al.: Optimized Artificial Intelligence Model for DDoS Detection in SDN Environment [30] S. Bhardwaj and S. N. Panda, ‘‘Performance Evaluation Using RYU SDN
Controller in Software-Defined Networking Environment,’’ Wireless Per-
sonal Communications, vol. 122, no. 1, pp. 701–723, Aug. 2021. HAMED S. AL-RAWESHIDY (M’92 - SM’97)
received the B.Eng. and M.Sc. degrees from the
University of Technology, Baghdad, Iraq, in 1977
and 1980, respectively, the Post Graduate Diploma
degree from Glasgow University, Glasgow, U.K.,
in 1987, and the Ph.D. degree from Strathclyde
University, Glasgow, in 1991. He was with the
Space and Astronomy Research Centre, Baghdad,
Iraq, PerkinElmer, Waltham, MA, USA, British
Telecom, London, U.K., Oxford University, Ox-
ford, U.K., Manchester Metropolitan University, Manchester, U.K., and Kent
University, Canterbury, U.K. He is currently the Director of the Wireless
Network Communications Centre, Brunel University London. pp
g
[31] Y. Bao and S. Yang, ‘‘Two Novel SMOTE Methods for Solving Imbalanced
Classification Problems,’’ IEEE Access, vol. 11, pp. 5816–5823, 2023. [32] S. Madichetty and S. M., ‘‘A Stacked Convolutional Neural Network for
Detecting the Resource Tweets During a Disaster,’’ Multimedia Tools and
Applications, vol. 80, no. 3, pp. 3927–3949, Sep. 2020. [33] M. Vilares Ferro, Y. Doval Mosquera, F. J. Ribadas Pena, and V. M. Darriba Bilbao, ‘‘Early stopping by correlating online indicators in neural
networks,’’ Neural Networks, vol. 159, pp. 109–124, 2023. [34] S. Verma, M. Pant, and V. Snasel, ‘‘A Comprehensive Review on NSGA-II
for Multi-Objective Combinatorial Optimization Problems,’’ IEEE Access,
vol. 9, pp. 57 757–57 791, 2021. [35] K. VOLUME 11, 2023 Yeturu, ‘‘Chapter 3 - Machine Learning Algorithms, Applications, and
Practices in Data Science,’’ in Principles and Methods for Data Science,
ser. Handbook of Statistics, A. S. Srinivasa Rao and C. Rao, Eds. Elsevier,
2020, vol. 43, pp. 81–206. [36] M. S. ElSayed, N.-A. Le-Khac, M. A. Albahar, and A. Jurcut, ‘‘A Novel
Hybrid Model for Intrusion Detection Systems in SDNs Based on CNN
and a New Regularization Technique,’’ Journal of Network and Computer
Applications, vol. 191, p. 103160, 2021. [37] H.-T. Thai, ‘‘Machine Learning for Structural Engineering: A State-of-
The-Art Review,’’ Structures, vol. 38, pp. 448–491, 2022. [38] Ö. Tonkal, H. Polat, E. Başaran, Z. Cömert, R. Kocaoğlu, ‘‘Machine Learn-
ing Approach Equipped with Neighbourhood Component Analysis for
DDoS Attack Detection in Software-Defined Networking,’’ Electronics,
vol. 10, no. 11, 2021. YOUSIF AL-DUNAINAWI received B.Sc. degree
and M. Sc. degree in mechatronics engineering
from Baghdad University, Baghdad - Iraq, in 2004
and 2007, respectively. He received the Ph.D. de-
gree in intelligent and control systems from Brunel
University London, U.K., in 2017. His current re-
search interests include fuzzy logic, neural net-
works, and evolutionary based algorithms. YOUSIF AL-DUNAINAWI received B.Sc. degree
and M. Sc. degree in mechatronics engineering
from Baghdad University, Baghdad - Iraq, in 2004
and 2007, respectively. He received the Ph.D. de-
gree in intelligent and control systems from Brunel
University London, U.K., in 2017. His current re-
search interests include fuzzy logic, neural net-
works, and evolutionary based algorithms. BILAL R. AL-KASEEM (M’15) received the
B.Sc. and M.Sc. degrees in computer engineer-
ing from Nahrain University, Baghdad, Iraq, in
2006 and 2010, respectively. He received the Ph.D. degree in telecommunication and network engi-
neering from Brunel University London, U.K., in
2017. His current research interests include wire-
less sensor networks, design and implementation
of smart sensors, software-defined networking,
network functioning virtualization, M2M commu-
nication, cloud computing, network security, and intelligent IoT applications. VOLUME 11, 2023 16 This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0
|
https://openalex.org/W1974320648
|
https://journals.plos.org/plosone/article/file?id=10.1371/journal.pone.0034481&type=printable
|
English
| null |
Fused Traditional and Geometric Morphometrics Demonstrate Pinniped Whisker Diversity
|
PloS one
| 2,012
|
cc-by
| 9,894
|
Carly C. Ginter1*, Thomas J. DeWitt1, Frank E. Fish2, Christopher D. Marshall1,3 1 Department of Wildlife and Fisheries Sciences, Texas A&M University, College Station, Texas, United States of America, 2 Department of Biology, West Chester University,
West Chester, Pennsylvania, United States of America, 3 Department of Marine Biology, Texas A&M University, Galveston, Texas, United States of America Department of Wildlife and Fisheries Sciences, Texas A&M University, College Station, Texas, United States of America, 2 Department of
West Chester, Pennsylvania, United States of America, 3 Department of Marine Biology, Texas A&M University, Galveston, Texas, United Abstract nding: Funding was provided by the Department of Marine Biology, Texas A&M University (http://www.tamug.edu/marb/) to CCG. The
dy design, data collection and analysis, decision to publish, or preparation of the manuscript. Funding: Funding was provided by the Department of Marine Biology, Texas A&M University (http://www.tamug.edu/marb/) to CCG
study design, data collection and analysis, decision to publish, or preparation of the manuscript. rovided by the Department of Marine Biology, Texas A&M University (http://www.tamug.edu/marb/) to CCG. The funders had no role in
tion and analysis, decision to publish, or preparation of the manuscript. Competing Interests: The authors have declared that no competing interests exist. * E-mail: ccginter@tamu.edu PLoS ONE | www.plosone.org Introduction exhibits the unusual beaded vibrissal morphology possessed by
most phocid seals. Since vision is a limited sensory modality in
aquatic habitats, it follows that marine mammals may have
experienced strong selection for compensatory sensory adaptations
that facilitate functions such as prey detection, particularly when
foraging at night, in turbid water or when diving deeply. Whereas
odontocetes (toothed whales) evolved echolocation, pinnipeds have
highly derived vibrissal sensory systems [14,18–21]. Reports of
healthy but blind seals foraging successfully in the wild suggest the
importance of this sensory system for the aquatic environment
[15,22]. Experimental evidence has shown that phocid seals rely
heavily upon their vibrissae to follow hydrodynamic trails
[19,20,23–25], as well as to orient themselves when vision is
restricted [26]. Vibrissae are not shed with the rest of the pelage in
an annual molt, but rather individually and irregularly throughout
the year [9,27]. The shortest vibrissae on the muzzles of phocid
species have been observed not to have beads but the development
of beading is not currently understood [17]. Stable isotope studies
of vibrissal growth rates found that otariids retain their vibrissae
for over two years, while phocid vibrissae appear to grow more
quickly and be replaced more often [28,29]. The periodic and
intermittent shedding and re-growth of vibrissae likely serves to Many organisms possess highly sensitive mechanosensory struc-
tures to monitor and detect physical cues in their environment. Mammalian vibrissae (whiskers) are finely tuned sensory structures. The vibrissae include a follicle-sinus complex (F-SC) with numerous
and various types of mechanoreceptors within its complex micro-
structure and a vibrissal hair shaft that transmits vibrotactile
environmental stimuli to these mechanoreceptors deep in the F-SC
[1–4]. Although most mammals possess vibrissae, the majority of our
knowledge regarding their function is limited to laboratory animals
[5–8]. The number, geometric arrangement, size, morphology, shape
and stiffness of vibrissae vary widely among mammals [9]. Pinnipeds
(seals, sea lions and walruses) possess the largest vibrissae among
mammals (e.g., Antarctic fur seals (Arctocephalus gazella) have vibrissae
up to 480 mm long; [10]) and exhibit a diversity of shapes in these
structures [9,11,12]. In particular, the mystacial vibrissae of phocid
seals, with the exception of bearded (Erignathus barbatus) and monk
(Monachus spp.) seals, show a repeating sequence of crests and troughs
along their length, giving them a beaded appearance [11–17]. Many mammals, terrestrial and aquatic, use vibrissae for active
touch and other discrimination tasks. Abstract Vibrissae (whiskers) are important components of the mammalian tactile sensory system, and primarily function as detectors
of vibrotactile information from the environment. Pinnipeds possess the largest vibrissae among mammals and their
vibrissal hair shafts demonstrate a diversity of shapes. The vibrissae of most phocid seals exhibit a beaded morphology with
repeating sequences of crests and troughs along their length. However, there are few detailed analyses of pinniped vibrissal
morphology, and these are limited to a few species. Therefore, we comparatively characterized differences in vibrissal hair
shaft morphologies among phocid species with a beaded profile, phocid species with a smooth profile, and otariids with a
smooth profile using traditional and geometric morphometric methods. Traditional morphometric measurements (peak-to-
peak distance, crest width, trough width and total length) were collected using digital photographs. Elliptic Fourier analysis
(geometric morphometrics) was used to quantify the outlines of whole vibrissae. The traditional and geometric
morphometric datasets were subsequently combined by mathematically scaling each to true rank, followed by a single
eigendecomposition. Quadratic discriminant function analysis demonstrated that 79.3, 97.8 and 100% of individuals could
be correctly classified to their species based on vibrissal shape variables in the traditional, geometric and combined
morphometric analyses, respectively. Phocids with beaded vibrissae, phocids with smooth vibrissae, and otariids each
occupied distinct morphospace in the geometric morphometric and combined data analyses. Otariids split into two groups
in the geometric morphometric analysis and gray seals appeared intermediate between beaded- and smooth-whiskered
species in the traditional and combined analyses. Vibrissal hair shafts modulate the transduction of environmental stimuli to
the mechanoreceptors in the follicle-sinus complex (F-SC), which results in vibrotactile reception, but it is currently unclear
how the diversity of shapes affects environmental signal modulation. Citation: Ginter CC, DeWitt TJ, Fish FE, Marshall CD (2012) Fused Traditional and Geometric Morphometrics Demonstrate Pinniped Whisker Diversity. PLoS
ONE 7(4): e34481. doi:10.1371/journal.pone.0034481
Editor: Thomas Wennekers, The University of Plymouth, United Kingdom Citation: Ginter CC, DeWitt TJ, Fish FE, Marshall CD (2012) Fused Traditional and Geometric Morphometrics Demonstrate Pinniped Whisker Diversity. PLoS
ONE 7(4): e34481. doi:10.1371/journal.pone.0034481 Editor: Thomas Wennekers, The University of Plymouth, United Kingdom Received September 14, 2011; Accepted March 5, 2012; Published April 3, 2012 Copyright: 2012 Ginter et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits
unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Introduction However, no other mammal April 2012 | Volume 7 | Issue 4 | e34481 1 PLoS ONE | www.plosone.org Pinniped Whisker Shape and Morphology Table 1. Pinnipeds analyzed in this study. Vibrissal Profile
Family
Species
Number of Individuals
Beaded
Phocidae
Gray seal (Halichoerus grypus)
8
Phocidae
Harbor seal (Phoca vitulina)
15
Phocidae
Harp seal (Pagophilus groenlandicus)
16
Phocidae
Ringed seal (Pusa hispida)
16
Phocidae
Spotted seal (Phoca largha)
9
Smooth
Phocidae
Bearded seal (Erignathus barbatus)
10
Otariidae
California sea lion (Zalophus californianus)
12
Otariidae
Guadalupe fur seal (Arctocephalus townsendi)
1
Otariidae
Northern fur seal (Callorhinus ursinus)
2
Otariidae
South American fur seal (Arctocephalus australis)
2
Otariidae
Steller sea lion (Eumetopias jubatus)
1
Individuals are classified by family, species and vibrissal profile. doi:10.1371/journal.pone.0034481.t001 Individuals are classified by family, species and vibrissal profile. doi:10.1371/journal.pone.0034481.t001 photograph, all vibrissae were dyed black (using Revlon ColorSilk
hair dye #10; Revlon Cons. Prod. Corp., N.Y., NY 10017). Vibrissae were coated with the dye mixture until the dye had
penetrated the hair shaft and would not rinse off with water. Vibrissae were placed on a white background with the laterally
flattened side down and held flat by a large glass slide to eliminate
shape distortion. The camera was mounted on a photographic
copy stand normal to the vibrissa and a remote shutter release was
used to trigger imaging. avoid a decrease in sensory capability if all vibrissae were shed
simultaneously or due to vibrissae loss or physical damage
[9,28,29]. The divergence of the vibrissae from the shedding
cycle of the pelage underscores the functional significance of this
sensory system. The distinctive shape of phocid seal vibrissae, with a sinusoidal
beaded profile, has been shown to decrease vibrations during
ambient flow while the seal is swimming, compared to smooth
vibrissae [30]. Previous morphological analyses of phocid vibrissae
demonstrated species-specific differences [17]. However, there are
few quantitative data regarding the morphology of phocid
vibrissae and, to our knowledge, no characterization of their
geometry has been conducted. Therefore, we conducted a
comparative study on the shape and morphology of pinniped
vibrissae. We tested the hypothesis that the beaded morphology of
phocid vibrissae is conserved with variants in peak-to-peak
distance and crest and trough width representing species-specific
differences. We also hypothesized that phocids with beaded
vibrissae, phocids with smooth vibrissae and otariids would each
occupy distinct morphospace from each other, which may
facilitate functional differences with ecological consequences. Introduction Traditional morphometric measures were collected using the
scaled digital photographs (27.0 pixels/mm) of whole vibrissae
following Ginter et al. [17]. Traditional morphometric measure-
ments collected were: peak-to-peak distance (linear distance
between successive crests along the top and bottom margins of
each vibrissa), crest width (width of beads), trough width (width of
constrictions), and curvilinear length of the entire vibrissa (Fig. 1). All measurements were made using ImageJ (version 1.41, National
Institutes of Health, Bethesda, MD, USA; http://rsbweb.nih.gov/
ij/). Each of the five measurement types was made multiple times
on each vibrissa. The number of measurements made on a vibrissa
depended on the number of crests along the length. The mean
number of crests along all beaded vibrissae was used as the
number of measurements taken, distributed evenly along smooth
vibrissae. Values were averaged for each individual vibrissa. The
ratio of mean crest width to mean trough width was computed for
each species. The geometric mean of all traditional morphometric
measures for each individual was used as a standard proxy for
overall size in statistical analysis of the traditional morphometrics
[31]. Level A data (age class and gender) were obtained from the
recovering stranding network for 45 individuals from five species. PLoS ONE | www.plosone.org Ethics Statement All samples were collected under a National Marine Fisheries
Service (NMFS) Southeast Regional Office salvage permit letter to
CDM and NMFS permits #358-1585 and 358-1787 issued to the
Alaska Department of Fish and Game. Mystacial vibrissal hair shafts external to the follicle (hereafter
simply termed vibrissae) from 11 pinniped species and 92
individuals were analyzed to quantify shape and morphological
differences among species with beaded and smooth vibrissal
profiles (Table 1). Samples were collected from dead, stranded
animals in New Jersey, New England and California, from legal
indigenous hunts in Alaska and opportunistically when shed by
captive animals. To standardize our comparisons, the longest
vibrissae which did not show any wear or breakage from each
individual were used. These vibrissae always occurred in the most
lateral portions of the lower rows of the mystacial vibrissal field. Scaled digital photographs (27.0 pixels/mm) of whole vibrissae
were taken with a Nikon D200 SLR camera. To maximize the
contrast between the background and the vibrissa in the g
g
p
For geometric morphometric characterization, we used outline-
based, rather than landmark-based analysis due to the lack of
homologous landmarks on vibrissae. The same images used for the
traditional morphometric measurements were thresholded using
ImageJ. TpsDig2 software [32] was used to fit an outline to each
thresholded vibrissal image (Fig. 2), calculate the area within the
outline, and save the series of X-Y coordinates for each outline. The coordinates were renumbered using a Microsoft Excel 2007
algorithm to standardize the location of ‘‘point 1’’ at the middle of
the base of each vibrissa. The renumbered coordinates were
imported to EFAWin software [33] for elliptic Fourier analysis
(EFA). Harmonics were added sequentially until digital recon-
structions of the vibrissae outlines were judged by eye to PLoS ONE | www.plosone.org April 2012 | Volume 7 | Issue 4 | e34481 April 2012 | Volume 7 | Issue 4 | e34481 2 Pinniped Whisker Shape and Morphology Figure 1. Locations of traditional morphometric measurements on beaded and smooth vibrissae. A ringed seal (top) and California sea
lion (bottom) vibrissa are shown with locations of three traditional morphometric measurements (peak-to-peak distance, crest width and trough
width) indicated by the black arrows. The measurements are labeled by the text between the two vibrissae. Peak-to-peak distance is not in reference
to any physical structures on smooth vibrissae; it is simply the distance between successive measurements of vibrissal width. Ethics Statement doi:10.1371/journal.pone.0034481.g001 Figure 1. Locations of traditional morphometric measurements on beaded and smooth vibrissae. A ringed seal (top) and California sea
lion (bottom) vibrissa are shown with locations of three traditional morphometric measurements (peak-to-peak distance, crest width and trough
width) indicated by the black arrows. The measurements are labeled by the text between the two vibrissae. Peak-to-peak distance is not in reference
to any physical structures on smooth vibrissae; it is simply the distance between successive measurements of vibrissal width. doi:10.1371/journal.pone.0034481.g001 completely represent the outlines of all vibrissae in this study. Good fit was apparent with 15 harmonics. The shape variables
that resulted were set to be invariant to size (area of the first
harmonic ellipse), orientation, and rotation. Size of the first ellipse
(area or length of the major axis) is a reasonable and accepted
standard for calculating subsequent harmonic coefficients [33,34]. However, area within the outline, measured in pixels using
tpsDig2, was used as a more biologically relevant measure of size
in our statistical analyses. is the number of components that collectively summarizes variance
above ‘‘noise’’ such as digitization error [36]. Such thresholds are
usually set to encompass 95 or 99% of total variance. True rank
scaling is accomplished by multiplying each data matrix by the
square root of the ratio of the data’s true rank divided by the
original variance of the data matrix. Data scaled in this manner
can be entered together into a single principal components
decomposition (T. J. DeWitt, In prep.). Pinniped Whisker Shape and Morphology for the traditional morphometric dataset and the correlation
matrix from the PCA for the geometric morphometric dataset. True rank was defined as the number of principal components
needed to summarize 99% of the variance in each dataset. Once
scaled, the datasets were subsequently combined and covariance
principal components were generated for the new data matrix. The principal components needed to summarize 99% of variance
in the combined dataset were used as dependent variables in a
MANOVA with species, geometric mean, log-transformed square
root of vibrissae area, interaction effects between species and each
size measure, an interaction effect between the two size measures
and a three-way interaction effect between species and both size
measures. To remove the effect of size in the dataset before DFA, a
second MANOVA was run with only geometric mean and log-
transformed square root of vibrissae area as the independent
variables and principal components as the dependent variables. Residuals from that analysis were used in linear and quadratic
DFA with the species effect. Quadratic, as opposed to linear, DFA allows for each taxon to be
predicted based on its own variance-covariance matrix in the
canonical space. Different covariance structures can result from
factors such as small sample sizes (e.g., ,30 individuals per taxon). We ran both types of DFA in each of the three analyses to provide
a complete comparative basis for each method. Results of all
statistical analyses were considered to be significant at p,0.05. Quadratic, as opposed to linear, DFA allows for each taxon to be
predicted based on its own variance-covariance matrix in the
canonical space. Different covariance structures can result from
factors such as small sample sizes (e.g., ,30 individuals per taxon). We ran both types of DFA in each of the three analyses to provide
a complete comparative basis for each method. Results of all
statistical analyses were considered to be significant at p,0.05. Traditional Morphometrics. All traditional morphometric
data
were
log
(x+1)
transformed
for
normality. This
transformation was used to avoid negative numbers that could
not be used in calculation of geometric mean size. Species-specific
differences
in
morphological
variation
were
tested
using
multivariate
analysis of variance
(MANOVA) with species,
geometric mean and an interaction effect between species and
geometric mean as the independent variables and the log-
transformed linear distances as the dependent variables. Pinniped Whisker Shape and Morphology To
remove the effect of size in the dataset before evaluating the
species effect using DFA, a second MANOVA was run with only
geometric mean as the independent variable and the log-
transformed
linear
distances
as
the
dependent
variables. Residuals from that analysis were subsequently used in both
linear and quadratic DFA with the species effect. One-Way
ANOVA with Tukey HSD post-hoc tests was used to evaluate
differences among species in morphometric measurements on the
vibrissae. Such use of ANOVA following MANOVA is referred to
as ‘‘protected’’ ANOVA, and is a common practice in ecology and
evolutionary biology [37]. However, significance levels for these
ANOVA tests should be taken less as precise statements and more
as useful heuristic devices, to help distinguish which original
variables contribute most strongly to differences between taxa in
the
multivariate
space. These
‘‘protected’’
p-values
are
demarcated herein using the approximation symbol ‘‘<’’. A
separate One-Way ANOVA was run to evaluate differences
among age classes and between genders for vibrissal length and
area. Traditional Morphometrics To maintain the comparative aspect of the study, we measured
smooth vibrissae, which do not have crests and troughs, at
approximate points analogous to the crest and trough locations on
beaded vibrissae (Fig. 1). The mean peak-to-peak distance
measured along the top and bottom margins of all beaded
vibrissae was used as a guide to measure width at multiple
locations on smooth vibrissae. These width measurements were
analogous to crest width on beaded vibrissae. Since smooth
vibrissae do not have a sinusoidal profile (i.e., changes in width
along the length of the vibrissa), our calculation of the ratio of crest
width to trough width was 1 by definition for otariids and bearded
seals (Table 2). Although we refer to width values of smooth
vibrissae as being crest and trough widths, this nomenclature was
used simply to enable comparison with beaded vibrissae and does
not describe the shape of smooth vibrissae. Geometric Morphometrics. The EFA coefficients were
compared among species using Principal Components Analysis
(PCA) on correlations. Principal components (PCs) on correlations,
rather than covariances, were used because small-scale differences
in vibrissal morphology were more important in discriminating
species than large-scale differences, such as overall curvature. PCs
on covariances tend to weight large-scale differences more heavily,
while PCs on correlations give all variables equal weight [38]. The
number of principal components necessary to summarize 99% of
the variation in the dataset was used in subsequent analyses. Area
within the vibrissal outline was used as the measure of size for the
geometric morphometric analysis. Since area is a squared (two
dimensional) measurement, the square root of area was taken to
convert this value to a linear measurement, which would be
comparable to the linear traditional measurements. The square
root of area was subsequently log transformed for normality. A
MANOVA was run with species, log-transformed square root of
vibrissae area and an interaction effect between species and
vibrissae
area
as
the
independent
variables
and
principal
components of Fourier coefficients as the dependent variables. To remove the effect of size in the dataset before conducting DFA,
a second MANOVA was run with only log-transformed square
root of vibrissae area as the independent variable and principal
components summarizing 99% of the variance as the dependent
variables. Residuals from that analysis were used in both linear
and quadratic DFA with the species effect. Statistical Analyses Elliptic Fourier harmonic coefficients and the log-transformed
traditional morphometric measures were combined to create a
more complete analysis of vibrissal form. However, since the two
methods measured vibrissae in different units, the scale of each
dataset was converted to match their true rank to be comparable. Matrix rank is the number of linearly independent columns or
rows in a data matrix [35]. In a more practical vein, the true rank All statistical analyses used JMP software (version 8.0.1, SAS
Institute, Inc., Cary, NC), with the exception of partial g2, which
was calculated using SPSS (version 14.0.1, SPSS Inc.). Discrim-
inant function analysis (DFA) provided a heuristic measure of how
well each morphometric method performed in correctly classifying
individual
pinnipeds
to
species
based
on
vibrissal
shape. Figure 2. Example outlines of a vibrissa from each species analyzed. Left column from top to bottom: gray, harp, ringed and spotted. Right
column from top to bottom: harbor, bearded, fur seal and sea lion. doi:10.1371/journal.pone.0034481.g002 Figure 2. Example outlines of a vibrissa from each species analyzed. Left column from top to bottom: gray, harp, ringed and spotted. Right
column from top to bottom: harbor, bearded, fur seal and sea lion. doi:10.1371/journal.pone.0034481.g002 PLoS ONE | www.plosone.org April 2012 | Volume 7 | Issue 4 | e34481 PLoS ONE | www.plosone.org 3 Traditional Morphometrics The total length of vibrissae analyzed in this study ranged from
60 mm to 110 mm. The number of beads per cm along the
vibrissa ranged from 1.1 (harp seal) to 3.4 (gray seal) and species
were significantly different in this variable (p<0.002; Table 2). For
beaded vibrissae, the mean values for the peak-to-peak distance
along the top and bottom vibrissal margins were nearly identical,
indicating that the distance between successive crests is not
affected by curvature of the overall vibrissa (Table 2, Fig. 3A). Bearded seals had the widest smooth vibrissae and spotted seals
(Phoca largha) showed the greatest mean crest and trough widths of
all beaded vibrissae. The lowest crest and trough widths were seen
in ringed seals (Pusa hispida; Table 2, Fig. 3B). However, ringed
seals had the highest crest width to trough width ratio, indicating
that these vibrissae have the most pronounced sinusoidal profile of
the five phocid species investigated with beaded vibrissae. Spotted
seals had the lowest crest width to trough width ratio, indicating
that these seals have the least pronounced sinusoidal profile of the
five beaded phocid species. Crest width to trough width ratios of
harp (Pagophilus groenlandicus) and harbor (Phoca vitulina) seals were
nearly identical to each other as were the crest width to trough
width ratios of spotted and gray (Halichoerus grypus) seals (Table 2). Quadratic DFA on the traditional morphometric measures
separated beaded species from non-beaded species on Canonical
axis 1 (Fig. 4A). In this analysis otariids separated into two groups
that slightly overlapped with one another. Bearded seals (smooth
vibrissae) overlapped with both gray seals from the beaded phocid
cluster and fur seals in the otariid group. Gray seals overlapped Combined Traditional and Geometric Morphometric
Data. To scale each of the morphometric datasets to its true
rank, we used eigenvalues of the covariance matrix from the PCA PLoS ONE | www.plosone.org April 2012 | Volume 7 | Issue 4 | e34481 4 Pinniped Whisker Shape and Morphology Table 2. Results of traditional and geometric morphometric measurements. Traditional Morphometrics Top Peak-to-Peak
Distance (mm)
Bottom Peak-to-
Peak Distance (mm)
Crest Width (mm)
Trough Width
(mm)
Crest Width/
Trough Width
Total Length (mm)
# of Beads/cm
Harp
3.9760.70
3.9460.69
0.8860.14
0.6960.13
1.28
80.6610.8ab
2.260.4bc
Harbor
3.2760.39
3.2660.40
0.9260.13
0.7360.12
1.26
86.5610.5ab
2.360.4abc
Ringed
3.5660.73
3.5360.72
0.7060.21
0.4960.21
1.44
79.6610.9ab
2.560.5ab
Spotted
4.0160.63
3.9960.63
1.0460.15
0.8560.15
1.22
91.267.9a
1.960.2c
Gray
3.4360.38
3.4160.39
0.7660.13
0.6360.11
1.21
73.8612.5b
2.760.5a
Bearded
3.6460.66
3.6260.65
1.1860.27
1.1860.27
0.99
77.269.5ab
0
Fur Seals
3.6460.66
3.6260.65
0.9660.12
0.9660.12
1.01
83.0620.3ab
0
Sea Lions
3.6460.66
3.6260.65
0.8460.19
0.8460.19
1.00
90.7611.3a
0
Mean 6 SD values for each species are given. Significant differences between species for total length and number of beads per cm are indicated by different letters. Significant differences between species for the other traditional morphometric measurements are given in Figure 3. Species with different letters are significantly
different from one another. doi:10.1371/journal.pone.0034481.t002 Mean 6 SD values for each species are given. Significant differences between species for total length and number of beads per cm are indicated by different letters. Significant differences between species for the other traditional morphometric measurements are given in Figure 3. Species with different letters are significantly
different from one another Mean 6 SD values for each species are given. Significant differences between species for total length and number of beads per cm are indicated by different letters. Significant differences between species for the other traditional morphometric measurements are given in Figure 3. Species with different letters are significantly
different from one another. doi:10 1371/journal pone 0034481 t002 Figure 3. Results of the traditional morphometric measurements. Vibrissae from five species of phocid seals with beaded vibrissae, one
species of phocid with smooth vibrissae (bearded) and otariids were analyzed. One-Way ANOVA’s were performed following overall significance of
species effects in MANOVA. Species with the same letter were not significantly different; species with different letters were significantly different for
the given measure (Tukey HSD; p,0.05). A) Mean (+SE) peak-to-peak distances along the top and bottom margins of the vibrissa. B) Mean (+SE) crest
and trough widths. Since smooth vibrissae do not have crests and troughs, both measurements simply describe the vibrissal width. doi:10.1371/journal.pone.0034481.g003 Figure 3. Results of the traditional morphometric measurements. Vibrissae from five species of phocid seals with beaded vibrissae, one
species of phocid with smooth vibrissae (bearded) and otariids were analyzed. One-Way ANOVA’s were performed following overall significance of
species effects in MANOVA. Pinniped Whisker Shape and Morphology only with harbor seals from the beaded phocid cluster. Spotted
and harp seals overlapped extensively, while ringed and harbor
seals were completely separated from each other. Canonical axis 2
was likely composed of more than one variable and may have
separated species primarily based on overall vibrissal length, with
harbor seals being further separated from the beaded cluster by
their lower peak-to-peak distances (Table 2). Interestingly, gray
seals had the highest mean number of beads per cm but were
located closer to the smooth-whiskered species in morphospace
(Fig. 4A). Quadratic DFA showed that the traditional morpho-
metric measures correctly classified 73 out of 92 individuals
(79.3%; Table 3). Figure 4. Centroid plots for the quadratic discriminant function
analyses (QDFA) on the three morphometric methodologies. Crosses mark the mean for each species; ellipses are 95% confidence
regions. A) Results of QDFA on the traditional morphometric measures. B) Results of QDFA on elliptic Fourier harmonic coefficients (geometric
morphometrics). C) Results of QDFA on the combined traditional and
geometric morphometrics. doi:10.1371/journal.pone.0034481.g004 There were significant differences between some species in each
of the traditional morphometric measurements. ANOVA with
Tukey post-hoc results showed that harp and spotted seals
significantly differed from harbor seals (p<0.004 and p<0.010,
respectively; Fig. 3A) for the log of peak-to-peak distance along the
top margin of the vibrissa. The other species did not significantly
differ from one another for this variable. Harp and spotted seals
also significantly differed from harbor seals (p<0.005 and
p<0.013, respectively) for the log of peak-to-peak distance along
the bottom margin of the vibrissa, and the other species did not
significantly differ from one another in this variable (Fig. 3A). Bearded seals differed from all species except spotted (p<0.964),
harbor (p<0.144) and fur seals (p<0.809) for the crest width
variable, and ringed seals differed from all species except sea lions
(p<0.258) and gray seals (p<0.948; Fig. 3B). Ringed seals differed
from all species except gray seals for the trough width variable
(p<0.066), and bearded seals differed from all species except fur
(p<0.835) and spotted seals (p<0.085; Fig. 3B). Again, bearded
seals, fur seals and sea lions do not have physical crests and troughs
so these comparisons refer to their vibrissal width values at points
analogous to crests and troughs on beaded vibrissae (Fig. 1). Pinniped Whisker Shape and Morphology Species were not significantly different in the log length of the
vibrissae variable, except for gray seals, which differed from sea
lions and spotted seals (p<0.021 and p<0.029, respectively;
Table 2). Intraspecific variation could not be effectively assessed
due to the low number of individuals per species. Individual animals were classified by the recovering stranding
network or indigenous group into one of five age classes: pup,
yearling, juvenile, subadult or adult. The five age classes did not
have significantly different vibrissal lengths (p = 0.058) or vibrissal
areas (p = 0.144). Additionally, males and females did not differ in
vibrissal length (p = 0.9169) or vibrissal area (p = 0.825). Traditional Morphometrics Species with the same letter were not significantly different; species with different letters were significantly different for
the given measure (Tukey HSD; p,0.05). A) Mean (+SE) peak-to-peak distances along the top and bottom margins of the vibrissa. B) Mean (+SE) crest
and trough widths. Since smooth vibrissae do not have crests and troughs, both measurements simply describe the vibrissal width. doi:10.1371/journal.pone.0034481.g003 PLoS ONE | www.plosone.org April 2012 | Volume 7 | Issue 4 | e34481 5 Geometric Morphometrics A good (identical by eye) outline fit of each vibrissa was obtained
with 15 harmonics. Four Fourier coefficients (A, B, C, D) were
generated to describe each of the 15 harmonics, plus the two
zeroth (A0 and C0) harmonics, for a total of 62 shape variables. Fifteen harmonics provided enough variability in shape definitions
that 18 PCs summarized 99% of the variance within the dataset. We therefore used these 18 variables as our metrics of shape in
subsequent analyses. Both linear and quadratic DFA provided
good discrimination among species based on shape (Table 3). Due
to the greater discriminatory ability of quadratic DFA we focused
on those results. The quadratic DFA on the geometric morpho-
metric measures separated otariids from phocids (Fig. 4B). The
phocids with beaded vibrissae clustered together and all species
overlapped with each other. Additionally, the otariids separated
into two non-overlapping groups, whereas these same groups
overlapped slightly in the traditional plot. Bearded seals were
positioned between beaded phocids and otariids, but were
completely separated from the phocid cluster and both otariid Figure 4. Centroid plots for the quadratic discriminant function
analyses (QDFA) on the three morphometric methodologies. Crosses mark the mean for each species; ellipses are 95% confidence
regions. A) Results of QDFA on the traditional morphometric measures. B) Results of QDFA on elliptic Fourier harmonic coefficients (geometric
morphometrics). C) Results of QDFA on the combined traditional and
geometric morphometrics. doi:10.1371/journal.pone.0034481.g004 Figure 4. Centroid plots for the quadratic discriminant function
analyses (QDFA) on the three morphometric methodologies. Crosses mark the mean for each species; ellipses are 95% confidence
regions. A) Results of QDFA on the traditional morphometric measures. B) Results of QDFA on elliptic Fourier harmonic coefficients (geometric
morphometrics). C) Results of QDFA on the combined traditional and
geometric morphometrics. doi:10.1371/journal.pone.0034481.g004 PLoS ONE | www.plosone.org April 2012 | Volume 7 | Issue 4 | e34481 6 Pinniped Whisker Shape and Morphology Table 3. Results of the MANOVA for the three analyses. F
DF
p
Partial g2
LDFA %
QDFA %
Traditional
Species
9.2
35,305.3
,0.001
0.45
64.1
79.3
Geomsize
6867.5
5,72
,0.001
Species*Geomsize
3.1
35,305.3
,0.001
Geometric
Species
2.5
126,397.7
,0.001
0.42
78.3
97.8
LogSqrtArea
3.4
18,59
,0.001
Species*LogSqrtArea
1.7
126,397.7
,0.001
Combined
Species
4.4
133,340.9
,0.001
0.50
84.8
100.0
LogSqrtArea
10.1
19,50
,0.001
Geomsize
8.9
19,50
,0.001
Species*LogSqrtArea
2.8
133,340.9
,0.001
Species*Geomsize
2.3
133,340.9
,0.001
Partial g2 is a measure of effect strength. Combined Traditional and Geometric Morphometric
Data Quadratic DFA on the combined dataset correctly classified
100% of individuals (Table 3), compared to 79.3% and 97.8% for
the traditional and geometric datasets, respectively. This demon-
strated that each method found differences between species that
the other method did not. The combined methodology incorpo-
rated all differences and was the best at correctly classifying
species. In all three analyses, quadratic DFA outperformed linear
DFA in correctly classifying individuals (Table 3). Since all factors
in the statistical models were significant, we compared the
proportion of partial variance explained by the main effect of
interest, species, using Wilks’ partial g2. This value was similar for
all three analyses but highest for the combined dataset (Table 3). This indicates that the species effect was relatively strongest in the
combined analysis, and relatively weakest in the geometric
morphometric analysis. However, the species effect still explained
approximately 50% of the variance in each of the three models. The traditional and geometric morphometric data exhibited
shared
and
unique
aspects
of
discriminatory
ability. The
redundancy (similar taxonomic discrimination) implies a com-
bined analysis is more appropriate, more complete and more
powerful. A PCA on the EFA harmonic coefficients showed that
18 PCs were required to summarize 99% of the variance in the
dataset compared to a PCA on the traditional dataset, which only
needed three PCs to summarize 99% of the variance in the data. Therefore, the true ranks of the geometric and traditional
morphometric datasets were 18 and three, respectively. If there
were no redundancy between the two datasets, a combined,
expected 21 PCs would be required to summarize 99% of the
variance in the dataset. However, the combined dataset required
19 PCs on correlations to summarize 99% of the variance in the
dataset. This showed that there was some overlap between the
traditional and geometric morphometric methodologies. The
MANOVA results for the combined data are summarized in
Table 3. Since two different measures of size were used in the
traditional and geometric morphometric analyses, both size
measures were included in the combined data model. Non-
significant interaction effects were removed from the model. Geometric Morphometrics LDFA% and QDFA% are the percentages of individuals that were correctly classified by the linear discriminant function analysis
and quadratic discriminant function analysis, respectively. Geomsize (geometric mean) and LogSqrtArea (log of the square root of the area) were the measures of
vibrissal size. doi:10.1371/journal.pone.0034481.t003 Partial g2 is a measure of effect strength. LDFA% and QDFA% are the percentages of individuals that were correctly classified by the linear discriminant function analysis
and quadratic discriminant function analysis, respectively. Geomsize (geometric mean) and LogSqrtArea (log of the square root of the area) were the measures of
vibrissal size. doi:10.1371/journal.pone.0034481.t003 doi:10.1371/journal.pone.0034481.t003 groups, in contrast to the traditional plot. Canonical axis 1
appeared to separate beaded vibrissal species from non-beaded
vibrissal species, as seen in the traditional plot. Canonical axis 2
may have separated species based on overall vibrissal length, since
sea lions and bearded seals had the longest and shortest smooth
vibrissae, respectively, and spotted and gray seals had the longest
and shortest beaded vibrissae, respectively (Table 2). Alternatively,
this axis may be detecting differences in cross-sectional shape, or
an
interaction
of
several
shape
characteristics. Subsequent
canonical axes made only minor contributions to discriminatory
ability. Quadratic DFA showed that the principal components of
elliptic Fourier harmonic coefficients facilitated classification of 90
out of 92 individuals (97.8%; Table 3). vibrissal length. This separation was clear for the smooth-
whiskered species, but the small-scale intricacies of the beaded
profile likely complicated the relationships between the beaded
phocids. For example, spotted seals had the longest vibrissae but
may have been pulled along Canonical axis 2 towards harp seals
and away from harbor seals by the peak-to-peak distance variables
(Table 2). Surprisingly, gray seals, with the highest mean number
of beads per cm, were again pulled towards the smooth-whiskered
species. In this analysis, harbor and gray seals did not overlap with
either spotted or harp seals, but did overlap with each other and
ringed seals. Harp and ringed seals overlapped considerably
(Fig. 4C). In contrast to the centroid plots for the geometric and
traditional morphometrics, otariids loaded higher than phocids on
Canonical axis 1 and bearded seals had the highest loading on
Canonical axis 2. Combined Traditional and Geometric Morphometric
Data PLoS ONE | www.plosone.org Pinniped Whisker Shape and Morphology This was surprising
since this species had the highest number of beads per cm and the
overall shape appeared quite similar to the rest of the beaded
phocids. Ginter et al. [17] initially found a different pattern in
traditional morphometric measurements along the vibrissae of
gray seals compared to harp and hooded (Cystophora cristata) seals,
but that difference was not maintained when additional samples
were added (C.C. Ginter, unpubl. data). Gray seals possess
different head morphology from other phocids. Cameron [39] and
King [11] described the nose of male gray seals as high and
arched, while females have a long, straight profile to the top of the
head. Gray seals are similar to hooded and elephant (Mirounga
spp.) seals in that the males have enlarged snouts used in visual
signaling [40]. The broader snout may change the position of
vibrissae on the muzzle. As a result of this difference in location,
vibrissae may have evolved an alternative morphology. While the
phocid species with beaded vibrissae always clustered together in
morphospace, the fact that all individuals could be correctly
classified to the species level demonstrated that the beaded profile
is not identical. We sampled a large number of Phocinae species,
but not all members of this subfamily were included in the study. However, based on the variation within this beaded vibrissal
group, we predict that the rest of family Phocidae would cluster
with the beaded phocids examined here. (Callorhinus ursinus) fetus with vibrissae as long as 63 mm and
Bonner’s [10] observation of a three week old Antarctic fur seal
(Arctocephalus gazella) with vibrissae up to about 80 mm long. A four
and a half month old Southern elephant seal (Mirounga leonina) fetus
had mystacial vibrissae up to 27 mm long [27]. A ringed seal pup
of the year in the present study had vibrissae that were slightly
longer than the mean value for that species. Although caveats are
certain to arise, the data analyzed here suggest that there are no
gender or age class effects on vibrissal morphology and shape. Although we focus on the shape of individual vibrissae here,
other factors also are likely to be important in vibrotactile sensory
perception, such as location, distribution of the vibrissae on the
muzzle, and innervation. The function of the entire mystacial
vibrissa pad is most likely an interaction between vibrissal hair
shaft shape, the geometry and location of the vibrissae. Pinniped Whisker Shape and Morphology The
interaction of morphology at these two scales is likely to be related
to foraging mode and strategy. Bearded seals and walruses
exemplify the importance of the geometry of mystacial vibrissae
location. The distribution of bearded seal vibrissae differs from
other phocids. Instead of lying along the lateral sides of the
rostrum, bearded seals have vibrissae widely distributed over the
anterior portion of a blunt muscular muzzle [9,16]. Bearded seals
forage for benthic invertebrates [42]. Walruses, another benthic
foraging specialist, exhibit a similar vibrissal distribution [43]. This
vibrissal arrangement is related to a benthic foraging mode and
appears to be convergent with the vibrissal arrangement on the
oral disk of sirenians, which also spend considerable time foraging
on the benthos [44–46]. Due to the orientation of bearded seal and
walrus vibrissae [16] it is unlikely that these benthic foragers are
protracting the vibrissae into the flow field, and therefore their
function will be different than either beaded or other smooth
vibrissae. Bearded seals, a phocid with smooth vibrissae, were positioned
between beaded phocids and otariids in morphospace. Interest-
ingly, the centroid ellipses for bearded seals and otariids (which
also possess smooth vibrissae) never overlapped in the geometric
morphometric or the combined data analysis. In fact, the centroid
ellipse for bearded seals was closer in morphospace to the beaded
phocids than to the smooth-whiskered otariids in the geometric
and combined analyses. This strongly suggests that the smooth
vibrissae of bearded seals are different from the smooth vibrissae of
otariids in this study. This difference may be related to cross-
sectional shape. Bearded seal vibrissae are almost rectangular in
cross-section and this shape differs considerably from the oval
cross-sectional shape of otariids and other phocids ([16], pers. obs.). A limitation of this comparative analysis is that the other
phocids with smooth vibrissae, monk seals (Monachus spp.), were
not included. However, based on personal observations and diet
studies in the literature, we predict that monk seals would not
occupy the same morphospace as bearded seals. Rather, they may
be intermediate between bearded seals and the beaded phocid
cluster or intermediate between bearded seals and otariids. The
differences among pinniped vibrissae without a beaded profile
were a surprising result of this study. Pinniped Whisker Shape and Morphology We have reported evidence
that there is variation in smooth vibrissal shape and morphology
among otariids, since the geometric morphometric approach
completely separated otariids into two groups, and the combined
data analysis demonstrated only a minimal overlap between these
groups. Clearly these differences in morphology and shape
previously have been overlooked and may have important
ecological consequences. Otariids have smooth vibrissae and feed on similar prey to
phocids with beaded vibrissae [47], which suggests that beaded
vibrissae are not critical in catching certain prey items. However,
otariids generally do not dive to the depths that phocids do in
search of prey. The greater amount of ambient light present in
shallower water may allow otariids to rely more heavily upon
vision for prey detection and capture or a combination of visual,
auditory and tactile cues [48]. Other ecological factors such as
time of day of foraging, actively swimming versus sedentary prey,
foraging on the bottom versus in the water column and water
turbidity may also affect the presence of beading along pinniped
vibrissae. Both California sea lions and harbor seals are able to
detect water velocities below those that would be generated by a
swimming fish using their vibrissae [19,49]. However, blindfolded
California sea lions could successfully track a hydrodynamic trail
using their vibrissae only 50% of the time when the signal made a
single turn. There was also a decrease in tracking ability when
there was a delay of more than a few seconds between the
generation of the trail and the beginning of the sea lion’s search for
it [48]. The animal always failed to find the hydrodynamic trail
when the vibrissae were covered by a stocking mask [48]. These
performance data suggest that vibrissae are an important sensory
modality in this species, but are not the only sensory system
involved in prey tracking, since it is unlikely that a chased fish will
swim in a straight line. It is important to note that although the comparisons of vibrissal
length and area between genders and among age classes are
interesting, full Level A data were only obtained for 45 out of 92
individuals from five species. More males than females were
obtained (34 males versus 16 females) and there were more known
adult males than known adult females, both of which may have
biased the results. Discussion Phocids with beaded vibrissae show species-specific variation on
a common sinusoidal beaded pattern. In all analyses, the phocids
possessing beaded vibrissae clustered together. However, it is
interesting to note that the only two congeneric phocids in the
study, harbor and spotted seals, did not overlap at all in the
combined data analysis and overlapped only partially in the
geometric morphometric and traditional morphometric analyses. Quadratic DFA again clearly separated phocids and otariids on
Canonical axis 1 with bearded seals occupying an intermediate
position between otariids and beaded phocids (Fig. 4C). Canonical
axis 2 again appeared to separate species in our dataset based on PLoS ONE | www.plosone.org April 2012 | Volume 7 | Issue 4 | e34481 April 2012 | Volume 7 | Issue 4 | e34481 7 PLoS ONE | www.plosone.org Pinniped Whisker Shape and Morphology Gray seals appeared to occupy an intermediate position in
morphospace between the other beaded phocids and smooth-
whiskered bearded seals in all three analyses. This was surprising
since this species had the highest number of beads per cm and the
overall shape appeared quite similar to the rest of the beaded
phocids. Ginter et al. [17] initially found a different pattern in
traditional morphometric measurements along the vibrissae of
gray seals compared to harp and hooded (Cystophora cristata) seals,
but that difference was not maintained when additional samples
were added (C.C. Ginter, unpubl. data). Gray seals possess
different head morphology from other phocids. Cameron [39] and
King [11] described the nose of male gray seals as high and
arched, while females have a long, straight profile to the top of the
head. Gray seals are similar to hooded and elephant (Mirounga
spp.) seals in that the males have enlarged snouts used in visual
signaling [40]. The broader snout may change the position of
vibrissae on the muzzle. As a result of this difference in location,
vibrissae may have evolved an alternative morphology. While the
phocid species with beaded vibrissae always clustered together in
morphospace, the fact that all individuals could be correctly
classified to the species level demonstrated that the beaded profile
is not identical. We sampled a large number of Phocinae species,
but not all members of this subfamily were included in the study. However, based on the variation within this beaded vibrissal
group, we predict that the rest of family Phocidae would cluster
with the beaded phocids examined here. (Callorhinus ursinus) fetus with vibrissae as long as 63 mm and
Bonner’s [10] observation of a three week old Antarctic fur seal
(Arctocephalus gazella) with vibrissae up to about 80 mm long. A four
and a half month old Southern elephant seal (Mirounga leonina) fetus
had mystacial vibrissae up to 27 mm long [27]. A ringed seal pup
of the year in the present study had vibrissae that were slightly
longer than the mean value for that species. Although caveats are
certain to arise, the data analyzed here suggest that there are no
gender or age class effects on vibrissal morphology and shape. Gray seals appeared to occupy an intermediate position in
morphospace between the other beaded phocids and smooth-
whiskered bearded seals in all three analyses. Pinniped Whisker Shape and Morphology fish [23]. As seen in California sea lions, harbor seals never found
the hydrodynamic trail when the vibrissae were covered with a
stocking mask [20,23–25]. The beaded profile of harbor seals’
vibrissae was shown to suppress self-induced vibrations caused by
ambient water flow during swimming [30]. It is likely that reduced
vibrations of the vibrissae allow detection of hydrodynamic trails
as prey turn away from the seal during escape maneuvers. Objects
of different sizes and shapes can be perceived based on
characteristics of their resulting hydrodynamic trail by harbor
seals’ vibrissae [50]. Since we have shown that morphology and
shape of beaded phocid vibrissae are species-specific variants on a
basic pattern, the performance data from harbor seals may not be
completely representative of all phocids with beaded vibrissae. research should investigate further the shape differences among
smooth vibrissae of otariids and phocids. Behavioral performance
data for additional phocids and otariids will help elucidate the
potential functional and ecological diversification that correlates
with the variation in vibrissal morphology and shape reported in
this study. Finally we hope to highlight the methodological insight
that geometric and traditional morphometrics should not be
treated as alternatives. It is fashionable to compete the two
approaches to see which is ‘‘best’’ (e.g., [51,52]). Rather, the two
types of analysis should generally be used in harmony, by fusing
the data as illustrated herein, to yield the most complete
understanding of morphology. In the present case, 100% of
vibrissae could be classified to taxon, which is a testament to both
the synthetic methodology and the biological diversity in vibrissal
shape. p
y
p
p
The different sizes and shapes of beaded vibrissae may have
functional consequences related to identifying the hydrodynamic
signal of prey that have not yet been explored. Different shaped
paddles created different flow fields that were quantified using
digital particle imaging velocimetry [50]. The species-specific
differences observed in beaded vibrissae may be related to subtle
changes in the flow fields created by prey of slightly different size
or shape. However, this possibility would best be explored using
live pinnipeds with different vibrissal morphologies chasing live
prey that create different flow fields. The biological mechanisms
responsible for beading and species-specific differences in beaded
morphologies
are
currently
not
understood. One
possible
mechanism affecting beading is periodic variation in keratin
deposition rate during vibrissal growth. References 1. Burgess P, Perl E (1973) Cutaneous mechanoreceptors and nociceptors. In:
Iggo A, ed. Handbook of Sensory Physiology 2: Somatosensory System. Berlin:
Springer-Verlag. pp 29–78. 13. Ognev S (1935) Mammals of USSR and adjacent countries, vol. III, Carnivora
(Fissipedia and Pinnipedia). Russian, English translation by Birron A and Coles
ZS for Israel Program for Scientific Translations, 1962. g
,
14. Hyva¨rinen H, Katajisto H (1984) Functional structure of the vibrissae of the
ringed seal (Phoca hispida Schr.). Acta Zoologica Fennica 171: 17–30. 2. Gottschaldt KM, Iggo A, Young DW (1973) Functional characteristics of
mechanoreceptors in sinus hair follicles of the cat. Journal of Physiology 235:
287–315. 14. Hyva¨rinen H, Katajisto H (1984) Functional structure of ringed seal (Phoca hispida Schr.). Acta Zoologica Fennica 171: 17–30. 15. Dehnhardt G, Kaminski A (1995) Sensitivity of the mystacial vibrissae of
harbour seals (Phoca vitulina) for size differences of actively touched objects. Journal of Experimental Biology 198: 2317–2323. 3. Dykes RW (1975) Afferent fibers from mystacial vibrissae of cats and seals. Journal of Neurophysiology 38: 650–662. 16. Marshall CD, Amin H, Kovacs KM, Lydersen C (2006) Microstructure and
innervation of the mystacial vibrissal follicle-sinus complex in bearded seals,
Erignathus barbatus (Pinnipedia: Phocidae). The Anatomical Record Part A 288A:
13–25. 4. Halata Z (1975) The mechanoreceptors of the mammalian skin ultrastructure
and morphological classification. Advances in Anatomy, Embryology, and Cell
Biology 50: 3–77. 5. Ahl AS (1986) The role of vibrissae in behavior: a status review. Veterinary
Research Communications 10: 245–268. 17. Ginter CC, Fish FE, Marshall CD (2010) Morphological analysis of the bumpy
profile of phocid vibrissae. Marine Mammal Science 26: 733–743. 6. Rice FL, Mance A, Munger BL (1986) A comparative light microscopic analysis
of the sensory innervation of the mystacial pad. I. Innervation of vibrissal follicle
sinus complexes. Journal of Comparative Neurology 252: 154–174. 18. Hyva¨rinen H (1989) Diving in darkness: whiskers as sense organs of the ringed
seal (Phoca hispida saimensis). Journal of Zoology 218: 663–678. 7. Rice FL, Munger BL (1986) A comparative light microscopic analysis of the
sensory innervation of the mystacial pad. II. The common fur between the
vibrissae. Journal of Comparative Neurology 252: 186–205. 19. Dehnhardt G, Mauck B, Bleckmann H (1998) Seal whiskers detect water
movements. Nature 394: 235–236. 20. Dehnhardt G, Mauck B, Hanke W, Bleckmann H (2001) Hydrodynamic trail-
following in harbor seals (Phoca vitulina). Science 293: 102–104. 8. Acknowledgments We would like to thank the New Jersey Marine Mammal Stranding Center,
Perry Habecker and the University of Pennsylvania New Bolton Center
Large Animal Pathology Laboratory staff, Keith Matassa and the
University of New England Marine Animal Rehabilitation Center
necropsy staff, Lauren Rust and the Marine Mammal Center, Diane
Olsen, Laura Hilbert and the animal care staff at Moody Gardens, Alex
Whiting of the Native Village of Kotzebue, Alaska, Kathy Frost of the
University of Alaska Fairbanks, Gay Sheffield, Lori Quakenbush and Anna
Bryan of Alaska Department of Fish and Game, David Rosen of the
University of British Columbia, Brandon Russell at Vancouver Aquarium,
and Stephen Raverty of Ministry of Agriculture and Lands, British
Columbia, Canada for assistance in vibrissae collection. Three anonymous
reviewers are gratefully acknowledged for their time and helpful comments
and suggestions to improve the manuscript. In summary, although the classic view is that pinniped vibrissae
exhibit two distinct vibrissal morphologies, beaded and smooth,
the morphology and shape of pinniped vibrissae within this study
fall into at least three distinct groups: phocids with beaded
vibrissae, phocids with smooth vibrissae, and otariids. A fourth
group may be identified if additional research substantiates (with
additional species and greater sample size) the division of otariid
vibrissal shape into two groups as shown in our analyses. Future Pinniped Whisker Shape and Morphology Regardless of the process,
the resulting shapes likely relate to functional differences that
should be investigated in live animals. Author Contributions Conceived and designed the experiments: CCG FEF CDM. Performed the
experiments: CCG TJD. Analyzed the data: CCG. Contributed reagents/
materials/analysis tools: TJD FEF CDM. Wrote the paper: CCG TJD FEF
CDM. Conceived and designed the experiments: CCG FEF CDM. Performed the
experiments: CCG TJD. Analyzed the data: CCG. Contributed reagents/
materials/analysis tools: TJD FEF CDM. Wrote the paper: CCG TJD FEF
CDM. 1.
Burgess P, Perl E (1973) Cutaneous mechanoreceptors and nociceptors. In:
Iggo A, ed. Handbook of Sensory Physiology 2: Somatosensory System. Berlin:
Springer-Verlag. pp 29–78. Pinniped Whisker Shape and Morphology However, additional support for the observed
lack of difference in vibrissal length among age classes comes from
Scheffer’s [41] observation of a full term Northern fur seal In contrast, harbor seals are able to follow a complex
hydrodynamic trail as long as 40 m with high accuracy, even
with glide phases in the trail, and can determine the direction of a
trail after delays up to 35 s [20,24,25]. Additionally, these seals are
able to follow a trail, even when they contact it at an obtuse angle,
by repeatedly crossing the trail and gradually narrowing the angle. Such a search method would be more successful in tracking fleeing PLoS ONE | www.plosone.org April 2012 | Volume 7 | Issue 4 | e34481 April 2012 | Volume 7 | Issue 4 | e34481 8 Pinniped Whisker Shape and Morphology Pinniped Whisker Shape and Morphology 25. Wieskotten S, Dehnhardt G, Mauck B, Miersch L, Hanke W (2010) The impact
of glide phases on the trackability of hydrodynamic trails in harbour seals (Phoca
vitulina). Journal of Experimental Biology 213: 3734–3740. 40. Miller EH, Boness DJ (1979) Remarks on display functions of the snout of the
grey seal, Halichoerus grypus (Fab.), with comparative notes. Canadian Journal of
Zoology 57: 140–148. gy
41. Scheffer VB (1962) Pelage and surface topography of the northern fur seal. North American Fauna 64: 1–206. 26. Wartzok D, Elsner R, Stone H, Kelly BP, Davis RW (1992) Under-ice
movements and the sensory basis of hole finding by ringed and Weddell seals. Canadian Journal of Zoology 70: 1712–1712. 42. Burns JJ (1981) Bearded seal Erignathus barbatus Erxleben, 1777. In: Ridgway SH, 42. Burns JJ (1981) Bearded seal Erignathus barbatus Erxleben, 1777. In: Ridgway SH,
Harrison RJ, eds. Handbook of Marine Mammals. London: Academic Press. pp
145–170. Harrison RJ, eds. Handbook of Marine Mammals. London: Academic Press. pp
145–170. 27. Ling JK (1966) The skin and hair of the southern elephant seal, Mirounga
leonina (Linn.) I. The facial vibrissae. Australian Journal of Zoology 14:
855–866. 43. Fay FH (1982) Ecology and biology of the Pacific walrus, Odobenus rosmarus
divergens Illiger. North American Fauna 74: 1–279. 28. Hirons AC, Schell DM, St Aubin DJ (2001) Growth rates of vibrissae of harbor
seals (Phoca vitulina) and Steller sea lions (Eumetopias jubatus). Canadian Journal of
Zoology 79: 1053–1061. 44. Marshall CD, Huth GD, Edmonds VM, Halin DL, Reep RL (1998) Prehensile
use of perioral bristles during feeding and associated behaviors of the Florida
manatee (Trichechus manatus latirostris). Marine Mammal Science 14: 274–289. gy
29. Greaves DK, Hammill MO, Eddington JD, Pettipas D, Schreer JF (2004)
Growth rate and shedding of vibrissae in the gray seal, Halichoerus grypus: A
cautionary note for stable isotope diet analysis. Marine Mammal Science 20:
296–304. 45. Marshall CD, Maeda H, Iwata M, Furuta M, Asano S, et al. (2003) Orofacial
morphology and feeding behaviour of the dugong, Amazonian, West African
and Antillean manatees (Mammalia: Sirenia): functional morphology of the
muscular-vibrissal complex. Journal of Zoology 259: 245–260. 30. Hanke W, Witte M, Miersch L, Brede M, Oeffner J, et al. (2010) Harbor seal
vibrissa morphology suppresses vortex-induced vibrations. Journal of Experi-
mental Biology 213: 2665–2672. 46. Pinniped Whisker Shape and Morphology Reep RL, Marshall CD, Stoll ML, Whitaker DM (1998) Distribution and
innervation of facial bristles and hairs in the Florida manatee (Trichechus manatus
latirostris). Marine Mammal Science 14: 257–273. 31. Darroch JN, Mosimann JE (1985) Canonical and principal components of
shape. Biometrika 72: 241–252. )
47. Pauly D, Trites AW, Capuli E, Christensen V (1998) Diet composition and
trophic levels of marine mammals. ICES Journal of Marine Science 55:
467–481. p
32. Rohlf FJ (2004) tpsDig2. Stony Brook, NY: Department of Ecology and
Evolution, State University of New York at Stony Brook. 48. Gla¨ser N, Wieskotten S, Otter C, Dehnhardt G, Hanke W (2011)
Hydrodynamic trail following in a California sea lion (Zalophus californianus). Journal of Comparative Physiology A: Neuroethology, Sensory, Neural, and
Behavioral Physiology 197: 141–151. 33. Rohlf FJ (1993) EFAWin. Stony Brook: Department of Ecology and Evolution,
State University of New York at Stony Brook. y
y
34. Kuhl FP, Giardina CR (1982) Elliptic Fourier features of a closed contour. Computer Graphics and Image Processing 18: 236–258. 49. Dehnhardt G, Mauck B (2008) The Physics and Physiology of Mechanorecep-
tion. In: Thewissen JGM, Nummela S, eds. Sensory Evolution on the Threshold:
Adaptations in Secondarily Aquatic Vertebrates. Berkeley, CA: University of
California Press. pp 287–293. 35. Pavlicev M, Wagner GP, Cheverud JM (2009) Measuring evolutionary
constraints through the dimensionality of the phenotype: Adjusted bootstrap
method to estimate rank of phenotypic covariance matrices. Evolutionary
Biology 36: 339–353. 50. Wieskotten S, Mauck B, Miersch L, Dehnhardt G, Hanke W (2011)
Hydrodynamic discrimination of wakes caused by objects of different size or
shape in a harbour seal (Phoca vitulina). Journal of Experimental Biology 214:
1922–1930. 36. Vickerman JC, Gilmore IS (2009) Surface analysis: the principal techniques. New York: John Wiley & Sons. 686 p. 37. Scheiner SM (2001) MANOVA: Multiple response variables and multispecies
interactions. In: Scheiner SM, Gurevitch J, eds. Design and Analysis of
Ecological Experiments. 2nd ed. New York: Oxford University Press. pp
99–115. 51. Parsons KJ, Robinson BW, Hrbek T (2003) Getting into shape: an empirical
comparison of traditional truss-based morphometric methods with a newer
geometric method applied to New World cichlids. Environmental Biology of
Fishes 67: 417–431. 38. Jackson DA (1993) Stopping rules in principal components analysis: A
comparison of heuristical and statistical approaches. Ecology 74: 2204–2214. 52. Mutanen M, Pretorius E (2007) Subjective visual evaluation vs. References Rice FL, Fundin BT, Arvidsson J, Aldskogius H, Johansson O (1997)
Comprehensive immunofluorescence and lectin binding analysis of vibrissal
follicle sinus complex innervation in the mystacial pad of the rat. Journal of
Comparative Neurology 385: 149–184. 21. Hanke FD, Scholtyssek C, Hanke W, Dehnhardt G (2011) Contrast sensitivity in
a harbor seal (Phoca vitulina). Journal of Comparative Physiology A: Neuroethol-
ogy, Sensory, Neural, and Behavioral Physiology 197: 203–210. 9. Ling JK (1977) Vibrissae of Marine Mammals. In: Harrison RJ, ed. Functional
Anatomy of Marine Mammals. London: Academic Press. pp 387–415. 22. Newby TC, Hart FM, Arnold RA (1970) Weight and blindness of harbor seals. Journal of Mammalogy 51: 152. 10. Bonner WN (1968) The fur seal of South Georgia. British Antarctic Survey
Scientific Reports 56: 1–81. 23. Schulte-Pelkum N, Wieskotten S, Hanke W, Dehnhardt G, Mauck B (2007)
Tracking of biogenic hydrodynamic trails in harbour seals (Phoca vitulina). Journal
of Experimental Biology 210: 781–787. 11. King J (1983) Seals of the World. London: British Museum (Natural History). 240 p. 24. Wieskotten S, Dehnhardt G, Mauck B, Miersch L, Hanke W (2010)
Hydrodynamic determination of the moving direction of an artificial fin by a
harbour seal (Phoca vitulina). Journal of Experimental Biology 213: 2194–2200. 12. Watkins WA, Wartzok D (1985) Sensory biophysics of marine mammals. Marine Mammal Science 1: 219–260. PLoS ONE | www.plosone.org 9 April 2012 | Volume 7 | Issue 4 | e34481 April 2012 | Volume 7 | Issue 4 | e34481 Pinniped Whisker Shape and Morphology traditional and
geometric morphometrics in species delimitation: a comparison of moth
genitalia. Systematic Entomology 32: 371–386. 39. Cameron AW (1967) Breeding behavior in a colony of western Atlantic gray
seals. Canadian Journal of Zoology 45: 161–173. PLoS ONE | www.plosone.org April 2012 | Volume 7 | Issue 4 | e34481 10
|
https://openalex.org/W3046288584
|
https://biblio.ugent.be/publication/8672667/file/8672668
|
English
| null |
Detrended fluctuation analysis of gait dynamics when entraining to music and metronomes at different tempi in persons with multiple sclerosis
|
Scientific reports
| 2,020
|
cc-by
| 10,742
|
Detrended fluctuation analysis
of gait dynamics when entraining
to music and metronomes
at different tempi in persons
with multiple sclerosis Lousin Moumdjian
1,2*, Pieter‑Jan Maes1, Simone Dalla Bella3,4,6,7, Leslie M. Decker
5,
Bart Moens
1 Peter Feys2 & Marc Leman1 In persons with multiple sclerosis (PwMS), synchronizing walking to auditory stimuli such as to music
and metronomes have been shown to be feasible, and positive clinical effects have been reported
on step frequency and perception of fatigue. Yet, the dynamic interaction during the process of
synchronization, such as the coupling of the steps to the beat intervals in music and metronomes,
and at different tempi remain unknown. Understanding these interactions are clinically relevant, as
it reflects the pattern of step intervals over time, known as gait dynamics. 28 PwMS and 29 healthy
controls were instructed to walk to music and metronomes at 6 tempi (0–10% in increments of 2%). Detrended fluctuation analysis was applied to calculate the fractal statistical properties of the gait
time-series to quantify gait dynamics by the outcome measure alpha. The results showed no group
differences, but significantly higher alpha when walking to music compared to metronomes, and when
walking to both stimuli at tempi + 8, + 10% compared to lower tempi. These observations suggest that
the precision and adaptation gain differ during the coupling of the steps to beats in music compared to
metronomes (continuous compared to discrete auditory structures) and at different tempi (different
inter-beat-intervals). The study of temporal correlations in step or stride intervals over time-also known as gait dynamics1- pro-
vides useful insights on the neural control of locomotion in young adults2,3, healthy older adults4, and patients
with movement disorders such as Parkinson’s disease5, Huntington’s disease5,6 or multiple sclerosis (MS)7. Gait
gives rise to non-stationary inter-step/stride-interval signals, with temporal fluctuations which can be analyzed
via non-linear methods1. An example of these methods, capable of capturing the complexity of time-evolving
behavior in the domain of gait analysis, is detrended fluctuation analysis (DFA). This analysis method can be
applied to a time series obtained from gait measurements such as inter-step/stride-intervals2,3,8. DFA is robust
to non-stationaries in the data, often observed in gait interval time series. This method scales the long-term
auto-correlations of non-stationary signals and quantifies the fluctuations in the time series using its self-similar
property8,9 with a value of fractal scaling index ‘alpha’8,9. Alpha provides an estimation of statistical ‘persistence’
or ‘anti-persistence’ in a time series1. www.nature.com/scientificreports www.nature.com/scientificreports www.nature.com/scientificreports/ Similar effects were also found in patients with
Parkinson’s Disease. Stimuli embedding biological variability (i.e., a metronome with variable inter-beat interval
following a fractal structure) resulted in a more persistent gait dynamics in Parkinson’s Disease compared to the
isochronous and random-variable (white noise) metronomes26. Thus, given that music generally embeds some
degree of long-range correlated temporal variability27,28 as compared to metronomes, a more persistent gait
dynamics should be observed when coupling to music. y
g
In terms of clinical application, the use of music is advantageous to implement in a rehabilitation setting
compared to metronomes, as seen in studies with persons with multiple sclerosis (PwMS)29–31, a disease where
walking impairments are prevalent32,33. Additionally, in PwMS, changes in spatiotemporal gait parameters were
shown to be dependent on walking impairment and speed instructions34. In previous studies, we have shown
that PwMS and healthy controls (HC) can sustain auditory-motor coupling30 and synchronize their walking to
music and metronomes across different tempi29. In addition, lower fatigue perception when walking to music
compared to metronomes in PwMS compared to HC were reported in these studies. However, little is known
about the gait dynamics as a result of the auditory-motor coupling and entrainment when walking to music
and metronomes at different tempi. The present study aims to fill this gap in PwMS and HC. That is by apply-
ing DFA to investigate the pattern of gait dynamics in PwMS compared to HC, when subjecting the walking to
the task-related context of coupling to music and metronomes at different tempi, and in turn understanding
the processes underlying auditory-motor entrainment. Notably, gait dynamics measured by DFA was shown
not to differ between mildly and moderately impaired PwMS and healthy controls7 during walking. Given this
evidence, as well as the results from our previous studies in regards to PwMS’s ability to synchronize to music
and metronomes29, we hypothesize that gait dynamics will not be different between PwMS and HC. We thus
assume that both groups will use the same underlying processes to engage in auditory-motor coupling. Rather,
we assume that the differences in coupling would be found when entraining to different stimuli (e.g. music and
metronomes) and at different tempi. f
p
We hypothesize that walking with music may result in a more persistent gait dynamics compared to walking
in time to isochronous metronomes. The type of auditory stimulus (e.g. www.nature.com/scientificreports/ the presence of statistical anti-persistence in inter-step-intervals, namely that the inter-step-intervals between
consecutive steps are varied and random at a long range1,10. In other words, small deviations are immediately
corrected on subsequent strides. the presence of statistical anti-persistence in inter-step-intervals, namely that the inter-step-intervals between
consecutive steps are varied and random at a long range1,10. In other words, small deviations are immediately
corrected on subsequent strides. Notably, persistent gait dynamics has been associated with a healthy gait, while the loss of persistency in
gait dynamics has been associated with aging and neurological disorders such as in Parkinson’s Disease1. Loss
of statistical persistence in these subgroups has also been related to increased risk of falling1,4. However, while
these findings may reflect the neural control mechanisms of locomotion, one needs to consider additional
mechanisms when interpreting the physiological processes underlying these properties, when gait is modulated
by specific task related factors or contexts11–13. An important effect of context on the statistical persistence of gait
dynamics is observed when walking is entrained to an auditory stimulus. Entrainment is a process that governs
the alignment and coupling of the auditory and motor domains such that the period of the steps align with the
period or phase of the auditory beat, to reach a state of synchronization14–16. This alignment can be understood
in terms of coupled oscillators that achieve synchronization by locking into each other’s period and/or phase15,17,
or alternatively, as the effect of minimizing prediction errors14,15,18. We therefore hypothesized, that the study
of the structure of fluctuations in gait series specifically during a process when the gait is coupled with auditory
stimuli would provide a reflection of the underlying predictive processes and timing control engaged during
auditory-motor coupling and entrainment.hl y
p
g
The structure of fluctuations in the gait time series when walking in time to rhythmic auditory cues has
been studied in healthy individuals and patient with Parkinson’s Disease. These studies reported a change in
gait dynamics from persistent to anti-persistent behavior when gait was paced by a fixed-tempo metronome as
compared to walking without auditory stimuli19–25. We assume that this anti-persistent behavior is a reflection of
the interaction between the auditory and motor systems during entrainment. Gait dynamics can be modulated
by the temporal structure of the stimulus. In healthy controls, the fractal temporal structure of gait dynamics
tends toward the statistical properties of the auditory signals23. www.nature.com/scientificreports/ continuous vs discrete) may influence
gait dynamics due to the underlying temporal mechanisms underpinning auditory-motor coupling. In a continu-
ous stimulus such as a sinusoidal pitch change or the complex temporal structure of music, the perceived beat
is a property of the stimulus’ energetic structure35. In contrast, a discrete stimulus includes singular events (e.g.,
sounds of a metronome) indicating discrete moments in time. Thus, entraining to the continuous structures in
music compared to the discrete structures in a metronome is assumed to affect gait dynamics differently. This
difference is supported by the theoretical frameworks of emergent and event-based timing36 based on discrete
and continuous movements37, demands on cognition38 and by models of the neuronal mechanisms in tracking
acoustic events in time, being the cerebellum and the basal-ganglia39,40. In addition, we also hypothesize that
walking to different tempi would differently influence the gait dynamics. In other words, we assume that the
gait dynamics would reflect the process of coupling the footfalls and beats given different inter-beat-intervals
presented in the experimental conditions (i.e. presented by the different tempi). This difference is supported by
evidence which report that the processing of inter-beat intervals of different durations engage different temporal
processing mechanisms18,36,39,40. Detrended fluctuation analysis
of gait dynamics when entraining
to music and metronomes
at different tempi in persons
with multiple sclerosis A healthy value of alpha in gait is between 0.5 and 1.0 (1 being highly
persistent), and indicates the presence of statistical persistence within the inter-step-intervals1,10. This means
that the inter-step-intervals between consecutive steps are non-random and constant at a long range, with small
deviations still being present across multiple consecutive strides. On the other hand, a value of alpha < 0.5 signifies 1IPEM Institute of Psychoacoustics and Electronic Music, Faculty of Arts and Philosophy, Ghent University,
Gent, Belgium. 2REVAL Rehabilitation Research Center, Faculty of Rehabilitation Sciences, Hasselt University,
Hasselt, Belgium. 3International Laboratory for Brain, Music and Sound Research (BRAMS), Montreal,
Canada. 4Department of Psychology, University of Montreal, Montreal, Canada. 5Normandie Univ, UNICAEN,
INSERM, COMETE, GIP CYCERON, Caen, France. 6Centre for Research on Brain, Language and Music (CRBLM),
Montreal, Canada. 7University of Economics and Human Sciences in Warsaw, Warsaw, Poland. *email:
lousin.moumddjian@ugent.be | https://doi.org/10.1038/s41598-020-69667-8 Scientific Reports | (2020) 10:12934 www.nature.com/scientificreports/ www.nature.com/scientificreports/ Average (AVG) and standard deviation (SD) values of the number of data points (i.e. inter-step-
intervals), resultant vector length, average baseline speed of participants, and the average speed of the study
participants when walking to music and metronomes at different tempi. Study participants comprised of
healthy controls (HC) n = 29, persons with multiple sclerosis (PwMS) n = 28. Table 2. Experimental results of the main and interaction effects of the statistical mixed model analysis of
variance tests and partial eta squared values when walking to music and metronome at different tempi on
co-efficient of variation, alpha and resultant vector length. Outcome measure
Statistical results
Group
Stimuli
Tempi
Assistive device
Baseline speed
Interactions
Co-efficient of variation (%)
P value
ns
0.0007
ns
ns
0.0018
Stimuli*assistive device (p = 0.0054)
Partial eta squared
–
0.00907 –
–
–
0.00575
Alpha (0–1)
P value
ns
< .0001
< 0.0001
ns
ns
Stimuli*baseline speed(p = 0.0350) ; tempi*baseline
speed (p < .0001)
Partial eta squared
–
0.03377 0.11035
–
–
0.00684; 0.03770
Resultant vector length (0–1)
P value
ns
< .0001
< 0.0001
ns
0.0434
ns
Partial eta squared
–
0.02266 0.02217
–
–
/
Percentage change in speed (%)
P value
ns
< .0001
< 0.0001
ns
ns
Group*stimuli (p < .0001); stimuli*baseline speed
(p < .0001)
Partial eta squared
–
0.014
0.0036
/
/
0.0467; 0.0237 Table 2. Experimental results of the main and interaction effects of the statistical mixed model analysis of
variance tests and partial eta squared values when walking to music and metronome at different tempi on
co-efficient of variation, alpha and resultant vector length. Table 2. Experimental results of the main and interaction effects of the statistical mixed model analysis of
variance tests and partial eta squared values when walking to music and metronome at different tempi on
co-efficient of variation, alpha and resultant vector length. Analysis of co‑variance. The analysis of co-variance showed no significances between the data-points col-
lected (i.e. the inter-step-intervals) and alpha (F(1,0.002) = 0.09, p = 0.7604, error 0.02). The average and standard
deviation of data points collected per participant when walking to music and metronomes a the different tempi
are reported in Table 1. Mixed model analysis of variances (ANOVAs). The experimental results of the below ANOVAs in
terms of the p and the partial eta squared values are reported in Table 2. Coefficient of variation (CoV) of inter‑step‑intervals. www.nature.com/scientificreports/ www.nature.com/scientificreports/ Table 1. Average (AVG) and standard deviation (SD) values of the number of data points (i.e. inter-step-
intervals), resultant vector length, average baseline speed of participants, and the average speed of the study
participants when walking to music and metronomes at different tempi. Study participants comprised of
healthy controls (HC) n = 29, persons with multiple sclerosis (PwMS) n = 28. Participants
Walking to music at different tempi
Walking to metronomes at different tempi
0
2
4
6
8
10
0
2
4
6
8
10
Baseline
speed
(m/s)
HC
1.21 ± 0.13
PwMS
0.96 ± 0.23
Average
speed
(m/s)
HC
1.16 ± 0.14
1.18 ± 0.15
1.195 ± 0.17
1.19 ± 0.18
1.21 ± 0.15
1.23 ± 0.18
1.18 ± 0.13
1.20 ± 0.15
1.23 ± 0.19
1.26 ± 0.15
1.25 ± 0.18
1.26 ± 0.19
PwMS
0.91 ± 0.26
0.96 ± 0.27
0.95 ± 0.27
0.96 ± 0.27
0.96 ± 0.27
0.96 ± 0.29
0.92 ± 0.23
0.94 ± 0.25
0.95 ± 0.25
0.97 ± 0.26
1.25 ± 0.18
0.96 ± 0.26
Resultant
vector
length
(0–1)
HC
0.90 ± 0.13
0.87 ± 0.18
0.87 ± 0.16
0.83 ± 0.26
0.80 ± 0.26
0.79 ± 0.23
0.94 ± 0.03
0.94 ± 0.02
0.94 ± 0.03
0.93 ± 0.05
0.90 ± 0.12
0.88 ± 0.17
PwMS
0.74 ± 0.31
0.77 ± 0.26
0.73 ± 0.31
0.72 ± 0.27
0.70 ± 0.32
0.64 ± 0.34
0.83 ± 0.21
0.79 ± 0.25
0.78 ± 0.27
0.77 ± 0.27
0.77 ± 0.26
0.72 ± 0.31
Inter-step-
intervals
(number
of data-
points)
HC
362 ± 22.42 363 ± 41.85 372 ± 32.40
371 ± 39.83 380 ± 45.80 389 ± 41.19 362 ± 20.79 366 ± 21.77 370 ± 35.53 384 ± 25.91 391 ± 21.08
396 ± 21.70
PwMS
324 ± 74.99 324 ± 67.29 329 ± 66.82
339 ± 69.87 322 ± 78.15 343 ± 75.00 326 ± 55.01 327 ± 64.78 335 ± 66.13 340 ± 62.77 348 ± 60.8
354 ± 62.30 Table 1. Average (AVG) and standard deviation (SD) values of the number of data points (i.e. inter-step-
intervals), resultant vector length, average baseline speed of participants, and the average speed of the study
participants when walking to music and metronomes at different tempi. Study participants comprised of
healthy controls (HC) n = 29, persons with multiple sclerosis (PwMS) n = 28. Table 1. Results
S 28 PwMS (mean age ± SD: 53.45 ± 10.61) and 29 healthy controls (HC) (mean age ± SD: 51.77 ± 11.40) completed
the study. PwMS were significantly more impaired than HC on the following motor function tests: the time 25
foot walk (PwMS 7.81 ± 2.11; HC 5.58 ± 0.80 s) and time up and go test (PwMS, 10.94 ± 4.36, HC 6.70 ± 1.17 s). https://doi.org/10.1038/s41598-020-69667-8 Scientific Reports | (2020) 10:12934 | www.nature.com/scientificreports/ Significant main effects were found for stim-
uli (F(1,577.8) = 11.5, p = 0.0007) and baseline speed (F(1,52.3) = 10.8, p = 0.0018). The main effect of stimuli
indicated a significantly higher CoV when walking to music compared to walking to metronomes (t = − 3.39,
p = 0.0007). A significant interaction effect was found for stimuli*assistive device (F(1,577.8) = 7.8, p = 0.0054)
as seen in Fig. 1A. The post-hoc test indicated significantly higher CoV in PwMS using assistive devices when
walking to music compared to walking to metronomes (t = − 3.29, p = 0.0058). A significant negative correla-
tion was found between baseline speed and CoV, indicating the lower the baseline speed, the higher the CoV
(t = − 0.43, p < 0.0001) as seen in Fig. 1B. Fractal dynamics in inter‑step‑intervals (DFA scaling exponent (α)). Significant main effects were
found for stimuli (F(1,563.8) = 23.1, p < 0.0001) indicating a significantly higher α value when walking to music
compared to metronomes, and tempi (F(5,557.2) = 16.7, p < 0.0001), but no significant effect of group was found as
shown in Fig. 2A. Significant interactions were found for tempi*baseline speed (F(5,559.2) = 5.4, p < 0.0001)) and
stimuli*baseline speed (F(1,571.8) = 4.5, p = 0.0350). The post-hoc tests indicated a significantly higher α value:
when walking (i) at the tempo + 8% compared to the tempi + 0, + 2, and + 4% (t = − 4.84, p < 0.0001; t = − 4.75,
p < 0.0001; t = − 4.37, p < 0.0001 respectively), and ii) at the tempo + 10% compared to the tempi + 0, + 2, + 4%
and + 6% (t = − 7.05, p < 0.0001; t = − 6.95, p < 0.0001; t = − 6.58, p < 0.0001; t = − 4.55, p = 0.0015 respectively). https://doi.org/10.1038/s41598-020-69667-8 Scientific Reports | (2020) 10:12934 | www.nature.com/scientificreports/ Figure 1. (A) Coefficient of variation of inter-step-intervals when waking to a metronomes or music at all
experimental tempi in persons with multiple sclerosis (PwMS), divided by patients walking with assistive
devices (n = 8) and patients walking without assistive devices (n = 20). Mean and standard errors of mean
are shown. (B) Co-efficient of variation of inter-step-intervals plotted against baseline speed (m/s) at all
experimental tempi in all participants (healthy controls and PwMS) when walking to both music and
metronomes. Figure 1. DFA scaling exponent (α): original vs. surrogate time series.
The following effects were found: (i) A significant effect for time series X2(1) = 24.25, p < 0.0001, indicating significantly higher alpha values
for the original time series.ifi (i) A significant effect for time series X2(1) = 24.25, p < 0.0001, indicating significantly higher alpha values
for the original time series.ifi (i) A significant effect for time series X2(1) = 24.25, p < 0.0001, indicating significantly higher alpha values
for the original time series.ifi g
(ii) A significant effect for stimuli in the original time series X2(1) = 14.41, p = 0.0001, but no significance for
stimuli in the surrogate time series X2(1) = 3.01, p = 0.00827. g
(ii) A significant effect for stimuli in the original time series X2(1) = 14.41, p = 0.0001, but no significance for
stimuli in the surrogate time series X2(1) = 3.01, p = 0.00827. g
p
(iii) A significant effect for music in the original time series compared to the surrogate time series
X2(1) = 29.41, p < 0.0001, but no significance for metronomes between the original and surrogate time
series X2(1) = 2.38, p = 0.1228.i g
p
(iii) A significant effect for music in the original time series compared to the surrogate time series
X2(1) = 29.41, p < 0.0001, but no significance for metronomes between the original and surrogate time
series X2(1) = 2.38, p = 0.1228. p
(iv) No significances for the groups between the original and surrogate time series X2(1) = 0.40, p = 0.5377. ( )
A i
ifi
t ff
t f
t
pi i th
i i
l ti
i
X2(5)
60 88 p<0 0001 b t
i
ifi
f p
(iv) No significances for the groups between the original and surrogate time series X2(1) = 0.40, p = 0.5377. (v) A significant effect for tempi in the original time series X2(5) = 60.88, p < 0.0001, but no significance for
tempi in the surrogate time series X2(5) = 1.64, p = 0.8969.if (iv) No significances for the groups between the original and surrogate time series X (1) = 0.40, p = 0.5377. www.nature.com/scientificreports/ (A) Coefficient of variation of inter-step-intervals when waking to a metronomes or music at all
experimental tempi in persons with multiple sclerosis (PwMS), divided by patients walking with assistive
devices (n = 8) and patients walking without assistive devices (n = 20). Mean and standard errors of mean
are shown. (B) Co-efficient of variation of inter-step-intervals plotted against baseline speed (m/s) at all
experimental tempi in all participants (healthy controls and PwMS) when walking to both music and
metronomes. The significant interaction between tempi and baseline speed is shown in Fig. 2B; at higher tempi, persons with
higher baseline speed are reported to have higher alpha. The significant interaction between stimuli and baseline
speed is shown in Fig. 2C; persons with higher speed are reported to have a higher alpha when walking to music
compared to metronomes. The significant interaction between tempi and baseline speed is shown in Fig. 2B; at higher tempi, persons with
higher baseline speed are reported to have higher alpha. The significant interaction between stimuli and baseline
speed is shown in Fig. 2C; persons with higher speed are reported to have a higher alpha when walking to music
compared to metronomes. DFA scaling exponent (α): original vs. surrogate time series.
The following effects were found: (v) A significant effect for tempi in the original time series X2(5) = 60.88, p < 0.0001, but no significance for
tempi in the surrogate time series X2(5) = 1.64, p = 0.8969.if g
(vi) Significant effects for tempi 8 and 10% between the original and surrogate time series X2(1) = 25.23,
p < 0.0001; X2(1) = 50.71, p < 0.0001 respectively, but no significance for tempi 0, 2, 4 and 6 between
the original and surrogate time series X2(1) = 0.85, p < 0.3562; X2(1) = 0.75, p < 0.3879; X2(1) = 0.0005,
p < 0.9852; X2(1) = 3.54, p < 0.0599 respectively. Figure 3 shows graphical representations of these results. Supplementary Fig. 2 additionally provides visu-
alization of the above data in three different perspectives. Synchronization consistency. Significant main effects were found for stimuli (F(1,579.5) = 27.2,
p < 0.0001), tempi (F(5,579.1) = 5.8, p < 0.0001), as previously reported29, and baseline speed (F(1,51.7) = 4.3,
p = 0.0434), but no significant effect of group was found. Post-hoc comparisons showed more consistent syn-
chronization: (i) at tempi 0 and + 2% compared to + 10% (t = 4.31, p = 0.0003; t = 4.08, p = 0.0007, respectively),
and (ii) with a metronome rather than with music (t = 5.22, p < 0.0001) as seen in Table 1. Percentage change in speed. Significant main effects were found for stimuli (F(1,591.1) = 18.9, p < 0.0001)
and tempi (F(5,591.1) = 13.9, p < 0.0001), as previously reported29, but no significant effect of group was found. Significant interaction effects were found for group*stimuli (F(1,591.0) = 16.3, p < 0.0001) and stimuli*baseline
speed (F(1,591.1) = 17.1, p < 0.0001). Post-hoc comparisons showed: (i) an increase of speed when walking Scientific Reports | (2020) 10:12934 | https://doi.org/10.1038/s41598-020-69667-8 www.nature.com/scientificreports/ Figure 2. (A) The DFA exponent (α) of inter-step-intervals when walking to a metronome or music at different
tempi in healthy controls and persons with multiple sclerosis (PwMS). Mean and standard errors of mean are
shown. (B) The DFA exponent (α) of inter-step-intervals plotted against baseline speed (m/s) at the different
experimental tempi in all participants when walking to both music and metronomes. Lines of fit are shown,
the figures show a linear regression with confidence intervals for the variables on the x and y axis. DFA scaling exponent (α): original vs. surrogate time series.
The following effects were found: (C) The
DFA exponent (α) of inter-step-intervals plotted against baseline speed (m/s) when walking to music and
metronomes for all participants. Figure 2. (A) The DFA exponent (α) of inter-step-intervals when walking to a metronome or music at different
tempi in healthy controls and persons with multiple sclerosis (PwMS). Mean and standard errors of mean are
shown. (B) The DFA exponent (α) of inter-step-intervals plotted against baseline speed (m/s) at the different
experimental tempi in all participants when walking to both music and metronomes. Lines of fit are shown,
the figures show a linear regression with confidence intervals for the variables on the x and y axis. (C) The
DFA exponent (α) of inter-step-intervals plotted against baseline speed (m/s) when walking to music and
metronomes for all participants. to metronomes compared to music (t = 4.35, p < 0.0001), (ii) a lower increase in speed when walking to 0%
compared to + 4, + 6, + 8 and + 10% (t = − 4.15, p = 0.0005; t = − 5.87, p < 0.0001; t = − 6.72, p < 0.0001; t = − 6.57,
p < 0.0001, respectively) and 2% compared to + 8 and + 10% (t = − 4.03, p = 0.0009; t = − 3.90, p = 0.0015) and (iii)
an increase of speed in healthy controls when walking to metronomes compared to walking to music (t = 5.97,
p < 0.0001), as seen in Fig. 4. p
g
Table 1 reports on the average and standard deviation values of the number of data points (i.e. inter-step-
intervals), resultant vector length, average baseline speed of participants, and the average speed of participants
when walking to music and metronomes at the six different tempi. Correlation analysis. No correlations were found between the resultant vector length and the DFA scaling
exponent (α) when walking to music or metronomes at all tempi in both HC and PwMS. Discussion In this study, we aimed to understand the underlying process of entraining steps to beats in music and met-
ronomes and at different tempi in PwMS compared to HC. To reach this end, we applied DFA on inter-step-
intervals to examine the gait dynamics when PwMS and HC were asked to synchronize their steps to metronomes
and music at their preferred walking cadence up to + 10%, in increments of 2%. The analysis of the original and
surrogate time series confirmed that the reported statistical persistency measures were not due to chance, but
were a reflection of participants’ gait dynamics. In addition, the variability of step duration (coefficient of vari-
ation; CoV), synchronization consistency and walking speed were examined.l y
y
g p
Studies investigating inter-stride-interval fluctuations reported the presence of anti-persistent gait dynam-
ics as a result of the task-related context. These studies explained that the anti-persistent gait dynamics were
observed as a result of deviations in the gait time series caused by participants applying fast corrections to their
gait because of the experimental task-related context12,13. Similarly, we can interpret our results on gait dynamics
in function of the imposed experimental task-related context. Our experimental design provided two manipula-
tions, of the type and of the tempi of the auditory stimulus. The task-related contexts therefore were: coupling Scientific Reports | (2020) 10:12934 | https://doi.org/10.1038/s41598-020-69667-8 www.nature.com/scientificreports/ Figure 3. The DFA exponent (α) of the original and surrogate inter-step-intervals time series when walking
to a metronome or music at different tempi in healthy controls and persons with multiple sclerosis. Mean and
standard errors of mean are shown. Figure 3. The DFA exponent (α) of the original and surrogate inter-step-intervals time series when walking
to a metronome or music at different tempi in healthy controls and persons with multiple sclerosis. Mean and
standard errors of mean are shown. Figure 3. The DFA exponent (α) of the original and surrogate inter-step-intervals time series when walking
to a metronome or music at different tempi in healthy controls and persons with multiple sclerosis. Mean and
standard errors of mean are shown. to music and metronomes (continuous compared to discrete auditory structures) and coupling to six different
tempi (different inter-beat-intervals).f f
In this study, we found no differences in gait dynamics between PwMS and HC when walking to the auditory
stimuli and tempi. Discussion To our knowledge, only one study investigated gait dynamics during walking using the DFA
method in PwMS with mild and moderate impairment and HC7. Albeit having included a small sample (10
persons per group), their results revealed no differences in gait dynamics between PwMS and HC7 during walk-
ing. Although we cannot directly compare the results of our respective studies due to the different task-related
constraint of our walking tasks, their results assist in the interpretation of our results. In other words, when gait
dynamics between PwMS and HC are not different at baseline, one can assume that the gait dynamics observed
in this study in both groups reflects the auditory-motor coupling process. Our first hypothesis therefore can be
endorsed, that both groups engage in auditory-motor coupling using the same underlying processes (explained
further in the following paragraphs). g
g
We have however found two patterns of gait dynamics in all participants depending on the auditory stimuli
and tempi used during the coupling. Given these differences, our second and third hypotheses can be supported;
the auditory-motor coupling processes differ when entraining to music and metronomes and to the different
tempi.i First, with metronomes, we found significantly lower alpha values compared to when walking to music, an
indication of a trend towards an anti-persistent gait dynamics. This pattern of gait dynamics indicated that the
inter-step-intervals between consecutive steps occurred as a non-structured pattern over time. Second, with
music, we found significantly higher alpha values compared to when walking to metronomes, indicating a
transition to a more persistent gait dynamics. This gait dynamics indicated that the inter-step-intervals between
consecutive steps occurred with some structured pattern over time. In addition, when walking to higher tempi
to both stimuli, we found a transition to a higher persistent gait dynamics indicated by the significantly higher
value of alpha at the + 8% and + 10% as compared to the lower percentages. This gait dynamics indicated that the
inter-step-intervals between consecutive steps occurred with more structure over time. Our results of the trend towards anti-persistent gait dynamics when walking to metronomes are consistent
with several studies which reported the same finding when gait was paced by metronomes19,20,22–25. Two explana-
tions could be given to our results. The first explanation concerns the presence of variability in inter-beat-intervals
in music versus the absence of variability in isochronous metronomes. Discussion Earlier evidence presented in several Scientific Reports | (2020) 10:12934 | https://doi.org/10.1038/s41598-020-69667-8 www.nature.com/scientificreports/ Figure 4. Percentage change in walking speed relative to baseline walking speed when walking to a metronome
or music at different tempi in healthy controls and persons with multiple sclerosis. Mean and standard errors of
mean are shown. Figure 4. Percentage change in walking speed relative to baseline walking speed when walking to a metronome
or music at different tempi in healthy controls and persons with multiple sclerosis. Mean and standard errors of
mean are shown. studies on HC and patients with Parkinson’s Disease reported that gait dynamics evolve towards a persistent state
once participants get paced with cues that have a random variability23, biological variability26 and interactive
biological variability41,42 compared to isochronous metronomes. The second explanation concerns the discrete
and continuous energies in the auditory signals. The period of a metronome tick consists of an energy burst
followed by a time interval without energy, while the period in music consists of a continuous acoustic energy
flow35. This energy flow may facilitate the persistence of the gait dynamics through entrainment43. The timing
mechanisms underlying the temporal processing of the discrete compared to continuous structures in the audi-
tory stream can be referred to as event-based versus emergent timing processes44 respectively.f y
g
g p
p
y
We propose that alpha, as an outcome measure, can explain differences in control mechanisms during the
auditory-motor coupling. Below, we provide an explanation for this proposal. During the coupling, a dynamical
process (termed entrainment) is engaged through which steps and beats become aligned. The alignment is made
possible through prediction error minimization mechanisms. Prediction error here can be seen as the timing
differences between steps and beats. Thus, prediction error minimization describes the dynamic process in which
the timing differences between steps and the beats become minimized, to reach an alignment. We propose that
this dynamic process of prediction error minimization during the coupling is reflected by the gait dynamics14,15,
measured by alpha. Our results imply that both PwMS and HC engage in this dynamic process of prediction
error minimization during the coupling as no group differences were observed in our results.f f
However, a difference in gait dynamics (alpha) was found between walking to music compared to metro-
nomes. www.nature.com/scientificreports/ In this case, there will be less immediate adaptation to the beat because those noisy
perceptions of the beat will be ignored. Consequently, in this case, we propose that the gait dynamics would
display a transition towards a more persistent structure, reflecting the fractal structure closer to a natural gait
dynamics1, not subjected to a task-related context. On the other hand, the adaptation gain could be high if both
the step-to-beat alignment and beat perception would have a low precision. In this case the adaptation to the beat
would be more immediate, which would display a transition towards a lower persistent gait dynamics. We there-
fore propose that at higher tempi, HC and PwMS responded with low adaptation gain due to high precision of
predicting and performing the step-to-beat alignment, reflected by a transition towards persistent gait dynamics. l
With our results, we note that the fractal properties of the inter-step-intervals provided valuable information
in regards to timing mechanisms employed during auditory-motor coupling. Specifically, it allowed to interpret
the interactive dynamics of a coupled system, whereas gait dynamics had a trend towards anti-persistence when
entraining steps to beats metronomes, a trend towards persistence when entraining steps to beats in music,
and a trend towards higher persistence when entraining steps to beats of both stimuli at the + 8% and + 10% of
preferred walking cadence tempo. For clinical practice, we recommend to couple steps to beats in music, as the predictive mechanisms engaged
when coupling to music resulted in a more persistent and a less rigid gait dynamics when compared to walking
to metronomes. In addition, the results of this study contributes to fundamental explanations of the process of
coupling walking to music and metronomes at different tempi in viewpoint of predictive coding and timing
control. These findings can be used to design, fine-tune, specify and individualize the content of auditory-motor
coupling interventions. p
g
Some methodological considerations apply for our experimental design, such as our square (4.5 by 6 m)
walking track. We acknowledge that this path involved an environmental constraint, which could have had an
effect on alpha48. Yet, we believe that this did not have an effect on the comparison of our experimental results
because the walking path was standardized for all experimental conditions (i.e. the effect of the environmental
constraint was consistent in all experimental conditions). www.nature.com/scientificreports/ PwMS had a less rigid gait, relying less on the assistive device for movement and more on their own walking
when entraining to music compared to metronomes.if Another finding from this study concerns gait dynamics (alpha) at different tempi. One explanation for the
differences of alpha observed between the high and low tempi is walking speed, as studies have shown an associa-
tion between gait speed, gait variability and gait dynamics46. Our results also showed that at higher tempi, those
participants with higher baseline speeds had higher alpha values. In addition, our results show that the percent-
age change of speed from baseline was significantly different for the conditions 0% compared to + 4, + 6, + 8
and + 10%, and for condition 2 compared to + 6, + 8 and + 10%. However, it is important to note, that a different
pattern was observed by alpha (significantly higher persistency observed at 8 and 10% compared to 0–6% tempi). Furthermore, if the differences of alpha observed were due to walking speed alone, one would expect finding
differences in alpha values between those participants with higher baseline speeds and lower baseline speeds
across all tempi, while this was not the case. We therefore suggest that walking speed, although an important
factor, does not provide a full explanation of our results. Below, we provide a complementary proposal based on
the process of prediction error minimization engaged during the coupling. p
p
g g
g
p
g
At higher tempi, we observed a transition to a higher persistent gait dynamics in all participants. At higher
tempi, perception of beats becomes more variable. This variance could thereby have an impact on the alignment
of steps-to-beats by affecting the individual adaptation processes that persons engage in to achieve step-to-beat
alignment. Here, the individual adaptation processes relates to the concepts of precision and adaptation gain in
predictive coding47. The adaptation gain determines the degree in which the step-to-beat alignment error drives
the adaptation towards alignment. The adaptation gain can be defined in terms of a balance between the preci-
sion of predicting and performing the step-to-beat alignment versus the precision of perceiving the beat. The
adaptation gain could be low if the step-to-beat alignment would have a high precision, while beat perception
would have a low precision. www.nature.com/scientificreports/ Nevertheless, for future work, we advise over-ground
walking conditions with avoidance of 90 degree or 180 sharp turns48. A second methodological consideration in
this exploratory study is the length of the time series. Obtaining longer time series by increasing the trial dura-
tions are advised in future studies. We also advise future studies investigating walking at different tempi to take
into account the possible difference in the length of the time series as a function of walking to different tempi
(especially when persons are instructed to synchronize). We acknowledge this is a methodological consideration
in this current study, and applied transparency by reporting the averages and standard deviations of the length
of the time series in our experimental conditions in Table 1. A third methodological consideration is the lack of
reporting on the expanded disability status scale due to an incomplete dataset because of the recruitment process. Discussion With metronomes, we believe that discrete metronome ticks led to clear-cut attention of errors in step-
to-beat alignment35 that got instantly corrected in order for alignment to occur35. This instant response was
reflected in the gait dynamics, and it corresponded to an anti-persistent structure of inter-step-intervals. As the
corrections occurred instantly, the intervals had a non-structured (more random) characteristic. This manner
of step-to-beat alignment can also explain the significantly higher synchronization constancy when coupling to
metronomes, higher walking speed with a lower variability of the step duration. With music, however, we believe that errors in step-to-beat alignment were less noticeable due to the con-
tinues inter-beat-interval structure, making the response to errors less immediate, but more structured over
time35. This structure was again reflected in the gait dynamics and it corresponded to a persistent structure of
inter-step-intervals. The explanations above are supported by existing evidence of different control mechanisms
of timing40, event based and emergent based control of timing44 given the discrete and continuous stimulus
structures35,37,45. These explanations could also explain our previous behavioral findings showing that walking
to music was perceived less cognitively fatiguing than walking to a metronome, even when the coupling was
sustained longer with music30. This manner of step-to-beat alignment can also explain the significant lower
synchronization constancy reported when coupling to music as compared to metronomes, the lower gait speed
and higher variability of step durations. It is noteworthy, that the reported level of synchronization consistency
when walking to music was adequately high. In addition, the variability of step durations were significantly
higher for those participants using an assistive device when walking to music. This finding could indicate that Scientific Reports | (2020) 10:12934 | https://doi.org/10.1038/s41598-020-69667-8 www.nature.com/scientificreports/ www.nature.com/scientificreports/ g
g
g
Persistent and anti-persistent structures in step interval time series were examined using Detrended Fluc-
tuation Analysis (DFA). DFA is a method that removes local trends, thus it is less likely to be affected by non-
stationarity in the time series2. DFA computes the mean square roots of (linearly) detrended residuals, F(n) , of
the integrated time series over a range of equal, non-overlapping window sizes n . The scaling exponent α is then
estimated from the slope of the linear relationship between log[F(n)] and log(n) . Restricted range of window
sizes were used, from n = 4 steps to n = N /4 steps53,54 (N being the total number of steps). An α > 0.5 indicates
statistical persistence, with fluctuations in one direction followed by fluctuations in the same direction. An α < 0.5
indicate anti-persistence, with subsequent fluctuations in the opposite direction. An α = 0.5 indicates uncorrelated
noise, with subsequent fluctuations equally in either direction. In the context of control, gait variables that are
not tightly regulated exhibit strong persistence, while those that are more tightly regulated exhibit either uncor-
related or anti-persistent fluctuations ( α ≤ ~ 0.5). Details of the methodology are published elsewhere2,8,12,55–57. To determine if relatively simple random processes might account for the statistical properties seen within the
experimental data, surrogate time series were generated by randomly shuffling each original time series collected
per participant and per experimental condition one hundred times58,59. In other words, one hundred surrogate
time-series were generated per participant and per experimental condition. Thereafter, per participant and
experimental condition, one mean alpha value was calculated from the distribution of the surrogate time series. Synchronization with the auditory stimuli. Synchronization to the stimulus beat was measured by cal-
culating the resultant vector length (RVL). The RVL is a measure that indicates the consistency of timing differ-
ences between two periodic signals. Here, these were the timing difference of the individual steps relative to their
closest beats. These timing differences were calculated and expressed as phase angles. Thereafter, the average of
the sine and cosine coordinates of all phase angles were calculated, in order to obtain the RVL60,61. The RVL rep-
resents how well the participants matched their steps to the beats over time. RVL is expressed as a number from
0 to 1. www.nature.com/scientificreports/ tered in the clinical trials.gov registry (NCT03281330). All methods were carried out in accordance with relevant
guidelines and regulations. tered in the clinical trials.gov registry (NCT03281330). All methods were carried out in accordance with relevant
guidelines and regulations. Materials and experimental procedure. Participants underwent testing in two sessions: in a clinical
descriptive testing session, and an experimental testing session, held one week apart. During the descriptive test-
ing session, descriptive demographic and the following clinical motor tests were collected: Time 25 Foot Walk
test49 for mobility and Time Up and Go test50 for balance. During the experimental session, first, participants
were asked to walk in their comfort tempo in a square of 4.5 by 6 m three times for one minute, to determine
the average preferred walking cadence of the day. This was followed by a familiarization task, using the song
‘Sanctum’ by the artist ‘Shades of the Abyss’ to instruct participants to synchronize by stepping to the beat. This
song was chosen because of its clear beats. A similar familiarization was conducted with metronomes as well. Participants then walked three minutes per six tempi to the beat of music and to isochronous metronomes
blocks. The tempi were: 0%, + 2%, + 4%, + 6%, + 8%, + 10% of their preferred walking cadence. The tempi and
the stimuli blocks were randomized. The D-jogger technology51 was used to provide the auditory stimuli at the
required conditions. This technology consisted of a software, headphones (Sennheiser, Germany) and two wire-
less inertial measurement units strapped at the ankles for measuring cadence and step times (iPod, Apple, USA),
sampling kinematic gait data at 100 Hz. Auditory stimuli. Participants were asked to choose one of six available genres to walk to in the music block
from: disco, instrumental, pop, softpop, poprock and variety. See Buhmann et al.52 for details of the music data-
base generation. Isochronous digital metronomes were used for the metronome block. Supplementary Fig. 1
illustrates the coefficient of variation and standard deviation of the inter-beat-intervals across time of the used
stimuli. Inter‑step‑intervals: data processing. Matlab (MathWorks Inc., USA) was used to process the data. The
onset of steps logged by the D-jogger system was used to derive step interval time series. The variability of inter-
step-intervals was obtained by calculating the coefficient of variation (CoV; a tempo-independent variable) of
each time series as a percentage using the following formula: ((standard deviation/mean)*100). Methods
Participants. Participants. 30 Persons with Multiple Sclerosis (PwMS) and 30 healthy controls (HC) were included, of
which 28 PwMS (mean age ± SD: 53.45 ± 10.61) and 29 HC (mean age ± SD: 51.77 ± 11.40) completed the study. Once recruited, the informed consents were signed and participants were tested for the inclusion criteria, which
were: (a) a diagnosis of MS (> 1 year), (b) no exacerbation in the last month, (c) an average comfortable walk-
ing speed between 0.4 and 1.4 m per second (m/s), and (d) being older than 18 years of age. Participants were
excluded if they were pregnant, had either hearing or cognitive impairment hindering the understanding of
instructions. This study is a secondary analysis on a previously reported observational non-blinded case–control
study, published in Neuro-Rehabilitation and Neural Repair29. The study was approved by the Medical Ethi-
cal Committees of universities Hasselt and Ghent (Belgium) and multiple sclerosis centers (The national MS
center,Rehabilitation and MS center Overpelt) on November 23rd, 2016 (B670201629797). The study was regis- https://doi.org/10.1038/s41598-020-69667-8 Scientific Reports | (2020) 10:12934 | www.nature.com/scientificreports/ www.nature.com/scientificreports/ factor; stimuli (music and metronome) experimental tempi (0, + 2, + 4, + 6, + 8 and + 10% of the preferred walk-
ing cadence) as within-subjects factor. In addition, two co-variates were added to the analysis, these were assis-
tive devices (yes (n = 8) and no (n = 20)) and the baseline walking speed of participants. These co-variates were
added to the analysis as confounding variables, as assistive devices and walking speed have shown to effect gait
variability46,63. Multiple comparisons Tukey’s tests were further performed as post-hoc tests. For the ANOVA’s
above, assumptions of normality and homoscedasticity have been verified by visually examining each model’s
residual quantile and predicted plots. In addition, six Kruskal–Wallis tests were conducted to statistically exam-
ine differences in alpha values between the time series (original and surrogate), stimuli (music and metronome),
group (HC and PwMS) and conditions (0, 2, 4, 6, 8, 10%). Bonferroni’s correction was applied to correct for mul-
tiple comparisons. The following were examined: comparison between the time series; comparison of the time
series between stimuli; comparison of the stimuli between the time series; comparison of the groups between
the time series; comparison of the tempi between the time series and; comparison of the time series between
the tempi. To further examine the relationship between synchronization and α, Spearman’s correlation coef-
ficients were calculated for PwMS and HC when walking to music or metronomes at the six different tempi. All
analyses were performed using SAS JMP Pro 13.2.0 (copyright SAS Institute Inc., USA). The significance level
was set at p < 0.05. Given the experimental design (i.e. the fixed time of walking three minutes per tempi) it was
expected for cadence to be modified across the different tempi, and as a result, the number of steps collected (i.e. data points) could differ across tempi. To control correctly for the discrepancy, prior to running the ANOVAs
described above, an analysis of covariance was applied by fitting a standard least squares to the outcome alpha by
the total number of data points collected per trial. Data availability
Th d
d The datasets generated during and/or analyzed during the current study are not publicly available due to privacy
regulations but are available from the corresponding author on reasonable request until 2022. Received: 31 October 2019; Accepted: 15 July 2020 Received: 31 October 2019; Accepted: 15 July 2020 Referencesf 1. Hausdorff, J. M. Gait dynamics, fractals and falls: Finding meaning in the stride-to-stride fluctuations of human walking. Hum. Mov. Sci. 26, 555–589. https://doi.org/10.1016/j.humov.2007.05.003 (2007).f 1. Hausdorff, J. M. Gait dynamics, fractals and falls: Finding meaning in the stride-to-stride fluctuations of human walking. Hum. Mov. Sci. 26, 555–589. https://doi.org/10.1016/j.humov.2007.05.003 (2007).f p
g
j
2. Hausdorff, J. M., Peng, C. K., Ladin, Z., Wei, J. Y. & Goldberger, A. L. Is walking a random walk? Evidence for long-range correla
tions in stride interval of human gait. J. Appl. Physiol. 1985(78), 349–358. https://doi.org/10.1152/jappl.1995.78.1.349 (1995).fl g
pp
y
p
g
j pp
3. Hausdorff, J. M. et al. Fractal dynamics of human gait: Stability of long-range correlations in stride interval fluctuations. J. Appl
Physiol. 1985(80), 1448–1457. https://doi.org/10.1152/jappl.1996.80.5.1448 (1996).f y
g
j
4. Hausdorff, J. M., Rios, D. A. & Edelberg, H. K. Gait variability and fall risk in community-living older adults: A 1-year prospective
study. Arch. Phys. Med. Rehabil. 82, 1050–1056. https://doi.org/10.1053/apmr.2001.24893 (2001).f 5. Hausdorff, J. M., Cudkowicz, M. E., Firtion, R., Wei, J. Y. & Goldberger, A. L. Gait variability and basal ganglia disorders: Stride-
to-stride variations of gait cycle timing in Parkinson’s disease and Huntington’s disease. Mov. Disord. 13, 428–437. https://doi. org/10.1002/mds.870130310 (1998).f g
(
)
6. Hausdorff, J. M. et al. Altered fractal dynamics of gait: Reduced stride-interval correlations with aging and Huntington’s disease
J. Appl. Physiol. 1985(82), 262–269. https://doi.org/10.1152/jappl.1997.82.1.262 (1997).f 7. Kaipust, J. P., Huisinga, J. M., Filipi, M. & Stergiou, N. Gait variability measures reveal differences between multiple sclerosis
patients and healthy controls. Mot. Control 16, 229–244 (2012). y
8. Peng, C. K. et al. Long-range anticorrelations and non-Gaussian behavior of the heartbeat. Phys. Rev. Lett. 70, 1343–1346. https
://doi.org/10.1103/PhysRevLett.70.1343 (1993).i g
y
(
)
9. Peng, C. K., Havlin, S., Stanley, H. E. & Goldberger, A. L. Quantification of scaling exponents and crossover phenomena in non-
stationary heartbeat time series. Chaos 5, 82–87. https://doi.org/10.1063/1.166141 (1995). y
p
g
0. Maraun, D., Rust, H. W. & Timmer, J. Tempting long-memory—on the interpretation of DFA results. Nonlinear Process. Geophy
11, 495–503. https://doi.org/10.5194/npg-11-495-2004 (2004). p
g
pg
(
)
11. Stergiou, N., Buzzi, U.H., Kurz, M.J., Heidel, J. in Innovative analyses of human movement. 63–90 (Human Kinetics, 2004).l 11. Stergiou, N., Buzzi, U.H., Kurz, M.J., Heidel, J. in Innovative a g
y
f
12. Dingwell, J. B. & Cusumano, J. P. www.nature.com/scientificreports/ Higher numbers indicate more consistent synchronization, meaning that the timing differences between
individual steps and beats become concentrated around a single value. Lower numbers indicate inconsistency of
synchronization, meaning the timing differences between individual steps and beats become highly variable over
time. Please refer to a methodological review for details16. Spatio‑temporal gait parameters. Participants were equipped with three OPAL wearable sensors
(Mobility lab, APDM, USA), two strapped on their ankles, and one strapped on the sternum, to measure the
spatiotemporal gait parameters62. Spatio-temporal gait measurements, e.g. cadence and speed were obtained
during the baseline walking conditions (when walking without any auditory stimuli) and the experimental walk-
ing conditions (when walking to the auditory stimuli). Thereafter, percentage change in speed was calculated as
a percentage change between the average speed of the participant at baseline compared to their average speed
during the experimental conditions. Statistical analysis. Four mixed model analysis of variances (ANOVA) were performed, one on the pri-
mary outcome measure (α value) of the original time series and three for secondary outcome measures (RVL,
CoV, percentage change in speed). The independent variables were: group (HC and PwMS) as between-subjects Scientific Reports | (2020) 10:12934 | https://doi.org/10.1038/s41598-020-69667-8 www.nature.com/scientificreports/ www.nature.com/scientificreports/ www.nature.com/scientificreports/ 3. Marmelat, V., Torre, K., Beek, P. J. & Daffertshofer, A. Persistent fluctuations in stride intervals under fractal auditory stimulation
PLoS ONE 9, e91949. https://doi.org/10.1371/journal.pone.0091949 (2014).l g
j
24. Terrier, P. Fractal fluctuations in human walking: Comparison between auditory and visually guided stepping. Ann. Biomed. 44, 2785–2793. https://doi.org/10.1007/s10439-016-1573-y (2016).l l
g
44, 2785–2793. https://doi.org/10.1007/s10439-016-1573-y (2016) p
g
y (
)
5. Choi, J. S., Kang, D. W., Seo, J. W. & Tack, G. R. Fractal fluctuations in spatiotemporal variables when walking on a self-paced
treadmill. J. Biomech. 65, 154–160. https://doi.org/10.1016/j.jbiomech.2017.10.015 (2017). l
p
p
ech. 65, 154–160. https://doi.org/10.1016/j.jbiomech.2017.10.015 g
j j
26. Dotov, D. G. et al. Biologically-variable rhythmic auditory cues are superior to isochronous cues in
in Parkinson’s disease. Gait Post. 51, 64–69. https://doi.org/10.1016/j.gaitpost.2016.09.020 (2017)ff 6. Dotov, D. G. et al. Biologically-variable rhythmic auditory cues are superior to isochronous cues in fostering natural gait variability
in Parkinson’s disease. Gait Post. 51, 64–69. https://doi.org/10.1016/j.gaitpost.2016.09.020 (2017).ff g
j g
7. Leow, L. A., Parrott, T. & Grahn, J. A. Individual differences in beat perception affect gait responses to low- and high-groove music
Front. Hum. Neurosci. 8, 811. https://doi.org/10.3389/fnhum.2014.00811 (2014).hl f
p
pf
urosci. 8, 811. https://doi.org/10.3389/fnhum.2014.00811 (2014).hl g
8. Hunt, N., McGrath, D. & Stergiou, N. The influence of auditory-motor coupling on fractal dynamics in human gait. Sci. Rep. 4
5879. https://doi.org/10.1038/srep05879 (2014). p
g
p
(
)
29. Moumdjian, L. et al. Walking to music and metronome at various tempi in persons with multiple sclerosis: A basis for Rehab
tion. Neurorehabil. Neural Repair 33, 464–475. https://doi.org/10.1177/1545968319847962 (2019).f 30. Moumdjian, L. et al. Continuous 12 min walking to music, metronomes and in silence: Auditory-motor coupling and its effects
on perceived fatigue, motivation and gait in persons with multiple sclerosis. Mult. Scler. Relat. Disord. 35, 92–99. https://doi. org/10.1016/j.msard.2019.07.014 (2019).hf g
j
31. Seebacher, B., Kuisma, R., Glynn, A. & Berger, T. The effect of rhythmic-cued motor imagery on walking, fatigue and quality of
life in people with multiple sclerosis: A randomised controlled trial. Mult. Scler. 23, 286–296. https://doi.org/10.1177/1352458516
644058 (2017). (
)
32. Martin, C. L. et al. Gait and balance impairment in early multiple sclerosis in the absence of clinical disability. Mult. Scler. 12,
620–628. https://doi.org/10.1177/1352458506070658 (2006).ii g
3. Filli, L. et al. Profiling walking dysfunction in multiple sclerosis: Characterisation, classification and progression over time. Sci
Rep. 8, 4984. https://doi.org/10.1038/s41598-018-22676-0 (2018).f p
p
g
4. Feys, P. et al. www.nature.com/scientificreports/ Spatio-temporal gait parameters change differently according to speed instructions and walking history in MS patients
with different ambulatory dysfunction. Mult. Scler. Relat. Disord. 2, 238–246. https://doi.org/10.1016/j.msard.2013.01.004 (2013) f
35. Hogan, N. & Sternad, D. On rhythmic and discrete movements: Reflections, definitions and implications for motor control. Exp. Brain Res. 181, 13–30. https://doi.org/10.1007/s00221-007-0899-y (2007). p
g
y
36. Lemoine, L. & Delignieres, D. Detrended windowed (lag one) autocorrelation: A new method for distinguishing between event-
based and emergent timing. Q. J. Exp. Psychol. (Hove) 62, 585–604. https://doi.org/10.1080/17470210802131896 (2009). g
g
p
y
p
g
7. Huys, R., Studenka, B. E., Rheaume, N. L., Zelaznik, H. N. & Jirsa, V. K. Distinct timing mechanisms produce discrete and continu- g
g
p
y
g
37. Huys, R., Studenka, B. E., Rheaume, N. L., Zelaznik, H. N. & Jirsa, V. K. Distinct timing mechanisms produce d
ous movements PLoS Comput Biol 4 e1000061 https://doi org/10 1371/journal pcbi 1000061 (2008) 37. Huys, R., Studenka, B. E., Rheaume, N. L., Zelaznik, H. N. & Jirsa, V. K. Distinct timing mechanisms produ
ous movements. PLoS Comput. Biol. 4, e1000061. https://doi.org/10.1371/journal.pcbi.1000061 (2008). 7. Huys, R., Studenka, B. E., Rheaume, N. L., Zelaznik, H. N. & Jirsa, V. K. Distinct timing mechanisms produce discrete and continu-
ous movements. PLoS Comput. Biol. 4, e1000061. https://doi.org/10.1371/journal.pcbi.1000061 (2008).l 38. Decker, L. M., Cignetti, F. & Stergiou, N. Executive function orchestrates regulation of task-relevant gait fluctuations. Gait Post. 38, 537–540. https://doi.org/10.1016/j.gaitpost.2012.12.018 (2013). p
g
j g
p
39. Schwartze, M., Tavano, A., Schroger, E. & Kotz, S. A. Temporal aspects of prediction in audition: Cortical and subcortical neural
mechanisms. Int. J. Psychophysiol. 83, 200–207. https://doi.org/10.1016/j.ijpsycho.2011.11.003 (2012). 0. Lewis, P. A. & Miall, R. C. Distinct systems for automatic and cognitively controlled time measurement: Evidence from neuroimag-
ing. Curr. Opin. Neurobiol. 13, 250–255 (2003). 41. Hove, M. J., Suzuki, K., Uchitomi, H., Orimo, S. & Miyake, Y. Interactive rhythmic auditory stimulation reinstates natura
timing in gait of Parkinson’s patients. PLoS ONE 7, e32600. https://doi.org/10.1371/journal.pone.0032600 (2012).h 42. Dotov, D. G. et al. The role of interaction and predictability in the spontaneous entrainment of movement. J. Exp. Psychol. Gen. 148, 1041–1057. https://doi.org/10.1037/xge0000609 (2019). p
g
g
43. Friston, K., FitzGerald, T., Rigoli, F., Schwartenbeck, P. & Pezzulo, G. Active inference: A process theory. Neural. Comput. 29,
(2017). (
)
44. Schoner, G. Timing, clocks, and dynamical systems. Brain Cogn. 48, 31–51. https://doi.org/10.1006/brcg.2001.1302 (2002). 45. Referencesf Re-interpreting detrended fluctuation analyses of stride-to-stride variability in human wal
Gait Posture 32, 348–353. https://doi.org/10.1016/j.gaitpost.2010.06.004 (2010).ff p
g
j g
p
(
)
13. Roerdink, M., de Jonge, C. P., Smid, L. M. & Daffertshofer, A. Tightening up the control of treadmill walking: Effects of maneuver-
ability range and acoustic pacing on stride-to-stride fluctuations. Front. Physiol. 10, 257. https://doi.org/10.3389/fphys.2019.00257
(2019). 14. Repp, B. H. & Su, Y. H. Sensorimotor synchronization: A review of recent research (2006–2012). Psychon Bull Rev 20, 403–452. https://doi.org/10.3758/s13423-012-0371-2 (2013).h 5. Leman, M. The Expressive Moment: How Interaction (with Music) Shapes Human Empowerment (MIT press, Boca Raton, 2016). d
h
ll
d
h
d
l d h
p
p
p
p
16. Moumdjian, L., Buhmann, J., Willems, I., Feys, P. & Leman, M. Entrainment and synchronization to auditory stimuli during
walking in healthy and neurological populations: A methodological systematic review. Front. Hum. Neurosci. 12, 263. https://doi. org/10.3389/fnhum.2018.00263 (2018).h g
7. Bennett, M., Schatz, M.F., Rockwood, H., Wiesenfeld, K. in Mathematical, Physical and Engineering Sciences. 563–579 (The Roya
Society).f y
18. Moumdjian, L. et al. A model of different cognitive processes during spontaneous and intentional coupling to music in mu
sclerosis. Ann. N. Y. Acad. Sci. 1445, 27–38. https://doi.org/10.1111/nyas.14023 (2019).f p
g
y
19. Delignieres, D. & Torre, K. Fractal dynamics of human gait: A reassessment of the 1996 data of Hausdorff et al. J Appl Physiol
(1985) 106, 1272–1279. https://doi.org/10.1152/japplphysiol.90757.2008 (2009).hf p
g
y
19. Delignieres, D. & Torre, K. Fractal dynamics of human gait: A reassessment of the 199
(1985) 106, 1272–1279. https://doi.org/10.1152/japplphysiol.90757.2008 (2009).hf 272–1279. https://doi.org/10.1152/japplphysiol.90757.2008 (2009 p
g
j pp p y
0. Sejdic, E., Fu, Y., Pak, A., Fairley, J. A. & Chau, T. The effects of rhythmic sensory cues on the temporal dynamics of human gait
PLoS ONE 7, e43104. https://doi.org/10.1371/journal.pone.0043104 (2012).l p
g
j pp p y
20. Sejdic, E., Fu, Y., Pak, A., Fairley, J. A. & Chau, T. The effects of rhythmic sensory cues on th
PLoS ONE 7, e43104. https://doi.org/10.1371/journal.pone.0043104 (2012). 21. Terrier, P. & Deriaz, O. Persistent and anti-persistent pattern in stride-to-stride variability of treadmill walking: Influence of
rhythmic auditory cueing. Hum. Mov. Sci. 31, 1585–1597. https://doi.org/10.1016/j.humov.2012.05.004 (2012).f y
y
g
g
j
22. Terrier, P. & Deriaz, O. Non-linear dynamics of human locomotion: Effects of rhythmic auditory cueing on local dynamic stab
Front. Physiol. 4, 230. https://doi.org/10.3389/fphys.2013.00230 (2013). https://doi.org/10.1038/s41598-020-69667-8 Scientific Reports | (2020) 10:12934 | www.nature.com/scientificreports/ Varlet, M., Marin, L., Issartel, J., Schmidt, R. C. & Bardy, B. G. Continuity of visual and auditory rhythms influences sensorim
coordination. PLoS ONE 7, e44082. https://doi.org/10.1371/journal.pone.0044082 (2012).l 6. Jordan, K., Challis, J. H. & Newell, K. M. Walking speed influences on gait cycle variability. Gait Post. 26, 128–134. https://doi
org/10.1016/j.gaitpost.2006.08.010 (2007). 7. Friston, K. Does predictive coding have a future?. Nat. Neurosci. 21, 1019–1021. https://doi.org/10.1038/s41593-018-0200-7 (2018)
8. Dotov, D. G., Bardy, B. G. & Dalla Bella, S. The role of environmental constraints in walking: Effects of steering and sharp turns ,
p
g
,
p
g
(
)
48. Dotov, D. G., Bardy, B. G. & Dalla Bella, S. The role of environmental constraints in walking: Effects of steering and sharp turns
on gait dynamics. Sci. Rep. 6, 28374. https://doi.org/10.1038/srep28374 (2016). 8. Dotov, D. G., Bardy, B. G. & Dalla Bella, S. The role of environmental constraints in walking: Effects of steering and sharp turns
on gait dynamics. Sci. Rep. 6, 28374. https://doi.org/10.1038/srep28374 (2016). l
l
l d
f h
d
f
lk
b l
f
f
l
l
l
l g
y
p
p
g
p
49. Motl, R. W. et al. Validity of the timed 25-foot walk as an ambulatory performance outcom
Scler. 23, 704–710. https://doi.org/10.1177/1352458517690823 (2017).f 9. Motl, R. W. et al. Validity of the timed 25-foot walk as an ambulatory performance outcome measure for multiple sclerosis. Mult
Scler. 23, 704–710. https://doi.org/10.1177/1352458517690823 (2017).f p
g
50. Sebastiao, E., Sandroff, B. M., Learmonth, Y. C. & Motl, R. W. Validity of the timed up and go test as a measure of functional
mobility in persons with multiple sclerosis. Arch. Phys. Med. Rehabil. 97, 1072–1077. https://doi.org/10.1016/j.apmr.2015.12.031
(2016). 51. Moens, B. et al. Encouraging spontaneous synchronisation with D-Jogger, an adaptive music player that aligns movement and
music. PLoS ONE 9, e114234. https://doi.org/10.1371/journal.pone.0114234 (2014). h
C
h
C
k
h
d
l C NE 9, e114234. https://doi.org/10.1371/journal.pone.0114234 (201 52. Buhmann, J., Masson, J.-B., Cochen De Cock, V., Damm, L., Leman, M. in Motor Behav uhmann, J., Masson, J.-B., Cochen De Cock, V., Damm, L., Leman, M. in Motor Behaviour and Emotion International Congress. ryce R M & Sprague K B Revisiting detrended fluctuation analysis Sci Rep 2 315 https://doi org/10 1038/srep00315 (2012) 52. Buhmann, J., Masson, J.-B., Cochen De Cock, V., Damm, L., Leman, M. in Motor Behaviour and Emotion International Cong 52. Acknowledgements
W
k
l d
h M h Acknowledgements
We acknowledge the Methusalem project (awarded by the Flemish Government) at UGent and the UHasselt
BOF grant (BOF16DOC41) for funding this study. g
We acknowledge the Methusalem project (awarded by the Flemish Government) at UGent and the UHasselt
BOF grant (BOF16DOC41) for funding this study. Author contributions
d L.M., P.F. and M.L. were responsible for conception and design of the study; B.M. developed the D-jogger soft-
ware used in the study and L.M. executed the study. L.M., PJ.M., S.D.B., L.D., B.M. computed and analyzed the
data. All authors contributed equally to interpretation of the data. L.M. wrote the manuscript, and all co-authors
participated equally to revising the manuscript. p
g
The authors declare no competing interests. p
g
The authors declare no competing interests. © The Author(s) 2020 www.nature.com/scientificreports/ Buhmann, J., Masson, J.-B., Cochen De Cock, V., Damm, L., Leman, M. in Motor Behaviour and Emotion International Congress. 53. Bryce, R. M. & Sprague, K. B. Revisiting detrended fluctuation analysis. Sci. Rep. 2, 315. https://doi.org/10.1038/srep00315 (2012). 54 K
t
N A & Rh
C K P
id
ti
f
th
li
ti
f d t
d d fl
t
ti
l
i i
it
i bilit
t di g
53. Bryce, R. M. & Sprague, K. B. Revisiting detrended fluctuation analysis. Sci. Rep. 2, 315. https://doi.org/10.1038/srep00315 (2012).l 53. Bryce, R. M. & Sprague, K. B. Revisiting detrended fluctuation analysis. Sci. Rep. 2, 315. https:/ 54. Kuznetsov, N. A. & Rhea, C. K. Power considerations for the application of detrended fluctuation analysis in gait variability st
PLoS ONE 12, e0174144. https://doi.org/10.1371/journal.pone.0174144 (2017). 55. Goldberger, A. L. et al. Fractal dynamics in physiology: Alterations with disease and aging. Proc. Natl. Acad. Sci. U.S.A. 99(Suppl
1), 2466–2472. https://doi.org/10.1073/pnas.012579499 (2002). h
l
l
f
b l
( d
)
(
l
) p
g
p
56. McGrath, D. in Nonlinear Analysis for Human Movement Variability (ed N. Stergiou) 261–300 (Taylor & Francis, 2016) 57. Dingwell, J. B. & Cusumano, J. P. Humans use multi-objective control to regulate lateral foot placement when walking. PLoS
Comput. Biol. 15, e1006850. https://doi.org/10.1371/journal.pcbi.1006850 (2019).hh 58. Theiler, J., Eubank, S., Longtin, A., Galdrikian, B. & Farmer, D. Testing for nonlinearity in time series: The method of surro
data. Phys. D 58, 77–94 (1992). y
59. Schreiber, T. & Schmitz, A. Surrogate time series. Phys. D 142, 346–382 (2000). g
y
60. Fisher, N. I. Statistical Analysis of Circular Data (Cambridge University Press, Cambridge, 1995). 61. Cremers, J. & Klugkist, I. One direction? A tutorial for circular data analysis using R with examples in cognitive psychology. Front. Psychol. 9, 2040. https://doi.org/10.3389/fpsyg.2018.02040 (2018). 61. Cremers, J. & Klugkist, I. One direction? A tutorial for circular dat
Psychol. 9, 2040. https://doi.org/10.3389/fpsyg.2018.02040 (2018) 62. Washabaugh, E. P., Kalyanaraman, T., Adamczyk, P. G., Claflin, E. S. & Krishnan, C. Validity and repeatability of inertial measure-
ment units for measuring gait parameters. Gait Post. 55, 87–93. https://doi.org/10.1016/j.gaitpost.2017.04.013 (2017).ff g g
g
j g
63. Socie, M. J., Motl, R. W., Pula, J. H., Sandroff, B. M. & Sosnoff, J. J. Gait variability and disability in multiple sclerosis. Gait
38, 51–55. https://doi.org/10.1016/j.gaitpost.2012.10.012 (2013). Scientific Reports | (2020) 10:12934 | https://doi.org/10.1038/s41598-020-69667-8 www.nature.com/scientificreports/ Additional information Supplementary information is available for this paper at https://doi.org/10.1038/s41598-020-69667-8. Correspondence and requests for materials should be addressed to L.M. Reprints and permissions information is available at www.nature.com/reprints. Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and
institutional affiliations. Open Access This article is licensed under a Creative Commons Attribution 4.0 International
License, which permits use, sharing, adaptation, distribution and reproduction in any medium or
format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the
Creative Commons license, and indicate if changes were made. The images or other third party material in this
article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the
material. If material is not included in the article’s Creative Commons license and your intended use is not
permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from
the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/. © The Author(s) 2020 Scientific Reports | (2020) 10:12934 | https://doi.org/10.1038/s41598-020-69667-8
|
https://openalex.org/W3215998758
|
https://journal.agrimetassociation.org/index.php/jam/article/download/181/139
|
English
| null |
Rainfall analysis across the north east Indian state of Tripura
|
Journal of Agrometeorology
| 2,021
|
cc-by
| 2,880
|
1ICAR Research Complex for NEH Region, Mizoram Centre, Kolasib, India
2ICAR Research Complex for NEH Region, Lembucherra, Tripura, India
3ICAR- Research Complex for NEH Region, Umiam, India
4ICAR Central Coastal Agricultural Research Institute, Old Goa, India
*Corresponding author email : sauravs.saha@gmail.com 1ICAR Research Complex for NEH Region, Mizoram Centre, Kolasib, India
2ICAR Research Complex for NEH Region, Lembucherra, Tripura, India
3ICAR- Research Complex for NEH Region, Umiam, India
4ICAR Central Coastal Agricultural Research Institute, Old Goa, India
*Corresponding author email : sauravs.saha@gmail.com (March-May), monsoon (June–September) and post-
monsoon (October–December) months. Standardised
Precipitation Index (SPI) follow the normal distribution
but rainfall series rarely. SPI based wetness analysis gained
major importance in recent years as the potential indicator
of drought occurrence and its intensity, permitting
comparisons over variable space and time (Saha et al.,
2015). The SPI was calculated for different timescales viz. 1 month (1-SPI), 2 months (2-SPI), 3 months (3-SPI), 4
months (4-SPI), 6 months (6-SPI), 12 months (12-SPI)
and 24 months (24-SPI). We tested null hypothesis (H0)
for having no trend in the time series against alternative
hypothesis (H1) for identifying significant increasing
or decreasing trend in the respective time series using
rank-based non-parametric Man Kendall test. Finally, the
spatial distribution of identified seasonal and annual SPI
trend was mapped under GIS environment (Arc GIS 9.3). (March-May), monsoon (June–September) and post-
monsoon (October–December) months. Standardised
Precipitation Index (SPI) follow the normal distribution
but rainfall series rarely. SPI based wetness analysis gained
major importance in recent years as the potential indicator
of drought occurrence and its intensity, permitting
comparisons over variable space and time (Saha et al.,
2015). The SPI was calculated for different timescales viz. 1 month (1-SPI), 2 months (2-SPI), 3 months (3-SPI), 4
months (4-SPI), 6 months (6-SPI), 12 months (12-SPI)
and 24 months (24-SPI). We tested null hypothesis (H0)
for having no trend in the time series against alternative
hypothesis (H1) for identifying significant increasing
or decreasing trend in the respective time series using
rank-based non-parametric Man Kendall test. Finally, the
spatial distribution of identified seasonal and annual SPI
trend was mapped under GIS environment (Arc GIS 9.3). The warm and humid tropical climate of Tripura
(Hyperthermic) in the Purvachal range manifests a
high degree of seasonal variability in rainfall (length of
growing period >300 days), which is the main support to
rainfed shifting cultivation and horticultural plantations
in the hill slopes and extensive rice cultivation along
with vegetable integration in the plain areas. al of Agrometeorology 23 (4) : 471-475 (December 2021) : https://doi.org/10.54386/jam.v23i4.181 al of Agrometeorology 23 (4) : 471-475 (December 2021) : https://doi.org/10.54386/jam.v23i4.181 1ICAR Research Complex for NEH Region, Mizoram Centre, Kolasib, India
2ICAR Research Complex for NEH Region, Lembucherra, Tripura, India
3ICAR- Research Complex for NEH Region, Umiam, India
4ICAR Central Coastal Agricultural Research Institute, Old Goa, India
*Corresponding author email : sauravs.saha@gmail.com The
spatiotemporal variability in meteorological parameters
like rainfall, temperature and wind speed were evident
over the NE region including Tripura, with increasing
irregularities in the seasonal rainfall distribution and
extremes over NE Indian region (Saha et al., 2016;
Chakraborty et al., 2017). Despite present concerns about
the potential consequences of global climate change on
the spatiotemporal variability of regional wetness trend
evaluation; such time series datasets are rarely available
for different northeast Indian states (Saha et al., 2016). Therefore, our present study focused on characterization
of the spatiotemporal variability of intrastate location-
specific wetness trend rather than rainfall pattern at
different time scales in Tripura. Long-term trend in monthly, seasonal and annual
rainfall unveiled a high degree of spatial variability
(Table 1). However, the decreasing trend in monthly
and seasonal rainfall was prominent for all the selected
rain gauge stations. Pre-monsoonal rainfall significantly
reduced at AD Nagar (P< 0.01) and Belonia (P< 0.05)
station; monsoon (P< 0.01) and post-monsoon (P< 0.05)
rainfall at Amarpur. Sonamura and Udaipur experienced
a reduction in pre-monsoon rainfall over the entire
period, including April rainfall in particular (P<0.1). The decline in pre-monsoon shower will adversely affect
the net water availability for growing rainfed crops in The rainfall dataset from twelve rain gauge
stations were available across Tripura from 1970 to 2016
(Fig. 1). The variability in average annual rainfall ranged
between 2153 mm (Agartala) to 2725 mm (Sabroom). The available rainfall datasets were unimodal and free
from any first-order autocorrelation. 1ICAR Research Complex for NEH Region, Mizoram Centre, Kolasib, India
2ICAR Research Complex for NEH Region, Lembucherra, Tripura, India
3ICAR- Research Complex for NEH Region, Umiam, India
4ICAR Central Coastal Agricultural Research Institute, Old Goa, India
*Corresponding author email : sauravs.saha@gmail.com The changes in rainfall dynamics may remould
the present pattern of stream flows and agricultural water
availability from runoff, antecedent soil moisture and
groundwater reserves (Jain and Kumar, 2012).The scatter
plot between 1- SPI and monthly rainfall deficit (%) for
all the selected raingauge stations are presented in Fig. 2. The monthly linear slope values (m) of the fitted curve
varied between 0.221-2.145 with statistically significant
R2 values (p<0.01).We observed that the steepness of
slope served as an effective and promising indicator for
characterization of the major rainfall receiving months in Tripura region. the linear fitted curves, it was evident
that major rainfall receiving months bear the linear slope
values > 1. The slope values of major rainfall receiving
months varied from 1.34 (April) to its peak at august
(m = 2.145) and then declined to 1.27 during October. From November to March the values of the slopes were
<1.0. Indian monsoon monitoring systems. However, the
pattern differed for the NE Indian states of Mizoram, so
as for Tripura also between April to October (Saha et al.,
2016). Therefore, it was clear that the slope of linear plot
between 1- SPI and rainfall deficit (%) would be simple
effective criteria for characterization of major rain-bearing
months in any location. The identified criteria still require
extensive supportive validation studies for getting wide
acceptablity from different very low to moderate rainfall
receiving regions of India (Saha et al., 2015). Tripura region. the linear fitted curves, it was evident
that major rainfall receiving months bear the linear slope
values > 1. The slope values of major rainfall receiving
months varied from 1.34 (April) to its peak at august
(m = 2.145) and then declined to 1.27 during October. From November to March the values of the slopes were
<1.0. Indian monsoon monitoring systems. However, the
pattern differed for the NE Indian states of Mizoram, so
as for Tripura also between April to October (Saha et al.,
2016). Therefore, it was clear that the slope of linear plot
between 1- SPI and rainfall deficit (%) would be simple
effective criteria for characterization of major rain-bearing
months in any location. The identified criteria still require
extensive supportive validation studies for getting wide
acceptablity from different very low to moderate rainfall
receiving regions of India (Saha et al., 2015). 1ICAR Research Complex for NEH Region, Mizoram Centre, Kolasib, India
2ICAR Research Complex for NEH Region, Lembucherra, Tripura, India
3ICAR- Research Complex for NEH Region, Umiam, India
4ICAR Central Coastal Agricultural Research Institute, Old Goa, India
*Corresponding author email : sauravs.saha@gmail.com Trend in seasonal
and annual rainfall pattern (mean values) for four distinct
seasons namely, winter (January–February), pre-monsoon Rainfall analysis across the north east Indian state of Tripura December 2021 472 Table 1: Observed trend in periodic rainfall pattern at different raingauge stations across the odic rainfall pattern at different raingauge stations across the Tripura State Table 1: Observed trend in periodic rainfall pattern at different raingauge stations across the Tripura State
Period
Agartala
Kailashar
AD Nagar Amarpur Dharmanagar Belonia Kamalpur Khowai Sabroom Sonamura Udaipur
January
-1.75^
-2.68**
0.54
0.72
0.31
0.44
0.08
1.92^
0.00
0.86
0.18
February
-0.59
-0.82
-0.56
-1.07
-0.40
-0.76
0.20
0.77
-0.82
0.35
-0.64
March
-1.03
-1.47
-0.37
-0.30
-0.78
0.30
0.70
0.23
0.30
0.44
0.32
April
-0.14
-0.64
-2.52**
-2.25**
-1.87^
-2.50**
-0.89
-1.60
-0.52
-1.68^
-1.86^
May
1.34
1.34
-1.63
-0.32
0.43
-0.99
0.82
-0.73
1.03
-1.22
-0.80
June
-0.98
-1.58
-1.16
-1.01
0.46
0.11
0.06
0.12
-0.59
0.16
-0.67
July
-1.17
-1.48
-1.69^
-2.65**
0.48
-2.17*
0.46
0.26
-0.71
-1.04
-1.28
August
-1.09
-0.50
-0.96
-1.68^
-0.86
-0.50
0.36
-1.51
-0.90
-1.09
-0.31
September
0.27
1.24
-0.78
0.89
0.92
0.20
-0.68
0.33
0.63
0.03
-0.54
October
0.26
-0.78
0.14
-0.62
0.38
0.76
1.54
2.39**
0.82
0.12
1.00
November
-1.25
-1.78^
-2.04*
-2.36**
-1.61
-2.61**
-1.56
-1.52
-1.98^
-1.61
-1.54
December
-0.80
-0.18
-0.83
-1.09
-0.97
-1.30
-0.99
-0.49
-0.43
-1.11
-1.59
Annual
-0.62
-0.86
-2.46**
-2.49**
0.11
-2.40**
-0.01
-1.49
-0.65
-1.65
-1.49
Winter
-1.16
-1.72^
-0.39
-0.80
0.25
-0.36
0.41
1.38
-0.52
0.55
-0.39
Pre monsoon
1.01
0.25
-2.54**
-1.40
-0.50
-2.22*
-0.03
-1.05
0.86
-1.72^
-1.44
Monsoon
-1.20
-1.01
-1.67
-2.49**
0.98
-1.22
-0.01
-1.42
-0.84
-1.18
-1.38
Post Monsoon
-0.53
-1.04
-0.66
-1.28
0.00
-0.50
0.59
1.11
-0.21
-0.69
0.15
(Note: ‘^’signifies p< 0.1; ‘*’signifies p< 0.05; ‘**’signifies p< 0.01) Tripura (Chakraborty et al., 2017). The annual rainfall
also indicated a significant declining trend at AD Nagar
(P< 0.01), Amarpur (P< 0.01) and Belonia (P< 0.05). Under limited availability of water harvesting structures
(jalkunds), supportive irrigation facilities, mulching
and water conservation facilities, reduction in annual
rainfall may take toll on rabi cultivation with limited soil
moisture availability (Saha et al., 2015). Rapid clustered
urbanization in the nearby areas of rain gauge stations
replaced the natural vegetation with agricultural croplands,
such localized human-induced land-use changes may be
resulted in such anomalies for the selected locations in
Tripura. 1ICAR Research Complex for NEH Region, Mizoram Centre, Kolasib, India
2ICAR Research Complex for NEH Region, Lembucherra, Tripura, India
3ICAR- Research Complex for NEH Region, Umiam, India
4ICAR Central Coastal Agricultural Research Institute, Old Goa, India
*Corresponding author email : sauravs.saha@gmail.com Site-specific SPI based time series over different
timescales were more robust and efficient to detect the
periodic progress pattern in seasonal wetness, than the
cumulative rainfall time series in Tripura. The significant
reduction in average winter wetness (2 SPI) may intensify
the water scarcity for rabi cultivation at Kailshar in
near future (Fig.3a). For the other nine stations, there Vol. 23, No. 4
Fig 1: Location details for the selected raingauge stations across the Tripura state
SAHA et al. SAHA et al. 473 Vol. 23, No. 4 Vol. 23, No. 4
ig 1: Location details for the selected raingauge stations across the Tripura state
ig. 2: Scatter plot between 1-SPI and monthly rainfall deficit (%) in Tripura
SAHA et al. Fig 1: Location details for the selected raingauge stations across the Tripura state Fig. 2: Scatter plot between 1-SPI and monthly rainfall deficit (%) in Tripura Fig. 2: Scatter plot between 1-SPI and mont Fig. 2: Scatter plot between 1-SPI and monthly rainfall deficit (%) in Tripura Rainfall analysis across the north east Indian state of Tripura December 20
Rainfall analysis across the north east Indian state of Tripura Rainfall analysis across the north east Indian state of Tripura Rainfall analysis across the north east Indian state of Tripura 474 December 2021 ig. 3: Spatial distribution of seasonal SPI trends: a. winter b. pre-monsoon c, monsoon d, post-monsoon e. Annual and Bi-annual. Up
and inverted filled (black) triangle indicate significant positive (wet) and negative (drying) trends respectively; white filled up
triangle and inverted triangle indicate non-significant positive (wet) and negative (drying) trends Fig. 3: Spatial distribution of seasonal SPI trends: a. winter b. pre-monsoon c, monsoon d, post-monsoon e. Annual and Bi-annual. Upright
and inverted filled (black) triangle indicate significant positive (wet) and negative (drying) trends respectively; white filled upright
triangle and inverted triangle indicate non-significant positive (wet) and negative (drying) trends SAHA et al. Vol. 23, No. 4 475 Disclaimer: The contents, opinions, and views expressed
in the research article published in the Journal of
Agrometeorology are the views of the authors and do not
necessarily reflect the views of the organizations they
belong to. was no significant change in winter wetness during the
study period. Pre-monsoon wetness (3-SPI) significantly
decreased at AD Nagar, Amarpur, Sonamura, and Belonia
(Fig.3b). 1ICAR Research Complex for NEH Region, Mizoram Centre, Kolasib, India
2ICAR Research Complex for NEH Region, Lembucherra, Tripura, India
3ICAR- Research Complex for NEH Region, Umiam, India
4ICAR Central Coastal Agricultural Research Institute, Old Goa, India
*Corresponding author email : sauravs.saha@gmail.com At Amarpur, the significant reduction in
post-monsoon shower was in synchrony to the significant
reduction in 3-SPI based post monsoon wetness that
may hamper the progress of rabi agriculture without
widespread adaptation of suitable soil water conservation
measures in near future (Table 1 and Fig.3d). Rest all other
stations, remained un-affected from localized seasonal
wetness variability over the present study period in
Tripura. Though the initial annual rainfall trend analysis
confirmed the significant reduction in annual rainfall at
AD Nagar, Amarpur and Belonia station only (Table 1);
but a significant reduction in annual wetness (12 SPI)
was evident at Khowai, Amarpur, AD Nagar, Udaipur,
Sonamura and Belonia (Fig.3e). This gradual significant
decline in seasonal and annual wetness over past four
decades poses a potential threat for agricultural water
availability and annual regional watershed hydrology in
Tripura (particularly at Amarpur, Belonia and AD Nagar;
Chakraborty et al., 2017). Widespread adaptation of
water-saving irrigation techniques, cover crop cultivation,
mulching, and selection of suitable crops or appropriate
water stress tolerant cultivars will encounter the adverse
impact of periodic water scarcity across identified
locations prone to periodic water scarcity in Tripura. Publisher’s Note: The periodical remains neutral with
regard to jurisdictional claims in published maps and
institutional affiliations. 1ICAR Research Complex for NEH Region, Mizoram Centre, Kolasib, India
2ICAR Research Complex for NEH Region, Lembucherra, Tripura, India
3ICAR- Research Complex for NEH Region, Umiam, India
4ICAR Central Coastal Agricultural Research Institute, Old Goa, India
*Corresponding author email : sauravs.saha@gmail.com For Amarpur station, a non-significant reducing
trend in pre-monsoon rainfall resulted in a significant
reduction in SPI based pre-monsoon wetness time series. In contrast to the significant reduction in monsoon
rainfall at Amarpur, we observed a significant reduction
in 4-SPI based monsoon wetness at AD Nagar, Amarpur
and Belonia station (Fig.3c). Crop management practices
facilitating diversification and shifting cropping pattern
i.e. from high requiring (rice) to low water requiring
(pulses) crops or cultivation of abiotic stress tolerant
cultivars may reduce the risk of crop failure during
monsoon months. At Amarpur, the significant reduction in
post-monsoon shower was in synchrony to the significant
reduction in 3-SPI based post monsoon wetness that
may hamper the progress of rabi agriculture without
widespread adaptation of suitable soil water conservation
measures in near future (Table 1 and Fig.3d). Rest all other
stations, remained un-affected from localized seasonal
wetness variability over the present study period in
Tripura. Though the initial annual rainfall trend analysis
confirmed the significant reduction in annual rainfall at
AD Nagar, Amarpur and Belonia station only (Table 1);
but a significant reduction in annual wetness (12 SPI)
was evident at Khowai, Amarpur, AD Nagar, Udaipur,
Sonamura and Belonia (Fig.3e). This gradual significant
decline in seasonal and annual wetness over past four
decades poses a potential threat for agricultural water
availability and annual regional watershed hydrology in
Tripura (particularly at Amarpur, Belonia and AD Nagar;
Chakraborty et al., 2017). Widespread adaptation of
water-saving irrigation techniques, cover crop cultivation,
mulching, and selection of suitable crops or appropriate
water stress tolerant cultivars will encounter the adverse
impact of periodic water scarcity across identified
locations prone to periodic water scarcity in Tripura. Conflict of Interest Statement: The author(s) declare(s) was no significant change in winter wetness during the
study period. Pre-monsoon wetness (3-SPI) significantly
decreased at AD Nagar, Amarpur, Sonamura, and Belonia
(Fig.3b). For Amarpur station, a non-significant reducing
trend in pre-monsoon rainfall resulted in a significant
reduction in SPI based pre-monsoon wetness time series. In contrast to the significant reduction in monsoon
rainfall at Amarpur, we observed a significant reduction
in 4-SPI based monsoon wetness at AD Nagar, Amarpur
and Belonia station (Fig.3c). Crop management practices
facilitating diversification and shifting cropping pattern
i.e. from high requiring (rice) to low water requiring
(pulses) crops or cultivation of abiotic stress tolerant
cultivars may reduce the risk of crop failure during
monsoon months. REFERENCES Chakraborty, D., Saha, S., Singh, R.K., Sethy, B.K.,
Kumar, A., Saikia, U.S., Das, S.K., Makdoh,
B., Borah, T.R., NomitaChanu, A., Walling,
I., Rolling Anal, P.S., Chowdhury, S., Das
Chaudhuri, D. (2017). Spatio-temporal trends
& and change point analysis detection of in the
rainfall observation series from different places
of North-eastern Indian states. J. Agrometeorol.,
19(2): 160-163. Jain, S.K., Kumar, V. (2012). Trend Analysis of Rainfall
and Temperature Data for India. Curr. Sci.,102(1):
37-49. Saha, S., Chakraborty, D., Choudhury, B.U., Singh,
S.B., Chinza, N., Lalzarliana, C., Dutta, S.K.,
Chowdhury, S., Boopathi, T., Lungmuana, Singh,
A.R., Ngachan, S.V. (2015). Spatial variability
in temporal trends of precipitation and its impact
on the agricultural scenario of the North Eastern
Indian state of Mizoram. Curr. Sci.,109 (12):
2278-2282. Saha, S., Chakraborty, D., Singh, S.B., Chowdhury,
S., Syiem, E.K., Dutta, S.K., Lungmuana,
Choudhury, B.U., Boopathi, T., Singh, A.R.,
Ramakrishna, Y., Roy, A. (2016). Analyzing the
trend in thermal discomfort and other bioclimatic
indices at Kolasib, Mizoram. J. Agrometeorol.,
18(1): 57-61. Conflict of Interest Statement: The author(s) declare(s)
that there is no conflict of interest. Conflict of Interest Statement: The author(s) declare(s)
that there is no conflict of interest. Received : December 2020 : Accepted : August 2021
|
https://openalex.org/W1517288763
|
https://peerj.com/articles/1039.pdf
|
English
| null |
A Bayesian approach to optimizing stem cell cryopreservation protocols
| null | 2,014
|
cc-by
| 6,866
|
ABSTRACT Cryopreservation is beset with the challenge of protocol alignment across a wide
range of cell types and process variables. By taking a cross-sectional assessment of
previously published cryopreservation data (sample means and standard errors) as
preliminary meta-data, a decision tree learning analysis (DTLA) was performed to
develop an understanding of target survival using optimized pruning methods based
on different approaches. Briefly, a clear direction on the decision process for selection
of methods was developed with key choices being the cooling rate, plunge tempera-
ture on the one hand and biomaterial choice, use of composites (sugars and proteins
as additional constituents), loading procedure and cell location in 3D scaffolding
on the other. Secondly, using machine learning and generalized approaches via the
Na¨ıve Bayes Classification (NBC) method, these metadata were used to develop
posterior probabilities for combinatorial approaches that were implicitly recorded
in the metadata. These latter results showed that newer protocol choices developed
using probability elicitation techniques can unearth improved protocols consistent
with multiple unidimensionally-optimized physical protocols. In conclusion, this
article proposes the use of DTLA models and subsequently NBC for the improvement
of modern cryopreservation techniques through an integrative approach. Subjects Bioengineering, Statistics, Computational Science
Keywords Decision-tree learning (DTL), Sugars, Meta-data, Na¨ıve Bayes Classifier (NBC),
3D cryopreservation, Mouse embryonic stem cells, Cell attachment, Modified alginates,
Encapsulation, CPA loading Subjects Bioengineering, Statistics, Computational Science
Keywords Decision-tree learning (DTL), Sugars, Meta-data, Na¨ıve Bayes Classifier (NBC),
3D cryopreservation, Mouse embryonic stem cells, Cell attachment, Modified alginates,
Encapsulation, CPA loading Submitted 10 December 2014
Accepted 30 May 2015
Published 25 June 2015
Corresponding author
Sammy Sambu,
Sambu@post.harvard.edu
Academic editor
Massimiliano Zanin
Additional Information and
Declarations can be found on
page 13
DOI 10.7717/peerj.1039
Copyright
2015 Sambu
Distributed under
Creative Commons CC-BY 4.0 Submitted 10 December 2014
Accepted 30 May 2015
Published 25 June 2015
Corresponding author
Sammy Sambu,
Sambu@post.harvard.edu
Academic editor
Massimiliano Zanin
Additional Information and
Declarations can be found on
page 13
DOI 10.7717/peerj.1039
Copyright
2015 Sambu
Distributed under
Creative Commons CC-BY 4.0 How to cite this article Sambu (2015), A Bayesian approach to optimizing cryopreservation protocols. PeerJ 3:e1039;
DOI 10.7717/peerj.1039 A Bayesian approach to optimizing
cryopreservation protocols Sammy Sambu Sammy Sambu
Nandi Hills, Kenya Subjects Bioengineering, Statistics, Computational Science
Keywords Decision-tree learning (DTL), Sugars, Meta-data, Na¨ıve Bayes Classifier (NBC),
3D cryopreservation, Mouse embryonic stem cells, Cell attachment, Modified alginates,
Encapsulation, CPA loading INTRODUCTION Cryopreservation of cells often results in cell survivals that are lower because of suboptimal
process variables. When cryopreservation is performed in biological constructs (including
encapsulated cells), the results are often worse due to the differentials in cryoprotectant
(CPA) concentration, CPA exposure times and cooling rates. These differences are
exaggerated for cells that are farthest from the surface which is typical for large specimens
(Muldrew et al., 2000). Conversely, for small capsules that are just above the upper limit
of the capsule diameter, differentials in temperature and concentration are meagre (Cui
et al., 2002) i.e., lower size leads to lower barriers to diffusion and heat transfer leading
to smaller CPA concentration and temperature (C–T) differentials—for single cells in
suspension, the challenge is further reduced as the cell is closer to C–T equilibrium
with its surroundings than encapsulated cells. The main challenge for capsules remains
the varying CPA concentrations and exposure times for cells encapsulated in them. Distributed under
Creative Commons CC-BY 4.0 OPEN ACCESS A spherical capsule, for example, will have different cell survivals at each radius (Sambu
et al., 2011). Furthermore, there are new variables arising from 3D cryopreservation
including cell encapsulation densities, cooling rates and the determiners for CPA
concentration profiles e.g., initial Concentration, temperature and construct size (Sambu
et al., 2011). The challenge is to minimize the deleterious effects of these differentials by
providing cryoprotection without mass transfer limitations. Therefore, new construct
variables arise. Hence, a re-prioritization of process and capsule geometry variables
becomes necessary—a DTLA approach can help resolve the tensions in prioritizing and
then recursively optimizing these variables. In using a decision-tree learning analysis (DTLA) approach, it is necessary to leverage
existing principles in cryopreservation. The entrapment of CPAs within the capsule
during cell resuspension in encapsulating material is an alternative strategy to the
modulation of cell location. However, intracellular CPAs are often toxic and encapsulation
is not lossless. Hence, when a potentially cytotoxic CPA such as DMSO is added to the
encapsulating material, cell survival is often diminished (Cui et al., 2002; Liu, 2007). Therefore, the success of the strategy relies on using a benign CPA as the “entrapped
CPA” followed by the loading of the potentially malign but fast-diffusing CPA into the
construct. Sambu (2015), PeerJ, DOI 10.7717/peerj.1039 INTRODUCTION For protein-based cryoprotection, mixing the proteins with polar CPAs also
has a deleterious effect on the ability of the protein to confer cryoprotection due to the
potential denaturation of the proteins prior to cooling (Meryman, 2007). Overall, the
protein-DMSO interaction is minimized by incorporating the protein in the encapsulating
material, and allowing a time-limited diffusion of the malign CPA into the construct prior
to cooling. Besides modifying the encapsulation process, slow-cooling procedures should
also be adopted to minimize the required concentrations of the potentially malign CPA. Besides using proteins and intracellular CPAs, synthetic non-penetrating polymers
(SNPP) can also provide cryoprotection within the scaffold, thereby bypassing the
limitations of diffusion in higher-dimensional cryopreservation. A subclass of these
SNPPs are the vitrifying polymers which have been used to encourage extracellular
vitrification of the cryopreservative during cooling thereby limiting ice crystal growth
(Gibson et al., 2009). Vinyl-derived polymers have also been shown to decrease ice crystal
size colligatively. Examples include polyethylene glycol (PEG), polyvinyl alcohol (PVA)
and hydroxyethyl starch. The key challenge in using these polymers is minimizing large
increases in the viscosity of the solution (which will make encapsulation difficult) and
minimizing the difficulty in post-thaw extraction of cells from the scaffold. Shifting the focus from solutes to encapsulates, changing the encapsulation material
(the matrix) properties by using polymer composites can be avoided if one entraps
lower molecular weight encapsulates. One alternative would be to use low molecular
weight polymers. Another would be to use sugars with cryoprotective properties. The
mechanism for cryoprotection conferred by sugars is three-fold: complex sugars can
inhibit the formation of ice (Roos & Karel, 1991; Shirakashi et al., 2003); they can replace
lost water molecules and can stabilize lipid bilayers from sudden phase transitions during
cooling (Arakawa et al., 1990; Rudolph & Crowe, 1985). They may prevent sudden changes Sambu (2015), PeerJ, DOI 10.7717/peerj.1039 2/15 in the lipid phase and ensuing phase separation through hydrogen bond formation
with phospholipid head groups (Prestrelski, Arakawa & Carpenter, 1993; Anchordoguy
et al., 1987). This interaction causes a reduction in the liquid crystal-gel transition
temperature and minimizes membrane fusion by separating one layer from another
(Arakawa & Timasheff, 1982). Sugars also stabilize proteins by encouraging a preferential
hydration of the proteins in solution (Huang et al., 2003). Granted that complex sugars
are non-permeating solutes, these benefits are extracellular unless a delivery mechanism is
engineered. INTRODUCTION Each of the techniques mentioned can be represented as random decision variables
(factors) in the abstract; each one taking on a probability state according to a protocol
developed collectively and based on categorical field-specific knowledge. In that way,
a DTLA will describe the epistemology of the factors and their relationship to a final
probability state describing the life/death or living (long-living)/apoptotic (short-living)
state of the hypothetical cell. However, the ability to generalize these DTL relationships
is often severely reduced unless sophisticated pruning algorithms are used (F¨urnkranz,
1997) (conversely local results are definitively optimized based on the learning set)
hence requiring the use of a more generalized approach taking the final hypothetical
cell state as discrete and each influencing factor as conditionally independent (given nodal
classification) in the decision-making process as well as in the execution of the protocol. Once the requirement for broader applicability of a learned model is defined, a more
general approach to making predictions on related but distinct data sets is required. In this latter case, the Na¨ıve Bayes Classifier (NBC) presents a more efficient and direct
approach to prediction. The assumption of class-conditional factor independence shrinks
the pre-prediction computational requirements significantly and speeds the overall
time-to-decision making. Nonetheless, empirical evidence must be invoked to support
a combined approach using both NBC and DTL (Rubinstein & Hastie, 1997). Overall,
from the decision-maker’s view, by combining the content-centric approach of the DTLA
and the boundary-focussed NBC, the opportunity cost is low and the decision process is
retrospectively holistic and exhaustive. This paper seeks to demonstrate the promising strategies for successful cryopreservation
of cells in 3D scaffolds by using a DTLA approach to develop a heuristic for approaching
cryopreservation across many subjects. Given this cryopreservation challenge specifically
concerns the opportunities in 3D cryopreservation; success is measured by a cell survival
percentage that is either equal to or better than suspension cryopreservation. Sambu (2015), PeerJ, DOI 10.7717/peerj.1039 Decision tree learning analysis Analysis was performed using Rpart package in the open source program R for recursive
partitioning and regression analysis (Therneau & Atkinson, 2015). Note, since the
technique allows both classification and regression, the cell survival could be represented
either as a class (categorical) or regressed as a continuous variable (continuous). In brief, the collected data (Sambu et al., 2011) and meta-data (Sambu et al., 2011; Heng,
Yu & Ng, 2004; Kashuba Benson, Benson & Critser, 2008; Miszta-Lane et al., 2007) was
imported into R, partitioned according to a variable that best describes the data at each
given partition level. The model is 5-fold cross-validated (giving >30 samples as test set
for each model instantiation) and the model that best describes the data set is selected after
evaluation for fidelity to the original data. For each node, sample numbers are provided. Additionally, the p-value is provided for each split showing whether the branching can be
supported at 95% confidence. The DTLA process involves the development of a loss function, computation of a
residual over which the learner model is fit. To minimize the loss function, a multiplier
is chosen such that through the reapplication of the learner on the training set, the loss
function successively diminishes until the tree is fully constructed (Friedman, 2001). The
overall governing equation is provided below:- Equation 1: Governing Equation for the DTLA algorithm where h is the regression
tree fitted onto the negative gradient 147 (middle term) containing the minimized loss
function L—drawn from Friedman (2001). h = −∂L(yi,F(xi))
∂F(xi)
= F(xi) −yi. Data collection Data was collected and categorized into the main decision factors namely: dimensionality,
cell location, use of a biomaterial, lyoprotection, CPA loading, use of integrins, CPA type,
containment, cell line, cooling rate and plunge temperature (11 features) against 153
instances. Where original data was not present, the mean and standard error were used to
redraw virtual samples from the original population described by the sample parameters. Sambu (2015), PeerJ, DOI 10.7717/peerj.1039 3/15 3/15 The predicted property is the survival (as a percent for DTLA with regression or as a
category for classification with NBC i.e., long-surviving or short-surviving) While there are many studies that could be incorporated, the key requirement was
the explicit description of decision factors that were continuous from study to study;
additionally, the studies were chosen to be different enough to incorporate diverse decision
factors but of sufficient relational value as to allow a successful DTLA and/or NBC analysis
without algorithmic or model instability—in this regard, many failures were met when
studies failed to have the same number of sufficient match critical factors (though of
overall good cryopreservation results) to allow acceptable predictions and classifications. Sambu (2015), PeerJ, DOI 10.7717/peerj.1039 Na¨ıve Bayes Classification (NBC) The e1071 package (Therneau & Atkinson, 2015) was used to run the NBC process using the
decision factors of dimensionality, location, biomaterial, sugars, step-loading, integrins
with the terminal decision variable as cell survival represented in two conditionally
independent classes: short-living (i.e., apoptotic cells expected shown to undergo cell
death soon after recovery) or long-living (i.e., live cells expected to survive and can be
cultured and expanded successfully). Therefore, there are 6 features and 125 instances. All cryopreservation ‘decision factors’ are drawn from the literature cited and matched Sambu (2015), PeerJ, DOI 10.7717/peerj.1039 4/15 Figure 1 The decision tree learning model. A recursive partitioning in detail showing the hot-spots at “Integrins,” “Sucrose” and “controlled
seeding.” The validation scores were 77% (match) and 23% (failed) for the DTL model. model. A recursive partitioning in detail showing the hot-spots at “Integrins,” “Sucrose” and “controlled
7% (match) and 23% (failed) for the DTL model. Figure 1 The decision tree learning model. A recursive partitioning in detail showing the hot-spots at “Integrins,” “Sucrose” and “controlled
seeding.” The validation scores were 77% (match) and 23% (failed) for the DTL model. semantically by author. The model is 10-fold cross-validated (giving >10 samples as test
set for each model instantiation based on a >100 instances for training the model) and the
model that best describes the data set is selected after evaluation for fidelity to the original
data. The governing equation is provided below:- Equation 2: Governing equation for the NBC algorithm where Y is the target vari-
able while X are the decision factors 161 drawn from Mitchell (1997). Conditional
independence enables the simpler representation within the product term. P(Y = yk|Xi) =
P(Y = yk)
N
i=1
P(Xi|Y = yk)
jP(Y = yj)
N
i=1
P(Xi|Y = yj)
. RESULTS AND DISCUSSION Sambu (2015), PeerJ, DOI 10.7717/peerj.1039 RESULTS AND DISCUSSION A DTL analysis focussed on material and process choices From Fig. 1, processes involving the incorporation of sugars improve survival as is
captured by the terminal leaves labelled “Trehalose” and “No Sugars” with a 9% delta
across both predictions coming after the introduction of a natural biomaterial. The Sambu (2015), PeerJ, DOI 10.7717/peerj.1039 5/15 5/15 incorporation of sucrose into hydrophilic biomaterials modulates porosity e.g., in the
alleviation of decreased porosity caused by material compression (Huang et al., 2003). In the context of cryopreservation, the addition of sucrose helped create further pore
interconnections in the final scaffold and/or create new voids. This is primarily because
of the inhibition of DMSO action by van der Waals (VWF) and hydrogen-bonding
interactions between sugars, water and DMSO. These sugar-DMSO and water-DMSO
interactions are expected because DMSO is prone to act as a hydrogen bond acceptor
with both sugar and water molecules (Huang et al., 2003). The addition of sucrose as
a component of the alginate capsule and as part of the cryopreservative improves the
post-thaw cell viability as previous studies of such systems have shown that the effective
concentration of water and the fraction of freezable water are diminished for alginate
concentrations greater that 0.5% (w/w) (Pongsawatmanit, Ikeda & Miyawaki, 1999). This
is consistent with diminished ice formation and freeze-concentration. The addition of
sugars to the alginate scaffold was thought to be necessary since mass transfer barriers are
eliminated and the cryoprotective benefits of sugar were expected to lower the required
minimum DMSO exposure times and intracellular DMSO content (MacGregor, 1967). Whether in the cryopreservative or in the scaffold, the mechanism for cryoprotection
involves hydrogen bonding between atoms within hydroxyl group of the sugar, DMSO
and water. Sucrose can act as a hydrogen bond donor to both DMSO and water. DMSO
is known to isolate water molecules and to induce hydrogen bonding (Sato et al., 2004). This may be related to the interaction between trehalose and alginate solutions; it has
already been established that sugars impact the swelling ratios of hydrogels and by the same
token, the survival of encapsulated cells (Selmer-Olsen et al., 1999). The higher swelling
ratio of the trehalose-alginate system is linked to the lower impact trehalose has on the
effective concentration of water available for alginate swelling. Sambu (2015), PeerJ, DOI 10.7717/peerj.1039 RESULTS AND DISCUSSION This will have an impact
in the cryopreservation of cells since the relatively higher effective concentration of water
in the alginate-trehalose system will increase the probability for ice formation which is
lethal to cells. Overall, it is expected that two measurable quantities i.e., the fraction of
freezable water and the effective concentration of water in alginate/trehalose scaffolds will
be higher (Martinsen, Storrø & Skj˚ark-Bræk, 1992); this will in turn mean that the level of
cryoprotection in these scaffolds will be lower than that observed for sucrose consistent
with experimental observations and DTL analysis. Given the differentials in cell survival for 3D capsules, the provision of optimized
cryoprotection throughout the scaffold has been shown to improve overall cell survival
and to minimize differentials along the construct radius (Cui et al., 2002); this will ensure
that excessively long exposure times are eliminated from the cryopreservation process
(Sambu et al., 2011). Comparable results indicate that cryopreservation requirements
place even greater diffusion limits than do nutrient and oxygen perfusion limits (<2 mm)
(Griffith et al., 2005); this is corroborated by the 16% (see “Random seeding” vs. “Controlled
Seeding” means in Fig. 1) increase caused by the controlled seeding of cells at “location” of
cells within the 3D construct for optimal nutrient, oxygen and CPA concentrations. Sambu (2015), PeerJ, DOI 10.7717/peerj.1039 6/15 Figure 2 Tracking the coefficient of determination as the DTL model is built. A line plot of the
R-squared for each branch added at each step in the generation of the decision tree learning model. The prediction is the cell survival (a continuous variable) after cryopreservation. Figure 2 Tracking the coefficient of determination as the DTL model is built. A line plot of the
R-squared for each branch added at each step in the generation of the decision tree learning model. The prediction is the cell survival (a continuous variable) after cryopreservation. Figure 2 Tracking the coefficient of determination as the DTL model is built. A line plot of the
R-squared for each branch added at each step in the generation of the decision tree learning model. The prediction is the cell survival (a continuous variable) after cryopreservation. Figure 2 Tracking the coefficient of determination as the DTL model is built. A line plot of the
R-squared for each branch added at each step in the generation of the decision tree learning model. RESULTS AND DISCUSSION The prediction is the cell survival (a continuous variable) after cryopreservation. From Fig. 1, there are competitive alternatives which are found on the right-most
termini of the decision tree with the use of RGD integrins and controlled seeding breaking
the 80% barrier and with low p-values. These alternatives are optimal in their own respect;
however, overall, the employment of integrins is by far the best approach. From Fig. 2, the DTL approach is locally optimal and can often be visualized by plotting
the prediction of nodes with very low p-values. However, there is a limit to the DTLA
approach in that generalization cannot be delivered as there are no predictions outside the
very local and very limited scope. As such, the proper diagnostic is the variable R-squared
at each branch as shown in Fig. 2. Sambu (2015), PeerJ, DOI 10.7717/peerj.1039 An augmented DTL analysis focussed on thermal history, material
and process choices Further analysis with augmented data (Kashuba Benson, Benson & Critser, 2008) via the
DTLA enabled further refinement of the decision process per Fig. 3 which shows that
the local maximum is sustained against new process variables related to thermal history
i.e., cooling rates and plunge temperatures. Given the branching mode, it is clear that
process variables lead to a different decision-making path than the matrix related variables. Of note, changes on process variables lead to significant changes in the survival (see
14% increase on Plunge Temperature <−60.5 ◦C). In Fig. 3, a meta-data analysis shows
a survival-optimized termini with RGD-modified natural biomaterials and controlled
seeding in capsules with >80% survival and low p-values. However, the limit in this
heuristic is in generalization i.e., no predictions can be made outside the local and very
limited original scope of the specific combinations within the meta-data. Sambu (2015), PeerJ, DOI 10.7717/peerj.1039 7/15 Figure 3 A Decision Tree Learning Analysis of cryopreservation data with branch-validation tracker underneath. (A) A metadata analysis of
cryopreservation data showing that the right-side analysis is conserved across methods—Temperatures are in ◦C and Cooling rates in ◦C/min; (B)
plot showing coefficient of determination during model validation; the corresponding classification scores for test data are at 84% correct match
and 16% misclassification. Figure 3 A Decision Tree Learning Analysis of cryopreservation data with branch-validation tracker underneath. (A) A metadata analysis of
cryopreservation data showing that the right-side analysis is conserved across methods—Temperatures are in ◦C and Cooling rates in ◦C/min; (B)
plot showing coefficient of determination during model validation; the corresponding classification scores for test data are at 84% correct match
and 16% misclassification. Figure 3 A Decision Tree Learning Analysis of cryopreservation data with branch-validation tracker underneath. (A) A metadata analysis of
cryopreservation data showing that the right-side analysis is conserved across methods—Temperatures are in ◦C and Cooling rates in ◦C/min; (B)
plot showing coefficient of determination during model validation; the corresponding classification scores for test data are at 84% correct match
and 16% misclassification. Before a generalization can be developed, an examination of patterns to follow in the
construction of the generalizable classifier is required— Fig. 4 captures a comparison of
the meta-data to be used in generating the generalizable classifier. From the graph, the
variables i.e., Step-loading, integrins and location are particularly pivotal in the attainment
of high post-thaw cell survival rates. Sambu (2015), PeerJ, DOI 10.7717/peerj.1039 An augmented DTL analysis focussed on thermal history, material
and process choices One additional factor, i.e., sugars, when intersected
with location also provided one of the 7 graphs with exceptional survival rates. In setting Sambu (2015), PeerJ, DOI 10.7717/peerj.1039 8/15 Figure 4 A summary scatter-plot analysis of key values for meta-data on main decision factors. A
scatter plot matrix capturing the summary statistics for the meta-data collected from previous analyses
used to develop a heuristic for predicting the posterior survival probabilities for cells for a given set
of process decisions (Sambu et al., 2011; Heng, Yu & Ng, 2004; Kashuba Benson, Benson & Critser,
2008; Miszta-Lane et al., 2007). Key decision variable combinations with 100% “live” cells are framed
in purple. The numbers on each axis represent the categories identified within a decision variable. Red
dots represent short-living cells while green dots represent long-living cells. Figure 4 A summary scatter-plot analysis of key values for meta-data on main decision factors. A
scatter plot matrix capturing the summary statistics for the meta-data collected from previous analyses
used to develop a heuristic for predicting the posterior survival probabilities for cells for a given set
of process decisions (Sambu et al., 2011; Heng, Yu & Ng, 2004; Kashuba Benson, Benson & Critser,
2008; Miszta-Lane et al., 2007). Key decision variable combinations with 100% “live” cells are framed
in purple. The numbers on each axis represent the categories identified within a decision variable. Red
dots represent short-living cells while green dots represent long-living cells. up the meta-data sample for the generalizable NBC classifier, the selection of both
robust technology-enabled cooling protocols e.g., using controlled-rate cooling and
techniques using cost-efficient techniques with validated cooling rates was essential to
avoid a pre-selection bias and to ensure that the predictive classifier could elicit the right
probabilistic nodes from root-to-leaf. Therefore, studies e.g., by Heng, Yu & Ng (2004) were
deliberately included with salient descriptors for protocol choices such as containment and
cooling rate specified as per authors’ note (these rates were verified via a thermal probe)
(Heng, Yu & Ng, 2004). The results in latter paragraphs were examined against these data
samples and the marginal descriptors introduced to capture methodical improvisations. Figure 5 shows how a change in survival against the plunge temperature set point
significantly changes the terminal/‘leaf’ value. This is the temperature at which the
cryovials will be immersed in liquid nitrogen for long-term storage. From Fig. An augmented DTL analysis focussed on thermal history, material
and process choices 5, the plunge
temperature set point is at −60 ◦C—while this may be applicable to the specific protocols
used (Kashuba Benson, Benson & Critser, 2008), there may be difficulties in generalization
outside of the cell type used unless the models developed are able to learn across this data
set. The importance of plunge temperature can be explained using the central crybiological up the meta-data sample for the generalizable NBC classifier, the selection of both
robust technology-enabled cooling protocols e.g., using controlled-rate cooling and
techniques using cost-efficient techniques with validated cooling rates was essential to
avoid a pre-selection bias and to ensure that the predictive classifier could elicit the right
probabilistic nodes from root-to-leaf. Therefore, studies e.g., by Heng, Yu & Ng (2004) were
deliberately included with salient descriptors for protocol choices such as containment and
cooling rate specified as per authors’ note (these rates were verified via a thermal probe)
(Heng, Yu & Ng, 2004). The results in latter paragraphs were examined against these data
samples and the marginal descriptors introduced to capture methodical improvisations. Figure 5 shows how a change in survival against the plunge temperature set point
significantly changes the terminal/‘leaf’ value. This is the temperature at which the
cryovials will be immersed in liquid nitrogen for long-term storage. From Fig. 5, the plunge
temperature set point is at −60 ◦C—while this may be applicable to the specific protocols
used (Kashuba Benson, Benson & Critser, 2008), there may be difficulties in generalization
outside of the cell type used unless the models developed are able to learn across this data
set. The importance of plunge temperature can be explained using the central crybiological up the meta-data sample for the generalizable NBC classifier, the selection of both
robust technology-enabled cooling protocols e.g., using controlled-rate cooling and
techniques using cost-efficient techniques with validated cooling rates was essential to
avoid a pre-selection bias and to ensure that the predictive classifier could elicit the right
probabilistic nodes from root-to-leaf. Therefore, studies e.g., by Heng, Yu & Ng (2004) were
deliberately included with salient descriptors for protocol choices such as containment and
cooling rate specified as per authors’ note (these rates were verified via a thermal probe)
(Heng, Yu & Ng, 2004). Sambu (2015), PeerJ, DOI 10.7717/peerj.1039 An augmented DTL analysis focussed on thermal history, material
and process choices The results in latter paragraphs were examined against these data
pl
d th
i
l d
ipt
i t
d
d t
pt
th di
l i
p
i ti Sambu (2015), PeerJ, DOI 10.7717/peerj.1039 9/15 9/15 Figure 5 A 2D plot zooming in on a classification boundary for the plunge temperature. (A) A change
in survival against the plunge temperature set point against which the prediction changes the ‘leaf’ value. (B) A 2D plot of the survival rate against plunge temperature was drawn from a DTL analysis. The cooling
rate is used to lower the temperature from CPA loading temperature (temperature when cells are exposed
to CPA for equilibration) to the plunge temperature—this rate is a separate decision factor (see methods
section). Figure 5 A 2D plot zooming in on a classification boundary for the plunge temperature. (A) A change
in survival against the plunge temperature set point against which the prediction changes the ‘leaf’ value. (B) A 2D plot of the survival rate against plunge temperature was drawn from a DTL analysis. The cooling
rate is used to lower the temperature from CPA loading temperature (temperature when cells are exposed
to CPA for equilibration) to the plunge temperature—this rate is a separate decision factor (see methods
section). 7717/peerj.1039
10/15 Figure 5 A 2D plot zooming in on a classification boundary for the plunge temperature. (A) A change
in survival against the plunge temperature set point against which the prediction changes the ‘leaf’ value. (B) A 2D plot of the survival rate against plunge temperature was drawn from a DTL analysis. The cooling
rate is used to lower the temperature from CPA loading temperature (temperature when cells are exposed
to CPA for equilibration) to the plunge temperature—this rate is a separate decision factor (see methods
section). Sambu (2015), PeerJ, DOI 10.7717/peerj.1039
10/15 Figure 5 A 2D plot zooming in on a classification boundary for the plunge temperature. (A) A change
in survival against the plunge temperature set point against which the prediction changes the ‘leaf’ value. (B) A 2D plot of the survival rate against plunge temperature was drawn from a DTL analysis. The cooling
rate is used to lower the temperature from CPA loading temperature (temperature when cells are exposed
to CPA for equilibration) to the plunge temperature—this rate is a separate decision factor (see methods
section). Sambu (2015), PeerJ, DOI 10.7717/peerj.1039 An augmented DTL analysis focussed on thermal history, material
and process choices Figure 5 A 2D plot zooming in on a classification boundary for the plunge temperature. (A) A change
in survival against the plunge temperature set point against which the prediction changes the ‘leaf’ value. (B) A 2D plot of the survival rate against plunge temperature was drawn from a DTL analysis. The cooling
rate is used to lower the temperature from CPA loading temperature (temperature when cells are exposed
to CPA for equilibration) to the plunge temperature—this rate is a separate decision factor (see methods
section). Figure 5 A 2D plot zooming in on a classification boundary for the plunge temperature. (A) A change
in survival against the plunge temperature set point against which the prediction changes the ‘leaf’ value. (B) A 2D plot of the survival rate against plunge temperature was drawn from a DTL analysis. The cooling
rate is used to lower the temperature from CPA loading temperature (temperature when cells are exposed
to CPA for equilibration) to the plunge temperature—this rate is a separate decision factor (see methods
section). Sambu (2015), PeerJ, DOI 10.7717/peerj.1039 Figure 6 A nested summary table of the NBC model. Prediction of survival using the Na¨ıve Bayes
Classifier on the following factors: dimensionality, location, biomaterial, sugars, step-loading & integrins. The training set has a 79% accuracy while the smaller test set has an 89% accuracy showing that the Na¨ıve
Bayes Classifier is a demonstrably accurate predictor even when reduced to a smaller test set where the
samples are all cryopreserved in 3D natural RGD-containing biomaterials using controlled slow-cooling
with sucrose for lyoprotection, in a two-step loading process. (Similar to the optimal leaves from the
recursive partitioning trees developed earlier.) Cross-validations are 10-fold (i.e., k = 0.1 for the NBC
model construct). Figure 6 A nested summary table of the NBC model. Prediction of survival using the Na¨ıve Bayes
Classifier on the following factors: dimensionality, location, biomaterial, sugars, step-loading & integrins. The training set has a 79% accuracy while the smaller test set has an 89% accuracy showing that the Na¨ıve
Bayes Classifier is a demonstrably accurate predictor even when reduced to a smaller test set where the
samples are all cryopreserved in 3D natural RGD-containing biomaterials using controlled slow-cooling
with sucrose for lyoprotection, in a two-step loading process. (Similar to the optimal leaves from the
recursive partitioning trees developed earlier.) Cross-validations are 10-fold (i.e., k = 0.1 for the NBC
model construct). Sambu (2015), PeerJ, DOI 10.7717/peerj.1039 An augmented DTL analysis focussed on thermal history, material
and process choices theory which states that for the cell to survive cryopreservation, intracellular ice formation
(IIF) must be suppressed while intracellular vitrification must be fostered through a
programmatic replacement of free water with cryoprotective chemistry—therefore, plunge
temperatures that are too high (above the homogenous nucleation temperature of water,
−40 ◦C) will result in IIF while plunge temperatures that are too low will result in cell
dehydration via ex-osmosis (Meryman, 2007). Figure 6 shows results from the NBC method which allows generalization across the
factors used whereby new combinations of otherwise conditionally independent (given
its classification) decision factors can be made and evaluated for posterior probabilities. The assumption of conditional independence stems from the observation that once the
classification is computed, then for any two factors, the value of one factor does little
to inform the model of the state of the other factor or its contribution to the target
variable (i.e., “survival”). These predictive syntheses demonstrate a generalized accuracy
of 79% across the meta-data while the accuracy for predictive analytics for a specified Sambu (2015), PeerJ, DOI 10.7717/peerj.1039 case whereby samples are all cryopreserved in 3D natural RGD-containing matrices, using
controlled slow-cooling with sucrose for lyoprotection, in a two-step loading process
shows an 89% accuracy which meets the needs for evaluation of most-likely protocol
choices. The “two step-loading” qualifier refers to the process used to get the CPA into
the cells whereby cells are exposed to a low CPA concentration and then a higher CPA
concentration prior to the commencement of cooling to cryogenic temperatures. 2-step
loading is preferred over 1-step loading because cells are less prone to damage from
sudden ex-osmosis of water and equilibration is ensured prior to cooling leading to better
results upon recovery (Santis et al., 2007). After these results were derived, an analysis
of recent published and peer-reviewed data showed that loss of RGD-responsive cell
attachments during cryopreservation is a leading cause of cell loss (Terry et al., 2007). Furthermore, there are strong indications that sucrose as a porogen helps provide in
vivo-like conditions in a step-wise process during tissue fabrication (Verhulsel et al., 2014). Synthesis A synthesis of the previous analyses shows that the use integrin ligands improve cell
survival sharply—this result has been verified in more recent publications along with 4
major decision variables mentioned in the methods section (Ahmad & Sambanis, 2013; Li
et al., 2015). However, where cost may be a limiting factor, the use of sugars and a careful
biomaterial choice may suffice as alternatives confirmed by recursive partitioning (though
only as far as survival is concerned; maintenance of cell character e.g. pluripotency may
be a different matter). This result is definitively captured by the decision tree model which
shows that the lead attributes for cryopreservation are: sugars, integrins, dimensionality
and location according to the information-gain these attributes afford at each cycle of
recursive testing. An augmented DTL analysis focussed on thermal history, material
and process choices Subsequent to the derivation of these initial results, further work has shown that expansion
and cryopreservation in 3D capsule matrices for cellular aggregates is demonstrable for
human embryonic stem cells further confirming that decision choices that encompass
cell attachment (similar in function to RGD-containing matrices), slow-cooling, matrix
permeability, natural biomaterials (e.g., alginate) enhance cell viability (Serra et al., 2011). Sambu (2015), PeerJ, DOI 10.7717/peerj.1039 Competing Interests The author declares there is no competing interests. Funding The author declares there was no funding for this work. Author Contributions • Sammy Sambu conceived and designed the analytic approach, performed the model
building, calibration and validation, wrote the paper, prepared figures and/or tables,
reviewed drafts of the paper. • Sammy Sambu conceived and designed the analytic approach, performed the model
building, calibration and validation, wrote the paper, prepared figures and/or tables,
reviewed drafts of the paper. Sambu (2015), PeerJ, DOI 10.7717/peerj.1039 CONCLUSION In conclusion, DTL and NBC were employed on cryopreservation meta-data with
NBC demonstrating greater generalizability. These results indicate that the handles
for improving cryopreservation outcomes are: integrin-mediated cryopreservation,
the modification (by entrapment of benign CPAs or other means) of the scaffolding
material and the modification of cell location in scaffolds. The DTL and NBC models
were demonstrated to be robust against tougher validation data from different cell types
thereby confirming that cryopreservation/bio-preservation technology can be improved
upon using these approaches. Sambu (2015), PeerJ, DOI 10.7717/peerj.1039 REFERENCES Ahmad HF, Sambanis A. 2013. Cryopreservation effects on recombinant myoblasts encapsulated
in adhesive alginate hydrogels. Acta Biomaterialia 9(6):6814–6822
DOI 10.1016/j.actbio.2013.03.002. Ahmad HF, Sambanis A. 2013. Cryopreservation effects on recombinant myoblasts encapsulated
in adhesive alginate hydrogels. Acta Biomaterialia 9(6):6814–6822
DOI 10.1016/j.actbio.2013.03.002. Anchordoguy TJ, Rudolph AS, Carpenter JF, Crowe JH. 1987. Modes of interaction of
cryoprotectants with membrane phospholipids during freezing. Cryobiology 24(4):324–331
DOI 10.1016/0011-2240(87)90036-8. Anchordoguy TJ, Rudolph AS, Carpenter JF, Crowe JH. 1987. Modes of interaction of
cryoprotectants with membrane phospholipids during freezing. Cryobiology 24(4):324–331
DOI 10.1016/0011-2240(87)90036-8. Arakawa T, Carpenter JF, Kita YA, Crowe JH. 1990. The basis for toxicity of certain
cryoprotectants: a hypothesis. Cryobiology 27(1):401–415 DOI 10.1016/0011-2240(90)90017-X. Arakawa T, Timasheff SN. 1982. Stabilization of protein structure by sugars. Biochemistry (Mosc)
21(25):6536–6544 DOI 10.1021/bi00268a033. Cui ZF, Dykhuizen RC, Nerem RM, Sembanis A. 2002. Modeling of cryopreservation of
engineered tissues with one-dimensional geometry. Biotechnology Progress 18(2):354–361
DOI 10.1021/bp0101886. Friedman JH. 2001. Greedy function approximation: a gradient boosting machine. The Annals of
Statistics 29(5):1189–1232 DOI 10.1214/aos/1013203451. F¨urnkranz J. 1997. Pruning algorithms for rule learning. Machine Learning 27(2):139–172
DOI 10.1023/A:1007329424533. Gibson MI, Barker CA, Spain SG, Albertin L, Cameron NR. 2009. Inhibition of ice crystal growth
by synthetic glycopolymers: implications for the rational design of antifreeze glycoprotein
mimics. Biomacromolecules 10(2):328–333 DOI 10.1021/bm801069x. Griffith CK, Miller C, Sainson RCA, Calvert JW, Jeon NL, Hughes CCW, George SC. 2005. Diffusion limits of an in vitro thick prevascularized tissue. Tissue Engineering 11(1–2):257–266
DOI 10.1089/ten.2005.11.257. Heng BC, Yu Y-JH, Ng SC. 2004. Slow-cooling protocols for microcapsule cryopreservation. Journal of Microencapsulation 21(4):455–467 DOI 10.1080/02652040410001729250. Huang YC, Connell M, Park Y, Mooney DJ, Rice KG. 2003. Fabrication and in vitro testing of
polymeric delivery system for condensed DNA. Journal of Biomedical Materials Research
67(4):1384–1392 DOI 10.1002/jbm.a.20036. Sambu (2015), PeerJ, DOI 10.7717/peerj.1039 13/15 Kashuba Benson CM, Benson JD, Critser JK. 2008. An improved cryopreservation method for a
mouse embryonic stem cell line. Cryobiology 56(2):120–130
DOI 10.1016/j.cryobiol.2007.12.002. Li N, Zhang Y, Xiu Z, Wang Y, Chen L, Wang S, Li S, Guo X, Ma X. 2015. The preservation of
islet with alginate encapsulation in the process of transportation. Biotechnology and Applied
Biochemistry. Epub ahead of print 7 Jan 2015 DOI 10.1002/bab.1295. Liu Y. 2007. Cryopreservation of Mouse embryonic Stem Cells. Master’s thesis,
Oxford University. MacGregor WS. 1967. The chemical and physical properties of DMSO. Annals of the New York
Academy of Sciences 141:3–12 DOI 10.1111/j.1749-6632.1967.tb34860.x. Martinsen A, Storrø I, Skj˚ark-Bræk G. Sambu (2015), PeerJ, DOI 10.7717/peerj.1039 REFERENCES 1992. Alginate as immobilization material: III. Diffusional
properties. Biotechnology and Bioengineering 39(2):186–194 DOI 10.1002/bit.260390210. Meryman HT. 2007. Cryopreservation of living cells: principles and practice. Transfusion (Paris)
47(5):935–945 DOI 10.1111/j.1537-2995.2007.01212.x. Miszta-Lane H, Gill P, Mirbolooki M, Lakey JRT. 2007. Effect of slow freezing versus vitrification
on the recovery of mouse embryonic stem cells. Cell Preservation Technology 5(1):16–24
DOI 10.1089/cpt.2006.9990. Mitchell TM. 1997. Machine learning. Vol. 45. Burr Ridge: McGraw Hill. Muldrew K, Novak K, Yang H, Zernicke R, Schachar NS, McGann LE. 2000. Cryobiology of
articular cartilage: ice morphology and recovery of chondrocytes. Cryobiology 40(2):102–109
DOI 10.1006/cryo.2000.2236. Pongsawatmanit R, Ikeda S, Miyawaki O. 1999. Effect of sucrose on physical properties of
alginate dispersed aqueous systems. Food Science and Technology Research 5(2):183–187
DOI 10.3136/fstr.5.183. Prestrelski SJ, Arakawa T, Carpenter JF. 1993. Separation of freezing- and drying-induced
denaturation of lyophilized proteins using stress-specific stabilization. II. Structural
studies using infrared spectroscopy. Archives of Biochemistry and Biophysics 303(2):465–473
DOI 10.1006/abbi.1993.1310. Roos Y, Karel M. 1991. Amorphous state and delayed ice formation in sucrose solutions. International Journal of Food Science and Technology 26(6):553–566 Rubinstein YD, Hastie T. 1997. Discriminative vs informative learning. In: KDD, vol. 5. 49–53. Available at http://www.aaai.org/Papers/KDD/1997/KDD97-008.pdf. Rudolph AS, Crowe JH. 1985. Membrane stabilization during freezing: the role of two natural
cryoprotectants, trehalose and proline. Cryobiology 22(4):367–377 Sambu S, Xu X, Schiffter H, Cui Z, Ye H. 2011. RGDS-fuctionalized alginates improve the survival
rate of encapsulated embryonic stem cells during cryopreservation. Cryoletters 32(5):389–401. Sambu S, Xu X, Ye H, Cui ZF. 2011. Predicting the survival rate of mouse embryonic stem
cells cryopreserved in alginate beads. Proceedings of the Institution of Mechanical Engineers
225(11):1092–1107 DOI 10.1177/0954411911418568. Santis LD, Coticchio G, Paynter S, Albertini D, Hutt K, Cino I, Iaccarino M, Gambardella A,
Flamigni C, Borini A. 2007. Permeability of human oocytes to ethylene glycol and their
survival and spindle configurations after slow cooling cryopreservation. Human Reproduction
22(10):2776–2783 DOI 10.1093/humrep/dem240. Sambu (2015), PeerJ, DOI 10.7717/peerj.1039 14/15 Sato Y, Kawabuchi S, Irimoto Y, Miyawaki O. 2004. Effect of water activity and solvent-ordering
on intermolecular interaction of high-methoxyl pectins in various sugar solutions. Food
Hydrocolloids 18(4):527–534 DOI 10.1016/j.foodhyd.2003.09.001. Selmer-Olsen E, Sørhaug T, Birkeland SE, Pehrson R. 1999. Survival of Lactobacillus helveticus
entrapped in Ca-alginate in relation to water content, storage and rehydration. Journal of
Industrial Microbiology and Biotechnology 23(2):79–85 DOI 10.1038/sj.jim.2900693. Serra M, Correia C, Malpique R, Brito C, Jensen J, Bjorquist P, Carrondo MJT, Alves PM. Sambu (2015), PeerJ, DOI 10.7717/peerj.1039 REFERENCES 2011. Microencapsulation technology: a powerful tool for integrating expansion and cryopreservation
of human embryonic stem cells. PLoS ONE 6(8):e23212 DOI 10.1371/journal.pone.0023212. Shirakashi R, M¨uller KJ, Sukhorukov VL, Zimmermann U. 2003. Effects of physiological isotonic
cryoprotectants on living cells during the freezing-thawing process and effects of their uptake
by electroporation: Sp 2 cells in alginate-trehalose solutions. Heat Transfer Asian Research
32(6):511–523 DOI 10.1002/htj.10105. Terry C, Hughes RD, Mitry RR, Lehec SC, Dhawan A. 2007. Cryopreservation-induced
nonattachment of human hepatocytes: role of adhesion molecules. Cell Transplantation
16(6):639–647 DOI 10.3727/000000007783465000. Therneau T, Atkinson B. 2015. port) BR (author of initial R. rpart: Recursive Partitioning and
Regression Trees. Available at http://cran.r-project.org/web/packages/rpart/index.html (accessed
13 February 2015). Verhulsel M, Vignes M, Descroix S, Malaquin L, Vignjevic DM, Viovy J-L. 2014. A review
of microfabrication and hydrogel engineering for micro-organs on chips. Biomaterials
35(6):1816–1832 DOI 10.1016/j.biomaterials.2013.11.021. 15/15
|
https://openalex.org/W2109846272
|
https://curis.ku.dk/ws/files/154439649/Jarvis_2015_Phylogenomic_analyses.pdf
|
English
| null |
Phylogenomic analyses data of the avian phylogenomics project
|
Gigascience
| 2,015
|
cc-by
| 5,638
|
university of copenhagen university of copenhagen
Phylogenomic analyses data of the avian phylogenomics project
Jarvis, Erich D; Mirarab, Siavash; Aberer, Andre J; Li, Bo; Houde, Peter; Li, Cai; Ho, Simon Y. W.; Faircloth, Brant C.; Nabholz, Benoit; Howard, Jason T.; Suh, Alexander; Weber, Claudia
C.; Rodrigues da Fonseca, Rute Andreia; Alfaro Nuñez, Luis Alonso; Narula, Nitish; Liu,
Liang; Burt, Dave; Ellegren, Hans; Edwards, Scott V.; Stamatakis, Alexandros; Mindell, David
P.; Cracraft, Joel; Braun, Edward L.; Warnow, Tandy; Wang, Jun; Gilbert, M. Thomas P.;
Zhang, Guojie
Published in:
GigaScience university of copenhagen Citation for published version (APA):
Jarvis, E. D., Mirarab, S., Aberer, A. J., Li, B., Houde, P., Li, C., Ho, S. Y. W., Faircloth, B. C., Nabholz, B.,
Howard, J. T., Suh, A., Weber, C. C., Rodrigues da Fonseca, R. A., Alfaro Nuñez, L. A., Narula, N., Liu, L., Burt,
D., Ellegren, H., Edwards, S. V., ... Zhang, G. (2015). Phylogenomic analyses data of the avian phylogenomics
project. GigaScience, 4, [4]. https://doi.org/10.1186/s13742-014-0038-1 Download date: 24. Oct. 2024 Phylogenomic analyses data of the avian phylogenomics project Jarvis, Erich D; Mirarab, Siavash; Aberer, Andre J; Li, Bo; Houde, Peter; Li, Cai; Ho, Simon Y. W.; Faircloth, Brant C.; Nabholz, Benoit; Howard, Jason T.; Suh, Alexander; Weber, Claudia
C.; Rodrigues da Fonseca, Rute Andreia; Alfaro Nuñez, Luis Alonso; Narula, Nitish; Liu,
Liang; Burt, Dave; Ellegren, Hans; Edwards, Scott V.; Stamatakis, Alexandros; Mindell, David
P.; Cracraft, Joel; Braun, Edward L.; Warnow, Tandy; Wang, Jun; Gilbert, M. Thomas P.;
Zhang, Guojie
Published in:
GigaScience DOI:
10.1186/s13742-014-0038-1 Document version
Publisher's PDF, also known as Version of record Document version
Publisher's PDF, also known as Version of record Document license:
CC BY Citation for published version (APA):
Jarvis, E. D., Mirarab, S., Aberer, A. J., Li, B., Houde, P., Li, C., Ho, S. Y. W., Faircloth, B. C., Nabholz, B.,
Howard, J. T., Suh, A., Weber, C. C., Rodrigues da Fonseca, R. A., Alfaro Nuñez, L. A., Narula, N., Liu, L., Burt,
D., Ellegren, H., Edwards, S. V., ... Zhang, G. (2015). Phylogenomic analyses data of the avian phylogenomics
project. GigaScience, 4, [4]. https://doi.org/10.1186/s13742-014-0038-1 Citation for published version (APA):
Jarvis, E. D., Mirarab, S., Aberer, A. J., Li, B., Houde, P., Li, C., Ho, S. Y. W., Faircloth, B. C., Nabholz, B.,
Howard, J. T., Suh, A., Weber, C. C., Rodrigues da Fonseca, R. A., Alfaro Nuñez, L. A., Narula, N., Liu, L., Burt,
D., Ellegren, H., Edwards, S. V., ... Zhang, G. (2015). Phylogenomic analyses data of the avian phylogenomics
project. GigaScience, 4, [4]. https://doi.org/10.1186/s13742-014-0038-1 Download date: 24. Oct. 2024 DATA NOTE Open Access Abstract Background: Determining the evolutionary relationships among the major lineages of extant birds has been one
of the biggest challenges in systematic biology. To address this challenge, we assembled or collected the genomes
of 48 avian species spanning most orders of birds, including all Neognathae and two of the five Palaeognathae
orders. We used these genomes to construct a genome-scale avian phylogenetic tree and perform comparative
genomic analyses. Findings: Here we present the datasets associated with the phylogenomic analyses, which include sequence alignment
files consisting of nucleotides, amino acids, indels, and transposable elements, as well as tree files containing gene trees
and species trees. Inferring an accurate phylogeny required generating: 1) A well annotated data set across species based
on genome synteny; 2) Alignments with unaligned or incorrectly overaligned sequences filtered out; and 3) Diverse data
sets, including genes and their inferred trees, indels, and transposable elements. Our total evidence nucleotide tree (TENT)
data set (consisting of exons, introns, and UCEs) gave what we consider our most reliable species tree when using the
concatenation-based ExaML algorithm or when using statistical binning with the coalescence-based MP-EST algorithm
(which we refer to as MP-EST*). Other data sets, such as the coding sequence of some exons, revealed other properties of
genome evolution, namely convergence. Conclusions: The Avian Phylogenomics Project is the largest vertebrate phylogenomics project to date that we are
aware of. The sequence, alignment, and tree data are expected to accelerate analyses in phylogenomics and other
related areas. Keywords: Avian genomes, Phylogenomics, Sequence alignments, Species tree, Gene trees, Indels, Transposable
elements Keywords: Avian genomes, Phylogenomics, Sequence alignments, Species tree, Gene trees, Indels, Transposable
elements * Correspondence: jarvis@neuro.duke.edu; warnow@illinois.edu;
wangj@genomics.org.cn; mtpgilbert@gmail.com; zhanggj@genomics.cn
†Equal contributors
1Department of Neurobiology, Howard Hughes Medical Institute and Duke
University Medical Center, Durham, NC 27710, USA
2Department of Computer Science, The University of Texas at Austin, Austin,
TX 78712, USA
4China National GeneBank, BGI-Shenzhen, Shenzhen 518083, China
6Centre for GeoGenetics, Natural History Museum of Denmark, University of
Copenhagen, Øster Voldgade 5-7, 1350 Copenhagen, Denmark
Full list of author information is available at the end of the article © 2015 Jarvis et al.; licensee BioMed Central. This is an Open Access article distributed under the terms of the Creative
Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and
reproduction in any medium, provided the original work is properly credited. Phylogenomic analyses data of the avian
phylogenomics project Erich D Jarvis1*†, Siavash Mirarab2†, Andre J Aberer3, Bo Li4,5,6, Peter Houde7, Cai Li4,6, Simon Y W Ho8, Brant C Faircloth9,10,
Benoit Nabholz11, Jason T Howard1, Alexander Suh12, Claudia C Weber12, Rute R da Fonseca6, Alonzo Alfaro-Núñez6,
Nitish Narula7,13, Liang Liu14, Dave Burt15, Hans Ellegren12, Scott V Edwards16, Alexandros Stamatakis3,17, David P Mindell18,
Joel Cracraft19, Edward L Braun20, Tandy Warnow2*, Wang Jun4,21,22,23,24*, M Thomas Pius Gilbert6,25*,
Guojie Zhang4,26* and The Avian Phylogenomics Consortium * Correspondence: jarvis@neuro.duke.edu; warnow@illinois.edu;
wangj@genomics.org.cn; mtpgilbert@gmail.com; zhanggj@genomics.cn
†Equal contributors
1Department of Neurobiology, Howard Hughes Medical Institute and Duke
University Medical Center, Durham, NC 27710, USA
2Department of Computer Science, The University of Texas at Austin, Austin,
TX 78712, USA
4China National GeneBank, BGI-Shenzhen, Shenzhen 518083, China
6Centre for GeoGenetics, Natural History Museum of Denmark, University of
Copenhagen, Øster Voldgade 5-7, 1350 Copenhagen, Denmark
Full list of author information is available at the end of the article Jarvis et al. GigaScience (2015) 4:4
DOI 10.1186/s13742-014-0038-1 Phylogenomic analyses data of the avia Phylogenomic analyses data of the avian
phylogenomics project
Jarvis et al. Jarvis et al. GigaScience (2015) 4:4
DOI 10.1186/s13742-014-0038-1 High and low variance introns and exons These four data sets represent the 10% subsets of the
8295 exons and their associated introns when available
(i.e. from the same genes) that had the highest and lowest
variance in GC3 (third codon position) content across spe-
cies. To calculate GC3 variance, we first calculated GC3
for each ortholog in each species, and then we used the
correlation coefficient R to calculate variance in GC3 for
each species. Orthologs were ranked by their GC3 variance
and we selected the top and bottom 10% for analyses. Abstract The Creative Commons Public Domain
Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article,
unless otherwise stated. Jarvis et al. GigaScience (2015) 4:4 2516 intron gene set This is an orthologous subset of introns from the 8295
protein-coding genes among 52 species (includes out-
groups). Introns with conserved annotated exon-intron
boundaries between chicken and another species (±1
codon) were chosen. We filtered out introns with
length < 50 bp or intron length ratio > 1.5 between
chicken and another species or another species and
chicken. This filtering resulted in a conservative subset
of introns that could be reliably identified and aligned. 8295 protein-coding exon gene set 8295 protein-coding exon gene set
This is an exon-coding sequence data set of 8295 genes
based on synteny-defined orthologs we identified and se-
lected from the assembled genomes of chicken and
zebra finch [8,9]. We required these loci to be present in
at least 42 of the 48 avian species and outgroups, which
allowed for missing data due to incomplete assemblies. To be included in the dataset, the exons in each genome
assembly had to be 30% or more of the full-length se-
quence of the chicken or zebra finch ortholog. Anno-
tated untranslated regions (UTRs) were trimmed off to
remove non-coding sequence, in order to infer a coding-
only sequence phylogeny. We note that 44 genes were iden-
tified with various problems such as gene annotation issues,
and we removed them in the phylogenetic analyses. How-
ever, we provide them here in the unfiltered alignments. Whole genome alignment
h l
l These are alignments of the translated peptide sequences
for the 8295 protein-coding gene data set. Whole genome alignments were first created by a
LASTZ + MULTIZ alignment [13,14] (http://www.bx.psu. edu/miller_lab/) across all 48 bird species and outgroups
using individual chromosomes of the chicken genome as
the reference (initial alignment 392,719,329 Mb). They
were filtered for segments with fewer than 42 avian species
(>5 missing bird species) and aberrant sequence align-
ments. The individual remaining segments of the MULTIZ
alignment were realigned with MAFFT. We did not use
SATé + MAFFT due to computational challenges (too
much input/output was required). Supergenes These are the concatenated sets of loci from various parti-
tions of the TENT dataset (exons, introns, and UCEs de-
scribed above), brought together using the statistical binning
approach. The statistical binning approach put together sets
of loci that were deemed “combinable”. Two genes were
considered combinable if their respective gene trees had no
pairs of incompatible branches that had bootstrap support
above a 50% threshold. Alignments of genes in the same bin
were concatenated to form supergenes, but boundaries of
genes were kept so that a gene-partitioned phylogenetic
analysis could be performed on each supergene. Data description overlapping sequences among the UCEs. The source
UCE sequences used to search the genomes were deter-
mined from sequence capture probes [10-12] aligned to
each avian genome assembly. Unlike the exon and intron
data sets, we required that all 42 avian species and the
alligator outgroup contain the UCEs. We found this re-
quirement to be sufficient, because the central portions
of UCEs are highly conserved across all species. Here we present FASTA files of loci, sequence align-
ments, indels, transposable elements, and Newick files of
gene trees and species trees used in the Avian Phyloge-
nomics Project [1-4]. We also include scripts used to
process the data. The 48 species from which we col-
lected these data span the phylogeny of modern birds,
including representatives of all Neognathae (Neoaves
and Galloanseres) and two of the five Palaeognathae or-
ders (Table 1) [5-7]. Explanation of various data sets used to infer gene and
species trees Here we describe each locus data set in brief. Additional
details are provided in Jarvis et al. [1]. 3679 UCE locus set 5.7 million insertions and deletions (indels) were scored as
binary characters locus by locus from the same intron,
exon, and UCE alignments as used in the TENT data set
on the principle of simple indel coding using 2Xread
[15,16] and then concatenated. Coding was verified using
GapCoder [17] and by visual inspection of alignments for a This is the ultraconserved element (UCE) data set with
1000 bp flanking sequence at the 3′ and 5′ ends. The
UCE dataset was filtered to remove overlap with the
above exon and intron data sets, other exons and introns
in
the
chicken
genome
assembly
version
3,
and Jarvis et al. FASTA files of loci datasets in alignments alignments with outgroups included We provide the above loci data sets as FASTA files of both
unfiltered and filtered sequence alignments. The alignments
were filtered for aberrant over- and under-aligned se-
quences, and for the presence of the loci in 42 of the 48
avian species. All multiple sequence alignments were per-
formed in two rounds. The first round was used to find
contiguous portions of sequences that we identified as aber-
rant, and the second round was used to realign the filtered
sequences. We used SATé [22,23] combined with either
MAFFT [18] or PRANK [24] alignment algorithms, de-
pending on the limitations of working with large datasets. Alignments without and with outgroups are made available. 3679 UCE locus set GigaScience (2015) 4:4 Table 1 Genomes used in the avian phylogenomics project (Continued)
Taeniopygia guttata
Zebra Finch
PRJNA17289
N.A. Tauraco erythrolophus
Red-crested Turaco
PRJNA212908
http://dx.doi.org/10.5524/101038
Tinamus guttatus
White-throated Tinamou
PRJNA212876
http://dx.doi.org/10.5524/101014
Tyto alba
Barn Owl
PRJNA212909
http://dx.doi.org/10.5524/101039
Listed are the scientific species name, English name, BioProject ID in the NCBI database for each genome (http://www.ncbi.nlm.nih.gov/bioproject), and GigaScience
deposited genome sequences and raw reads. Full details are in [1,2]. Table 1 Genomes used in the avian phylogenomics project (Continued) to have annotation issues and we removed them from
the phylogenetic analyses (the list is provided in the file
FASTA_files_of_loci_datasets/Filtered_sequence_alignments/
8295_Exons/42-exon-genes-removed.txt). Two more genes
were removed because a gene tree could not be esti-
mated for them. The first round of alignment was per-
formed using SATé + PRANK, and the second round
was performed using SATé + MAFFT. Before align-
ment, the nucleotide sequences were converted to
amino acid sequences, and then reverted back to nu-
cleotide sequences afterwards. small subset of data. Intron indels were scored on align-
ments that excluded non-avian outgroups (48 taxa), UCE
indels were scored on alignments that included Alligator
(49 taxa), and exons were scored on alignments that in-
cluded all non-avian outgroups (52 taxa). Individual introns
of the same gene were scored independently to avoid creat-
ing artifactual indels between concatenated intron or whole
genome segments, whereas exons were concatenated as
complete unigenes before scoring. For exons, indels >30 bp
were excluded to avoid scoring missing exons as indels. Transposable element markers 8295 Exons 8295 Exons These are 61 manually curated presence/absence loci of
transposable elements (TEs) present in the Barn Owl
genome that exhibit presence at orthologous positions in
one or more of the other avian species. The TE markers
were identified by eye after a computational screening of
3,671 TguLTR5d retroposon insertions from the Barn
Owl. For each TguLTR5d locus, we conducted BLASTn
searches of TE-flanking sequences (1 kb per flank)
against the remaining avian species and generated multi-
species sequence alignments using MAFFT [18]. Redun-
dant or potentially paralogous loci were excluded from
analysis and the remaining marker candidates were care-
fully inspected using strict standard criteria for assigning
presence/absence character states [19-21]. 42-exon-genes-removed.txt: list of 42 genes removed
due to various issues pep2cds-filtered-sate-alignments-noout.tar.gz: DNA
alignments (Amino acid alignments translated to
DNA) without outgroups pep2cds-filtered-sate-alignments-original.zip: DNA
alignments (Amino acid alignments translated to
DNA) with outgroups included Intron loci alignments These are filtered alignments of introns from 2516
genes. Both rounds of alignment were performed using
SATé + MAFFT, because SATé + PRANK was too com-
putationally expensive on long introns. 2516 Introns introns-filtered-sate-alignments-with-and-without-
outgroups.tar.gz: Includes both alignments with and
without outgroups 3679 UCE locus set GigaScience (2015) 4:4 Table 1 Genomes used in the avian phylogenomics project
Species
English name Table 1 Genomes used in the avian phylogenomics project
Species
English name
BioProject ID
Acanthisitta chloris
Rifleman
PRJNA212877
Anas platyrhynchos
Pekin Duck
PRJNA46621
Antrostomus carolinensis
Chuck-will’s-widow
PRJNA212888
Apaloderma vittatum
Bar-tailed Trogon
PRJNA212878
Aptenodytes forsteri
Emperor Penguin
PRJNA235982
Balearica regulorum
Grey Crowned-crane
PRJNA212879
Buceros rhinoceros
Rhinoceros Hornbill
PRJNA212887
Calypte anna
Anna’s Hummingbird
PRJNA212866
Cariama cristata
Red-legged Seriema
PRJNA212889
Cathartes aura
Turkey Vulture
PRJNA212890
Chaetura pelagica
Chimney Swift
PRJNA210808
Charadrius vociferus
Killdeer
PRJNA212867
Chlamydotis macqueenii
MacQueen’s Bustard
PRJNA212891
Colius striatus
Speckled Mousebird
PRJNA212892
Columba livia
Pigeon
PRJNA167554
Corvus brachyrhynchos
American Crow
PRJNA212869
Cuculus canorus
Common Cuckoo
PRJNA212870
Egretta garzetta
Little Egret
PRJNA232959
Eurypyga helias
Sunbittern
PRJNA212893
Falco peregrinus
Peregrine Falcon
PRJNA159791
Fulmarus glacialis
Northern Fulmar
PRJNA212894
Gallus gallus
Chicken
PRJNA13342
Gavia stellata
Red-throated Loon
PRJNA212895
Geospiza fortis
Medium Ground-finch
PRJNA156703
Haliaeetus albicilla
White-tailed Eagle
PRJNA212896
Haliaeetus leucocephalus
Bald Eagle
PRJNA237821
Leptosomus discolor
Cuckoo-roller
PRJNA212897
Manacus vitellinus
Golden-collared Manakin
PRJNA212872
Meleagris gallopavo
Turkey
PRJNA42129
Melopsittacus undulatus
Budgerigar
PRJNA72527
Merops nubicus
Carmine Bee-eater
PRJNA212898
Mesitornis unicolor
Brown Mesite
PRJNA212899
Nestor notabilis
Kea
PRJNA212900
Nipponia nippon
Crested ibis
PRJNA232572
Opisthocomus hoazin
Hoatzin
PRJNA212873
Pelecanus crispus
Dalmatian Pelican
PRJNA212901
Phaethon lepturus
White-tailed Tropicbird
PRJNA212902
Phalacrocorax carbo
Great Cormorant
PRJNA212903
Phoenicopterus ruber
American Flamingo
PRJNA212904
Picoides pubescens
Downy Woodpecker
PRJNA212874
Podiceps cristatus
Great Crested Grebe
PRJNA212905
Pterocles gutturalis
Yellow-throated Sandgrouse
PRJNA212906
Pygoscelis adeliae
Adelie Penguin
PRJNA235983
Struthio camelus
Common Ostrich
PRJNA212875 p y
g
p
j
English name
BioProject ID
Rifleman
PRJNA212877
Pekin Duck
PRJNA46621
Chuck-will’s-widow
PRJNA212888
Bar-tailed Trogon
PRJNA212878
Emperor Penguin
PRJNA235982
Grey Crowned-crane
PRJNA212879
Rhinoceros Hornbill
PRJNA212887
Anna’s Hummingbird
PRJNA212866
Red-legged Seriema
PRJNA212889
Turkey Vulture
PRJNA212890
Chimney Swift
PRJNA210808
Killdeer
PRJNA212867
MacQueen’s Bustard
PRJNA212891
Speckled Mousebird
PRJNA212892
Pigeon
PRJNA167554
American Crow
PRJNA212869
Common Cuckoo
PRJNA212870
Little Egret
PRJNA232959
Sunbittern
PRJNA212893
Peregrine Falcon
PRJNA159791
Northern Fulmar
PRJNA212894
Chicken
PRJNA13342
Red-throated Loon
PRJNA212895
Medium Ground-finch
PRJNA156703
White-tailed Eagle
PRJNA212896
Bald Eagle
PRJNA237821
Cuckoo-roller
PRJNA212897
Golden-collared Manakin
PRJNA212872
Turkey
PRJNA42129
Budgerigar
PRJNA72527
Carmine Bee-eater
PRJNA212898
Brown Mesite
PRJNA212899
Kea
PRJNA212900
Crested ibis
PRJNA232572
Hoatzin
PRJNA212873
Dalmatian Pelican
PRJNA212901
White-tailed Tropicbird
PRJNA212902
Great Cormorant
PRJNA212903
American Flamingo
PRJNA212904
Downy Woodpecker
PRJNA212874
Great Crested Grebe
PRJNA212905
Yellow-throated Sandgrouse
PRJNA212906
Adelie Penguin
PRJNA235983
Common Ostrich
PRJNA212875 http://dx.doi.org/10.5524/101026
http://dx.doi.org/10.5524/100040
http://dx.doi.org/10.5524/101027
http://dx.doi.org/10.5524/101040
http://dx.doi.org/10.5524/101028
http://dx.doi.org/10.5524/101010
N.A. http://dx.doi.org/10.5524/101026
http://dx.doi.org/10.5524/100040 http://dx.doi.org/10.5524/101027 http://dx.doi.org/10.5524/101040 http://dx.doi.org/10.5524/100059
http://dx.doi.org/10.5524/101029 http://dx.doi.org/10.5524/100059 http://dx.doi.org/10.5524/101029 http://dx.doi.org/10.5524/101030 http://dx.doi.org/10.5524/101031 http://dx.doi.org/10.5524/101003 http://dx.doi.org/10.5524/101011 http://dx.doi.org/10.5524/101032 http://dx.doi.org/10.5524/101033 http://dx.doi.org/10.5524/101034 http://dx.doi.org/10.5524/101035 http://dx.doi.org/10.5524/101012 http://dx.doi.org/10.5524/101036 Jarvis et al. 8295 Amino Acids pep-filtered-sate-alignments-noout.tar.gz: Amino
acid alignments with outgroups removed pep-filtered-sate-alignments-original.zip: Amino acid
alignments with outgroups included Concatenated alignments used in ExaML analyses uce-filtered-alignments-without-gator.tar.gz: UCE
individual alignments with outgroups Exon.AminoAcid.ExaML.partitioned
Exon.c123. ExaML.partitioned
Exon.c123. ExaML.unpartitioned
Exon.c1.ExaML.unpartitioned
Exon.c2.ExaML.unpartitioned
Exon.c12.ExaML.unpartitioned
Exon.c123-RY.ExaML.unpartitioned
Exon.c3.ExaML.unpartitioned
Intron
TEIT.RAxML
TENT + c3.ExaML
TENT + outgroup.ExaML
TENT.ExaML.100%
TENT.ExaML.25%
TENT.ExaML.50%
TENT.ExaML.75%
WGT.ExaML Exon.AminoAcid.ExaML.partitioned Unfiltered loci sequence alignments These are individual loci alignments of the above data
sets, before filtering. Concatenated alignments used in RAxML analyses
UCE concatenated alignments with and without the alligator Amino.Acid.unfiltered Amino.Acid.unfiltered pep-unfiltered-alignments-original.zip: unfiltered
SATé + Prank alignments used for the filtering step
Exon.c123.unfiltered: pep-unfiltered-alignments-original.zip: unfiltered
SATé + Prank alignments used for the filtering step uce-filtered-alignments-w-gator-concatenated.phylip.gz uce-filtered-alignments-without-gator-concatenated. phylip.gz Exon.c123.unfiltered: Exon.c123.unfiltered: pep2cds-unfiltered-alignemtns-original.zip:
unfiltered SATé + Prank alignments used for the
filtering step Supergenes generated from statistical binning These are concatenated alignments for each of our 2022
supergene alignments. We note that although super-
genes are concatenated loci, we estimated supergene
trees using partitioned analyses where each gene was put
in a different partition. Thus, we also provide the bound-
aries between genes in text files (these can be directly
used as partition input files to RAxML). supergene-alignments.tar.bz2: supergene alignments
with partition files showing genes put in each bin
and their boundaries in the concatenated alignment 3769 UCE + 1000 flanking bp WGT.unfiltered These are uploaded as part of the comparative genomics
paper [2] data note [25], and a link is provided here
https://github.com/gigascience/paper-zhang2014. uce-probes-used.fasta.gz: Probes targeting UCE loci
shared among vertebrate taxa. uce-raw-genome-slices-of-probe-matches.tar: Probe +
flank slices around locations matching probes
targeting UCE loci. FASTA files of concatenated datasets in alignments We provide FASTA files of concatenated sequence align-
ments of the above filtered loci datasets. These are
concatenated alignments that were used in the ExaML
and RAxML analyses [3]. uce-raw-lastz-results-of-probe-matches.tar: LASTZ
results of mapping probes onto genome assemblies. uce-assembled-loci-from-probe-matches.tar: UCE loci
assembled from probe + flank slices from each genome. uce-filtered-alignments-w-gator.tar.gz: UCE
individual alignments without outgroups Exon loci alignments These are alignments of UCEs and their surrounding
1000 bp from 3769 loci after filtering. Both rounds of
alignment were performed using SATé + MAFFT. These are filtered alignments of exons from 8295 genes. Of these 8295, there were 42 genes that were identified Jarvis et al. GigaScience (2015) 4:4 3769 UCE + 1000 flanking bp Newick files for 32 species trees using different genomic
partitions and methods Newick files for 32 species trees using different genomic
partitions and methods Partition files showing which loci make up which
bins for MP-EST analyses Exon.AminoAcid.ExaML.partitioned.tr
Exon.c123.ExaML.partitioned.tre
Exon.c123.ExaML.unpartititoned.tre
Exon.c123-RY.ExaML.unpartitioned.tr
Exon.c12.ExaML.partitioned.tre
Exon.c12.ExaML.unpartitioned.tre
Exon.c1.ExaML.unpartitioned.tre
Exon.c2.ExaML.unpartitioned.tre
Exon.c3.ExaML.unpartitioned.tre
Exon.RAxML.heterogenous.c123.tre
Exon.RAxML.heterogenous.c12.tre
Exon.RAxML.homogenous.c123.tre
Exon.RAxML.homogenous.c12.tre
Intron.RAxML.heterogenous.tre.txt
Intron.RAxML.homogenous.tre.txt
Intron.RAxML.partitioned.tre
Intron.RAxML.unpartitioned.tre
Intron.MP-EST.binned.tre
Intron.MP-EST.unbinned.tre
TEIT.RAxML.tre
TENT + c3.ExaML.tre List of scripts used in avian phylogenomics project
We also deposit the key scripts used in this project in
GigaDB, which include: Script for filtering amino acid alignments Script for filtering nucleotide sequence alignments Script for mapping names from 5-letter codes to full
names Scripts related to indel analyses We provide readme files in the script directories de-
scribing the usage of the scripts. Clocklike exon alignment Concatenated c12 (1st + 2nd codons) DNA sequence
alignments from the 1156 clocklike genes were used for
the dating analyses. These are alignments of the first and
second codon positions of clock-like genes among the
8295 exon orthologs: Intron.unfiltered introns-unfiltered-alignments-original.zip: intron SATé
alignments before filtering with outgroups included
introns-unfiltered-alignments-noout.zip: intron SATé
alignments before filtering with outgroups included
UCE.unfiltered introns-unfiltered-alignments-original.zip: intron SATé
alignments before filtering with outgroups included c12.DNA.alignment.1156.clocklike.zip c12.DNA.alignment.1156.clocklike.txt introns-unfiltered-alignments-noout.zip: intron SATé
alignments before filtering with outgroups included c12.DNA.alignment.clocklike.readme.txt c12.DNA.alignment.clocklike.txt.zip UCE.unfiltered UCE.unfiltered High and low variance exons and their associated introns uce-unfiltered-alignments-w-gator.tar.gz: UCE
alignments before filtering with alligator outgroup High variance exons: Jarvis et al. GigaScience (2015) 4:4 Exon.heterogeneous.c123
Exon.heterogenous.c12
Low variance exons:
Exon.homogeneous.c123. Exon.homogenous.c12
High variance introns: These are heterogenous introns
concatIntronNooutMSAlow.fasta.gz
Low variance introns: These are homogenous introns
concatIntronNooutMSAhigh.fasta.gz
Indel sequence alignments
This is a concatenated alignment of indels from exons, in-
trons, and UCEs. A README file describes the content. Transposable element markers
owl_TE_marker_Table.txt
Species and gene tree files
Species trees (Newick format) were generated with ei-
ther RAxML, an improved ExaML version for handling
large alignments, or MP-EST* [4]. We deposit both the
maximum likelihood and bootstrap replicate trees. Newick files for 32 species trees using different genomic
partitions and methods
Exon.AminoAcid.ExaML.partitioned.tre
Exon.c123.ExaML.partitioned.tre
Exon.c123.ExaML.unpartititoned.tre
Exon.c123-RY.ExaML.unpartitioned.tre
Exon.c12.ExaML.partitioned.tre
Exon.c12.ExaML.unpartitioned.tre
Exon.c1.ExaML.unpartitioned.tre
Exon.c2.ExaML.unpartitioned.tre
Exon.c3.ExaML.unpartitioned.tre
Exon.RAxML.heterogenous.c123.tre
Exon.RAxML.heterogenous.c12.tre
Exon.RAxML.homogenous.c123.tre
Exon.RAxML.homogenous.c12.tre
Intron.RAxML.heterogenous.tre.txt
Intron.RAxML.homogenous.tre.txt
Intron.RAxML.partitioned.tre
Intron.RAxML.unpartitioned.tre
Intron.MP-EST.binned.tre
Intron.MP-EST.unbinned.tre
TEIT.RAxML.tre
TENT + c3.ExaML.tre
T
T
T
T
T
U
W
W
Newick
C
C
C
C
C
C
C
C
C
C
C
Newick
with 5-
M
B
M
B
M
P
b
List of
We al
GigaD
S
S
S
n
S
We
scribin
Availa
Proje
Proje
pape
page
pape
Ope Exon.heterogeneous.c123
Exon.heterogenous.c12 Chronogram01.TENT.ExAML.tre
Chronogram02.TENT.ExAML.max865.tre
Chronogram03.TENT.ExAML.Allig247.tre
Chronogram04.TENT.ExAML.no-outgroup.tre
Chronogram05.TENT.ExAML.no-outgroup.max865.tre
Chronogram06.TENT.MP-EST.tre
Chronogram07.WGT.ExAML.alternative.tre
Chronogram08.WGT.ExAML.best.tre
Chronogram09.Intron.ExAML.unpartitioned.tre
Chronogram10.UCE.RAxML.tre
Chronogram11.Exon.c123.RaXML.partitioned.tre Bootstrap replicates of supergene trees used in
MP-EST analyses Availability and requirements Any restrictions to use by non-academics: none Any restrictions to use by non-academics: none Any restrictions to use by non-academics: none Availability of supporting data Other data files presented in this data note for the
majority of genomes are available in the GigaScience
repository, GigaDB [26] (Table 1), as well as NCBI
(Table 1), ENSEMBL, UCSC, and CoGe databases. ENSEMBL: http://avianbase.narf.ac.uk/index.html UCSC:
(http://genome.ucsc.edu/cgi-bin/hgGateway; under verte-
brate genomes) CoGe: (https://genomevolution.org/wiki/
index.php/Bird_CoGe). Authors’ contributions Coordinated the project EDJ, TW, MTPG, and GZ; Wrote the paper and
co-supervised the project EDJ, SM, AJA, PH, TW, MTPG, GZ, ELB, JC, SE, ASt,
DPM; Sample coordination and collections JH, EDJ, MTPG, AAN; Alignments
SM, AJA, TW, ASt, RdF, MTPG, CL, GZ, BCF, EDJ; Species trees and gene trees
AA, SM, ASt, BCF, TW, CL, CCW; Indels PH, NN, AJA; Transposable Elements
ASu, HE; Fossil-calibrated chronograms SYWH, PH, MTPG, JC, DM, SE. The
contribution information for all authors is provided in Additional file 1. All authors read and approved the final manuscript. Received: 26 November 2014 Accepted: 16 December 2014 References 1. Jarvis ED, Mirarab S, Aberer AJ, Li B, Houde P, Li C, et al. Whole genome
analyses resolve the early branches in the tree of life of modern birds. Science. 2014;346(6215):1320–31. 1. Jarvis ED, Mirarab S, Aberer AJ, Li B, Houde P, Li C, et al. Whole genome
analyses resolve the early branches in the tree of life of modern birds. Science. 2014;346(6215):1320–31. The authors declare that they have no competing interests. The authors declare that they have no competing interests. Acknowledgements Princeton, N.J.: Princeton University Press; 2006. 8. Hillier LW, Miller W, Birney E, Warren W, Hardison RD, Ponting CP, et al. Sequence and comparative analysis of the chicken genome provide unique
perspectives on vertebrate evolution. Nature. 2004;432:695–716. 8. Hillier LW, Miller W, Birney E, Warren W, Hardison RD, Ponting CP, et al. Sequence and comparative analysis of the chicken genome provide unique
perspectives on vertebrate evolution. Nature. 2004;432:695–716. 9. Warren WC, Clayton DF, Ellegren H, Arnold AP, Hillier LW,
The genome of a songbird. Nature. 2010;464:757–62. The full author list of The Avian Phylogenomics Consortium is provided at
the end of the data note. The genome of a songbird. Nature. 2010;464:757–62. The full author list of The Avian Phylogenomics Consortium is provided at
the end of the data note. 10. Faircloth BC, McCormack JE, Crawford NG, Harvey MG, Brumfield RT, Glenn
TC. Ultraconserved elements anchor thousands of genetic markers spanning
multiple evolutionary timescales. Syst Biol. 2012;61:717–26. Abbreviations
bl Abbreviations
TE: Transposable element; TENT: Total evidence Nucleotide tree; TEIT: Total
evidence indel tree; WGT: Whole genome tree; UCE: Ultra conserved
element; c123: 1st, 2nd, and 3rd codons of exons. TE: Transposable element; TENT: Total evidence Nucleotide tree; TEIT: Total
evidence indel tree; WGT: Whole genome tree; UCE: Ultra conserved
element; c123: 1st, 2nd, and 3rd codons of exons. Acknowledgements 2. Zhang G, Li C, Li Q, Li B, Larkin DM, Lee C, et al. Comparative genomics
reveal insights into avian genome evolution and adaption. Science. 2014;346(6215):1311–20. 2. Zhang G, Li C, Li Q, Li B, Larkin DM, Lee C, et al. Comparative genomics
reveal insights into avian genome evolution and adaption. Science. 2014;346(6215):1311–20. The majority of genome sequencing and annotation was supported by
internal funding from BGI. Additional significant support is from the
coordinators of the project: E.D.J. from the Howard Hughes Medical Institute
(HHMI) and NIH Directors Pioneer Award DP1OD000448. S.M. from an HHMI
International Student Fellowship. G.Z. from Marie Curie International
Incoming Fellowship grant (300837); T.W. from NSF DEB 0733029, NSF DBI
1062335, NSF IR/D program; and M.T.P.G. from a Danish National Research
Foundation grant (DNRF94) and a Lundbeck Foundation grant (R52-A5062). We thank the following Centers that allowed us to conduct the
computationally intensive analyses for this study: Heidelberg Institute for
Theoretical Studies (HITS); San Diego Supercomputer Center (SDSC), with
support by an NSF grant; SuperMUC Petascale System at the Leibniz
Supercomputing Center; Technical University of Denmark (DTU); Texas
Advanced Computing Center (TACC); Georgia Advanced Computing
Resource Center (GACRC), a partnership between the University of Georgia’s
Office of the Vice President for Research and Office of the Vice President for
Information Technology; Amazon Web Services (AWS); BGI; Nautilus
supercomputer at the National Institute for Computational Sciences of the
University of Tennessee and Smithsonian; and Duke University Institute for
Genome Sciences and Policy. 3. A Stamatakis, AJ Aberer. Novel parallelization schemes for large-scale
likelihood-based phylogenetic inference. IEEE 27th International Symposium
on Parallel and Distributed Processing, 1195–1204. 2013 3. A Stamatakis, AJ Aberer. Novel parallelization schemes for large-scale
likelihood-based phylogenetic inference. IEEE 27th International Symposium
on Parallel and Distributed Processing, 1195–1204. 2013 4. Mirarab S, Bayzid MS, Boussau B, Warnow T. Statistical binning enables an
accurate coalescent-based estimation of the avian tree. Science. 2014;346
(6215):1–9. 5. J Cracraft, in The Howard and Moore Complete Checklist of the Birds of the
World, E. C. Dickinson, J. V. Remsen, Eds. Eastbourne, U.K.: Aves Press;
2013. pp. xxi-xliii 6. Dickinson EC, Remsen JV. Eds. Aves Press: The H
Complete Checklist of Birds of the World; 2013. 6. Dickinson EC, Remsen JV. Eds. Aves Press: The Howard and Mo Complete Checklist of Birds of the World; 2013. 7. Gill F, Wright M. Birds of the World: Recommended English Names. Availability and requirements Project name: Avian Phylogenomic Project scripts Project name: Avian Phylogenomic Project scripts
Project
home
page:
https://github.com/gigascience/
paper-jarvis2014; also see companion paper home
page for related data https://github.com/gigascience/
paper-zhang2014 Operating system: Unix Jarvis et al. GigaScience (2015) 4:4 Programming language: R, Perl, python
License: GNU GPL v3. Any restrictions to use by non-academics: none Programming language: R, Perl, python
License: GNU GPL v3. Programming language: R, Perl, python
License: GNU GPL v3. Programming language: R, Perl, python Heidelberg, Germany. 4China National GeneBank, BGI-Shenzhen, Shenzhen
518083, China. 5College of Medicine and Forensics, Xi’an Jiaotong University,
Xi’an 710061, China. 6Centre for GeoGenetics, Natural History Museum of
Denmark, University of Copenhagen, Øster Voldgade 5-7, 1350 Copenhagen,
Denmark. 7Department of Biology, New Mexico State University, Las Cruces,
NM 88003, USA. 8School of Biological Sciences, University of Sydney, Sydney,
NSW 2006, Australia. 9Department of Ecology and Evolutionary Biology,
University of California Los Angeles, Los Angeles, CA 90095, USA. 10Department of Biological Sciences, Louisiana State University, Baton Rouge,
LA 70803, USA. 11CNRS UMR 5554, Institut des Sciences de l’Evolution de
Montpellier, Université Montpellier II, Montpellier, France. 12Department of
Evolutionary Biology, Uppsala University, SE-752 36 Uppsala, Sweden. 13Biodiversity and Biocomplexity Unit, Okinawa Institute of Science and
Technology Onna-son, Okinawa 904-0495, Japan. 14Department of Statistics
and Institute of Bioinformatics, University of Georgia, Athens 30602, USA. 15Department of Genomics and Genetics, The Roslin Institute and Royal
(Dick) School of Veterinary Studies, University of Edinburgh, Easter Bush
Campus, Midlothian EH25 9RG, UK. 16Department of Organismic and
Evolutionary Biology and Museum of Comparative Zoology, Harvard
University, Cambridge, MA, USA. 17Institute of Theoretical Informatics,
Department of Informatics, Karlsruhe Institute of Technology, D- 76131
Karlsruhe, Germany. 18Department of Biochemistry & Biophysics, University of
California, San Francisco, CA 94158, USA. 19Department of Ornithology,
American Museum of Natural History, New York, NY 10024, USA. 20Department of Biology and Genetics Institute, University of Florida,
Gainesville, FL 32611, USA. 21Department of Biology, University of
Copenhagen, Ole Maaløes Vej 5, 2200 Copenhagen, Denmark. 22Princess Al
Jawhara Center of Excellence in the Research of Hereditary Disorders, King
Abdulaziz University, Jeddah 21589, Saudi Arabia. 23Macau University of
Science and Technology, Avenida Wai long, Taipa, Macau 999078, China. 24Department of Medicine, University of Hong Kong, Hong Kong, Hong
Kong. 25Trace and Environmental DNA Laboratory Department of
Environment and Agriculture, Curtin University, Perth, WA 6102, Australia. 26Centre for Social Evolution, Department of Biology, Universitetsparken 15,
University of Copenhagen, DK-2100 Copenhagen, Denmark. Competing interests p
g
The authors declare that they have no competing interests. Author details
1
f 11. McCormack JE, Faircloth BC, Crawford NG, Gowaty PA, Brumfield RT, Glenn
TC. Ultraconserved elements are novel phylogenomic markers that resolve
placental mammal phylogeny when combined with species-tree analysis. Genome Res. 2012;22:746–54. 1Department of Neurobiology, Howard Hughes Medical Institute and Duke
University Medical Center, Durham, NC 27710, USA. 2Department of
Computer Science, The University of Texas at Austin, Austin, TX 78712, USA. 3Scientific Computing Group, Heidelberg Institute for Theoretical Studies, Jarvis et al. GigaScience (2015) 4:4 12. Dimitrieva S, Bucher P. UCNEbase–a database of ultraconserved non-coding
elements and genomic regulatory blocks. Nucleic Acids Res. 2013;41:D101–9. 13. Harris RS. Improved pairwise alignment of genomic DNA. Ph.D. Thesis. 2007 14. Blanchette M, Faircloth BC, Crawford NG, Gowaty PA, Brumfield RT, Glenn
TC. Aligning multiple genomic sequences with the threaded blockset
aligner. Genome Res. 2004;14:708–15. g
15. Simmons MP, Ochoterena H. Gaps as characters in sequence-based
phylogenetic analyses. Syst Biol. 2000;49:369–81. phylogenetic analyses. Syst Biol. 2000;49:369–81. 16. D. P. Liitle. 2xread: a simple indel coding tool. Program distribu 16. D. P. Liitle. 2xread: a simple indel coding tool. Program distributed by the
author . 2005. http://www.nybg.org/files/scientists/2xread.html. author . 2005. http://www.nybg.org/files/scientists/2xread.htm 17. Young ND, Healy J. GapCoder automates the use of inde 17. Young ND, Healy J. GapCoder automates the use of ind
phylogenetic analysis. BMC Bioinformatics. 2003;4:6. phylogenetic analysis. BMC Bioinformatics. 2003;4:6. 18. Katoh K, Toh H. Recent developments in the MAFFT multiple sequence
alignment program. Brief Bioinform. 2008;9:286–98. alignment program. Brief Bioinform. 2008;9:286–98. 19. Suh A, Paus M, Kiefmann M, Churakov G, Franke FA, Brosius J, et al. Mesozoic retroposons reveal parrots as the closest living relatives of
passerine birds. Nat Commun. 2011;2:443. 20. Suh A, Kriegs JO, Donnellan S, Brosius J, Schmitz J. A universal method for
the study of CR1 retroposons in nonmodel bird genomes. Mol Biol Evol. 2012;29:2899–903. 21. Suh A, Churakov G, Ramakodi MP, Platt RN 2nd, Jurka J, Kojima KK, et al. Multiple lineages of ancient CR1 retroposons shaped the early genome
evolution of amniotes. Genome Biol Evol. 2015;7:205–17. 22. Liu K, Warnow TJ, Holder MT, Nelesen SM, Yu J, Stamatakis AP. SATe-II: very
fast and accurate simultaneous estimation of multiple sequence alignments
and phylogenetic trees. Syst Biol. 2012;61:90–106. 23. Liu K, Raghavan S, Nelesen S, Linder CR, Warnow T. Rapid and accurate
large-scale coestimation of sequence alignments and phylogenetic trees. Science. 2009;324:1561–4. 24. Löytynoja A, Goldman N. Author details
1
f An algorithm for progressive multiple alignment
of sequences with insertions. Proc Natl Acad Sci U S A. 2005;102:10557–62. 25. Zhang G, Li B, Li C, Gilbert MTP, Jarvis ED, Wang J. The Avian genome
Consortium, Wang J: Comparative genomic data of the Avian Phylogenomics
Project. GigaSci Database 2014, http://dx.doi.org/10.5524/101000 j
g
p
g
26. Jarvis ED, Mirarab S, Aberer A, Houde P, Li C, Ho S, et al. Phylogenomic
analyses data of the avian phylogenomics project. GigaScience Database. 2014. http://dx.doi.org/10.5524/101041 Submit your next manuscript to BioMed Central
and take full advantage of:
• Convenient online submission
• Thorough peer review
• No space constraints or color figure charges
• Immediate publication on acceptance
• Inclusion in PubMed, CAS, Scopus and Google Scholar
• Research which is freely available for redistribution
Submit your manuscript at
www.biomedcentral.com/submit Submit your next manuscript to BioMed Central
and take full advantage of:
• Convenient online submission
• Thorough peer review
• No space constraints or color figure charges
• Immediate publication on acceptance
• Inclusion in PubMed, CAS, Scopus and Google Scholar
• Research which is freely available for redistribution
Submit your manuscript at
www.biomedcentral.com/submit • Convenient online submission
|
https://openalex.org/W4304144018
|
https://www.researchsquare.com/article/rs-2002214/latest.pdf
|
English
| null |
Matrix Protein Tenascin-C Promotes Kidney Fibrosis via STAT3 activation in Response to Tubular Injury
|
Research Square (Research Square)
| 2,022
|
cc-by
| 9,121
|
Matrix Protein Tenascin-C Promotes Kidney Fibrosis
via STAT3 activation in Response to Tubular Injury
Chuanming Hao ( chuanminghao@fudan.edu.cn )
Department of Nephrology Huashan Hospital, Fudan University
Qionghong Xie
Min Zhang
Huashan Hospital, Fudan University
Xiaoyi Mao
Huashan Hospital,Fudan University
Mingyue Xu
Huashan Hospital,Fudan University
Shaojun Liu
Huashan Hospital of Fudan university
Da Shang
Huashan Hospital,Fudan University
Yunyu Xu
Huashan Hospital,Fudan University
Ruiying Chen
Huashan Hospital,Fudan University
Yi Guan
Huashan Hospital,Fudan University
Xinzhong Huang
A¨liated Hospital of Nantong University
Roy Zent
Vanderbilt University Medical Center
Ambra Pozzi
Vanderbilt University Medical Center Matrix Protein Tenascin-C
via STAT3 activation in Res
Chuanming Hao ( chuanminghao@fudan.edu.cn )
Department of Nephrology Huashan Hospital, Fudan
Qionghong Xie
Min Zhang
Huashan Hospital, Fudan University
Xiaoyi Mao
Huashan Hospital,Fudan University
Mingyue Xu
Huashan Hospital,Fudan University
Shaojun Liu
Huashan Hospital of Fudan university
Da Shang
Huashan Hospital,Fudan University
Yunyu Xu
Huashan Hospital,Fudan University
Ruiying Chen
Huashan Hospital,Fudan University
Yi Guan
Huashan Hospital,Fudan University
Xinzhong Huang
A¨liated Hospital of Nantong University
Roy Zent
Vanderbilt University Medical Center
Ambra Pozzi
Vanderbilt University Medical Center
Article Department of Nephrology Huashan Hospital, Fudan University Shaojun Liu
Huashan Hospital of Fudan university Da Shang
Huashan Hospital,Fudan University Yunyu Xu
Huashan Hospital,Fudan University Ruiying Chen
Huashan Hospital,Fudan University License: This work is licensed under a Creative Commons Attribution 4.0 International License.
Read Full License License: This work is licensed under a Creative Commons Attribution 4.0 International License. Corresponding author: E-mail: chuanminghao@fudan.edu.cn Article License: This work is licensed under a Creative Commons Attribution 4.0 International License. Read Full License License: This work is licensed under a Creative Commons Attribution 4.0 International License. Read Full License Matrix Protein Tenascin-C Promotes Kidney Fibrosis via STAT3 activation in
Response to Tubular Injury
Qionghong Xie*1, Min Zhang*1, Xiaoyi Mao*1, Mingyue Xu1, Shaojun Liu1, Da Shang1,
Yunyu Xu1, Ruiying Chen1, Yi Guan1, Xinzhong Huang2, Roy Zent3,4, Ambra Pozzi3,4 and
Chuan-Ming Hao1
1Division of Nephrology, Huashan Hospital, Fudan University, Shanghai China
2Division of Nephrology, Affiliated Hospital of Nantong University, Nantong China
3Division of Nephrology, Vanderbilt University Medical Center, Nashville, Tennessee
4Department of Veterans Affairs, Nashville, Tennessee
*These authors contributed equally to the study
Corresponding author:
Chuan-Ming Hao, MD, PhD
12 Wulumuqi Road (middle)
Division of Nephrology, Huashan Hospital, Fudan University
Shanghai 200040, China
Tel: 86-21-52888303
Fax: 86-21-52888304
E-mail: chuanminghao@fudan.edu.cn
Abstract
Accumulating evidence indicates that the extracellular matrix (ECM) is not only a
consequence of fibrosis but also contributes to the progression of fibrosis by creating a Matrix Protein Tenascin-C Promotes Kidney Fibrosis via STAT3 activation in
Response to Tubular Injury
Qionghong Xie*1, Min Zhang*1, Xiaoyi Mao*1, Mingyue Xu1, Shaojun Liu1, Da Shang1,
Yunyu Xu1, Ruiying Chen1, Yi Guan1, Xinzhong Huang2, Roy Zent3,4, Ambra Pozzi3,4 and
Chuan-Ming Hao1
1Division of Nephrology, Huashan Hospital, Fudan University, Shanghai China
2Division of Nephrology, Affiliated Hospital of Nantong University, Nantong China
3Division of Nephrology, Vanderbilt University Medical Center, Nashville, Tennessee
4Department of Veterans Affairs, Nashville, Tennessee
*These authors contributed equally to the study
Corresponding author:
Chuan-Ming Hao, MD, PhD
12 Wulumuqi Road (middle)
Division of Nephrology, Huashan Hospital, Fudan University
Shanghai 200040, China
Tel: 86-21-52888303
Fax: 86-21-52888304
E-mail: chuanminghao@fudan.edu.cn
Abstract
Accumulating evidence indicates that the extracellular matrix (ECM) is not only a
consequence of fibrosis, but also contributes to the progression of fibrosis, by creating a Response to Tubular Injury Qionghong Xie*1, Min Zhang*1, Xiaoyi Mao*1, Mingyue Xu1, Shaojun Liu1, Da Shang1,
Yunyu Xu1, Ruiying Chen1, Yi Guan1, Xinzhong Huang2, Roy Zent3,4, Ambra Pozzi3,4 and
Chuan-Ming Hao1 1Division of Nephrology, Huashan Hospital, Fudan University, Shanghai China
2Division of Nephrology, Affiliated Hospital of Nantong University, Nantong China
3Division of Nephrology, Vanderbilt University Medical Center, Nashville, Tennessee
4Department of Veterans Affairs, Nashville, Tennessee Abstract Accumulating evidence indicates that the extracellular matrix (ECM) is not only a
consequence of fibrosis, but also contributes to the progression of fibrosis, by creating a pro-fibrotic microenvironment. Tenascin-C (TNC) is an ECM glycoprotein that contains
multiple functional domains. We showed that following kidney injury, TNC was markedly
induced in fibrotic areas in the kidney from both mouse models and humans with kidney
diseases. Genetically deletion of TNC in mice significantly attenuated unilateral ureteral
obstruction-induced kidney fibrosis. Further studies showed that TNC promoted the
proliferation of kidney interstitial cells via STAT3 activation. To identify and characterize
TNC expressing cells, we generated a TNC promoter driven CreER2-IRES-eGFP knock-
in mouse line and found that the TNC reporter eGFP was markedly induced in cells around
injured tubules that had lost epithelial markers, suggesting TNC was induced in response
to epithelium injury. Most of the eGFP positive cells were both NG2 and PDGFRβ positive,
only half of which were αSMA positive. These cells did not carry markers of progenitor cells
or macrophages. In normal kidney, TNC is constitutively expressed by the renal medullary
interstitial cells (RMICs) in renal papilla that has been reported to contain stem cells. Cell
lineage tracing study revealed that the cells expressing TNC in renal cortex after fibrosis
were not originated from these RMICs. In conclusion, this study provides strong evidence
that matrix protein TNC contributes to kidney fibrosis. TNC pathway may serve as a
potential therapeutic target for interstitial fibrosis and progression of chronic kidney disease. Key words: Tenascin C, kidney fibrosis, extracellular matrix, STAT3 Patients To characterize the distribution of TNC expression in normal human kidney, specimens
from healthy parts of human carcinoma nephrectomy were obtained. TNC mRNA was
measured by qPCR in fresh tissue, and the distribution of TNC protein was investigated in
paraffin-embedded tissue by immunohistochemistry (IHC). To examine the induction of
TNC in diseased kidneys, TNC protein was detected in paraffin-embedded sections from
patients with biopsy-proven IgA nephropathy (IgAN, n=15) and diabetic nephropathy (DM,
n=6) by IHC. This study was approved by the ethics committee of Huashan Hospital, Fudan
University. Background Kidney fibrosis is the final common pathway to end stage renal disease and characterized
by extracellular matrix (ECM) accumulation and destruction of normal structure [1, 2]. Accumulating evidence indicates that the ECM is not only a consequence of fibrosis, but
also contributes to the progression of fibrosis, by creating a pro-fibrotic microenvironment
or a fibrogenic niche [3, 4]. It has been documented that the ECM not only provides physical scaffolds to cells by It has been documented that the ECM not only provides physical scaffolds to cells by forming a three-dimensional network, but also regulates many biologic processes including
proliferation, migration, differentiation, survival and morphogenesis during development or
in certain physiological/pathophysiological conditions [5-7]. Matricellular proteins (or non-
structural matrix proteins) are a group of ECM proteins that are characterized by dynamical
expressions and regulatory roles[8]. Rather than serving as stable structural elements, they
are usually transiently expressed during development or after injury [9]. They contain
multiple functional domains, such as binding sites for other ECM proteins, ligands for cell
surface receptors, and sites that can sequester specific growth factors, playing important
roles in regulating cellular processes [10]. Tenascin-C (TNC), a hexametric glycoprotein, is a member of the matricellular proteins
that is expressed during development and is at low levels in most of the adult tissues, but
re-induced following injury and associated with the severity of diseases and prognosis[11,
12]. The monomer of TNC contains an N-terminal assembly domain, epidermal growth
factor like (EGF-L) repeats, fibronectin type-III (FNIII) like repeats and a C-terminal
fibrinogen-like globe [13, 14]. These functional domains have been suggested to interact
with specific cell-surface receptors, such as epidermal growth factor receptor (EGFR),
integrins and toll like receptor 4 (TLR4) [13]. They can also recruit cytokines and growth
factors, and present them to or prevent them from cell-surface receptors depending on
their relative locations [14]. Mice lacking TNC develop normally, suggesting that TNC is not
indispensable for the development. However, accumulating evidence showed that TNC
was markedly increased in the fibrotic tissues and associated with organ fibrosis[15-17]. Recently, TNC was also found to promote renal interstitial fibroblast proliferation via integrin/focal adhesion kinase/mitogen-activated protein kinase pathway [18]. In the present study, we characterized the cells that expressed TNC and examined
their role in renal fibrosis. This study provided novel information about the role of TNC in
renal fibrosis. Animals and Models Mice. TNCCreER2-eGFP/+ mouse line was generated as described previously [19]. In brief, an
inducible CreER2 gene with an eGFP reporter was knocked into the 2nd exon of TNC at
the site of starting codon ATG. The insertion of this cassette also resulted in TNC deletion
(Figure 2A). The TNC deletion (TNC-/-) mice were fertile, and developed normally. Male
TNC-/- mice and their wild-type littermates, aged 8-12 weeks, were used to examine the
role of TNC in fibrosis. R26tdTomato reporter mice were purchased from the Jackson Laboratory (stock number, 007909). The bi-transgenic TNCCreER2-eGFP/+;R26tdTomato/+
reporter mice were generated by crossing TNCCreER2-eGFP/+ with R26tdTomato mice. All of the
mice were maintained in the animal facility of Fudan University and allowed free access to
standard rodent chow and water. All of the animal experiments were approved by the
Animal Care and Use Committees of Fudan University. TNC reporter mice. In the TNCCreER2-eGFP/+ mice, eGFP is driven by endogenous TNC
promoter (TNCp-CreER2-IRES-eGFP), and is thus used as a TNC reporter. In the bi-
transgenic TNCCreER2-eGFP/+;R26tdTomato/+ reporter mice, the recombination is induced by
TNC promoter driven CreER2 in the presence of tamoxifen, and then tdTomato (red
fluorescence protein) will label the TNC expressing cells and their daughter cells
permanently. Cell lineage experiments. To determine whether the constitutive TNC expressing cells in
renal medulla were precursors of the interstitial cells or TNC producing cells in the cortex
after fibrosis, a cell lineage tracing experiment was conducted, as shown in Figure 8A and
supplement Figure 3A. The bi-transgenic mice TNCCreER2-eGFP/+;R26tdTomato/+ were given
tamoxifen 1.5mg/d for 5 days to induce recombination at 8 weeks of age. After 12 weeks’
washout time, these mice were used to perform UUO or unilateral kidney ischemia
reperfusion (UIR) operation. When they were sacrificed at the end of the experiments, the
tdTomato positive cells we observed were originated from the cells that constitutively
expressed TNC in the renal medulla. In the other set of animals, if these mice were given
a second series of tamoxifen at the time of UUO or UIR operation, tdTomato would be also
expressed in the cells where TNC was newly induced after operation. The mice with sham Models. UUO induced kidney fibrosis model was used in our study. The mice were
anesthetized with chloral hydrate (400 mg/kg body weight) by intra-peritoneal injection and
body temperatures were maintained at 36.5-37.5℃ throughout the procedure. Animals and Models The left
ureter was exposed via a flank incision and ligated with two 3.0 silk ties at the level of the
lower renal pole. The mice were sacrificed at day 7, 10 or 14 after the operation. The EdU
cooperation assay was used to assess cell proliferation of the kidney: 0.1mg of EdU was
injected (i.p.) at day 3 and 5 after UUO operation and the mice were sacrificed at day 7. operation but similar tamoxifen exposure was used as controls. Cell Experiments Cell lines. The TNC expressing cells (TNC-Cell) were obtained by sorting the tdTomato
positive cells in the UUO kidneys of TNCCreER2-eGFP/+;R26tdTomato/+ mice using flow cytometry,
and immortalized using SV40 T lentivirus (Figure 9B). This cell line was cultured in
polylysine (20μg/mL, Sigma) coated dish with DMEM/F12 media containing 10% FBS,
1ng/mL basic-FGF (Peprotech) and 5ng/mL insulin (Sigma). The normal rat kidney
fibroblast NRK49F was purchased from American Type Culture Collection (ATCC). Both
NRK49F and mouse embryo fibroblast NIH3T3 were cultured in DMEM media (GIBCO,
Invitrogen Co, CA, USA) containing 10% heat-inactivated fetal bovine serum (FBS, GIBCO,
USA), 100U/ml penicillin and 100μg/ml streptomycin. Cells were incubated at 37℃ in a 5%
CO2, humidified atmosphere. Intervention. The cultured cells at 60% confluence were incubated with 1% FBS containing
media for 24 hours followed by treatment of human TNC (Millipore, CAT. CC605, 5μg/mL
or 10μg/mL) or TGFβ (Sigma, 2ng/mL to 10ng/mL). The STAT3 inhibitor Stattic (Selleck, Cell Proliferation. After TNC treatment for 24 hours, the number and viability of the cultured
cells were determined using the Cell Counting Kit-8 (CCK-8, Dojindo Laboratories)
following the manufacturer’s instructions. Briefly, water-soluble tetrazolium salt WST-8 is
reduced by dehydrogenase in the cells and turns into an orange formazan dye which is
soluble in the culture media, and the amount of the formazan dye is proportional to the
number of living cells. The cell proliferation was determined by detecting the proportion of
EdU positive cells using the Click-iT® EdU Imaging Kits (Invitrogen) according to the
manufacturer’s instructions. EdU is detected based on a copper-catalyzed covalent
reaction between an alkyne (contained in the EdU) and azide (contained in the Alexa Fluor
dye). 2μM) was added to the media 30 minutes before TNC treatment. Histology and Pathology Tissue Preparation. The mice were anesthetized with chloral hydrate (500 mg/kg body
weight) intra-peritoneal injection before sacrifice and immediately perfused with 60 ml ice-
cold PBS via the left ventricle. The kidneys were hemi-sectioned horizontally and fixed in
4% paraformaldehyde (PFA) on ice for 1 hour. Then half of each kidney was incubated in
30% sucrose-PBS at 4℃ overnight and embedded in OCT, and the other half was
processed and embedded in paraffin. Immunohistochemistry. Paraffin-embedded 2μm-thick sections of human kidney were
deparaffinized and rehydrated in water. After incubation with 3% H2O2 for 30 minutes at
room temperature (RT), the tissues were given microwave antigen retrieval in citrate buffer
pH 6.0 (100% power for 10 minutes), blocked with 5% BSA in PBS for 1 hour at RT, and followed by the incubation with anti-human TNC antibody (Sigma rabbit polyclonal, 1:500)
at 4℃ overnight. After 3 washes with PBS, the samples were incubated in a horseradish
peroxidase (HRP) conjugated secondary antibody for 45 minutes at RT, followed by
coloration with 3,3- diaminobenzidine solution (DAB). followed by the incubation with anti-human TNC antibody (Sigma rabbit polyclonal, 1:500)
at 4℃ overnight. After 3 washes with PBS, the samples were incubated in a horseradish
peroxidase (HRP) conjugated secondary antibody for 45 minutes at RT, followed by
coloration with 3,3- diaminobenzidine solution (DAB). Immunofluorescence. Frozen tissue embedded in OCT was cut in 5μm thick section. For
IF analysis, the sections were fixed in cold acetone for 3 minutes, washed in PBS, blocked
in 5% bovine serum albumin (BSA) for 30 minutes and then incubated with primary
antibodies in blocking buffer overnight at 4℃. The primary antibodies included: anti-GFP
(Abcam rabbit polyclonal, 1:200; Aves Labs chicken antibodies, 1:500;), anti-AQP2 (Santa
Cruz goat polyclonal, 1:100), anti-CD31 (Abcam rat monoclonal, 1:100), anti-PDGFRβ
(eBioscience rat monoclonal, 1:100), anti-CD34 (Abcam rat monoclonal, 1:200), anti-CD44
(Abcam rat monoclonal, 1:200), anti-α-SMA (Sigma mouse monoclonal, 1:100), anti-FSP-
1 (Abcam rabbit polyclonal, 1:200), anti-NG2 (Millipore rabbit polyclonal, 1:100), anti-TNC
(Sigma rat monoclonal, 1:100), anti-F4/80 (AbD serotec rat monoclonal, 1:200), and anti-
CD68 (AbD serotec rat monoclonal, 1:200), anti-THP (Santa Cruz rabbit polyclonal, 1:200)
antibody. After 3 washes in PBS, the tissue was incubated with FITC- or Cy3- conjugated
anti-IgG/IgY secondary antibody (Millipore, 1:200) for 1 hour at room temperature (RT),
followed by 3 washes in PBS and then covered by Vectashield Mounting Medium with
DAPI (Vector Labs). Quantitative RT-PCRs Renal cortex and papilla isolated from mice kidneys were used to measure TNC mRNA
and TNC reporter eGFP mRNA to determine the expression of TNC in different regions of
kidney. The entire kidney from UUO model was homogenized to exam the TNC expression
and evaluate the severity of fibrosis by detecting the mRNA of TNC, collagen Ⅰα,
fibronectin and PAI-1. Total mRNA was extracted using TRIzol Reagent (Invitrogen),
reversely transcribed using RevertAid RT reagent Kit (Takara) according to the
manufacturer’s protocol. Levels of mRNA were determined by real time qPCR using SYBR
Premix Ex Taq (Takara) and normalized to the eukaryotic 18s rRNA or β-actin. The primers
used were TNC: F 5’-CAA CTG TGC CCT GTC CTA C-3’, R 5’-AAC GCC CTG ACT GTG
GTT A-3’; eGFP: F 5’-CCT CAA GGA CGA CGG CAA C-3’, R 5’-CTC GAT GCG GTT CAC
CAG-3’; collagenⅠα: F 5’-TGA CTG GAA GAG CGG AGA G-3’, R 5’-GAC GGC TGA GTA
GGG AAC A-3’; fibronectin: F 5’-TGG GAG CAT TGT TGT GTC-3’, R 5’-AGC GGT GTC
ACT ACT CTG T-3’; PAI-1: F 5’-ACT TTA CCC CTC CGA GAA-3’, R 5’-CCT GCT GAA
ACA CTT TTA C-3’. Histology and Pathology Fluorescein lotus tetragonolobus lectin (LTL, Vector Labs, 1:100) was
used to label proximal tubules. EdU detection. Paraffin-embedded renal tissue from the mice which had received EdU
injection after operation was used for EdU detection according to the manufacturer’s instructions. EdU positive cells were calculated per high power field to compare the
proliferation between wildtype and TNC-/- mice. instructions. EdU positive cells were calculated per high power field to compare the
proliferation between wildtype and TNC-/- mice. Western Blot and Immunoprecipitation Total protein was extracted from the whole kidney or cultured cells using RIPA buffer
(0.05M Tris, 0.15M NaCl, 1% Triton X-100, 1% sodium deoxycholate, 0.1% SDS, pH7.4)
with PMSF, protease inhibitor cocktail and phosphatase inhibitor cocktail (Roche). The
concentration of the protein was determined using the BCA protein assay kit (Beyotime). For Western blot, equal amount of protein (20-50μg per lane) was loaded in a 7.5% or 10%
SDS-PAGE mini-gel and transferred to a PVDF membrane (Millipore). The membrane was
blocked with 5% BSA in PBST buffer (100mM PBS, pH 7.5, 0.1% Tween-20) and then
incubated in primary antibody diluted with blocking buffer overnight at 4℃. The primary
antibodies included: anti-TNC (IBL), anti-collagen Iα (Novus), anti-αSMA (Sigma), anti-
STAT3 (Αbcam), anti-phospho-STAT3 (Tyr 705, Αbcam), anti-GAPDH (CST) and anti-β-
actin (Sigma) antibody. Membranes were then incubated with appropriate secondary
antibodies and subjected to chemiluminescence detection using ECL Reagent (Millipore). Statistical Analysis Statistical analysis was performed using Prism 5.0 (GraphPad Software Inc.). Comparison
of two factors was performed by two-tailed t test and comparison of two factors with multiple
levels was performed by two-way ANOVA test. A P value <0.05 was considered significant. The results were presented as means and error bars indicate ±SEM. Results TNC was constitutively expressed by the renal medullary stromal cells in normal
kidney TNC was constitutively expressed by the renal medullary stromal cells in normal
kidney To determine the distribution of TNC in non-fibrosis kidneys, we examined TNC expression
using immunohistochemistry (IHC) in non-fibrosis human kidneys and found that TNC
immuno-protein was primarily located in the interstitium of renal medulla, but rarely
detected in the renal cortex (Figure 1A, Supplement Figure 1A&B). This result was
supported by TNC mRNA measurement in human kidneys which were resected because of renal carcinoma (Supplement Figure 1C) and mouse kidneys (Figure 2B) [19]. Since TNC is an extracellular matrix protein, IHC will not be able to identify cells that
express TNC. Therefore, we developed a TNC promoter driven eGFP reporter mice (Figure
2A) [19]. Co-staining for TNC reporter eGFP and AQP2, a marker of collecting ducts,
showed that eGFP positive cells were not epithelial cells. They were primarily expressed
by the renal medullary interstitial cells (RMICs) in normal mouse kidney (Figure 2C&D and
Supplement Figure 2C&D) [20, 21]. To characterize the cell types that expressed TNC, we
co-stained eGFP with specific cell markers in normal kidney and found that eGFP positive
cells expressed platelet-derived growth factor receptor beta (PDGFRβ), a marker of
stromal cells of mesenchymal origin [22] (Figure 2E and Supplement Figure 2E). They did
not express CD31 (a marker of endothelial cells [23]), CD34 (a marker of endothelial cells
and early hematopoietic stem cells and progenitors [24]), or CD44 (a marker of progenitors
[25, 26]), suggesting that TNC was neither expressed in endothelial cells nor progenitor
cells (Figure 2F&G&H and Supplement Figure 2F&G&H). In the normal renal medulla,
there were only a few cells positive for αSMA (a marker of myofibroblasts[27]), FSP1
(recognized as a specific marker of fibroblasts [28], but subsequently identified as a marker
of macrophages[29]), or NG2 (a transmembrane glycoprotein used to label pericytes [30]). We did not find eGFP was colocalized with these three markers (Figure 2I&J&K and
Supplement Figure 2I&J&K). TNC was markedly induced in the fibrotic area of diseased kidney
We previously reported that TNC was significantly increased in injured glomeruli of IgA TNC was markedly induced in the fibrotic area of diseased kidney We previously reported that TNC was significantly increased in injured glomeruli of IgA
nephropathy and expressed by PDGFRβ positive mesangial cells[31]. TNC was also markedly induced in the fibrotic interstitial area of renal cortex of biopsy specimen from
patients with IgA nephropathy and from patients with diabetic nephropathy (Figure 1B-F). To further examine the expression of TNC in kidney fibrosis, we generated a unilateral
ureteral obstruction (UUO) mouse model and found that TNC mRNA and protein were
significantly increased in the fibrotic kidneys (Figure 3A-C). Transgenic TNC reporter mice
also showed a remarkable increase of eGFP mRNA expression and eGFP positive cells in
the UUO kidneys (Figure 3B&D). Similar results were also found in the ischemia
reperfusion (IR) induced kidney fibrosis model (Supplement Figure 3). markedly induced in the fibrotic interstitial area of renal cortex of biopsy specimen from
patients with IgA nephropathy and from patients with diabetic nephropathy (Figure 1B-F). To further examine the expression of TNC in kidney fibrosis, we generated a unilateral
ureteral obstruction (UUO) mouse model and found that TNC mRNA and protein were
significantly increased in the fibrotic kidneys (Figure 3A-C). Transgenic TNC reporter mice
also showed a remarkable increase of eGFP mRNA expression and eGFP positive cells in
the UUO kidneys (Figure 3B&D). Similar results were also found in the ischemia
reperfusion (IR) induced kidney fibrosis model (Supplement Figure 3). TNC deficiency reduced kidney fibrosis in animal models To investigate the role of TNC induction in kidney fibrosis, we generated a TNC
homozygous knockout mice. Deletion of TNC (TNC-/-) significantly attenuated the induction
of collagen Ⅰ expression in the kidneys assessed by IF following UUO by approximate
30% at day 7 and 10 compared with their wild type littermates (Figure 4A, p<0.05). Consistently, mRNAs of fibrosis markers, such as collagen Ⅰα, fibronectin and
plasminogen activator inhibitor-1 (PAI-1), were significantly lower in the TNC-/- mice after
UUO at day 7 and 10 (Figure 4B, p<0.05). Western blot showed that the proteins of
collagen Ⅰα and α-SMA were also significantly reduced in TNC-/- mice after UUO at day
7 (Figure 4C, n=7, p<0.05). TNC expressing cells in the fibrotic kidneys were not epithelial cells To confirm the above results that TNC was significantly induced in fibrosis, we generated
another TNC mouse reporter TNCCreER2-eGFP/+;R26tdTomato/+, in which the cells that express
TNC will be labeled with red fluorescence protein tdTomato permanently following tamoxifen (Figure 5A). There are some differences between tdTomato reporter mice and
eGFP reporter mice. First, the tdTomato reporter not only labels the cells which are
expressing TNC but also the cells once expressed TNC, while the eGFP reporter only
labels the cells which are expressing TNC. Second, tdTomato is only expressed in the
presence of tamoxifen while eGFP not. Third, the expression of tdTomato is driven by the
promoter Rosa26 and the expression level is high with strong autofluorescence. The
tdTomato expression does not reflect TNC expression, while eGFP does. Fourth, the
downstream IRES-eGFP protein may be less efficiently translated than the coding
sequence of Cre which is directly adjacent to the promoter. However, tamoxifen induction
efficiency as well as excision rate vary among tissue types. tamoxifen (Figure 5A). There are some differences between tdTomato reporter mice and
eGFP reporter mice. First, the tdTomato reporter not only labels the cells which are
expressing TNC but also the cells once expressed TNC, while the eGFP reporter only
labels the cells which are expressing TNC. Second, tdTomato is only expressed in the
presence of tamoxifen while eGFP not. Third, the expression of tdTomato is driven by the
promoter Rosa26 and the expression level is high with strong autofluorescence. The
tdTomato expression does not reflect TNC expression, while eGFP does. Fourth, the
downstream IRES-eGFP protein may be less efficiently translated than the coding
sequence of Cre which is directly adjacent to the promoter. However, tamoxifen induction
efficiency as well as excision rate vary among tissue types. Using this bi-transgenic reporter mice TNCCreER2-eGFP/+;R26tdTomato/+, we confirmed that
TNC expressing cells were primarily expressed in the medulla of normal kidney (Figure
5B&D). Then we generated UUO model in this mice and administered tamoxifen right after
operation, and also found that the tdTomato positive cells were significantly increased in
the cortex of fibrotic kidneys (Figure 5B-E). These tdTomato positive cells did not stained
for Lotus tetragonolobus lectin (LTL, proximal tubule) nor expressed AQP2 (collecting duct),
suggesting that they were not epithelial cells (Figure 5F&G and Supplement Figure 4). The expanded TNC expressing cells in the fibrotic renal cortex were not originated
from the constitutive TNC expressing cells in renal papilla The expanded TNC expressing cells in the fibrotic renal cortex were not originated
from the constitutive TNC expressing cells in renal papilla TNC expressing cells in the fibrotic kidneys were mostly PDGFRβ+NG2+ stroma To characterize the cells that express TNC in the fibrotic kidneys, we co-labeled TNC
reporter eGFP with specific markers in UUO model. We did not use tdTomato as reporter
because its autofluorescence is too strong that it can be detected in the green channel. We found that the eGFP positive cells were positive for PDGFRβ, consistent with stromal
cells (Figure 6A and Supplement Figure 5A). Some of the eGFP positive cells were αSMA
positive, and not all αSMA positive myofibroblasts expressed TNC (Figure 6B&I and
Supplement Figure 5B&I). In contrast, most of the eGFP positive cells were NG2 (a marker
of pericyte) positive (Figure 6C&J and Supplement Figure 5C&J). None of the eGFP
positive cells were positive for FSP1 (Figure 6D and Supplement Figure 5D). In addition,
the eGFP positive cells were neither CD34 nor CD44 positive (Figure 6E&F and
Supplement Figure 5E&F), and were neither CD68 nor F4/80 (markers of macrophages)
positive (Figure 6G&H and Supplement Figure 5G&H). Interestingly, the TNC reporter eGFP or tdTomato positive cells in UUO models were
predominantly located adjacent to the tubular structures that were dilated and negative for
LTL, THP (Tamm-Horsfall protein, a marker of the thick ascending limb), AQP2, and CD31
(Figure 7A-C and Supplement Figure 6A-C). To examine whether these tubular structures
surrounded by TNC reporter positive cells were injured renal tubules, we co-labeled TNC
reporter with LTL or antibodies against THP or AQP2 on serial sections. This serial section
study showed that the tubules surrounded by TNC expressing cells had continuation to the
tubules with positive AQP2 staining, supporting that TNC expressing cells were localized
surrounding the injured tubular epithelial cells in UUO kidneys (Figure 7C and Supplement
Figure 6C). from the constitutive TNC expressing cells in renal papilla It has been reported that renal papilla contains stem cells[32], and TNC has also been reported to be expressed in stem cells[33]. To determine whether the TNC expressing cells
in renal medulla were progenitors or source of the expanded stromal cells in the renal
cortex after fibrosis, we conducted a cell linage tracing study in TNCCreER2-
eGFP/+;R26tdTomato/+ mice. If not exposed to tamoxifen, there were only a few tdTomato
positive cells in normal or fibrotic kidneys, suggesting that the reporter mice had little
leakage and could be used for lineage tracing analysis (Supplement Figure 7). As detailed
in Figure 8A, the bi-transgenic mice were treated with tamoxifen (1.5mg/d for 5 days) at 8
weeks old and subjected to UUO operation after 12 weeks’ washout (Figure 8A). The cells
that were labeled by tdTomato in the kidney were examined 7 days after UUO: the
tdTomato positive cells were limited to the renal medulla, although they were significantly
increased (Figure 8B&C). However, if a second series of tamoxifen was given after UUO
operation, the tdTomato positive cells were found throughout the kidney (Figure 8D). These
results suggested that following UUO, the original TNC expressing cells in renal medulla
proliferated, but they were neither the source of TNC nor the progenitors of expanded
interstitial cells in the cortex. Similar results were found in IR induced fibrosis model
(Supplement Figure 8). reported to be expressed in stem cells[33]. To determine whether the TNC expressing cells
in renal medulla were progenitors or source of the expanded stromal cells in the renal
cortex after fibrosis, we conducted a cell linage tracing study in TNCCreER2-
eGFP/+;R26tdTomato/+ mice. If not exposed to tamoxifen, there were only a few tdTomato
positive cells in normal or fibrotic kidneys, suggesting that the reporter mice had little
leakage and could be used for lineage tracing analysis (Supplement Figure 7). As detailed
in Figure 8A, the bi-transgenic mice were treated with tamoxifen (1.5mg/d for 5 days) at 8
weeks old and subjected to UUO operation after 12 weeks’ washout (Figure 8A). The cells
that were labeled by tdTomato in the kidney were examined 7 days after UUO: the
tdTomato positive cells were limited to the renal medulla, although they were significantly
increased (Figure 8B&C). However, if a second series of tamoxifen was given after UUO
operation, the tdTomato positive cells were found throughout the kidney (Figure 8D). from the constitutive TNC expressing cells in renal papilla These
results suggested that following UUO, the original TNC expressing cells in renal medulla
proliferated, but they were neither the source of TNC nor the progenitors of expanded
interstitial cells in the cortex. Similar results were found in IR induced fibrosis model
(Supplement Figure 8). TNC enhanced fibrosis by promoting interstitial cells proliferation via STAT3 To explore the mechanism by which TNC enhanced fibrosis, we examined the effect of
TNC on interstitial cell proliferation and the potential signaling mechanism. Firstly, we
evaluated the cell proliferation using EdU incorporation study. In vivo, EdU positive cells
were markedly increased and predominantly located in the renal interstitium after UUO, and TNC deletion significantly reduced the number of EdU positive cells in the obstructed
kidney compared with wild-type mouse (Figure 9A). To further examine the effect of TNC
on cell proliferation, we cultured TNC-Cell (an interstitial cell line obtained from fibrotic
kidney, Figure 9B), NIH3T3 and NRK49F cells. Exogenous TNC dose dependently
increased the number of these cultured cells as assessed by CCK8 kit (Figure 9C-E). Exogenous TNC also markedly increased EdU incorporation in NRK49F cells, consistent
with increasing cell proliferation (Figure 9F). and TNC deletion significantly reduced the number of EdU positive cells in the obstructed
kidney compared with wild-type mouse (Figure 9A). To further examine the effect of TNC
on cell proliferation, we cultured TNC-Cell (an interstitial cell line obtained from fibrotic
kidney, Figure 9B), NIH3T3 and NRK49F cells. Exogenous TNC dose dependently
increased the number of these cultured cells as assessed by CCK8 kit (Figure 9C-E). Exogenous TNC also markedly increased EdU incorporation in NRK49F cells, consistent
with increasing cell proliferation (Figure 9F). To determine the signaling mechanism by which TNC promoted cell proliferation, we
screened 45 phospho-kinases in cultured fibroblast treated with or without TNC, and
identified STAT3 as one of the candidates that responded to TNC treatment. The potential
significance of STAT3 signaling in mediating the effect of TNC was further validated in UUO
kidney and cultured cells. In vivo, the STAT3 and phosphor-STAT3 levels were markedly
increased after UUO, and TNC deletion significantly reduced the phospho-STAT3 levels
(Figure 10A&B). In cultured cells (TNC-Cell), TNC markedly increased the phosphorylation
of STAT3, peaking at 45 minutes (Figure 10C). The effect of TNC on cell proliferation was
blocked by the STAT3 inhibitor Stattic (Figure 10D). Discussion Fibrosis is a common event that leads to irreversible loss of organ function and organ
failure, and so far has no effectively therapy [34]. The excessive deposition of ECM is a
key feature of fibrosis. Accumulating evidence has revealed that ECM is not only the
product of fibrosis, but also actively drives the persistence of fibrosis [4, 6, 35]. TNC is an ECM glycoprotein that is under tight temporal and spatial regulation [36, 37]. A recent study
suggests that TNC played an important role in kidney fibrosis [18]. The present studies
showed that (1) the TNC expression levels and the number of TNC expressing cells
dramatically increased during fibrosis in the kidney; (2) deletion of TNC attenuated kidney
fibrosis; (3) TNC facilitated fibrosis, at least in part, by promoting the phosphorylation of
STAT3; (4) the cells expressing TNC were mainly localized surrounding the injured tubules,
suggesting stromal cells in response to epithelium damage; (5) the cells that produced
TNC were a special population of stromal cells, 80% of which were positive for NG2 and
50% positive for αSMA. Our study provided strong evidence that TNC plays an important role in promoting
kidney fibrosis using a TNC knock-out mouse line, consistent with a previous study using
siRNA-TNC injection approach [18]. In chronic kidney disease progression, tubular
epithelia have been reported to sense and response to damage [38-41], followed by the
interstitial stroma activation [42]. Our study showed that the TNC expressing cells (eGFP
positive cells) in the fibrotic cortex were primarily localized around the injured tubules that
had lost their epithelial markers. This specific localization may indicate that TNC plays an
important role in linking tubule injury and interstitial cell activation and fibrosis. Interestingly,
TNC was reported to be expressed only in the mesenchyme surrounding the epithelia
undergoing differentiation during embryonic development [43]. TNC was induced by
epithelial-mesenchymal interactions where epithelia are undergoing differentiation or
restoration [44]. TGFβ, a well-established driver of fibrosis, was found to be upregulated in
the injured tubular epithelium [45, 46]. Our study showed that TNC was, at least partially, induced by TGFβ (Supplement Figure 9). Taken together, these data highlight a potential
role of TNC in linking tubule damage to persistence of interstitial fibrosis. TNC is a matricellular protein with multiple domains including EGF-like repeats, FNⅢ
repeats and fibrinogen-like globe [13]. These domains can serve as ligands of cell-surface
receptors and activate intracellular signals. Discussion These studies
suggest TNC is expressed in early stages of activated interstitial cells that will be
transdifferentiated to collagen producing myofibroblast in response to injury. It has been reported that renal papilla contains stem cells or progenitors which
repopulate after kidney injury[32]. TNC is constitutively expressed in the renal medullary
interstitial cells[19]. To examine whether these TNC expressing cells in renal medulla were
progenitors that migrated to renal cortex after injury, we used a genetic mouse model that
specifically tracked and fate-mapped renal medullary interstitial TNC expressing cells. The
results demonstrated that the cortical interstitial cells that express TNC during fibrosis were
not originated from papilla. Discussion Receptors for TNC include EGFR, integrins
and toll-like receptors [47-50]. These receptors have an enzymatic activity in intracellular
domain and usually regulate downstream kinase cascade. Therefore, we screened the
kinases and identified STAT3 as a potential signaling pathway mediating the effect of TNC. In UUO induced kidney fibrosis models, deficiency of TNC gene significantly reduced the
phosphorylation of STAT3. In cultured cells, TNC promoted fibroblast proliferation by
upregulating the phosphorylation of STAT3, because STAT3 inhibitor blocked the cells
proliferation induced by TNC. STAT3 is a transcription factor and has been reported as a
downstream target of EGFR [51, 52]. Mounting evidence has indicated that STAT3
activation is associated with cell proliferation [53] and the development of fibrosis [54, 55]. Myofibroblasts, characterized by positive staining for αSMA, are the key contributor of
kidney fibrosis [56]. It is well known that myofibroblasts are the major source of ECM [57]. Our study showed some of the αSMA positive cells expressed TNC, but not all TNC was
expressed by αSMA positive myofibroblasts. This result was consistent with an early study
in myocardial infarction which suggests that TNC was induced in the very early stage after
injury even before the appearance of myofibroblasts [58]. NG2 is a transmembrane
proteoglycan and usually used to label pericytes (NG2+PDGFRβ+) which had been
suggested to be a precursor of myofibroblasts [59]. However, NG2 expression has also been identified in several types of immature cells (such as progenitor and tumor cells) and
is downregulated when cells become mature and quiescent [30, 60, 61]. Similarly, TNC
expression has also been found in developing cells, such as progenitors during
development, tumor stroma and mesenchymal stem cells [33, 44, 62]. In our study, most
of the TNC expressing cells were positive for both NG2 and PDGFRβ. These studies
suggest TNC is expressed in early stages of activated interstitial cells that will be
transdifferentiated to collagen producing myofibroblast in response to injury. been identified in several types of immature cells (such as progenitor and tumor cells) and
is downregulated when cells become mature and quiescent [30, 60, 61]. Similarly, TNC
expression has also been found in developing cells, such as progenitors during
development, tumor stroma and mesenchymal stem cells [33, 44, 62]. In our study, most
of the TNC expressing cells were positive for both NG2 and PDGFRβ. Conclusions In conclusion, our study provides strong evidence that non-structural matrix protein TNC
that contains multiple functional domains plays an important role in kidney fibrosis. TNC is
mostly expressed by NG2+PDGFRβ+ cells around the injured tubules and STAT3 is a
signaling pathway at least partially mediating the profibrotic effect of TNC. TNC pathway
may serve as a potential therapeutic target to treat interstitial fibrosis and progression of
kidney injury. Ethics approval and consent to participate Human study was approved by the ethics committee of Huashan Hospital, Fudan
University, approval number KY2016-394. Informed consent was signed for each patient. Animal study was approved by the Animal Care and Use Committees of Fudan University,
approval number 20150588A274. Funding This work was supported by Natural Science Foundation of China grants NSFC: This work was supported by Natural Science Foundation of China grants NSFC:
81520108006 and 81930120 (C. M. H.), and 82170735 (Q. X.), and 82170750, and VA
Merit Awards I01 BX002196 (R. Z.) and I01 BX002025 (A. P.) as well as R01 DK069921
(R. Z.) and R01 DK119212 (A. P.), and P30-DK114809 (A. P. and R. Z.). A. P. is the
recipient of a Senior Research Career Scientist award from the Veteran Affairs. The
funders had no role in study design, data collection and analysis, decision to publish, or
preparation of the manuscript. Competing interests Investigators declare no other conflicts of financial interest for participation in this
research. CCK8, cell counting kit-8
TGFβ, transforming growth factor β CCK8, cell counting kit-8 TGFβ, transforming growth factor β List of abbreviations ECM, extracellular matrix TNC, tenascin-C
STAT3, signal transducer and activator of transcription 3
CreER2, a Cre recombinase fused to a human estrogen receptor ligand binding domain
IRES, internal ribosome entry site
eGFP/GFP, enhanced green fluorescent protein/ green fluorescent protein
NG2, neuron glial antigen
PDGFRβ, platelet derived growth factor receptor β
αSMA, smooth muscle actin
RMICs, renal medullary interstitial cells
EGF-L, epidermal growth factor like
EGFR, epidermal growth factor receptor
TLR, toll like receptors
IHC, immunohistochemistry
IF, immunofluorescence
AQP2, aquaporin 2
FSP1, fibroblast-specific protein 1
UUO, unilateral ureteral obstruction
IR, ischemia reperfusion
LTL, lotus tetragonolobus lectin
THP, Tamm-Horsfall protein STAT3, signal transducer and activator of transcription 3 CreER2, a Cre recombinase fused to a human estrogen receptor ligand binding domain
IRES, internal ribosome entry site Authors’ contributions Conceptualization, C.M.H., Q.H., M.Z., and X.M; Methodology, Q.H., C.M.H., X.M., and S.L;
Investigation, Q.H., M.Z., X.M., and S.L;
Writing – Original Draft, C.M.H., Q.H., R.Z. and A.P;
Funding Acquisition, C.M.H, R.Z. and A.P;
Resources, D.S., Y.X., R.C., Y.G., and X.H. Methodology, Q.H., C.M.H., X.M., and S.L;
Investigation, Q.H., M.Z., X.M., and S.L;
Writing – Original Draft, C.M.H., Q.H., R.Z. and A.P;
Funding Acquisition, C.M.H, R.Z. and A.P;
Resources, D.S., Y.X., R.C., Y.G., and X.H. Methodology, Q.H., C.M.H., X.M., and S.L;
Investigation, Q.H., M.Z., X.M., and S.L;
Writing – Original Draft, C.M.H., Q.H., R.Z. and A.P;
Funding Acquisition, C.M.H, R.Z. and A.P;
Resources, D.S., Y.X., R.C., Y.G., and X.H. Acknowledgements We acknowledge that all authors had access to the data and played a role in writing this
article. Reference 1. Kuncio GS, Neilson EG, Haverty T: Mechanisms of tubulointerstitial fibrosis. Kidney Int 1991,
39(3):550-556. 2. Ben Salem C, Slim R, Fathallah N: Fibrosis--A Common Pathway to Organ Injury and Failure. N
Engl J Med 2015, 373(1):95. 3. Herrera J, Henke CA, Bitterman PB: Extracellular matrix as a driver of progressive fibrosis. J
Clin Invest 2018, 128(1):45-53. 4. Wight TN, Potter-Perigo S: The extracellular matrix: an active or passive player in fibrosis? American Journal of Physiology-Gastrointestinal and Liver Physiology 2011, 301(6):G950-G955. 5. Frangogiannis NG: The extracellular matrix in myocardial injury, repair, and remodeling. J Clin
Invest 2017, 127(5):1600-1612. 6. Bonnans C, Chou J, Werb Z: Remodelling the extracellular matrix in development and disease. Nature Reviews Molecular Cell Biology 2014, 15(12):786-801. 7. Clause KC, Barker TH: Extracellular matrix signaling in morphogenesis and repair. Curr Opin
Biotechnol 2013, 24(5):830-833. 8. Bornstein P, Sage EH: Matricellular proteins: extracellular modulators of cell function. Curr
Opin Cell Biol 2002, 14(5):608-616. 9. Murphy-Ullrich JE, Sage EH: Revisiting the matricellular concept. Matrix Biol 2014, 37:1-14. 10. Theocharis AD, Skandalis SS, Gialeli C, Karamanos NK: Extracellular matrix structure. Adv Drug
Deliv Rev 2016, 97:4-27. 11. Xu Y, Li N, Gao J, Shang D, Zhang M, Mao X, Chen R, Zheng J, Shan Y, Chen M et al: Elevated
Serum Tenascin-C Predicts Mortality in Critically Ill Patients With Multiple Organ Dysfunction. Front Med (Lausanne) 2021, 8:759273. 12. Gao W, Li J, Ni H, Shi H, Qi Z, Zhu S, Hao C, Xie Q, Luo X, Xie K: Tenascin C: A Potential Biomarker
for Predicting the Severity of Coronary Atherosclerosis. J Atheroscler Thromb 2019, 26(1):31-
38. 13. Midwood KS, Hussenet T, Langlois B, Orend G: Advances in tenascin-C biology. Cell Mol Life Sci 5. Frangogiannis NG: The extracellular matrix in myocardial injury, repair, and remodeling. J Clin
Invest 2017, 127(5):1600-1612. 6. Bonnans C, Chou J, Werb Z: Remodelling the extracellular matrix in development and disease
Nature Reviews Molecular Cell Biology 2014, 15(12):786-801. 6. Bonnans C, Chou J, Werb Z: Remodelling the extracellular matrix in development and disease. Nature Reviews Molecular Cell Biology 2014, 15(12):786-801. 7. Clause KC, Barker TH: Extracellular matrix signaling in morphogenesis and repair. Curr Opin
Biotechnol 2013, 24(5):830-833. 7. Clause KC, Barker TH: Extracellular matrix signaling in morphogenesis and repair. Curr Opin
Biotechnol 2013, 24(5):830-833. Bornstein P, Sage EH: Matricellular proteins: extracellular modulators of cell function. Curr
Opin Cell Biol 2002, 14(5):608-616. Reference Murphy-Ullrich JE, Sage EH: Revisiting the matricellular concept. Matrix Biol 2014, 37:1-14. 10. Theocharis AD, Skandalis SS, Gialeli C, Karamanos NK: Extracellular matrix structure. Adv Drug
Deliv Rev 2016, 97:4-27. 11. Xu Y, Li N, Gao J, Shang D, Zhang M, Mao X, Chen R, Zheng J, Shan Y, Chen M et al: Elevated
Serum Tenascin-C Predicts Mortality in Critically Ill Patients With Multiple Organ Dysfunction. Front Med (Lausanne) 2021, 8:759273. 12. Gao W, Li J, Ni H, Shi H, Qi Z, Zhu S, Hao C, Xie Q, Luo X, Xie K: Tenascin C: A Potential Biomarker
for Predicting the Severity of Coronary Atherosclerosis. J Atheroscler Thromb 2019, 26(1):31-
38. 13. Midwood KS, Hussenet T, Langlois B, Orend G: Advances in tenascin-C biology. Cell Mol Life Sci 2011, 68(19):3175-3199. 14. De Laporte L, Rice JJ, Tortelli F, Hubbell JA: Tenascin C promiscuously binds growth factors via
its fifth fibronectin type III-like domain. PLoS One 2013, 8(4):e62076. 15. Shimojo N, Hashizume R, Kanayama K, Hara M, Suzuki Y, Nishioka T, Hiroe M, Yoshida T,
Imanaka-Yoshida K: Tenascin-C may accelerate cardiac fibrosis by activating macrophages via
the integrin alphaVbeta3/nuclear factor-kappaB/interleukin-6 axis. Hypertension 2015,
66(4):757-766. 16. El-Karef A, Yoshida T, Gabazza EC, Nishioka T, Inada H, Sakakura T, Imanaka-Yoshida K:
Deficiency of tenascin-C attenuates liver fibrosis in immune-mediated chronic hepatitis in
mice. The Journal of Pathology 2007, 211(1):86-94. 17. Bhattacharyya S, Wang W, Morales-Nebreda L, Feng G, Wu M, Zhou X, Lafyatis R, Lee J,
Hinchcliff M, Feghali-Bostwick C et al: Tenascin-C drives persistence of organ fibrosis. Nat
Commun 2016, 7:11703. 18. Fu H, Tian Y, Zhou L, Zhou D, Tan RJ, Stolz DB, Liu Y: Tenascin-C Is a Major Component of the
Fibrogenic Niche in Kidney Fibrosis. J Am Soc Nephrol 2017, 28(3):785-801. 19. He W, Xie Q, Wang Y, Chen J, Zhao M, Davis LS, Breyer MD, Gu G, Hao CM: Generation of a
tenascin-C-CreER2 knockin mouse line for conditional DNA recombination in renal medullary
interstitial cells. PLoS One 2013, 8(11):e79839. 20. Fushimi K, Uchida S, Hara Y, Hirata Y, Marumo F, Sasaki S: Cloning and expression of apical
membrane water channel of rat kidney collecting tubule. Nature 1993, 361(6412):549-552. 21. Nielsen S, DiGiovanni SR, Christensen EI, Knepper MA, Harris HW: Cellular and subcellular
immunolocalization of vasopressin-regulated water channel in rat kidney. Proc Natl Acad Sci
U S A 1993, 90(24):11663-11667. 22. mesangial cell proliferation and matrix protein production in IgA nephropathy. Nephrology
(Carlton) 2022, 27(5):458-466. 32. Oliver JA, Maarouf O, Cheema FH, Martens TP, Al-Awqati Q: The renal papilla is a niche for
adult kidney stem cells. J Clin Invest 2004, 114(6):795-804. 33. Chiquet-Ehrismann R, Orend G, Chiquet M, Tucker RP, Midwood KS: Tenascins in stem cell
niches. Matrix Biol 2014, 37:112-123. 34. Rockey DC, Bell PD, Hill JA: Fibrosis--a common pathway to organ injury and failure. N Engl J
Med 2015, 372(12):1138-1149. 35. Herrera J, Henke CA, Bitterman PB: Extracellular matrix as a driver of progressive fibrosis. Journal of Clinical Investigation 2018, 128(1):45-53. 36. Jones FS, Jones PL: The tenascin family of ECM glycoproteins: structure, function, and
regulation during embryonic development and tissue remodeling. Dev Dyn 2000, 218(2):235-
259. 37. Chiquet-Ehrismann R, Chiquet M: Tenascins: regulation and putative functions during
pathological stress. J Pathol 2003, 200(4):488-499. 38. Kang HM, Ahn SH, Choi P, Ko YA, Han SH, Chinga F, Park AS, Tao J, Sharma K, Pullman J et al:
Defective fatty acid oxidation in renal tubular epithelial cells has a key role in kidney fibrosis
development. Nat Med 2015, 21(1):37-46. 39. Wu H, Lai CF, Chang-Panesso M, Humphreys BD: Proximal Tubule Translational Profiling during
Kidney Fibrosis Reveals Proinflammatory and Long Noncoding RNA Expression Patterns with
Sexual Dimorphism. J Am Soc Nephrol 2020, 31(1):23-38. 40. Gewin LS: Renal fibrosis: Primacy of the proximal tubule. Matrix Biol 2018, 68-69:248-262. 41. Qi R, Yang C: Renal tubular epithelial cells: the neglected mediator of tubulointerstitial
fibrosis after injury. Cell Death Dis 2018, 9(11):1126. 42. Humphreys BD: Mechanisms of Renal Fibrosis. Annu Rev Physiol 2018, 80:309-326. 43. Aufderheide E, Chiquet-Ehrismann R, Ekblom P: Epithelial-mesenchymal interactions in the
developing kidney lead to expression of tenascin in the mesenchyme. J Cell Biol 1987,
105(1):599-608. 44. Midwood KS, Orend G: The role of tenascin-C in tissue injury and tumorigenesis. Journal of
Cell Communication and Signaling 2009, 3(3-4):287-310. 45. Meng XM, Nikolic-Paterson DJ, Lan HY: TGF-beta: the master regulator of fibrosis. Nat Rev
Nephrol 2016, 12(6):325-338. 46. Yamamoto T, Noble NA, Cohen AH, Nast CC, Hishida A, Gold LI, Border WA: Expression of
transforming growth factor-beta isoforms in human glomerular diseases. Kidney Int 1996,
49(2):461-469. 47. Iyer AK, Tran KT, Borysenko CW, Cascio M, Camacho CJ, Blair HC, Bahar I, Wells A: Tenascin
cytotactin epidermal growth factor-like repeat binds epidermal growth factor receptor with
low affinity. J Cell Physiol 2007, 211(3):748-758. 48. Reference Heldin CH, Westermark B: Mechanism of action and in vivo role of platelet-derived growth
factor. Physiol Rev 1999, 79(4):1283-1316. 23. Favaloro EJ, Moraitis N, Koutts J, Exner T, Bradstock KF: Endothelial cells and normal circulating
haemopoietic cells share a number of surface antigens. Thromb Haemost 1989, 61(2):217-
224. 24. Andrews RG, Singer JW, Bernstein ID: Precursors of colony-forming cells in humans can be
distinguished from colony-forming cells by expression of the CD33 and CD34 antigens and
light scatter properties. J Exp Med 1989, 169(5):1721-1731. 25. Hostettler KE, Gazdhar A, Khan P, Savic S, Tamo L, Lardinois D, Roth M, Tamm M, Geiser T:
Multipotent mesenchymal stem cells in lung fibrosis. PLoS One 2017, 12(8):e0181946. 26. Morath I, Hartmann TN, Orian-Rousseau V: CD44: More than a mere stem cell marker. Int J
Biochem Cell Biol 2016, 81(Pt A):166-173. 27. Hinz B, Phan SH, Thannickal VJ, Galli A, Bochaton-Piallat ML, Gabbiani G: The myofibroblast:
one function, multiple origins. Am J Pathol 2007, 170(6):1807-1816. 28. Strutz F, Okada H, Lo CW, Danoff T, Carone RL, Tomaszewski JE, Neilson EG: Identification and
characterization of a fibroblast marker: FSP1. J Cell Biol 1995, 130(2):393-405. 29. Osterreicher CH, Penz-Osterreicher M, Grivennikov SI, Guma M, Koltsova EK, Datz C, Sasik R,
Hardiman G, Karin M, Brenner DA: Fibroblast-specific protein 1 identifies an inflammatory
subpopulation of macrophages in the liver. Proc Natl Acad Sci U S A 2011, 108(1):308-313. 30. Stallcup WB: The NG2 Proteoglycan in Pericyte Biology. Adv Exp Med Biol 2018, 1109:5-19. 31. Yan M, Liu S, Zhang M, Lai L, Xie Q, Hao CM: Mesangial cell-derived tenascin-C contributes to mesangial cell proliferation and matrix protein production in IgA nephropathy. Nephrology
(Carlton) 2022, 27(5):458-466. mesangial cell proliferation and matrix protein production in IgA nephropathy. Nephrology
(Carlton) 2022, 27(5):458-466. 59. Humphreys BD, Lin SL, Kobayashi A, Hudson TE, Nowlin BT, Bonventre JV, Valerius MT,
McMahon AP, Duffield JS: Fate tracing reveals the pericyte and not epithelial origin of
myofibroblasts in kidney fibrosis. Am J Pathol 2010, 176(1):85-97. 60. Stallcup WB: The NG2 proteoglycan: past insights and future prospects. J Neurocytol 2002,
31(6-7):423-435. 61. Stallcup WB, Huang FJ: A role for the NG2 proteoglycan in glioma progression. Cell Adh Migr
2008, 2(3):192-201. 62. Chiquet-Ehrismann R, Mackie EJ, Pearson CA, Sakakura T: Tenascin: an extracellular matrix
protein involved in tissue interactions during fetal development and oncogenesis. Cell 1986,
47(1):131-139. 51. Abdelhamed S, Ogura K, Yokoyama S, Saiki I, Hayakawa Y: AKT-STAT3 Pathway as a
Downstream Target of EGFR Signaling to Regulate PD-L1 Expression on NSCLC cells. J Cancer
2016, 7(12):1579-1586. 52. Wu J, Patmore DM, Jousma E, Eaves DW, Breving K, Patel AV, Schwartz EB, Fuchs JR, Cripe TP,
Stemmer-Rachamimov AO et al: EGFR-STAT3 signaling promotes formation of malignant
peripheral nerve sheath tumors. Oncogene 2014, 33(2):173-180. 53. Daniel JM, Dutzmann J, Bielenberg W, Widmer-Teske R, Gunduz D, Hamm CW, Sedding DG:
Inhibition of STAT3 signaling prevents vascular smooth muscle cell proliferation and
neointima formation. Basic Res Cardiol 2012, 107(3):261. 54. Seo HY, Jeon JH, Jung YA, Jung GS, Lee EJ, Choi YK, Park KG, Choe MS, Jang BK, Kim MK et al:
Fyn deficiency attenuates renal fibrosis by inhibition of phospho-STAT3. Kidney Int 2016,
90(6):1285-1297. 55. Chakraborty D, Sumova B, Mallano T, Chen CW, Distler A, Bergmann C, Ludolph I, Horch RE,
Gelse K, Ramming A et al: Activation of STAT3 integrates common profibrotic pathways to
promote fibroblast activation and tissue fibrosis. Nat Commun 2017, 8(1):1130. 56. Meran S, Steadman R: Fibroblasts and myofibroblasts in renal fibrosis. Int J Exp Pathol 2011,
92(3):158-167. 57. Klingberg F, Hinz B, White ES: The myofibroblast matrix: implications for tissue repair and
fibrosis. J Pathol 2013, 229(2):298-309. 58. Imanaka-Yoshida K, Hiroe M, Nishikawa T, Ishiyama S, Shimojo T, Ohta Y, Sakakura T, Yoshida T:
Tenascin-C modulates adhesion of cardiomyocytes to extracellular matrix during tissue
remodeling after myocardial infarction. Lab Invest 2001, 81(7):1015-1024. 59. Humphreys BD, Lin SL, Kobayashi A, Hudson TE, Nowlin BT, Bonventre JV, Valerius MT,
McMahon AP, Duffield JS: Fate tracing reveals the pericyte and not epithelial origin of
myofibroblasts in kidney fibrosis. Am J Pathol 2010, 176(1):85-97. 60. Stallcup WB: The NG2 proteoglycan: past insights and future prospects. J Neurocytol 2002,
31(6-7):423-435. mesangial cell proliferation and matrix protein production in IgA nephropathy. Nephrology
(Carlton) 2022, 27(5):458-466. Swindle CS, Tran KT, Johnson TD, Banerjee P, Mayes AM, Griffith L, Wells A: Epidermal growth
factor (EGF)-like repeats of human tenascin-C as ligands for EGF receptor. J Cell Biol 2001,
154(2):459-468. 49. Tucker RP, Chiquet-Ehrismann R: Tenascin-C: Its functions as an integrin ligand. Int J Biochem
Cell Biol 2015, 65:165-168. 50. Midwood K, Sacre S, Piccinini AM, Inglis J, Trebaul A, Chan E, Drexler S, Sofat N, Kashiwagi M Orend G et al: Tenascin-C is an endogenous activator of Toll-like receptor 4 that is essential
for maintaining inflammation in arthritic joint disease. Nat Med 2009, 15(7):774-780. Orend G et al: Tenascin-C is an endogenous activator of Toll-like receptor 4 that is essential
for maintaining inflammation in arthritic joint disease. Nat Med 2009, 15(7):774-780. Orend G et al: Tenascin-C is an endogenous activator of Toll-like receptor 4 that is essential
for maintaining inflammation in arthritic joint disease. Nat Med 2009, 15(7):774-780. 51. Abdelhamed S, Ogura K, Yokoyama S, Saiki I, Hayakawa Y: AKT-STAT3 Pathway as a
Downstream Target of EGFR Signaling to Regulate PD-L1 Expression on NSCLC cells. J Cancer
2016, 7(12):1579-1586. 52. Wu J, Patmore DM, Jousma E, Eaves DW, Breving K, Patel AV, Schwartz EB, Fuchs JR, Cripe TP,
Stemmer-Rachamimov AO et al: EGFR-STAT3 signaling promotes formation of malignant
peripheral nerve sheath tumors. Oncogene 2014, 33(2):173-180. 53. Daniel JM, Dutzmann J, Bielenberg W, Widmer-Teske R, Gunduz D, Hamm CW, Sedding DG:
Inhibition of STAT3 signaling prevents vascular smooth muscle cell proliferation and
neointima formation. Basic Res Cardiol 2012, 107(3):261. 54. Seo HY, Jeon JH, Jung YA, Jung GS, Lee EJ, Choi YK, Park KG, Choe MS, Jang BK, Kim MK et al:
Fyn deficiency attenuates renal fibrosis by inhibition of phospho-STAT3. Kidney Int 2016,
90(6):1285-1297. 55. Chakraborty D, Sumova B, Mallano T, Chen CW, Distler A, Bergmann C, Ludolph I, Horch RE,
Gelse K, Ramming A et al: Activation of STAT3 integrates common profibrotic pathways to
promote fibroblast activation and tissue fibrosis. Nat Commun 2017, 8(1):1130. 56. Meran S, Steadman R: Fibroblasts and myofibroblasts in renal fibrosis. Int J Exp Pathol 2011,
92(3):158-167. 57. Klingberg F, Hinz B, White ES: The myofibroblast matrix: implications for tissue repair and
fibrosis. J Pathol 2013, 229(2):298-309. 58. Imanaka-Yoshida K, Hiroe M, Nishikawa T, Ishiyama S, Shimojo T, Ohta Y, Sakakura T, Yoshida T:
Tenascin-C modulates adhesion of cardiomyocytes to extracellular matrix during tissue
remodeling after myocardial infarction. Lab Invest 2001, 81(7):1015-1024. Figures Figures mesangial cell proliferation and matrix protein production in IgA nephropathy. Nephrology
(Carlton) 2022, 27(5):458-466. 61. Stallcup WB, Huang FJ: A role for the NG2 proteoglycan in glioma progression. Cell Adh Migr
2008, 2(3):192-201. 62. Chiquet-Ehrismann R, Mackie EJ, Pearson CA, Sakakura T: Tenascin: an extracellular matrix
protein involved in tissue interactions during fetal development and oncogenesis. Cell 1986,
47(1):131-139. 62. Chiquet-Ehrismann R, Mackie EJ, Pearson CA, Sakakura T: Tenascin: an extracellular matrix
protein involved in tissue interactions during fetal development and oncogenesis. Cell 1986,
47(1):131-139. Figure 1 See image above for ¦gure legend. Figure 2 See image above for ¦gure legend. Figure 3 See image above for ¦gure legend. Figure 4 See image above for ¦gure legend. Figure 5 See image above for ¦gure legend. See image above for ¦gure legend. Figure 6 See image above for ¦gure legend. See image above for ¦gure legend. Figure 7 See image above for ¦gure legend. Figure 8 See image above for ¦gure legend. Figure 10 See image above for ¦gure legend. Figure 9 See image above for ¦gure legend. supplemental¦gures.pdf Supplementary Files This is a list of supplementary ¦les associated with this preprint. Click to download. supplemental¦gures.pdf
|
https://openalex.org/W2911105534
|
http://qspace.qu.edu.qa/bitstream/10576/11255/1/nanomaterials-09-00122.pdf
|
English
| null |
Impaired Liver Size and Compromised Neurobehavioral Activity are Elicited by Chitosan Nanoparticles in the Zebrafish Embryo Model
|
Nanomaterials
| 2,019
|
cc-by
| 10,097
|
Received: 21 December 2018; Accepted: 15 January 2019; Published: 19 January 2019 Abstract: The use of chitosan nanoparticles (ChNPs) in various biological and environmental
applications is attracting great interest. However, potential side effects related to ChNP toxicity
remain the major limitation hampering their wide application. For the first time, we investigate the
potential organ-specific (cardiac, hepatic, and neuromuscular) toxicity of ChNPs (size 100–150 nm)
using the zebrafish embryo model. Our data highlight the absence of both acute and teratogenic
toxic effects of ChNPs (~100% survival rate) even at the higher concentration employed (200 mg/L). Although no single sign of cardiotoxicity was observed upon exposure to 200 mg/L of ChNPs, as
judged by heartbeat rate, the corrected QT interval (QTc, which measures the time between the start
of the Q wave and the end of the T wave in the heart's electrical cycle), maximum cardiac arrest,
and ejection fraction assays, the same dosage elicited the impairment of both liver size (decreased
liver size, but without steatosis and lipid yolk retention) and neurobehavioral activity (increased
movement under different light conditions). Although the observed toxic effect failed to affect embryo
survival, whether a prolonged ChNP treatment may induce other potentially harmful effects remains
to be elucidated. By reporting new insights on their organ-specific toxicity, our results add novel and
useful information into the available data concerning the in vivo effect of ChNPs. Keywords: zebrafish; chitosan; nanoparticles; organs specific toxicity nanomaterials nanomaterials nanomaterials nanomaterials Impaired Liver Size and Compromised
Neurobehavioral Activity are Elicited by Chitosan
Nanoparticles in the Zebrafish Embryo Model Haissam Abou-Saleh 1
, Nadin Younes 2,3, Kashif Rasool 4, Manaf H. Younis 5,
Rafael M. Prieto 6
, Hadi M. Yassine 3
, Khaled A. Mahmoud 4
, Gianfranco Pintus 2,3,7,†,*
and Gheyath K. Nasrallah 2,3,†,* Haissam Abou-Saleh 1
, Nadin Younes 2,3, Kashif Rasool 4, Manaf H. Younis 5,
Rafael M. Prieto 6
, Hadi M. Yassine 3
, Khaled A. Mahmoud 4
, Gianfranco Pintus 2,3,7,†,*
and Gheyath K. Nasrallah 2,3,†,* Doha 2713, Qatar; hasaleh@qu.edu.qa p
g
Q
y
Q
ny1204022@student.qu.edu.qa
3
Biomedical Research Center, Qatar University, Doha 2713, Qatar; hyassine@qu.edu.qa
4
Qatar Environment and Energy Research Institute (QEERI), Hamad Bin Khalifa University (HBKU),
Qatar Foundation, Doha 2713, Qatar; krasool@hbku.edu.qa (K.R.); kmahmoud@hbku.edu.qa (K.A.M.)
5
Department of Public Health, Qatar University, Doha 2713, Qatar; my1512859@student.qu.edu.qa
6
ZeClinics SL, PRBB (Barcelona Biomedical Research Park), 08003 Barcelona, Spain; zeminyana@gmail.com
7
Department of Biomedical Sciences, University of Sassari, 07100 Sassari, Italy
*
Correspondence: gpintus@qu.edu.qa (G.P.); gheyath.nasrallah@qu.edu.qa (G.K.N.)
†
E
l
ib
i *
Correspondence: gpintus@qu.edu.qa (G.P.); gheyath.nasrallah@qu.edu.qa (G.K.N.) www.mdpi.com/journal/nanomaterials 1. Introduction Chitosan nanoparticles (ChNPs) have emerged as a pivotal instrument in many applicative
fields, including chemistry, water treatment, aquatic herbicide, bioengineering, disease detection, and
drug delivery [1–5]. Indeed, among their several intrinsic properties, ChNPs have shown to possess
phytoremediation, antibacterial, and anti-pesticide effects [3–8], which make them a prominent tool in
applied environmental science. In addition, due to their many advantages, including simple preparation, Nanomaterials 2019, 9, 122; doi:10.3390/nano9010122 www.mdpi.com/journal/nanomaterials 2 of 13 Nanomaterials 2019, 9, 122 physicochemical stability, and improved bioavailability, these polymer-based nanomaterials have
attracted ample attention in the biomedical theragnostic field [9,10]. In fact, because of their nanosized
and chemical versatility, chemically engineered tissue-targeted ChNPs can readily cross cell membranes,
conductive tissues (such as blood and lymphatic vessels), and the brain barrier [10]. In addition, ChNPs
have been shown to exert antitumor activity, either by boosting the body’s immune system or by
interfering with the tumor growth [11,12]. The wide range of ChNP applications urges the need to accurately investigate potential ChNP
toxicity on both aquatic life and high vertebrate animals, including humans. Although the safe use of
ChNPs is still an area of debate, many studies consider chitosan a safe polymer that can be employed
in several in vivo applications, such as tissue engineering, drug delivery, enzyme immobilization,
and wound dressing [2,13]. Indeed, low molecular weight (MW) chitosan can be directly cleared by the
kidney in vivo, while high MW chitosan is first enzymatically degraded and then eventually cleared
from the body as the low MW chitosan [12]. Nevertheless, some extent of toxicity has been reported
in other studies, both in vitro and in vivo [14], thus emphasizing the necessity of further studies to
unravel this specific aspect of Chitosan. Zebrafish and their embryos are extensively used as tools to assess the toxicity of a compound,
not only in the environmental context (aquatic in particular) but also to predict the toxic effects of
compounds in humans [15,16]. This animal model has been widely used in toxicological research to
assess the toxicity of novel drugs in vivo [12]. Because of the great similarities in cellular structure,
signaling processes, anatomy, and physiology between zebrafish and other high-order vertebrates,
particularly in the early stages of development [16], this model can be useful in providing toxicological
information translatable to humans [17]. 1. Introduction Although some studies on ChNP toxicity in Zebrafish
have been performed [18–20], to our knowledge, there are no reported data concerning potential
organ-specific and long-term toxicity of ChNPs. Therefore, using zebrafish as a model system,
this study aims to design and implement a number of organ-specific (heart, muscles, nervous system,
and liver) toxicity assays to comprehensively evaluate any potential in vivo adverse effects of ChNPs. In this regard, we decided to investigate a range of concentrations (25–200 mg/L) consistent with the
acute toxicity rating scale provided by the U.S. Fish and Wildlife Service (USFWS) [21]. 2.1. Materials All materials were purchased from Sigma Aldrich (St. Louis, MO, United States) unless indicated
otherwise. Low MW (LMW) chitosan (50–190 KDa, based on viscosity) powder with 75%–85% degree
of deacetylation (Cat No., 448869-250G) was used. Dimethyl sulfoxide (DMSO) was used as negative
control (NC) in all assays; dimethylaminobenzaldehyde (DEAB), a competitive inhibitor of aldehyde
dehydrogenases known to cause mortality and teratogenic phenotype in zebrafish embryos, was used
as positive control (PC) in acute toxicity assays [22]. Haloperidol, an antipsychotic drug, was used
as a PC in cardiotoxicity assays [22–24]. In addition, 1-Methyl-4-phenyl-1,2,3,6-tetrahydropyridine
hydrochloride (MPTP), a neurotoxic drug that causes permanent symptoms of Parkinson's disease,
was used as a PC in neurotoxicity assays [25]. Paracetamol (APAP) was used as a PC in hepatotoxicity
assays [26,27]. A 60X stock solution of embryo media (E3 media) was prepared from 8.765 g NaCl,
380 mg KCl, 1.19 g MgSO4, and 1.765 g CaCl2 dissolved in 0.5 liters of MilliQ water. 2.2. Chitosan Nanoparticle Preparation ChNPs were synthesized by ionic crosslinking reaction at different initial chitosan concentrations
of 0.05%, 0.1%, 0.25%, and 0.5 %. LMW chitosan was dissolved in an aqueous solution of 1%
(w/w) acetic acid to form a 1.0 mg/mL chitosan solution and filtered (pore size 0.45 µm, Millipore,
Burlington, MA, United States) to remove insoluble particles. Tripolyphosphate (TPP 0.5 mg/mL)
was dissolved in ultrapure water and filtered. TPP solution was added dropwise to the chitosan 3 of 13 Nanomaterials 2019, 9, 122 solutions, having different initial chitosan concentrations and TPP ratios. The reaction was carried
out for 10 min, and the resulting suspension was washed three times with deionized (DI) water to
remove any remaining TPP. Finally, the obtained ChNPs pellet was resuspended into DI water for the
needed applications. 2.5. Acute Toxicity Assays The toxicology of ChNPs was investigated by acute toxicity assays adapted by the Organization
of Economic Co-operation and Development (OECD) guidelines for testing chemical toxicity (N◦203
and 236). Healthy dechorionated embryos were transferred to a six-well plate containing the following
treatments prepared in fresh E3 media: (i) four different concentrations (25, 50, 100, 200 mg/L) of
ChNPs, (ii) the PC DEAB (0.1, 1, 10, 100, and 1.0 µM), and (iii) the NC 0.1% DMSO. Embryos were
incubated at 28 ◦C from 24 hpf up to 96 hpf. Cumulative survival was recorded at three time point
intervals (48, 72, and 96-hpf). In addition, all embryos were observed every day for teratogenic
abnormalities, such as pigmentation, body deformities, heart edema, heartbeats, yolk edema, scoliosis,
and movement. The median lethal dose (LC50) was calculated by fitting a sigmoidal curve to mortality
data with a 95% confidence interval. A total of 100 embryos were used for each tested dose condition
of the ChNPs, and 20 embryos for each DEAB and 0.1% DMSO concentrations. 2.4. Zebrafish Embryo Culture Wild-type and transgenic green fluorescent proteins (GFP) AB zebrafish strains (Danio rerio) were
used. The fish were left in a 14 h light/10 h dark cycle with a water temperature of 28 ◦C. Fertilized
eggs were rinsed and were collected in Petri dishes containing E3 egg medium. After 3–4 h, unfertilized
and unhealthy embryos were discarded. At 24 h post-fertilization (hpf), the healthy fertilized embryos
were dechorionated by adding 1.0 mg/mL of pronase. The embryos were incubated for 10 min and
swirled, so the chorion becomes soft. Then the embryos were washed three times with fresh E3 media
and moved to six-well plates containing a fresh E3 medium. Finally, the E3 medium was replaced
by fresh E3 medium containing the different concentrations of drugs to be tested in different assays. In this study, we followed the national and international guidelines for the use of zebrafish embryos to
carry out all the experiments in accordance with animal protocol guidelines required by the ZeClinics,
Spain, and Qatar University and Policy on Zebrafish Research established by Department of Research
in the Ministry of Public Health, Qatar. 2.3. Chitosan Nanoparticle Characterization ChNPs were characterized by Zetasizer, X-ray diffraction (XRD), Fourier-transform infrared
spectroscopy (FTIR), scanning electron microscopy (SEM), and UV-Vis spectroscopy. A Zeta phase
analysis light-scattering (ZetaPALS) analyzer (Malvern Instruments, Malvern, United Kingdom)
was used to measure the electrophoretic mobilities (EPMs) of the ChNPs. The zeta potentials were
calculated from the average EPMs. The ChNPs sizes were measured by dynamic light scattering (DLS)
using a ZetaPALS analyzer. To determine how the ionic strength impacted the nanoparticles size,
we prepared nanoparticle suspensions containing different salt concentrations and immediately put
them into disposable cuvettes. XRD was recorded using a Bruker D8 Advance (Bruker AXS, Billerica,
MA, United States). SEM analyses were performed using an FEI Quanta 650 FEG scanning electron
microscope after samples gold sputter coatings. 2.8. Hepatotoxicity Evaluation The hepatotoxicity assays were performed using the Tg[cmLc:GRP] transgenic AB strain of
zebrafish. This strain expresses the RFP in the hepatocytes, thus allowing a good quality of liver
imaging. To evaluate potential ChNP hepatotoxicity, the following parameters were assessed in the
zebrafish liver: liver size, to measure necrosis and hepatomegaly, and yolk retention and steatosis,
to measure liver lipid metabolism. At 96 hpf, embryos were incubated for an additional 32 h at 28
◦C with the following treatment: (i) the NC DMSO; (ii) the PC 2% EtOH (for steatosis assessment),
2640 µM APAP, or 2% EtOH (yolk retention and necrosis assessment); and (iii) 200 mg/L ChNPs. 2.8.1. Liver Area Analysis For the liver size measurement, embryos were fixed in 4% paraformaldehyde for 4 h at RT and
then washed three times with PBS. Fixed embryos’ fluorescent livers were imaged with a fluorescence
stereomicroscope (Olympus MVX10) using a digital camera (Olympus DP71). RFP filtered images
were analyzed using cellˆD and the FIJI software. 2.6. Neurobehavioral Toxicity Evaluation In order to determine the effect of the ChNPs on the embryos’ nervous systems, we assessed
their swimming activity (distance traveled) in response to environmental stimuli, such as dark and
light exposure [28]. Briefly, 24-hpf healthy embryos were allowed to develop normally until 96 hpf at
28 ◦C, which is the time when the proper swimming activity is reached [29]. At 96 hpf, embryos were
transferred to a 96-well plate (one embryo per well) containing the following treatments: (i) the NC 4 of 13 Nanomaterials 2019, 9, 122 DMSO, (ii) the PC 100 µM MPTP, and (iii) 200 mg/L ChNPs. Next, the embryos were incubated with
each treatment until 120 hpf, and the locomotion activity was recorded up to 144 hpf. The potential
neurobehavioral toxicity of ChNPs was determined by locomotion assessment using the EthoVision XT
11.5 tracking software and the DanioVision device (Noldus, Wageningen, Netherlands), which allows
the tracking of individual zebrafish embryos during alternating dark and light cycles. The 96-well
plate was placed into the DanioVision chamber, and the embryos were left for 20 min with light for
acclimation. Next, embryo movements were recorded during a 10-min alternate light/dark cycle for a
total of 50 min. Neurotoxicity was evaluated by comparing the above parameters between the treated
and control groups. Due to circadian rhythms, all locomotion assays were performed from 13:00 pm
onwards to ensure the steady activity of embryos [30]. 2.7. Cardiotoxicity Assays The cardiotoxicity assays were performed using the transgenic cardiac myosin light chain
2 gene-GFP zebrafish (Tg[cmLc:GFP]) strain. This strain expresses the green fluorescence proteins
(GFP) in the cardiac myocytes, thus allowing a good quality of cardiac imaging. Briefly, 24-hpf healthy
embryos were allowed to develop normally until 96 hpf, which is the time where the heart is usually
fully developed [31]. At 96 hpf, embryos were incubated for 4 h at 28 ◦C with the following treatments:
(i) the NC DMSO, (ii) the PC 10 µM haloperidol, and (iii) 200 mg/L ChNPs. Treated embryos were
then anesthetized by immersion in 0.7 µM Tricaine methanesulfonate/E3 solution. For imaging, every
embryo was positioned under the microscope using an agarose-based mold, and the fluorescent heart
was recorded for 60 s [15]. Videos were acquired by high-speed cameras and analyzed with the
ZeCardio®software for the presence of any heart dysfunctions. In particular, the following cardiac
parameters were assessed: heart rate; QTc corrected interval (Framingham formula: QTc = QT + 0.154
(1 −RR)) [32]; cardiac arrest and ejection fraction (Ef%) = ((DD −SD)/DD) × 100. DD); ventricle
diastolic diameter (max dilatation); SD; and ventricle systolic diameter (max contraction)]. 2.9. Statistical Analysis Results were expressed as average ± SEM (standard error of the mean). Statistical evaluation of
differences between experimental group means was performed using one-way analysis of variance
(ANOVA), followed by the Dunnet test. The Chi-square test was used for the hepatotoxicity assays
(steatosis and yolk retention) to compare the significance between the percentages. Outliers were
eliminated by using the Graph Pad software. Significance (*): p < 0.05; (**): p < 0.01. 2.8.2. Detection of Steatosis and Yolk Retention Steatosis and yolk retention were evaluated by Oil Red O staining (Sigma-Aldrich, St. Louis,
MO, United States). Zebrafish embryos were stained as previously described [33]. Briefly, the skin
pigment of the fixed embryos was removed by incubation with 0.1 mL of 5% sodium hypochlorite
bleaching solution for 20 min, followed by five washes with PBS at RT. Then, the bleached embryos
were submerged in 85% propylene glycol (PG) (Sigma-Aldrich, United States) for 10 min, and then in
100% PG for another 10 min, before staining them with Oil Red 0.5% in 100% PG (overnight, at RT and
with a gentle rocking). Oil Red O stained embryos were then washed in 100% PG for 30 min, 85% PG
in PBS for 50 min, and finally 85% PG in PBS for 40 min. Next, the embryos were washed with 1x 5 of 13 Nanomaterials 2019, 9, 122 PBS before adding 80% glycerol. Finally, bright field images were taken to detect both steatosis and
yolk retention. For the analysis of steatosis, embryos were considered to be positive when three or
more round lipid droplets were visible within the hepatic parenchyma [22]. Then, the percentage of
steatosis was calculated by dividing the number of embryos showing steatosis with the total number
of embryos observed. For yolk retention, embryos showing a strong red signal in the yolk area were
considered positive. The percentage was calculated by dividing the number of positive embryos with
the total number of embryos. 3.1. Chitosan Nanoparticle Characterization Nanomaterials 2018, 8, x FOR PEER REVIEW
7 of 14 Figure 2. (A) SEM image, (B) FTIR spectra, and (C) X-ray diffraction patterns of ChNPs. (D) Impact
of ionic strength on the stability of ChNPs, measured as hydrodynamic diameter by dynamic light
scattering (DLS). A
B
C
D
20.5o
9.8o
Figure 2. (A) SEM image, (B) FTIR spectra, and (C) X-ray diffraction patterns of ChNPs. (D) Impact
of ionic strength on the stability of ChNPs, measured as hydrodynamic diameter by dynamic light
scattering (DLS). A B A B A C C C
20.5o
9.8o C D D Figure 2. (A) SEM image, (B) FTIR spectra, and (C) X-ray diffraction patterns of ChNPs. (D) Impact
of ionic strength on the stability of ChNPs, measured as hydrodynamic diameter by dynamic light
scattering (DLS). Figure 2. (A) SEM image, (B) FTIR spectra, and (C) X-ray diffraction patterns of ChNPs. (D) Impact
of ionic strength on the stability of ChNPs, measured as hydrodynamic diameter by dynamic light
scattering (DLS). les Do Not Affect Zebrafish Embryos Survival
Table 1. Physicochemical characteristics of ChNPs. hitosan Nanoparticles Do Not Affect Zebrafish Embryos Survival
Table 1. Physicochemical characteristics of ChNPs. Chitosan Nanoparticles Do Not Affect Zebrafish Embryos Survival
Embryos are more sensitive to external compounds and chemicals than larval or adult zebrafi
Therefore, we chose the embryonic period from 24 to 96 hpf as the administration time to stud
ntial ChNP toxicity. The percentage of cumulative survival was measured at 96 hpf, which is t
mmended observation time [22]. According to Figure 3B, the no-observed-effect concentratio
CE) (i.e., <20% mortality) for DEAB was 0.1 µM, as this concentration of DEAB showed
ulative mortality of only 10% (2 out of 20 embryos were dead), and the rest of the embryos (90%
not show any teratogenic effects (e.g., morphological or physiological abnormalities). While t
Table 1. Physicochemical characteristics of ChNPs. Properties
Technique
Unit
Value
Primary size
TEM
nm
100–150
Particle size in DI water
DLS
nm
210 ± 6
Phase and structure
XRD
-
Semi-crystalline
Shape/morphology
TEM
-
Spherical, hexagonal
Surface area
BET
m2/g
11.7
pHPZC
DLS
-
28 ± 2.6 low-observed-effect concentration (LOEC) (i.e., =>20% mortality) for DEAB was 1.0 µM, showing
cumulative mortality of 60% (12 of 20 embryos were dead). 3.1. Chitosan Nanoparticle Characterization Figure 1 depicts the hydrodynamic diameter and zeta potential of the synthesized ChNPs at
different initial chitosan concentrations measured by dynamic light scattering. As the chitosan
concentration was increased from 0.05% to 0.5%, the ChNPs hydrodynamic diameter increased
from 120 nm to 646 nm. Below a certain concentration of chitosan (2.0 mg/mL as reported),
the intermolecular hydrogen bonding attraction and the intermolecular electrostatic repulsion are
in equilibrium. Therefore, in this concentration range, as chitosan concentration increases, chitosan
molecules approach each other with a limit, leading to a limited increase in intermolecular cross-linking;
thus, larger but still nanoscale particles are formed. Nanomaterials 2018, 8, x FOR PEER REVIEW
6 of 14 Figure 1. Hydrodynamic diameter and zeta potential of synthesized chitosan nanoparticles (ChNPs)
at different initial chitosan concentrations. Figure 1. Hydrodynamic diameter and zeta potential of synthesized chitosan nanoparticles (ChNPs) at
different initial chitosan concentrations. Figure 1. Hydrodynamic diameter and zeta potential of synthesized chitosan nanoparticles (ChNPs)
at different initial chitosan concentrations. Figure 1. Hydrodynamic diameter and zeta potential of synthesized chitosan nanoparticles (ChNPs) at
different initial chitosan concentrations. The primary size of selected ChNPs was in the range of 100–150 nm, as measured by SEM (Figure
2A). Figure 2B shows the FTIR spectra of the ChNPs. A characteristic band at 3435 cm−1 was attributed
to the –NH2 and –OH groups stretching vibrations. The bands at 1640 cm−1, were attributed to the –
NH2 bending vibration, whereas 1396 and 1078 cm−1 were assigned to CH3 symmetrical deformation,
and the C–O stretching vibrations (C–O–C) of Ch, respectively. Figure 2C depicts the XRD spectra of
The primary size of selected ChNPs was in the range of 100–150 nm, as measured by SEM
(Figure 2A). Figure 2B shows the FTIR spectra of the ChNPs. A characteristic band at 3435 cm−1 was
attributed to the –NH2 and –OH groups stretching vibrations. The bands at 1640 cm−1, were attributed
to the –NH2 bending vibration, whereas 1396 and 1078 cm−1 were assigned to CH3 symmetrical
deformation, and the C–O stretching vibrations (C–O–C) of Ch, respectively. Figure 2C depicts 6 of 13 Nanomaterials 2019, 9, 122 the XRD spectra of ChNPs. The broad diffraction peaks at 2θ = ~9.80◦and ~20.5◦in the XRD
pattern of ChNPs showed the semi-crystalline nature of the synthesized nanoparticles. Table 1 gives
physicochemical characteristics of ChNPs, with the smallest size selected for further characterization
and toxicity studies. 3.2. Chitosan Nanoparticles Do Not Affect Zebrafish Embryos Survival Embryos are more sensitive to external compounds and chemicals than larval or adult
zebrafish [35]. Therefore, we chose the embryonic period from 24 to 96 hpf as the administration time
to study potential ChNP toxicity. The percentage of cumulative survival was measured at 96 hpf,
which is the recommended observation time [22]. According to Figure 3B, the no-observed-effect
concentration (NOCE) (i.e., <20% mortality) for DEAB was 0.1 µM, as this concentration of DEAB
showed a cumulative mortality of only 10% (2 out of 20 embryos were dead), and the rest of the embryos
(90%) did not show any teratogenic effects (e.g., morphological or physiological abnormalities). While the low-observed-effect concentration (LOEC) (i.e., =>20% mortality) for DEAB was 1.0 µM,
showing cumulative mortality of 60% (12 of 20 embryos were dead). At 10 µM of DEAB, the embryos
showed severe teratogenic effects, such as deformities in the heart and yolk sac (Figure 3A), and were
unable to survive up to 96 hpf. According to the sigmoidal curve (Figure 3C), the LC50 value for
DEAB was calculated to be 0.665 µM (r2 ≈0.999). Concerning the ChNPs treatments, no mortality,
morphological, or physiological abnormalities were observed (Figure 3D) at any dose and time
employed, therefore NOEC, LOEC, and LC50 for ChNPs could not be calculated. These results suggest
that the hypothetical LC50 for ChNPs would be much higher than 200 mg/L (Figure 3C). The current
data demonstrated that ChNPs, at the least at concentrations as high as 200 mg/L, failed to affect
zebrafish gross embryonic development. Nanomaterials 2018, 8, x FOR PEER REVIEW
8 of 14 0.0
99.7
100.0
100.0
100.0
100.0
0.0
10.0
20.0
30.0
40.0
50.0
60.0
70.0
80.0
90.0
100.0
DEAB 10µM
DMSO 0.1%
ChNPs 200
µg/mL
ChNPs 100
µg/mL
ChNPs 50
µg/mL
ChNPs 25
µg/mL
% Survival
Treatment concentration (mg/L)
ChNPs 200
mg/L
ChNPs 100
mg/L
ChNPs 50
mg/L
ChNPs 25
mg/L
ChNPs 200
mg/L Figure 3. (A) Acute toxicity representative pictures obtained with a stereomicroscope at magnification
X = 0.63. Note the deformed embryos in dimethylaminobenzaldehyde (DEAB): short size, scoliosis
heart edema (closed arrow), and yolk edema (open arrow). (B) Cumulative survival/mortality. (C)
Mortality response curve of different concentrations of DEAB and ChNPs. (D) Teratogenic
phenotypes analysis. DEAB: Dimethylaminobenzaldehyde, positive control for toxicity and
teratogenicity. 3.2. Chitosan Nanoparticles Do Not Affect Zebrafish Embryos Survival A)
DMSO 0.1%
ChNPs 200 mg/L
DEAB 10 μM
C)
B)
D)
-2
-1
0
1
2
3
4
50
100
150
Log [Conc (µg/mL)]
% Mortality
Log(LC50)
96h concentration- mortality responce: ChNPs
DEAB
ChNPs
0.0
99.7
100.0
100.0
100.0
100.0
0.0
10.0
20.0
30.0
40.0
50.0
60.0
70.0
80.0
90.0
100.0
DEAB 10µM
DMSO 0.1%
ChNPs 200
µg/mL
ChNPs 100
µg/mL
ChNPs 50
µg/mL
ChNPs 25
µg/mL
% Survival
Treatment concentration (mg/L)
ChNPs 200
mg/L
ChNPs 100
mg/L
ChNPs 50
mg/L
ChNPs 25
mg/L
ChNPs 200
mg/L
Body deformity
Escoliosis
Pigmentation
Yolk size
Heart edema
Heart beat
Movement
0.000
100.000
DEAB 10 µM
DMSO 0.1%
ChNPs 200 µg/mL
ChNPs 100 µg/mL
ChNPs 50 µg/mL
ChNPs 25 µg/mL
ChNPs 200
ChNPs 100
ChNPs 50
ChNPs 10
DMSO 0.1%
DEAB 10 µM
Log [Conc (mg/L)]
Figure 3. (A) Acute toxicity representative pictures obtained with a stereomicroscope at magnification X
= 0.63. Note the deformed embryos in dimethylaminobenzaldehyde (DEAB): short size, scoliosis heart
edema (closed arrow), and yolk edema (open arrow). (B) Cumulative survival/mortality. (C) Mortality
response curve of different concentrations of DEAB and ChNPs. (D) Teratogenic phenotypes analysis. DEAB: Dimethylaminobenzaldehyde, positive control for toxicity and teratogenicity. A)
DMSO 0.1%
ChNPs 200 mg/L
DEAB 10 μM
B) B) B) A) 96h concentration- mortality responce: ChNPs 96h concentration- mortality responce: ChNPs C)
-2
-1
0
1
2
3
4
50
100
150
Log [Conc (µg/mL)]
% Mortality
Log(LC50)
96h concentration- mortality responce: ChNPs
DEAB
ChNPs
Log [Conc (mg/L)] D)
(
g
)
Body deformity
Escoliosis
Pigmentation
Yolk size
Heart edema
Heart beat
Movement
0.000
100.000
DEAB 10 µM
DMSO 0.1%
ChNPs 200 µg/mL
ChNPs 100 µg/mL
ChNPs 50 µg/mL
ChNPs 25 µg/mL
ChNPs 200
ChNPs 100
ChNPs 50
ChNPs 10
DMSO 0.1%
DEAB 10 µM C) D) D) Figure 3. (A) Acute toxicity representative pictures obtained with a stereomicroscope at magnification
X = 0.63. Note the deformed embryos in dimethylaminobenzaldehyde (DEAB): short size, scoliosis
heart edema (closed arrow), and yolk edema (open arrow). (B) Cumulative survival/mortality. (C)
Mortality response curve of different concentrations of DEAB and ChNPs. (D) Teratogenic
phenotypes analysis. DEAB: Dimethylaminobenzaldehyde, positive control for toxicity and
teratogenicity. Figure 3. (A) Acute toxicity representative pictures obtained with a stereomicroscope at magnification X
= 0.63. Note the deformed embryos in dimethylaminobenzaldehyde (DEAB): short size, scoliosis heart
edema (closed arrow), and yolk edema (open arrow). (B) Cumulative survival/mortality. 3.1. Chitosan Nanoparticle Characterization At 10 µM of DEAB, the embryos showed
severe teratogenic effects, such as deformities in the heart and yolk sac (Figure 3A), and were unable
to survive up to 96 hpf. According to the sigmoidal curve (Figure 3C), the LC50 value for DEAB was
calculated to be 0.665 µM (r2 ≈ 0.999). Concerning the ChNPs treatments, no mortality, morphological,
or physiological abnormalities were observed (Figure 3D) at any dose and time employed, therefore
NOEC, LOEC, and LC50 for ChNPs could not be calculated. These results suggest that the
hypothetical LC50 for ChNPs would be much higher than 200 mg/L (Figure 3C) The current data
Zeta potential is an important tool for understanding the state of the nanoparticle surface and
predicting the long-term stability of the nanoparticle. However, there was no significant effect
of chitosan concentration found on the zeta potential, which was about 28 ± 2 for all chitosan
compositions. The magnitude of the zeta potential is predictive of colloidal stability [34]. Nanoparticles
with zeta potential values greater than +25 mV or less than −25 mV typically have high degrees of
stability. Therefore, our results revealed the ChNPs as stable in colloidal suspension with a zeta
potential of more than +25.0. 7 of 13 Nanomaterials 2019, 9, 122 The physicochemical properties and colloidal stability of the nanoparticles probably largely
determine their mobility, bioavailability, and toxicity to the living organisms. The effect of the salts’
concentration on the stability of the ChNPs was examined by varying the initial sodium chloride
concentrations and maintaining the solution pH and ChNPs concentration at 7 ± 02 and 20 mg/L,
respectively. It was found that synthesized nanomaterials were stable at a higher salt concentration
and there was no impact on the nanomaterials’ hydrodynamic diameter (Figure 2D). 3.2. Chitosan Nanoparticles Do Not Affect Zebrafish Embryos Survival (C) Mortality
response curve of different concentrations of DEAB and ChNPs. (D) Teratogenic phenotypes analysis. DEAB: Dimethylaminobenzaldehyde, positive control for toxicity and teratogenicity. 8 of 13 Nanomaterials 2019, 9, 122 It is believed that ChNPs’ toxicity is related to particles size and concentration, chemical
composition, and ambient conditions. Indeed, nanoparticles’ toxicity appears to increase with the
decrease in their size [36]. To the best of our knowledge, only three articles studying the effect of ChNPs
on zebrafish embryos have been published [18–20], and none of them performed a comprehensive
investigation of potential organ-specific toxic effects. Indeed, Hu et al. tested the potential ChNP
toxicity in relation to the particle sizes (200 and 340 nm), showing that the 200 nm ChNPs are more
toxic toward zebrafish embryos compared to the 340 nm. The 200 nm ChNPs were able to cause
100% mortality and severe teratogenic phenotypes at a concentration as low as 40 mg/L, while the
340 nm ChNPs, although at less of an extent as compared to the 200 nm, were able anyway to promote
significant mortality and teratogenic phenotypes as compared to the untreated embryos. Our current
results are in disagreement with those obtained by Hu et al. In fact, although we employed smaller
ChNPs sizes (100−150 nm) than those used by Hu et al., we were unable to record any toxic effects
or teratogenic phenotypes at concentrations as high as 200 mg/L. On the other hand, our current
results are consistent with those reported by Wang et al., indicating that 200 mg/L of ChNPs failed to
cause significant embryos mortality (<10%), even when a smaller ChNP size (84.86 nm) was employed. Under our current experimental conditions, the LOEC for ChNPs could not be determined. Therefore,
we used the highest tested concentration (200 mg/L) in the subsequent toxicity assays performed. 3.3. Chitosan Nanoparticles Induce Neurobehavioral Impairment In Zebrafish Embryos Zebrafish are widely used as a model for studying neurological diseases and behavioral
analysis [37]. To gain insight on ChNPs’ potential neurobehavioral toxicity, we investigated their effect
of on the embryos’ nervous system development by quantifying the swimming activity of the embryo
exposed to two illumination conditions. Consistently with a previous article [20], our data indicate that
ChNPs-treated embryos display an abnormal increase in the locomotion behavior (hyperactivity)
compared to the untreated ones (Figure 4B,C). However, the data reported by Yuan et al. 3.2. Chitosan Nanoparticles Do Not Affect Zebrafish Embryos Survival [20]
along with our current findings both show increased rather than a decreased behavior activity,
suggesting that ChNPs can adversely affect the embryos’ nervous system, but is unlikely to affect their
muscle development. Nanomaterials 2018, 8, x FOR PEER REVIEW
9 of 14
the embryos’ nervous system development by quantifying the swimming activity of the embryo
exposed to two illumination conditions. Consistently with a previous article [20], our data indicate
that ChNPs-treated embryos display an abnormal increase in the locomotion behavior (hyperactivity)
compared to the untreated ones (Figure 4B,C). However, the data reported by Yuan et al. [20] along
with our current findings both show increased rather than a decreased behavior activity, suggesting
that ChNPs can adversely affect the embryos’ nervous system, but is unlikely to affect their muscle
development. Figure 4. Locomotion assay. (A) Scheme of the experimental procedure. (B) Average of distance
travelled (mm) per min measured by DanioVision. A group of 14 embryos was used for each
treatment. (C) Average of cumulative distance travelled (mm) by the treated group after exposure to
50 min dark/light cycle. MPTP: 1-Methyl-4-phenyl-1,2,3,6-tetrahydropyridine hydrochloride, positive
control for neurotoxicity. 3 4 Chitosan Nanoparticles Do Not Affect Cardiac Functions
0
500
1000
1500
2000
2500
3000
3500
DMSO 1%
MPTP 100uM
ChNP
Total distance Travelled (mm)
Treatments concentrations
ChNPs 200 mg/L
MPTP 100 µM
DMSO 0.1%
B)
C)
96 hpf
Drug treatment
120 hpf
144 hpf
Locomotion
evaluation
A)
Light Dark Light Dark light
**
0
20
40
60
80
100
120
140
160
1
3
5
7
9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49
Distance Travelled (mm)
DMSO 1%
MPTP 100uM
ChNP
DMSO 0.1%
MPTP 100 µM
*
ChNP 200µg/mL
MPTP 100 µM
Time (min)
ChNPs 200 mg/L
Figure 4. Locomotion assay. (A) Scheme of the experimental procedure. (B) Average of distance
travelled (mm) per min measured by DanioVision. A group of 14 embryos was used for each treatment. (C) Average of cumulative distance travelled (mm) by the treated group after exposure to 50 min
dark/light cycle. MPTP: 1-Methyl-4-phenyl-1,2,3,6-tetrahydropyridine hydrochloride, positive control
for neurotoxicity. 3.4. Chitosan Nanoparticles Do Not Affect Cardiac Functions 3.4. Chitosan Nanoparticles Do Not Affect Cardiac Functions Among the advantages of zebrafish as an experimental model is the similarity of their
cardiovascular system to that of mammals [38]. At 120 hpf, zebrafish embryos were exposed
to 200 mg/L ChNPs and DMSO, and 10 µM haloperidol was used as positive control (PC) for
cardiotoxicity. Four parameters were assessed for the cardiotoxicity evaluation: (i) heart rate in
beat per min (BPM); (ii) the corrected QT interval (QTc), which measures the time between the start of
the Q wave and the end of the T wave in the heart's electrical cycle; (iii) cardiac arrest; and (iv) ejection
fraction. A compound such as the ChNPs, for instance, would be considered cardiotoxic if able to
induce a significant difference from the negative control (0.1% DMSO) in one of the above-mentioned
parameters. The positive control haloperidol was able to induce a significant decrease in the heart
rate as compared to negative control DMSO (Figure 5A). This result is consistent with previous data
showing that haloperidol is able to induce bradycardia in humans and zebrafish [24]. Similar to
the ability of haloperidol to produce QT prolongation in humans [39], our current data indicates a
significant prolongation of this cardiac parameter in ChNPs-treated zebrafish as compared to the
untreated ones (Figure 5B). However, ChNPs failed to affect the normal heartbeat frequency of both
the ventricle and the atrium, and were unable to elicit any impairment of the QTc when compared to
untreated embryos. As expected, haloperidol-treated embryos showed significantly longer cardiac
arrests as compared to DMSO (Figure 5C). Concerning ChNPs, the longest cardiac arrests at the
dose of 200 mg/L were 0.32 and 0.4 s for the ventricle and the atrium, respectively, which were both
not significant in comparison to DMSO. Regarding the ejection fraction (Figure 5D), no significant
difference was detected between DMSO and 200 mg/L ChNPs. In summary, ChNPs were unable to
induce any cardiotoxicity or developmental delay even at high concentrations or after long exposure
time (four days). Nanomaterials 2018, 8, x FOR PEER REVIEW
10 of 14
of 200 mg/L were 0.32 and 0.4 s for the ventricle and the atrium, respectively, which were both not
significant in comparison to DMSO. Regarding the ejection fraction (Figure 5D), no significant
difference was detected between DMSO and 200 mg/L ChNPs. 3.2. Chitosan Nanoparticles Do Not Affect Zebrafish Embryos Survival 96 hpf
Drug treatment
120 hpf
144 hpf
Locomotion
evaluation
A) A) B)
Light Dark Light Dark light
0
20
40
60
80
100
120
140
160
1
3
5
7
9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49
Distance Travelled (mm)
DMSO 1%
MPTP 100uM
ChNP
DMSO 0.1%
MPTP 100 µM
ChNP 200µg/mL
MPTP 100 µM
Time (min)
ChNPs 200 mg/L B) Total distance Travelled (mm)
C 0
500
1000
1500
2000
2500
3000
3500
DMSO 1%
MPTP 100uM
ChNP
Total distance Travelled (mm)
Treatments concentrations
ChNPs 200 mg/L
MPTP 100 µM
DMSO 0.1%
C)
**
* C) Figure 4. Locomotion assay. (A) Scheme of the experimental procedure. (B) Average of distance
travelled (mm) per min measured by DanioVision. A group of 14 embryos was used for each
treatment. (C) Average of cumulative distance travelled (mm) by the treated group after exposure to
50 min dark/light cycle. MPTP: 1-Methyl-4-phenyl-1,2,3,6-tetrahydropyridine hydrochloride, positive
control for neurotoxicity. Figure 4. Locomotion assay. (A) Scheme of the experimental procedure. (B) Average of distance
travelled (mm) per min measured by DanioVision. A group of 14 embryos was used for each treatment. (C) Average of cumulative distance travelled (mm) by the treated group after exposure to 50 min
dark/light cycle. MPTP: 1-Methyl-4-phenyl-1,2,3,6-tetrahydropyridine hydrochloride, positive control
for neurotoxicity. Nanomaterials 2019, 9, 122 9 of 13 3.4. Chitosan Nanoparticles Do Not Affect Cardiac Functions In summary, ChNPs were unable to
induce any cardiotoxicity or developmental delay even at high concentrations or after long exposure
time (four days). Figure 5. Cardiotoxicity assays. (A) Heartbeat rate. (B) QTc corrected interval. (C) Maximum cardiac
arrest. (D) Ejection fraction. Ten embryos were used for each treatment concentration. 0
0.4
0.8
1.2
1.6
2
DMSO 0.1%
Haloperidol 10µm ChNPs 200µg/mL
Longest cardiac arrest (sec)
Treatment
Ventricle
Atrium
0
40
80
120
160
200
DMSO 0.1%
Haloperidol 10µm
ChNPs 200µg/mL
Qtc (ms)
Treatment
A)
**
B)
C)
*
D)
**
0
40
80
120
160
200
DMSO 0.1%
Haloperidol
10µm
ChNPs
200µg/mL
Beats per minute
Treatment
Ventriuclar BPM
Atrial BPM
0
5
10
15
20
25
30
35
DMSO 0.1%
Haloperidol 10µm
ChNPs 200µg/mL
Ejection Fraction (%)
Treatment
DMSO 0.1%
Haloperidol 10 µM
ChNPs 200 mg/L
ChNPs 200 mg/L
DMSO 0.1%
Haloperidol 10 µM
DMSO 0.1%
Haloperidol 10 µM
ChNPs 200 mg/L
DMSO 0.1%
Haloperidol 10 µM
ChNPs 200 mg/L
Figure 5. Cardiotoxicity assays. (A) Heartbeat rate. (B) QTc corrected interval. (C) Maximum cardiac
arrest. (D) Ejection fraction. Ten embryos were used for each treatment concentration. 0
40
80
120
160
200
DMSO 0.1%
Haloperidol 10µm
ChNPs 200µg/mL
Qtc (ms)
Treatment
B)
**
M
ChNPs 200 mg/L
DMSO 0.1%
Haloperidol 10 µM A)
**
B
0
40
80
120
160
200
DMSO 0.1%
Haloperidol
10µm
ChNPs
200µg/mL
Beats per minute
Ventriuclar BPM
Atrial BPM
DMSO 0.1%
Haloperidol 10 µM
ChNPs 200 mg/L B) B) A) 0
0.4
0.8
1.2
1.6
2
DMSO 0.1%
Haloperidol 10µm ChNPs 200µg/mL
Longest cardiac arrest (sec)
Treatment
Ventric
Atrium
C)
*
Treatment
DMSO 0.1%
Haloperidol 10 µM
ChNPs 200 mg/L e
Treatment
D)
0
5
10
15
20
25
30
35
DMSO 0.1%
Haloperidol 10µm
ChNPs 200µg/mL
Ejection Fraction (%)
Treatment
DMSO 0.1%
Haloperidol 10 µM
ChNPs 200 mg/L D) C) D) Figure 5. Cardiotoxicity assays. (A) Heartbeat rate. (B) QTc corrected interval. (C) Maximum cardiac
arrest. (D) Ejection fraction. Ten embryos were used for each treatment concentration. Figure 5. Cardiotoxicity assays. (A) Heartbeat rate. (B) QTc corrected interval. (C) Maximum cardiac
arrest. (D) Ejection fraction. Ten embryos were used for each treatment concentration. 3 5 Chitosan Nanoparticles Induce Liver Size Impairment in Zebrafish Embryos
3.5. Chitosan Nanoparticles Induce Liver Size Impairment in Zebrafish Embryos 3.5. 3.4. Chitosan Nanoparticles Do Not Affect Cardiac Functions Chitosan Nanoparticles Induce Liver Size Impairment in Zebrafish Embryos
Due to the close similarity of mammals' molecular and cellular processes, Zebrafish embryos
have been recently used in liver research, including hepatotoxicology [26,27,40,41]. Since most
nanoparticles accumulate in the liver [42], hepatotoxicity tests are essential to assess potential ChNP
toxicity. After 120 hpf, the embryos’ liver is fully functional and able to metabolize external nutrients
Due to the close similarity of mammals' molecular and cellular processes, Zebrafish embryos
have been recently used in liver research, including hepatotoxicology [26,27,40,41]. Since most
nanoparticles accumulate in the liver [42], hepatotoxicity tests are essential to assess potential ChNP
toxicity. After 120 hpf, the embryos’ liver is fully functional and able to metabolize external nutrients 10 of 13 Nanomaterials 2019, 9, 122 from the environment [43]. The yolk sac of the zebrafish embryos consists of 70% lipids, which are
mainly metabolized in the liver [44]. Thus, excess lipid accumulation in the yolk can be used as
an indirect indication of impaired liver function. If the liver function is impaired, the yolk lipid
metabolism and absorption will be delayed, which in turn will result in lipid retention in the yolk [45]. Hepatotoxicity is primarily caused by metabolic process dysfunction, which requires a certain period
of time to occur. Therefore, we performed our experiments 32 h after embryos’ treatment with ChNPs
(at 128-hpf). Here, we performed three different hepatotoxicity assays. First, we assessed changes in
the liver size in response to different treatments (Figure 6A,B). Similar to the hepatotoxicity positive
control ethanol, 200 µg/mL ChNPs-treated embryos showed a significant decrease in both liver size
and shape as compared to DMSO. Further, embryo exposure to a different positive control (APAP) [46]
also displayed severe changes in liver size compared to DMSO, indicating severe liver necrosis. Second, we also assessed the percentage of embryos that developed abnormal ORO yolk retention in
response to the treatments. No significant difference in ORO yolk retention was found between the
DMSO and the ChNP-treated embryos (Figure 6D). However, consistent with a previous study [22],
APAP-treated embryos showed a significant increase in ORO yolk retention when compared to DMSO. Nanomaterials 2018, 8, x FOR PEER REVIEW
11 of 14
by a steatosis assay, which measures the amount of lipid retention in the liver in response to
treatments. The ChNP-treated embryos (Figure 6E) did not show significant signs of steatosis
compared to the untreated ones. 3.4. Chitosan Nanoparticles Do Not Affect Cardiac Functions Thirty embryos were used for each treatment in
every experiment. *, significantly different from DMSO (p < 0.05); **, significantly different from DMSO
(p < 0.01). Figure 6. (A) Representative images for the RFP liver area after each treatment. (B) Quantification and
measurement of the RFP liver area (mm2) for the controls and all the tested compounds. (C)
Representative images for steatosis and yolk retention. Arrow () in EtOH treatment indicate liver
steatosis (absorbed ORO stain by the liver), and arrows () in paracetamol (APAP) and ChNP
treatments indicate increased yolk retention (absorbed ORO stain by the yolk). (D) Percentages of
embryos showed yolk retention and (E) steatosis. Thirty embryos were used for each treatment in
every experiment. *, significantly different from DMSO (p < 0.05); **, significantly different from
DMSO (p < 0.01). Figure 6. (A) Representative images for the RFP liver area after each treatment. (B) Quantification
and measurement of the RFP liver area (mm2) for the controls and all the tested compounds. (C) Representative images for steatosis and yolk retention. Arrow (↗) in EtOH treatment indicate
liver steatosis (absorbed ORO stain by the liver), and arrows (↗) in paracetamol (APAP) and ChNP
treatments indicate increased yolk retention (absorbed ORO stain by the yolk). (D) Percentages of
embryos showed yolk retention and (E) steatosis. Thirty embryos were used for each treatment in
every experiment. *, significantly different from DMSO (p < 0.05); **, significantly different from DMSO
(p < 0.01). 4. Conclusions
Our study for the first time presents data that comprehensively investigates the organ-specific
toxicity of ChNPs in vivo in the zebrafish embryo model. The performed toxicity test indicated that
ChNPs were unable to induce teratogenic phenotypes and embryo death at the used concentrations. Moreover, ChNP-treated embryos displayed normal heart physiology, thus indicating the absence of
cardiotoxic effects. Nevertheless, ChNPs exerted remarkable neurotoxic effects and significantly
impaired the liver size, suggesting these nanoparticles are potentially harmful to both the brain and
Steatosis directly reflects an impairment of the normal synthesis and elimination of triglycerides
within the hepatocytes [47]. Thus, the effect of ChNPs on hepatocyte lipid metabolism was performed
by a steatosis assay, which measures the amount of lipid retention in the liver in response to treatments. The ChNP-treated embryos (Figure 6E) did not show significant signs of steatosis compared to the
untreated ones. 3.4. Chitosan Nanoparticles Do Not Affect Cardiac Functions However, 56% of the ethanol-treated embryos showed signs of
steatosis, a result consistent with data presented by Passeri et al. [48]. A)
C)
ChNPs
ChNPs Figure 6. (A) Representative images for the RFP liver area after each treatment. (B) Quantification and
measurement of the RFP liver area (mm2) for the controls and all the tested compounds. (C)
Representative images for steatosis and yolk retention. Arrow () in EtOH treatment indicate liver
steatosis (absorbed ORO stain by the liver), and arrows () in paracetamol (APAP) and ChNP
treatments indicate increased yolk retention (absorbed ORO stain by the yolk). (D) Percentages of
embryos showed yolk retention and (E) steatosis. Thirty embryos were used for each treatment in
every experiment. *, significantly different from DMSO (p < 0.05); **, significantly different from
DMSO (p < 0.01). 0
2
4
6
8
10
12
14
16
18
20
DMSO
0.1%
EtOH 2%
APAP
2640uM
ChNP
200µg/ml
Yolk retention (%)
DMSO
0.1%
0
10
20
30
40
50
60
DMSO
EtOH 2%
APAP
2640uM
ChNPs
200µg/mL
Steatosis (%)
A)
B)
*
C)
**
D)
E)
0
0.01
0.02
0.03
DMSO
APAP
EtOH
ChNP
Liver area (mm2)
**
*
**
DMSO
0.1%
APAP
2640 µM
EtOH 2%
ChNPs 200
µg/mL
APAP
2640 µM
EtOH 2%
ChNPs 200
µg/mL
ChNPs 200
mg/L
ChNPs 200
mg/L
ChNPs 200
mg/L
APAP 2640
µM
APAP 2640
µM
ChNPs 200
mg/L
ChNPs
ChNPs
Figure 6. (A) Representative images for the RFP liver area after each treatment. (B) Quantification
and measurement of the RFP liver area (mm2) for the controls and all the tested compounds. (C) Representative images for steatosis and yolk retention. Arrow (↗) in EtOH treatment indicate
liver steatosis (absorbed ORO stain by the liver), and arrows (↗) in paracetamol (APAP) and ChNP
treatments indicate increased yolk retention (absorbed ORO stain by the yolk). (D) Percentages of
embryos showed yolk retention and (E) steatosis. Thirty embryos were used for each treatment in
every experiment. *, significantly different from DMSO (p < 0.05); **, significantly different from DMSO
(p < 0.01). 3.4. Chitosan Nanoparticles Do Not Affect Cardiac Functions A) C)
ChNPs C) 0
2
4
6
8
10
12
14
16
18
20
DMSO
0.1%
EtOH 2%
APAP
2640uM
ChNP
200µg/ml
Yolk retention (%)
DMSO
0.1%
0
10
20
30
40
50
60
DMSO
EtOH 2%
APAP
2640uM
ChNPs
200µg/mL
Steatosis (%)
*
**
D)
E)
hNP
*
hNPs 200
g/mL
APAP
2640 µM
EtOH 2%
ChNPs 200
µg/mL
NPs 200
g/L
ChNPs 200
mg/L
ChNPs 200
mg/L
APAP 2640
µM
APAP 2640
µM
ChNPs 200
mg/L 0
2
4
6
8
10
12
14
16
18
20
DMSO
0.1%
EtOH 2%
APAP
2640uM
ChNP
200µg/ml
Yolk retention (%)
DMSO
0.1%
0
10
20
30
40
50
60
DMSO
EtOH 2%
APAP
2640uM
ChNPs
200µg/mL
Steatosis (%)
B)
*
**
D)
E)
0
0.01
0.02
0.03
DMSO
APAP
EtOH
ChNP
Liver area (mm2)
**
*
**
DMSO
0.1%
APAP
2640 µM
EtOH 2%
ChNPs 200
µg/mL
APAP
2640 µM
EtOH 2%
ChNPs 200
µg/mL
ChNPs 200
mg/L
ChNPs 200
mg/L
ChNPs 200
mg/L
APAP 2640
µM
APAP 2640
µM
ChNPs 200
mg/L E) D) Figure 6. (A) Representative images for the RFP liver area after each treatment. (B) Quantification and
Figure 6. (A) Representative images for the RFP liver area after each treatment. (B) Quantification Figure 6. (A) Representative images for the RFP liver area after each treatment. (B) Quantification and
measurement of the RFP liver area (mm2) for the controls and all the tested compounds. (C)
Representative images for steatosis and yolk retention. Arrow () in EtOH treatment indicate liver
steatosis (absorbed ORO stain by the liver), and arrows () in paracetamol (APAP) and ChNP
treatments indicate increased yolk retention (absorbed ORO stain by the yolk). (D) Percentages of
embryos showed yolk retention and (E) steatosis. Thirty embryos were used for each treatment in
every experiment. *, significantly different from DMSO (p < 0.05); **, significantly different from
DMSO (p < 0.01). Figure 6. (A) Representative images for the RFP liver area after each treatment. (B) Quantification
and measurement of the RFP liver area (mm2) for the controls and all the tested compounds. (C) Representative images for steatosis and yolk retention. Arrow (↗) in EtOH treatment indicate
liver steatosis (absorbed ORO stain by the liver), and arrows (↗) in paracetamol (APAP) and ChNP
treatments indicate increased yolk retention (absorbed ORO stain by the yolk). (D) Percentages of
embryos showed yolk retention and (E) steatosis. 3.4. Chitosan Nanoparticles Do Not Affect Cardiac Functions However, 56% of the ethanol-treated embryos showed signs of steatosis, a result
consistent with data presented by Passeri et al. [48]. 11 of 13 Nanomaterials 2019, 9, 122 11 of 13 4. Conclusions Our study for the first time presents data that comprehensively investigates the organ-specific
toxicity of ChNPs in vivo in the zebrafish embryo model. The performed toxicity test indicated that
ChNPs were unable to induce teratogenic phenotypes and embryo death at the used concentrations. Moreover, ChNP-treated embryos displayed normal heart physiology, thus indicating the absence
of cardiotoxic effects. Nevertheless, ChNPs exerted remarkable neurotoxic effects and significantly
impaired the liver size, suggesting these nanoparticles are potentially harmful to both the brain and
liver physiology of zebrafish embryos. We believe that our results add new insights into the available
data concerning the in vivo effect of ChNPs and provide useful information concerning the use of
ChNPs in polymer-based nanomaterials technology. Whether ChNPs may affect other organs at
higher concentrations or after prolonged exposition needs to be determined, and will be the object of
future study. Author Contributions: Conceptualization: G.P. and G.K.N.; formal analysis: K.R., M.H.Y., H.M.Y., K.A.M., G.P.,
and G.K.N.; funding acquisition: G.K.N.; investigation: H.A-.S., N.Y., and G.K.N.; methodology: R.M.P. and
G.K.N.; project administration: G.K.N.; resources: K.R., H.M.Y., and K.A.M.; supervision: G.K.N.; visualization:
M.H.Y.; writing (original draft): H.A-.S., G.P., and G.K.N.; writing (review and editing): H.A-.S., N.Y., R.M.P., G.P.,
and G.K.N. Author Contributions: Conceptualization: G.P. and G.K.N.; formal analysis: K.R., M.H.Y., H.M.Y., K.A.M., G.P.,
and G.K.N.; funding acquisition: G.K.N.; investigation: H.A-.S., N.Y., and G.K.N.; methodology: R.M.P. and
G.K.N.; project administration: G.K.N.; resources: K.R., H.M.Y., and K.A.M.; supervision: G.K.N.; visualization:
M.H.Y.; writing (original draft): H.A-.S., G.P., and G.K.N.; writing (review and editing): H.A-.S., N.Y., R.M.P., G.P.,
and G.K.N. Funding: This work was supported by the NPRP grant [#9-254-2-120] from the Qatar National Research Fund,
a Member of Qatar Foundation. The study was also partially supported by the grants [GCC-2017-001] given to
G.K.N. and [QUCG-CHS-2018\2019-1] given to G.P. Acknowledgments: The article processing charge (APC) for the publication of this article was funded by the
[Qatar National Library]. The authors would like to thank Rola Salem and Nadeema Haj Ali for their technical
support, and Robert Monticone for his help in proofreading the manuscript. Conflicts of Interest: The authors declare no conflicts of interest. Conflicts of Interest: The authors declare no conflicts of interest. References 1. Qilin, T.; Huang, G.; Ran, R.; Huang, Y.; Lou, Q.; Chen, Y. The Application of Chitosan and Its Derivatives as
Nanosized Carriers for the Delivery of Chemical Drugs and Genes or Proteins. Curr. Drug Targets 2016, 17, 811–816. 1. Qilin, T.; Huang, G.; Ran, R.; Huang, Y.; Lou, Q.; Chen, Y. The Application of Chitosan and Its Derivatives as
Nanosized Carriers for the Delivery of Chemical Drugs and Genes or Proteins. Curr. Drug Targets 2016, 17, 811–816. 2. Wang, J.J.; Zeng, Z.W.; Xiao, R.Z.; Xie, T.; Zhou, G.L.; Zhan, X.R.; Wang, S.L. Recent advances of chitosan
nanoparticles as drug carriers. Int. J. Nanomed. 2011, 6, 765–774. [CrossRef] 2. Wang, J.J.; Zeng, Z.W.; Xiao, R.Z.; Xie, T.; Zhou, G.L.; Zhan, X.R.; Wang, S.L. Recent advances of chitosan
nanoparticles as drug carriers. Int. J. Nanomed. 2011, 6, 765–774. [CrossRef] 3. Xu, Q.J.; Nian, Y.G.; Jin, X.C.; Yan, C.Z.; Liu, J.; Jiang, G.M. Effects of chitosan on growth of an aquatic plant
(Hydrilla verticillata) in polluted waters with different chemical oxygen demands. J. Environ. Sci. 2007, 19, 217–221. [CrossRef] 3. Xu, Q.J.; Nian, Y.G.; Jin, X.C.; Yan, C.Z.; Liu, J.; Jiang, G.M. Effects of chitosan on growth of an aquatic plant
(Hydrilla verticillata) in polluted waters with different chemical oxygen demands. J. Environ. Sci. 2007, 19, 217–221. [CrossRef] 4. Mosleh, Y.Y.; Paris-Palacios, S.; Ahmed, M.T.; Mahmoud, F.M.; Osman, M.A.; Biagianti-Risbourg, S. Effects of
chitosan on oxidative stress and metallothioneins in aquatic worm Tubifex tubifex (Oligochaeta, Tubificidae). Chemosphere 2007, 67, 167–175. [CrossRef] [PubMed] 4. Mosleh, Y.Y.; Paris-Palacios, S.; Ahmed, M.T.; Mahmoud, F.M.; Osman, M.A.; Biagianti-Risbourg, S. Effects of
chitosan on oxidative stress and metallothioneins in aquatic worm Tubifex tubifex (Oligochaeta, Tubificidae). Chemosphere 2007, 67, 167–175. [CrossRef] [PubMed] 5. Crini, G.; Badot, P.-M. Application of chitosan, a natural aminopolysaccharide, for dye removal from
aqueous solutions by adsorption processes using batch studies: A review of recent literature. Prog. Polym. Sci. 2008, 33, 399–447. [CrossRef] 5. Crini, G.; Badot, P.-M. Application of chitosan, a natural aminopolysaccharide, for dye removal from
aqueous solutions by adsorption processes using batch studies: A review of recent literature. Prog. Polym. Sci. 2008, 33, 399–447. [CrossRef] 6. Shariatinia, Z.; Fazli, M. Mechanical properties and antibacterial activities of novel nanobiocomposite films
of chitosan and starch. Food Hydrocoll. 2015, 46, 112–124. [CrossRef] 6. Shariatinia, Z.; Fazli, M. Mechanical properties and antibacterial activities of novel nanobiocomposite films
of chitosan and starch. Food Hydrocoll. References 2015, 46, 112–124. [CrossRef] 7. Choudhary, R.C.; Kumaraswamy, R.V.; Kumari, S.; Sharma, S.S.; Pal, A.; Raliya, R.; Biswas, P.; Saharan, V. Cu-chitosan nanoparticle boost defense responses and plant growth in maize (Zea mays L.). Sci. Rep. 2017, 7, 9754. [CrossRef] [PubMed] 7. Choudhary, R.C.; Kumaraswamy, R.V.; Kumari, S.; Sharma, S.S.; Pal, A.; Raliya, R.; Biswas, P.; Saharan, V. Cu-chitosan nanoparticle boost defense responses and plant growth in maize (Zea mays L.). Sci. Rep. 2017, 7, 9754. [CrossRef] [PubMed] 8. Vasconcelos, M.W. Chitosan and chitooligosaccharide utilization in phytoremediation and biofortification
programs: Current knowledge and future perspectives. Front. Plant Sci. 2014, 5, 616. [CrossRef] 8. Vasconcelos, M.W. Chitosan and chitooligosaccharide utilization in phytoremediation and biofortification
programs: Current knowledge and future perspectives. Front. Plant Sci. 2014, 5, 616. [CrossRef] 9. Fathi, M.; Majidi, S.; Zangabad, P.S.; Barar, J.; Erfan-Niya, H.; Omidi, Y. Chitosan-based multifunctional
nanomedicines and theranostics for targeted therapy of cancer. Med. Res. Rev. 2018, 38, 2110–2136. [CrossRef] 9. Fathi, M.; Majidi, S.; Zangabad, P.S.; Barar, J.; Erfan-Niya, H.; Omidi, Y. Chitosan-based multifunctional
nanomedicines and theranostics for targeted therapy of cancer. Med. Res. Rev. 2018, 38, 2110–2136. [CrossRef] 10. Nagpal, K.; Singh, S.K.; Mishra, D.N. Chitosan nanoparticles: A promising system in novel drug delivery. Chem. Pharm. Bull. 2010, 58, 1423–1430. [CrossRef] 10. Nagpal, K.; Singh, S.K.; Mishra, D.N. Chitosan nanoparticles: A promising system in novel drug delivery. Chem. Pharm. Bull. 2010, 58, 1423–1430. [CrossRef] 12 of 13 12 of 13 Nanomaterials 2019, 9, 122 11. Kumar, N.; Salar, R.K.; Prasad, M.; Ranjan, K. Synthesis, characterization and anticancer activity of vincristine
loaded folic acid-chitosan conjugated nanoparticles on NCI-H460 non-small cell lung cancer cell line. Egypt. J. Basic Appl. Sci. 2017. [CrossRef] 2. Wang, S.; Liu, K.; Wang, X.; He, Q.; Chen, X. Toxic effects of celastrol on embryonic development of zebra
(Danio rerio). Drug Chem. Toxicol. 2011, 34, 61–65. [CrossRef] [PubMed] 13. Vunain, E.; Mishra, A.K.; Mamba, B.B. Fundamentals of chitosan for biomedical applications. In Chitosan
Based Biomaterials; Jennings, J.A., Bumgardner, J.D., Eds.; Woodhead Publishing: Sawston, UK, 2017; Volume
1, pp. 3–30. pp
14. Kean, T.; Thanou, M. Biodegradation, biodistribution and toxicity of chitosan. Adv. Drug Deliv. Rev. 2010, 62, 3–11. [CrossRef] 15. Eimon, P.M.; Rubinstein, A.L. The use of in vivo zebrafish assays in drug toxicity screening. Expert Opin. Drug Metab. Toxicol. 2009, 5, 393–401. [CrossRef] [PubMed] 16. Lieschke, G.J.; Currie, P.D. Animal models of human disease: Zebrafish swim into view. Nat. References Rev. Genet. 2007, 8, 353–367. [CrossRef] [PubMed] 17. Bai, W.; Tian, W.; Zhang, Z.; He, X.; Ma, Y.; Liu, N.; Chai, Z. Effects of copper nanoparticles on the
development of zebrafish embryos. J. Nanosci. Nanotechnol. 2010, 10, 8760–8766. [CrossRef] 18. Hu, Y.-L.; Qi, W.; Han, F.; Shao, J.-Z.; Gao, J.-Q. Toxicity evaluation of biodegradable chitosan nanop
using a zebrafish embryo model. Int. J. Nanomed. 2011, 6, 3351–3359. [CrossRef] 19. Wang, Y.; Zhou, J.; Liu, L.; Huang, C.; Zhou, D.; Fu, L. Characterization and toxicology evaluation of chitosan
nanoparticles on the embryonic development of zebrafish, Danio rerio. Carbohydr. Polym. 2016, 141, 204–210. [CrossRef] 20. Yuan, Z.; Li, Y.; Hu, Y.; You, J.; Higashisaka, K.; Nagano, K.; Tsutsumi, Y.; Gao, J. Chitosan nanoparticles and
their Tween 80 modified counterparts disrupt the developmental profile of zebrafish embryos. Int. J. Pharm. 2016, 515, 644–656. [CrossRef] 21. El-Harbawi, M. Toxicity Measurement of Imidazolium Ionic Liquids Using Acute Toxicity Test. Procedia
Chem. 2014, 9, 40–52. [CrossRef] 22. Cornet, C.; Calzolari, S.; Miñana-Prieto, R.; Dyballa, S.; van Doornmalen, E.; Rutjes, H.; Savy, T.; D’Amico, D.;
Terriente, J. ZeGlobalTox: An Innovative Approach to Address Organ Drug Toxicity Using Zebrafish. Int. J. Mol. Sci. 2017, 18, 864. [CrossRef] 23. Milan, D.J.; Jones, I.L.; Ellinor, P.T.; MacRae, C.A. In vivo recording of adult zebrafish electrocardiogram and
assessment of drug-induced QT prolongation. Am. J. Physiol. Heart Circ. Physiol. 2006, 291, H269–H273. [CrossRef] 24. Dhillon, S.S.; Dóró, É.; Magyary, I.; Egginton, S.; Sík, A.; Müller, F. Optimisation of Embryonic and Larval
ECG Measurement in Zebrafish for Quantifying the Effect of QT Prolonging Drugs. PLoS ONE 2013, 8, e60552. [CrossRef] [PubMed] 25. Wen, L.; Wei, W.; Gu, W.; Huang, P.; Ren, X.; Zhang, Z.; Zhu, Z.; Lin, S.; Zhang, B. Visualization of
monoaminergic neurons and neurotoxicity of MPTP in live transgenic zebrafish. Dev. Biol. 2008, 314, 84–92. [CrossRef] [PubMed] 26. Vliegenthart, A.D.B.; Tucker, C.S.; Del Pozo, J.; Dear, J.W. Zebrafish as model organisms for studying
drug-induced liver injury. Br. J. Clin. Pharm. 2014, 78, 1217–1227. [CrossRef] 7. Goessling, W.; Sadler, K.C. Zebrafish: An Important Tool for Liver Disease Research. Gastroentero
2015, 149, 1361–1377. [CrossRef] [PubMed] 28. Crosby, E.B.; Bailey, J.M.; Oliveri, A.N.; Levin, E.D. Neurobehavioral Impairments Caused by Developmental
Imidacloprid Exposure in Zebrafish. Neurotoxicol. Teratol. 2015, 49, 81–90. [CrossRef] 29. McLean, D.L.; Fetcho, J.R. Using Imaging and Genetics in Zebrafish to Study Developing Spinal Circuits In
Vivo. Dev. Neurobiol. 2008, 68, 817–834. References [CrossRef] 30. MacPhail, R.C.; Brooks, J.; Hunter, D.L.; Padnos, B.; Irons, T.D.; Padilla, S. Locomotion in larval zebrafish:
Influence of time of day, lighting and ethanol. Neurotoxicology 2009, 30, 52–58. [CrossRef] [PubMed] Influence of time of day, lighting and ethanol. Neurotoxicology 2009, 30, 52–58. [CrossRef] [PubMed]
31. Denvir, M.A.; Tucker, C.S.; Mullins, J.J. Systolic and diastolic ventricular function in zebrafish embryos:
Influence of norepenephrine, MS-222 and temperature. BMC Biotechnol. 2008, 8, 21. [CrossRef] [PubMed] 31. Denvir, M.A.; Tucker, C.S.; Mullins, J.J. Systolic and diastolic ventricular function in zebrafish embryos:
Influence of norepenephrine, MS-222 and temperature. BMC Biotechnol. 2008, 8, 21. [CrossRef] [PubMed] 32. Postema, P.G.; Wilde, A.A.M. The Measurement of the QT Interval. Curr. Cardiol. Rev. 2014, 10, 287–294. [CrossRef] [PubMed] 13 of 13 13 of 13 Nanomaterials 2019, 9, 122 33. Howarth, D.L.; Yin, C.; Yeh, K.; Sadler, K.C. Defining hepatic dysfunction parameters in two models of fatty
liver disease in zebrafish larvae. Zebrafish 2013, 10, 199–210. [CrossRef] [PubMed] 34. Carvalho, P.M.; Felício, M.R.; Santos, N.C.; Gonçalves, S.; Domingues, M.M. Application of Light Scattering
Techniques to Nanoparticle Characterization and Development. Front. Chem. 2018, 6, 237. [CrossRef]
[PubMed] 35. Yan, H.; Teh, C.; Sreejith, S.; Zhu, L.; Kwok, A.; Fang, W.; Ma, X.; Nguyen, K.T.; Korzh, V.; Zhao, Y. Functional
mesoporous silica nanoparticles for photothermal-controlled drug delivery in vivo. Angew. Chem. Int. Ed. Engl. 2012, 51, 8373–8377. [CrossRef] [PubMed] 36. Napierska, D.; Thomassen, L.C.; Rabolli, V.; Lison, D.; Gonzalez, L.; Kirsch-Volders, M.; Martens, J.A.;
Hoet, P.H. Size-dependent cytotoxicity of monodisperse silica nanoparticles in human endothelial cells. Small 2009, 5, 846–853. [CrossRef] 37. Lange, M.; Norton, W.; Coolen, M.; Chaminade, M.; Merker, S.; Proft, F.; Schmitt, A.; Vernier, P.; Lesch, K.P.;
Bally-Cuif, L. The ADHD-susceptibility gene lphn3.1 modulates dopaminergic neuron formation and
locomotor activity during zebrafish development. Mol. Psychiatry 2012, 17, 946–954. [CrossRef] [PubMed] 38. Hsu, C.-H.; Wen, Z.-H.; Lin, C.-S.; Chakraborty, C. The zebrafish model: Use in studying cellular mechanism
for a spectrum of clinical disease entities. Curr. Neurovasc. Res. 2007, 4, 111–120. [CrossRef] 39. Hatta, K.; Takahashi, T.; Nakamura, H.; Yamashiro, H.; Asukai, N.; Matsuzaki, I.; Yonezawa, Y. The association
between intravenous haloperidol and prolonged QT interval. J. Clin. Psychopharmacol. 2001, 21, 257–261. [CrossRef] 40. Hill, A.J.; Teraoka, H.; Heideman, W.; Peterson, R.E. Zebrafish as a Model Vertebrate for Investigating
Chemical Toxicity. Toxicol. Sci. 2005, 86, 6–19. [CrossRef] 41. Menke, A.L.; Spitsbergen, J.M.; Wolterbeek, A.P.M.; Woutersen, R.A. © 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/). References Normal anatomy and histology of the
adult zebrafish. Toxicol. Pathol. 2011, 39, 759–775. [CrossRef] 42. van Pomeren, M.; Brun, N.R.; Peijnenburg, W.; Vijver, M.G. Exploring uptake and biodistribution of polystyrene
(nano)particles in zebrafish embryos at different developmental stages. Aquat. Toxicol. 2017, 190, 40–45. [CrossRef] [PubMed] 43. Chu, J.; Sadler, K.C. A New School in Liver Development: Lessons from Zebrafish. Hepatology 2009, 50, 1656–1663. [CrossRef] [PubMed] 44. Jones, K.S.; Alimov, A.P.; Rilo, H.L.; Jandacek, R.J.; Woollett, L.A.; Penberthy, W.T. A high throughput live
transparent animal bioassay to identify non-toxic small molecules or genes that regulate vertebrate fat
metabolism for obesity drug development. Nutr. Metab. (Lond.) 2008, 5, 23. [CrossRef] [PubMed] 45. He, J.-H.; Guo, S.-Y.; Zhu, F.; Zhu, J.-J.; Chen, Y.-X.; Huang, C.-J.; Gao, J.-M.; Dong, Q.-X.; Xuan, Y.-X.;
Li, C.-Q. A zebrafish phenotypic assay for assessing drug-induced hepatotoxicity. J. Pharm. Toxicol Methods
2013, 67, 25–32. [CrossRef] [PubMed] 46. North, T.E.; Babu, I.R.; Vedder, L.M.; Lord, A.M.; Wishnok, J.S.; Tannenbaum, S.R.; Zon, L.I.; Goessling, W. PGE2-regulated wnt signaling and N-acetylcysteine are synergistically hepatoprotective in zebrafish
acetaminophen injury. Proc. Natl. Acad. Sci. USA 2010, 107, 17315–17320. [CrossRef] [PubMed] 47. Rabinowich, L.; Shibolet, O. Drug Induced Steatohepatitis: An Uncommon Culprit of a Common Disease. Biomed. Res. Int. 2015. [CrossRef] [PubMed] 48. Passeri, M.J.; Cinaroglu, A.; Gao, C.; Sadler, K.C. Hepatic Steatosis in Response to Acute Alcohol Exposure
in Zebrafish requires Srebp Activation. Hepatology 2009, 49, 443–452. [CrossRef] © 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).
|
https://openalex.org/W4200296649
|
https://www.frontiersin.org/articles/10.3389/fbioe.2021.754266/pdf
|
English
| null |
The Development Tendency of 3D-Printed Bioceramic Scaffolds for Applications Ranging From Bone Tissue Regeneration to Bone Tumor Therapy
|
Frontiers in bioengineering and biotechnology
| 2,021
|
cc-by
| 10,686
|
REVIEW published: 20 December 2021
doi: 10.3389/fbioe.2021.754266 The Development Tendency of
3D-Printed Bioceramic Scaffolds for
Applications Ranging From Bone
Tissue Regeneration to Bone Tumor
Therapy
Zhixiang Fang 1†, Jihang Chen 2†, Jiangxia Pan 3, Guoqiang Liu 1 and Chen Zhao 2*
1Department of Orthopedics, The Second Hospital of Shaoxing, Shaoxing, China, 2Department of Orthopedics, Zhejiang
Provincial People’s Hospital, Affiliated People’s Hospital of Hangzhou Medical College, Hangzhou, China, 3Nursing Department,
Affiliated Hospital of Shaoxing University, Shaoxing, China Three-dimensional (3D) printing concept has been successfully employed in regenerative
medicine to achieve individualized therapy due to its benefit of a rapid, accurate, and
predictable production process. Traditional biocomposites scaffolds (SCF) are primarily
utilised for bone tissue engineering; nevertheless, over the last few years, there has already
been a dramatic shift in the applications of bioceramic (BCR) SCF. As a direct
consequence, this study focused on the structural, degeneration, permeation, and
physiological activity of 3D-printed BCR (3DP-B) SCF with various conformations and
work systems (macros, micros, and nanos ranges), as well as their impacts on the
mechanical, degeneration, porosity, and physiological activities. In addition, 3DP-B SCF
are highlighted in this study for potential uses applied from bone tissue engineering (BTE) to
bone tumor treatment. The study focused on significant advances in practical 3DP-B SCF
that can be utilized for tumor treatment as well as bone tissue regeneration (BTR). Given
the difficulties in treating bone tumors, these operational BCR SCF offer a lot of promise in
mending bone defects caused by surgery and killing any remaining tumor cells to
accomplish bone tumor treatment. Furthermore, a quick assessment of future
developments in this subject was presented. The study not only summarizes recent
advances in BCR engineering, but it also proposes a new therapeutic strategy focused on
the extension of conventional ceramics’ multifunction to a particular diagnosis. Edited by:
Ayşen Tezcaner,
Middle East Technical University,
Turkey Reviewed by:
Francesco Baino,
Politecnico di Torino, Italy
Sanzhong Xu,
Zhejiang University, China *Correspondence:
Chen Zhao
zhaochen@hmc.edu.cn
†These authors have contributed
equally to this work *Correspondence:
Chen Zhao
zhaochen@hmc.edu.cn Specialty section:
This article was submitted to
Biomaterials,
a section of the journal
Frontiers in Bioengineering and
Biotechnology Received: 06 August 2021
Accepted: 04 October 2021
Published: 20 December 2021 Keywords: 3D-printed bioceramic scaffolds, bone regeneration, bone tumor, scaffolds, bone Citation: Traffic accidents, old age, bone tumors, and other causes of bone tissue abnormalities have serious
consequences for one’s healthcare as well as living standards (Bădilă et al., 2021) . Large bone
abnormalities usually necessitate intermediation treatment in order to recover. Nonetheless, because
autogenous bone transplant or autograft is the “gold standard” graft material, supply is limited. As a
result, several studies (Wang et al., 2017; Zhu et al., 2017; Lai et al., 2018; Sui et al., 2019) are focusing
on developing innovative tissues engineering methodologies for bones tissues regeneration. Scaffold
(SCF) are important in bone tissue creation because they provide a three-dimensional (3D) Fang Z, Chen J, Pan J, Liu G and
Zhao C (2021) The Development
Tendency of 3D-Printed Bioceramic
Scaffolds for Applications Ranging
From Bone Tissue Regeneration to
Bone Tumor Therapy. Front. Bioeng. Biotechnol. 9:754266. doi: 10.3389/fbioe.2021.754266 December 2021 | Volume 9 | Article 754266 Frontiers in Bioengineering and Biotechnology | www.frontiersin.org Development Tendency of 3D-Printed Bioceramic Scaffolds Fang et al. FIGURE 1 | Microstructure and composition based bone tissue engineering. Reproduced with permission from (Hao et al., 2017). GURE 1 | Microstructure and composition based bone tissue engineering. Reproduced with permission from (Hao et al., 2017). background for cells connection and propagation. Gas frothing
(Mathieu et al., 2006), Thaw (Li and Feng, 2005), fibre
companionship, particulate/salt leachate (Cao and Kuboyama,
2010; Marques et al., 2017), emulsification (Bohner et al., 2005),
phase isolation, and other traditional manufacturing processes
are still unable to regulate the porous texture, architectural design,
permeability, or interconnectedness of the SCF, and can therefore
explicitly and sufficiently enhance cell progression and tissue
redevelopment (Butscher et al., 2011). 3D printing (3DP)
technique was developed to enterprise and develop SCF with
well-ordered
chemistries,
intended
contours,
and
linked
perviousness
using
computer-aided
design
(CAD)
and
computer-aided
manufacturing
(CAM)
to
address
the
limitations
of
traditional
manufacturing
techniques
(Bose
et al., 2013; Brunello et al., 2016; Guvendiren et al., 2016;
Bădilă et al., 2021). structures have more contact area and ruggedness, which
encourages osteoblast adhesion to the scaffold surfaces (Park
et al., 2007). Few studies of 3D-printed ceramics SCF about bones tissues
engineering have been published (Nowicki et al., 2017),
summarizing just 3DP technologies and various types of
ceramic-SCF. The goal of current review, on the other hand, is
to sketch out the development trend of 3DP-B SCF for
applications ranges from bones tissues regeneration to bone
tumor treatment. Citation: The prior step was to demonstrate 3DP-B
SCF with various configurations and structural organization
topologies
(macros-,
micros-,
and
nanos-scales). Conventionally, 3DP-B SCF were manufactured and primarily
used to regenerate bone tissue. This study, on the other hand,
emphases on current progresses in functional 3DP-B SCF that
could be utilized for both cancer therapy and BTE. Such
serviceable BCR SCF offer a lot of promise in terms of fixing
surgically produced bone defects and killing any remaining tumor
cells in order to attain bone tumor treatment. Lastly, a quick
assessment of forthcoming developments in this topic is
presented
(Zhang
et
al.,
2017a). Bones,
like
natural
nanocomposites
in
the
human
body,
have
multi-layer
architectures with complicated compositions, as seen in Figure 1. A standard scaffold for bone tissue engineering (BTE) should
mimic the shape and physiological activities of typical bone tissue
in based on its chemical mixtures, rigid hierarchy, and features. Ceramic-based SCF [e.g., Ca-salts-ceramics, Ca-Si (CS) ceramics,
and bioactive-glasses (BGs)] are drawing considerable interest to
be used in BTE due to their resemblance to indigenous bone
chemical compositions, bio—compatibility, wettability, biological
activities, osteoinduction, and printability (Woodard et al., 2007). Bioceramic (BCR) frameworks with such a hierarchical system
that
comprises
macroscopic,
microscopic,
and
nanoscale
structures have also been created. SCF have indeed been
intended
with
various
macrostructures
(e.g.,
pore
size,
permeability,
and
pore
connectivity)
to
enable
tissue
expanding and appropriate transportation of nutrients, oxygen,
waste, and growth factors, and to also encourage effective tissue
regeneration cell development from the fringes to the interior of
the SCF(Bose et al., 2013). SCF with submicron and nanoscale Bioceramic (BCR) frameworks with such a hierarchical system
that
comprises
macroscopic,
microscopic,
and
nanoscale
structures have also been created. SCF have indeed been
intended
with
various
macrostructures
(e.g.,
pore
size,
permeability,
and
pore
connectivity)
to
enable
tissue
expanding and appropriate transportation of nutrients, oxygen,
waste, and growth factors, and to also encourage effective tissue
regeneration cell development from the fringes to the interior of
the SCF(Bose et al., 2013). SCF with submicron and nanoscale 3-DIMENSIONAL PRINTING
MANUFACTURING TECHNIQUES
BENEFITS The filaments is dissolved and ejected from a
throttling valve (Luo et al., 2013), that is typically made of
ceramics plus binders. In 3DP, ceramics are diluted with water
binders to promote binding among powder particles and SCF,
allowing them to form their desired forms (Pham and Gault, 1998). The advantages of fabricating bio-ceramic SCF using the 3DP
approach are as follows: 1) SCF’ inner and exterior architectures,
such as pore form, perviousness, and interconnectedness, could
be precisely manipulated to achieve extraordinary organizational
complication, elasticity, and patient-particular burdens (Jin Woo
and Dong-Woo, 2015). 2) To restore bone tissue, optimal 3DP
SCF deliver a growth-directing construction for cell migration
and proliferation (Vaezi et al., 2013). 3) 3DP enables rapid
construction while reducing experimental mistakes (Vaezi
et al., 2013). The use of mesoporous bioactive glasses (MBGs) to fabricate
scaffolds via 3D printing has been identified as having significant
promise for the creation of bioactive synthetic bone replacements. The obvious advantages of additive manufacturing make it an
attractive
strategy
for
circumventing
the
most
common
limitations of traditional fabrication processes for glass and
ceramic scaffolds, such as poor reproducibility and control
over the final 3D structure, as well as low mechanical strength. Additionally, the highly tailorable textural characteristics of
mesoporous materials were discovered to be critical in the
construction
of
multifunctional
systems
capable
of
concurrently supporting bone repair and therapeutic action via
local delivery of medicines and/or biologically active ions (Baino
and Fiume, 2020). Table 1 listed some of the key studies that
reported MBGs along with the details. The advantages of fabricating bio-ceramic SCF using the 3DP
approach are as follows: 1) SCF’ inner and exterior architectures,
such as pore form, perviousness, and interconnectedness, could
be precisely manipulated to achieve extraordinary organizational
complication, elasticity, and patient-particular burdens (Jin Woo
and Dong-Woo, 2015). 2) To restore bone tissue, optimal 3DP
SCF deliver a growth-directing construction for cell migration
and proliferation (Vaezi et al., 2013). 3) 3DP enables rapid
construction while reducing experimental mistakes (Vaezi
et al., 2013). 3-DIMENSIONAL PRINTING
MANUFACTURING TECHNIQUES
BENEFITS Traditional fabrication procedures such as gas flocculation, fibre
bonding, freeze dryness, phases isolation, and particle rinsing were
used to create most SCF at first (Lu et al., 2021) . Conventional
strategies, but at the other hand, are unable to regulate the SCF’ December 2021 | Volume 9 | Article 754266 Frontiers in Bioengineering and Biotechnology | www.frontiersin.org 2 Development Tendency of 3D-Printed Bioceramic Scaffolds Fang et al. TABLE 1 | Selected studies using 3D-printed mesoporous bioactive glasses material, their parameters, manufacturing and other structural details. Material
Binding agent
Mesopore
SDA
BET surface area
(m2/g)
Dispensing
pressure (kPa)
Size
(µm)
Reference
SiO2-CaO-P2O5 MBG/PCL
composite
poly(caprolactone)
F-127
520
Not reported
190
Yun et al. (2011)
SiO2-P2O5 MBG
methyl cellulose
F-127
152–310
Adapted during
printing
400
García et al. (2011)
CSH/SiO2-CaO-P2O5 MBG
poly(caprolactone)
P123
4.5–12.8
220–360
350
Qi et al. (2017)
Carboxylic-modified SiO2-
CaO-P2O5 MBG
poly(3-hydroxybutyrate-co-3-
hydroxyhexanoate)
P123
63–330
Not reported
250
Zhu et al. (2015)
SiO2-CaO-P2O5 MBG/alginate
composite
Alginate
P123
Not reported
180–250
344–415
Luo et al. (2013) ted studies using 3D-printed mesoporous bioactive glasses material, their parameters, manufacturing and other structural details. porosity
structure,
morphology,
permeability,
or
interconnectedness, and the resulting SCF cannot be specifically
tailored to facilitate angiogenesis and tissue regeneration (Li et al.,
2014). Conventional
manufactured
products
for
the
manufacturing of BCR SCF have been extensively used to
overcome the limitations of 3DP techniques based on CAD and
CAM (Jin Woo and Dong-Woo, 2015). Professor Sachs of MIT
(Massachusetts Institute of Technology) was the first to create the
term 3D printing innovation, which really is a form of rapid
prototyping technology (Liu et al., 2013). There seem to be a
variety of 3DP techniques, which may be classified depending on
the
method
of
construction. The
methods
utilized
are
stereolithography (SLA), selective laser sintering (SLS), fused
deposition modeling (FDM), and binder-based 3DP3DP (Hwa
et al., 2017). SLA SCF are created by splitting a based model sliced
by sliced first from head to tail, then solidifying it by using UV laser
beam. Instead of utilizing a UV laser, SLS uses sintered approach to
produce materials. In SLS, the laser is utilized to polymerize the
granules and fuse them collectively (Yves-Christian et al., 2010). SCF are constructed in FDM by layering filaments on the
workstation. The filaments is dissolved and ejected from a
throttling valve (Luo et al., 2013), that is typically made of
ceramics plus binders. 3-DIMENSIONAL PRINTING
MANUFACTURING TECHNIQUES
BENEFITS In 3DP, ceramics are diluted with water
binders to promote binding among powder particles and SCF,
allowing them to form their desired forms (Pham and Gault, 1998). Th
d
f f b i
i
bi
i SCF
i
h 3DP (Woodard et al., 2007). They promote wound healing by modifying
the in vivo conditions. The exterior of biocomposites SCF absorbed
osteoinductive
chemicals
and/or
ions
[e.g.,
calcium
(Ca),
phosphate
(PHPH),
and
silicate
(Si)]
ions
from
the
surroundings, encouraging MSC growth (Asa’ad et al., 2016). 3DP-B SCF should have the following characteristics: 1) Porous
structure which can be regulated and is consistent, such as pores,
size distribution, direction, and interconnectedness. Such a
permeable assembly is advantageous for nutrient delivery and
cell migration into SCF. 2) Sufficient machine-driven asset that
can be accurately regulated to fulfil a variety of clinic requirements. 3) Bio-compatibility and bio-degradability, which ensures that SCF
decay into harmless compounds with minimum inflammatory
reactions. 4) Bioactivity, which results in the development of
biochemical linkages between SCF and physiological system by
generating specific bio-chemical responses at the boundary
between materials and biological tissues. Furthermore, 3DP SCF
can be bifunctional, allowing them to be employed for both tumor
treatment and bone regeneration. porosity
structure,
morphology,
permeability,
or
interconnectedness, and the resulting SCF cannot be specifically
tailored to facilitate angiogenesis and tissue regeneration (Li et al.,
2014). Conventional
manufactured
products
for
the
manufacturing of BCR SCF have been extensively used to
overcome the limitations of 3DP techniques based on CAD and
CAM (Jin Woo and Dong-Woo, 2015). Professor Sachs of MIT
(Massachusetts Institute of Technology) was the first to create the
term 3D printing innovation, which really is a form of rapid
prototyping technology (Liu et al., 2013). There seem to be a
variety of 3DP techniques, which may be classified depending on
the
method
of
construction. The
methods
utilized
are
stereolithography (SLA), selective laser sintering (SLS), fused
deposition modeling (FDM), and binder-based 3DP3DP (Hwa
et al., 2017). SLA SCF are created by splitting a based model sliced
by sliced first from head to tail, then solidifying it by using UV laser
beam. Instead of utilizing a UV laser, SLS uses sintered approach to
produce materials. In SLS, the laser is utilized to polymerize the
granules and fuse them collectively (Yves-Christian et al., 2010). SCF are constructed in FDM by layering filaments on the
workstation. Calcium-(poly (3-hydroxybutyrate-
co-3-hydroxyhexanoate) Bioceramic
Scaffold Different reports have been revealed that some Ca-PHPH (poly
(3-hydroxybutyrate-co-3-hydroxyhexanoate)
ceramics
are
osteoinductive in the absent of supplementation (Samavedi
et al., 2013). Moreover, the chemical nature and charges of
calcium PHPH ceramics impact their osteoinduction capacity,
that might change permeation rate and, as a consequence,
endorse angiogenesis via cell–extracellular matrix interactions
(Samavedi et al., 2013). Calcium PHPH ceramics’ structural
quality in the context roadblock to their broad clinical
application. As little more than a consequence, Ca-PHPH
ceramics are used as non-load-bearing implantation in middle
ear operations and to replace bone defects in the oral cavity and
bones, as well as as a coating for dentistry and orthopaedic
metallic
implants. The
fracture
toughness
of
Ca-PHPH
ceramics is related to having primary ionic bonding. To
enhance biocompatibility and good mechanical of calcium
PHPH-based SCF, Awad et al. developed a special equation
comprising
8.75 wt
percent
phosphoric
acid,
and
then
manufactured a Ca PHPH-based scaffolding at low temp. Such
combinations improved the scaffold’s cytocompatibility and bone
regeneration while also enhancing its mechanical properties in a
significantly sized femur fracture (Inzana et al., 2014). Tricalcium PHPH (TCP) is another typical Ca-PHPH
ceramic, because to its quick breakdown degree and capability
to establish a strong bone–calcium PHPH connection. Certain
biomimetic implants, including HA, have indeed been proven to
deteriorate more rapidly as TCP ceramics. Porosity, stiffness, and
tenacity (the min particle size and parting space of the struts, that
could be exactly regulated by a 3DP) of 3DP TCP-SCF are all
influenced
by
particle
diameter,
depowering
effectiveness,
adhesive droplet size, and scaffolding shape (Tarafder et al.,
2015). Most favoured type of TCP scaffolding is b-TCP and
for its structural chemical resistance durability. Following
American physician Fred Houdlette Albee’s original effort in
1920 to implant b-TCP as a synthetic material to mend
operationally induced cracks in rabbit’s skeletons, b-TCP
ceramic
gained
increasing
interest
due
to
its
excellent
biocompatibility and osteogenesis (Eliaz and Metoki, 2017). Peer groups studies demonstrate that BCP ceramics have a
reduced biodegradation, higher biocompatibility, and improved
bone regeneration ability than pure HA and pure b-TCP
SCF(Yamada et al., 1997; Ryu et al., 2004; Wu et al., 2011). g
y
Hydroxyapatite (HA) is a kind of calcium PHPH. Benefits of 3-Dimensional Printing-B
Scaffolds Owing to their resemblance to the synthetic formulation, impact
strength,
wettability,
bioactivity,
bio
-
compatibility,
osteoconductivity, and possible osteoinduction of the bone, bio-
ceramics have become the subject of intense research for years Bio-ceramics for bone tissue manufacturing include Ca-PHPH
ceramics, CS ceramics, and BGs, among others. The utilisation of
3DP-B SCF in BTRis determined by their printable, mechanical, December 2021 | Volume 9 | Article 754266 Frontiers in Bioengineering and Biotechnology | www.frontiersin.org 3 Development Tendency of 3D-Printed Bioceramic Scaffolds Fang et al. TABLE 2 | Examples of commercially available phosphocalcic cements (Monma, 1988). Name of
cement
Manufacturer
Solid-phase
composition
Liquid-phase
composition
S/L
α—BSM (Biobon, Embarc)
ETEX
ACP + DCPD
Physiological serum
0.67 (g/cm3)
VitalOs
CalciOS
1: βTCP + Na 2H2P2O7
1: water
2: MCPM + CaSO 4, 2H2O
2: water + H3PO4
Eurobone
Kasios
βTCP + Na4P2O7
water + H2SO4
Bonesource
Stryker—Leibinger Corporation
TTCP + DCPA
Na
0.25 (g/cm3)
2HPO4 +
NaH
2PO4
Calcibon
Biomet
αTCP + DCPA + CaCO3 + PHA
Na2SO4
ChronOS Synthes
βTCP + MCPM + MgHPO4 + MgSO4
Na2H2P2O7 Hyaluronate de Sodium
Cerapaste (Primafix)
NGK Spark Plug
TTCP + DCPA
Sodium and biological characteristics. Lewis et al. have published
numerous papers on the construction of complicated 3D-
printed SCF with improved performance. They also gave a
thorough
overview
of
the
various
characteristics
and
parameters of 3DP methods on ceramic constructions. Pure HA, on the other hand, has a slow degradation rate, low
mechanical strength, and low fracture toughness, which prevents
full bone regeneration and may raise the danger of contamination (Lu
et al., 2002; Ni et al., 2008). As a result, HA was mixed with other
alignments such as zirconium oxide (Rapacz-Kmita et al., 2006),
carbon fibre (Suping et al., 2004), and Al2O3 (Lee et al., 2007) in order
to increase mechanical qualities and boost HA’s capabilities for bone
tissue regeneration. However, bioinert materials such as ZrO2, carbon
fibre, and Al2O3 greatly decrease the bioactivity of HA (Inagaki and
Kameyama, 2007). Furthermore, when natural polymers with quicker
kinetics were added, a nanocomposite scaffold having a low rate of
decomposition was created. Polylactide-co-glycolide acid (PLGA) is
one of the most extensively used polymeric biomaterials, with
excellent biodegradability and biocompatibility. As a result, Yang
et al. used a 3DP method to construct porous PLGA/HA SCF (Yang
et al., 2016). Frontiers in Bioengineering and Biotechnology | www.frontiersin.org Calcium-(poly (3-hydroxybutyrate-
co-3-hydroxyhexanoate) Bioceramic
Scaffold HA
(Ca10(PO4)6(OH)2) is an utmost studied Ca PHPH bio-ceramics
in bone tissues engineering based on the chemical arrangement,
which is comparable to that of the primary bone components,
resulting
in
beneficial
effects
on
osteoblast
adhesion
and
proliferation (Kumar et al., 2019). This may be used to optimize
the material behavior in biological settings by developing HA with a
particular surface structure and electric charge (Salinas et al., 2013). Calcium-Silicon Bioceramic Scaffold Calcium-Silicon Bioceramic Scaffold
CS (CaSiO3) ceramics, a form of unique bioactive substance, have
been inspected for bone regeneration (De Aza et al., 2000). These
composites can stimulate the development of a carbonated HA December 2021 | Volume 9 | Article 754266 Frontiers in Bioengineering and Biotechnology | www.frontiersin.org 4 Development Tendency of 3D-Printed Bioceramic Scaffolds Fang et al. FIGURE 2 | The impact of hollow tubes on the delivery of stem cells. (A) Scanning electron microscope images of adhering rBMSCs on the scaffold surface. In the
hollow channels of the BRT-H scaffolds, cells may be seen. (B) After 7 days of in vitro cultivation, the cells in the hollow channels were still alive. (C) Cell viability of rBMSCs
planted in scaffolds on day 1 and day 7. (D) Mice were implanted with scaffolds containing luciferase-labeled BMSCs, which were identified using in vivo fluorescence
imaging. (E) Relative fluorescence intensity statistical findings. Reproduced with permission from (Zhang et al., 2017b). FIGURE 2 | The impact of hollow tubes on the delivery of stem cells. (A) Scanning electron microscope images of adhering rBMSCs on the scaffold surface. In the
hollow channels of the BRT-H scaffolds, cells may be seen. (B) After 7 days of in vitro cultivation, the cells in the hollow channels were still alive. (C) Cell viability of rBMSCs
planted in scaffolds on day 1 and day 7. (D) Mice were implanted with scaffolds containing luciferase-labeled BMSCs, which were identified using in vivo fluorescence
imaging. (E) Relative fluorescence intensity statistical findings. Reproduced with permission from (Zhang et al., 2017b). FIGURE 2 | The impact of hollow tubes on the delivery of stem cells. (A) Scanning electron microscope images of adhering rBMSCs on the scaffold surface. In the
hollow channels of the BRT-H scaffolds, cells may be seen. (B) After 7 days of in vitro cultivation, the cells in the hollow channels were still alive. (C) Cell viability of rBMSCs
planted in scaffolds on day 1 and day 7. (D) Mice were implanted with scaffolds containing luciferase-labeled BMSCs, which were identified using in vivo fluorescence
imaging. (E) Relative fluorescence intensity statistical findings. Reproduced with permission from (Zhang et al., 2017b). coating by immersing in simulated bodily fluid (SBF), which can
generate a strong chemical link between the biopolymer and the
bone tissue along the biomaterial (Kokubo and Takadama, 2006;
Xu et al., 2008). Zhao et al. Calcium-Silicon Bioceramic Scaffold (2005) developed a very consistent CS
SCF with a pore-controlled assembly and good mechanical
features using the 3DP technique. The 3DP CS SCF had such
a reasonably easy manufacturing process, outstanding flexural
modulus, satisfying apatite-mineralization capacity in SBF, and
an elevated rejuvenating of bone deficiencies, asserting that they
would have huge potential. Furthermore, because of its ability to induce angiogenesis in
response to the action of vascular endothelial growth factor
(VEGF), BG is a viable substitute to conventional scaffolding
biomaterials (Kentleach et al., 2006). Preceding research has
shown that BGs may be used as sintering aids to improve
BCRs’ mechanical and biotic properties. Winkel et al. (2012)
sintered a BG-reinforced HA amalgamated SCF with an
orientated
pore
at
a
truncated
temperature. Lin
et
al. (Suwanprateeb et al., 2009) produced 45S5 BG-reinforced
wollastonite permeable ceramics with a compressive asset of
109 MPa. Different Doped Elements Hench was the one who discovered BGs. Na2O, CaO, SiO2, and
P2O5 are the major components of bioactive glasses (BGs). BGs
offer a lot of promise in terms of repairing and regenerating bone
deficiencies. BGs
have
osteoconductiveness
and
oseteoproductiveness,
which
improves
progenitor
cell
proliferation
and
differentiation
(Rainer
et
al.,
2008). Biotic apatite’s, a chemical component of bone, contain a
definite quantity of Zn++, Sr++, Mn++, and Mg++ in
addition to HA. These ions have their own roles, and by
gently
releasing
these
therapeutic
ions,
they
stimulate
osteogenesis and angiogenesis. December 2021 | Volume 9 | Article 754266 Frontiers in Bioengineering and Biotechnology | www.frontiersin.org Fang et al. Development Tendency of 3D-Printed Bioceramic Scaffolds FIGURE 3 | Tissue development in channels with various cross-sections: triangular, square, hexagonal, and circular. (A) Phalloidin-FITC stained actin stress fibres
for various geometries. (B) A computational simulation of tissue development depicting several tissue evolutions. Reproduced with permission from (Gariboldi and Best,
2015). FIGURE 3 | Tissue development in channels with various cross-sections: triangular, square, hexagonal, and circular. (A) Phalloidin-FITC stained actin stress fibres
for various geometries. (B) A computational simulation of tissue development depicting several tissue evolutions. Reproduced with permission from (Gariboldi and Best,
2015). second is a brushite, corresponding to H3PO4 or DCPD (Mestres
and Ginebra, 2011; Tamimi et al., 2012). Some phosphocalcic
cements are put on the market, some cements and their
formulations are shown in Table 1 for the usage of these bio-
ceramics is confined by their low mechanical attributes. Table 2
list various commercially available phosphocalcic cement. The integration of Zn++ in HA ceramics increases the dissociation
property, encourages osteoblastic activity from bone marrow stromal
cells in vitro, and obviously boosts new bone formation in vivo (Li
et al., 2009) by repressing osteoclast differentiation by antagonizing
NF-jB activation motivated by TNFα, an inhibitor of bone formation
in vitro and in vivo (Li et al., 2009). Manganese (Mn), an abundant mineral component for regular
human metabolism (e.g., biosynthesis in bone tissues), has many
benefits as a dopant factor for tissue engineering (Mondal et al.,
2010), which include excitation of angiogenesis process (Armulik
et al., 2000), augmentation of the bioactivity of cartilage
oligomeric matrix protein (Liu et al., 2018), and advancement
of osteogenic activity in vitro (Srivastava et al., 2012). Different Doped Elements Mn ions
might even improve the quantity of BMSC repositories in the
bone tissue, that are otherwise stem cells which may grow into
osteoblasts, chondrocytes, and adipocytes (Wang et al., 2014). The development of new vascular system in the interior regions of
the transplanted SCF is related with tissue engineering and a
modest rate of vasculature (Santos and Reis, 2010). Phosphocalcic Cement p
Historically, Legeros et al. (1982) was the first to suggest
biomaterial phosphocalcic cement. Since then a range of
preparations
have
been
produced,
analyzed
and
commercialized (Mirtchi et al., 1989). Cements are a class of
biomaterials distinguished through their galenic composition. These are the alternative to high thermostable ceramics. Therefore they can be assumed that the cement is made up of
pulverized liquid and solid phases, when combined in the correct
ratios, is fixed and hardened. Different phosphocalcium cements,
currently employed in orthopaedic and dental-surgery. Mineral
hydraulic (MH) cements could be distinguished from polymeric
cements by containing mineral fillers). They contain two main
categories: the former is apathetic, like CaCO3 or TTCP, and 3-DPRINTING-BSCAFFOLD WITH VARIED
MACRO/MICROSTRUCTURES FOR BONE
TISSUE ENGINEERING Among the most challenging problems encountered in area of
biomedical applications is the fabrication of hierarchical porous
SCF that imitate the conformational shape and features of actual
bone. To demonstrate features, nutrition delivery, and cell–matrix
interactions, the architecture must be developed at the macros,
micros, and nanos dimensions (Figure 2). Macrostructure is
important in BTE because it is closely linked to the degree of
bone ingrowth. Porous structure, pore diameter, pore shape, and
pore interconnectivity, in particularly, have a significant influence on
scaffold function, cell penetration, and adhesion to scaffold
macropores, all of which play a role in osseointegration. Frontiers in Bioengineering and Biotechnology | www.frontiersin.org The Influence of Macrostructure on
Physico-chemical Assets BCR SCF are being reported extensively in bone regeneration due to
their resemblence to bone inorganic compounds, bioactivity, bio -
compatibility,
osteoinduction,
and
probable
osteoinductivity. Nevertheless, in a number of cases, their mechanical properties
limited their application. As a consequence, scientists attempted to
modify the porosities of BCR SCF to generate mechanical properties
which suited the requirements of the ultimate implementation. December 2021 | Volume 9 | Article 754266 Frontiers in Bioengineering and Biotechnology | www.frontiersin.org 6 Fang et al. Fang et al. Development Tendency of 3D-Printed Bioceramic Scaffolds FIGURE 4 | (A) 3D-printed pure bioceramics (BC), DOPA-BC (2 mg/ml), DOPA-BC (4 mg/ml), DOPA-BC (6 mg/ml) scaffolds; (B) Optical photos of pure BC (b) and
4 mg/ml DOPA-BC (C) scaffolds on top view; (G) SEM images of pure BC (D) and 2 mg/ml DOPA-BC (E), 4 mg/ml. Reproduced with permission from (Ma et al., 2016a). FIGURE 4 | (A) 3D-printed pure bioceramics (BC), DOPA-BC (2 mg/ml), DOPA-BC (4 mg/ml), DOPA-BC (6 mg/ml) scaffolds; (B) Optical photos of pure BC (b) and
4 mg/ml DOPA-BC (C) scaffolds on top view; (G) SEM images of pure BC (D) and 2 mg/ml DOPA-BC (E), 4 mg/ml. Reproduced with permission from (Ma et al., 2016a). FIGURE 4 | (A) 3D-printed pure bioceramics (BC), DOPA-BC (2 mg/ml), DOPA-BC (4 mg/ml), DOPA-BC (6 mg/ml) scaffolds; (B) Optical photos of pure BC (b) and
4 mg/ml DOPA-BC (C) scaffolds on top view; (G) SEM images of pure BC (D) and 2 mg/ml DOPA-BC (E), 4 mg/ml. Reproduced with permission from (Ma et al., 2016a). Roohani-Esfahani et al. developed Sr-HT Gahnite, a BG ceramic
with a novel triphasic microstructure comprising of strontium (Sr)-
doped hardystonite (Ca2ZnSi2O7, HT) grains, clustering of
submicron gahnite (ZnAl2O4) crystals, and a glass phase. In
attempt to construct Sr-HT Gahnite 3DP SCF with improved
tensile strength, they employed a range of pore geometrie with a
range of porosities (Roohani-Esfahani et al., 2016). Roohani-Esfahani et al. developed Sr-HT Gahnite, a BG ceramic
with a novel triphasic microstructure comprising of strontium (Sr)-
doped hardystonite (Ca2ZnSi2O7, HT) grains, clustering of
submicron gahnite (ZnAl2O4) crystals, and a glass phase. In
attempt to construct Sr-HT Gahnite 3DP SCF with improved
tensile strength, they employed a range of pore geometrie with a
range of porosities (Roohani-Esfahani et al., 2016). Cells want a modulus of elasticity that is much larger than their
own radius of curvature (Zadpoor, 2015). The Influence of Macrostructure on
Physico-chemical Assets In reaction to
curvature, murine osteoblast-like cells migrated, proliferated,
or differentiated, ensuing the early tissue creation at the angles
[96]. High curvature induces mechanical stresses in cells, as
evidenced by the formation of actin stress fibres at the
tissue–fluid interface, which drive future tissue growth. Wu et al. employed a customized 3DP methodology to a
multichannel structure, that addresses the constraints of the
standard
3DP
method
(Feng
et
al.,
2017). They
were
influenced by the root of the genuine lotus plant. These
bioinspired SCF’ physicochemical properties may be tweaked. The lotus root-like form of the biomimetic materials aids
nutrition transport and oxygen distribution in the inner part
of the scaffold. Furthermore, by encouraging circulatory system
and new bone tissues to grow into the inner portion of the
biomimetic materials, this lotus root-like structure may reliably
and efficiently promote bone defect healing. The angiogenic and
osteogenic stimulatory ability of conventional 3DP materials is
lower than that of lotus root-like biomimetic materials. In comparison to another pore types, SCF with hexagonal designs
displayed the peak tensile strength anywhere at particular porosity,
leading in a substantially anisotropy morphology and enhanced oad
transmission. Because as average diameter grew, the compressive
strength of scaffolding with equal pore geometries decreased (450,
550, 900, and 1,200 lm). Sr-HT Gahnite SCF (hexagonal, 50%
porosity) have a compressive strength of 180 MPa, demonstrating
that they might be utilised to repair load-bearing bone abnormalities
(Roohani-Esfahani et al., 2016). Freeze-Drying Technique y
g
q
Using 3DP technologies and the thawing process, Xu et al. constructed porous BCR-silk composite SCF with hierarchical
pore topologies, that were helpful for osteogenesis. By freeze-
drying hierarchial composite SCF, ordered macropores of BCR
SCF and micropores of silk networks were produced. The
composite SCF have ordered hierarchical pore structures,
excellent apatite-mineralization capabilities, and mechanical
characteristics with a compressive strength of 25 MPa. The
foliage of sheet-shaped polymeric SCF generated mesopores,
and the highly porous functions as a net bag to restrict cell
leaking from the mesopores of ceramic scaffolding during cell
growth. A particular range of enhanced cell density, that is
advantageous
to
cell
development,
proliferation,
and
differentiation,
aids
intercellular
interconnections
and
paracrine communication (Xu et al., 2015). Influence of Macrostructure on Biological
Features The geometry of macropores can influence cell network creation
and tissues implantation (Figure 3) (Gariboldi and Best, 2015). The use of innovative HSP BCR SCF with customized ceramic
particles and multioriented hollow channel designs to promote December 2021 | Volume 9 | Article 754266 Frontiers in Bioengineering and Biotechnology | www.frontiersin.org 7 Development Tendency of 3D-Printed Bioceramic Scaffolds Fang et al. FIGURE 6 | Illustration showing (A) CuFeSe2 nanocrystals on BG
scaffolds and (B) their bifunction of bone tumor therapy and tissue
regeneration. Reproduced with permission from (Dang et al., 2018). FIGURE 5 | Bifunctional scaffold with nanolayer developed with 3D
Printing and dopamine modification. Reproduced with permission from (Ma
et al., 2016a). FIGURE 5 | Bifunctional scaffold with nanolayer developed with 3D
Printing and dopamine modification. Reproduced with permission from (Ma
et al., 2016a). FIGURE 6 | Illustration showing (A) CuFeSe2 nanocrystals on BG
scaffolds and (B) their bifunction of bone tumor therapy and tissue
regeneration. Reproduced with permission from (Dang et al., 2018). FIGURE 6 | Illustration showing (A) CuFeSe2 nanocrystals on BG
scaffolds and (B) their bifunction of bone tumor therapy and tissue
regeneration. Reproduced with permission from (Dang et al., 2018). FIGURE 5 | Bifunctional scaffold with nanolayer developed with 3D
Printing and dopamine modification. Reproduced with permission from (Ma
et al., 2016a). been endeavored. The following are a few instances of
common questions. bone regeneration yielded the following findings (Luo et al.,
2015). Initially, hollow channel designs induced fast Si ion
migration
from
of
the
SCF. Secondly,
the
increased
degradation of the HSP scaffolding allowed for greater area for
the establishment of new bone tissue. Thirdly, the hollow tube
assisted bone formation by acting as a conduit for oxygen and
nutrition supply as well as cell movement. In contrast to potential
osteogenic effects, the bioactive ions generated by hollow-pipe-
packed Si BCR SCF can endorse angiogenesis by promoting
endothelial cell migration. Further significantly, the hollow
pipes seem to enable the infiltrating of host blood arteries into
the hollow channels, and also the transfer of stem cells and
growth hormones, both of which help tissue regeneration (Zhang
et al., 2017b). (Figure 2). EFFECT OF MICRO/NANOSTRUCTURES
ON BONE TISSUE REGENERATION 3DP SCF offer greater interconnectedness, ordered crystallinity,
and
tensile
stability
than
SCF
made
using
conventional
techniques,
making
them
appropriate
for
bone
tissue
production. Cell leakage from the large hole in 3DP-BSCF is,
however, one of their major disadvantages. The creation of
hierarchically 3DP-BSCF was detailed in order to imitate the
structure of nature. Cell expansions can be provided with
anchored sites by the hierarchial porous structure of 3DP SCF,
enabling them to multiply as well as permeate the biological
material. In addition, the osteogenic activity and bioactivity of
hierarchically porous materials is much improved. Several
methods for producing hierarchically porous 3DP SCF have Spin Coating Technique Zhang et al. created MBG-modified b-TCP SCF with a
functioning MBG nanolayer and a hierarchy pore structure
evenly covered on the scaffolding columns using 3DP and
spin coating methods (Zhang et al., 2015). Micro defects and
grain boundaries on the strut faces of bTCP SCF were plugged
using MBG nanolayers, possibly increasing the composite SCF′
compressive strength. Luo et al. created a hierarchy SCF
comprised of an MBG/alginate composites with size-controlled
mesopores and very well planned nanochannels for drug delivery. Within that system, the MBG particulates enhanced the
mechanical properties and apatitemineralization capability of
alginate. Alginate is used for a variety of reasons, including its
high biocompatibility, biodegradation, and capacity to act as a
drug delivery system (Luo et al., 2012). Self-Assembly Method y
A hierarchy BCR scaffolding with regulated mesopores and
mussel-inspired exterior nanoparticles was produced using a
purely self approach (Xu et al., 2013). The self-assembling Ca-
P/polydopamine complex nano-layer enhanced the surface
coarseness and wettability of amalgamated SCF (Figure 4). On the one side, the presence of OH–and NH2– in the Ca-P/
polydopamine composite nanolayer may help increase the
surface hydrophilicity of composite BCRs. But at the other
side, pure b-TCP BCR SCF have a thick surface. On the other
contrary,
the
surface
of
composite
SCF
with
a
Ca-P/
polydopamine
nanolayer
is
loose,
which
increases December 2021 | Volume 9 | Article 754266 Frontiers in Bioengineering and Biotechnology | www.frontiersin.org 8 Fang et al. Development Tendency of 3D-Printed Bioceramic Scaffolds FIGURE 7 | Optical micrographs of b-TCP, b-TCP–4Fe–GO, and b-TCP–8Fe–GO scaffolds (A), b-TCP–4Fe–GO (B), and b-TCP–8Fe–GO scaffolds (C). The
fracture surfaces of b-TCP–GO (D), b-TCP–4Fe–GO (E), and b-TCP–8Fe–GO (F) scaffolds revealed sandwich-like layers in the magnetic scaffolds. FIGURE 7 | Optical micrographs of b-TCP, b-TCP–4Fe–GO, and b-TCP–8Fe–GO scaffolds (A), b-TCP–4Fe–GO (B), and b-TCP–8Fe–GO scaffolds (C). The
fracture surfaces of b-TCP–GO (D), b-TCP–4Fe–GO (E), and b-TCP–8Fe–GO (F) scaffolds revealed sandwich-like layers in the magnetic scaffolds. wettability. In contrast, calcium PHPH chelated polydopamine
nanolayers significantly improved adhesion strength, cell
proliferation, bone-related gene expression, ALP activity,
and osteogenic protein synthesis in BMSCs when comparing
to pure TCP SCF. The boosted biotic retort of cells is largely due
to the hybrid nanostructures and bioactive functional groups. Through enhancing surface quality and furthermore affecting
serum protein adherence, the nanostructure may increase cell
proliferation and differentiation, hence changing substrate–cell
interaction
and
thus
enhancing
cell
proliferation
and
differentiation (Ma et al., 2016a). illustrating
functionality
in
raising
compressive
strength,
prohibiting rapid pH increases, and balancing biodegradation. Moreover, the MCS component may load dexamethasone into
particular regions and release it at a slow rate. BTE might benefit
from the SCF (Pei et al., 2017) . Nano-Stereolithography “Cha et al. utilized the nanostereolithography (NSTL) technology
to integrate micro patterns on the scaffolding. They examined the
effects of micropillar and microridge patterns on cell adhesion,
proliferation, and osteogenic differentiation on 3Dprinted SCF. They observed that SCF with micropatterns seemed to improve
cell
adherence
when
compared
to
non-patterned
SCF. Investigators also perceived that bone-related gene expression,
such as Runx2 and ALP, was much greater in micro patterns
(micro-pillar and micro-ridge) than in non-patterned SCF(Cha
et al., 2012)”. 3-DIMENSIONAL PRINTING-B SCAFFOLD
FOR BONE TUMOR THERAPY A
significant
integer
of
3DP-BSCF
with
predetermined
configuration and construction have been created and shown
to
have
good
physicochemical
and
biotic
presentation,
suggesting that they might be used in BTE. In addition to its
use in bone tissue engineering, new functional BCR SCF have
been shown to have the capacity to treat tumors and regenerate
bone tissue at the same time. As a result, these practical SCF may
be used to mend bone defects caused by surgery while also
killing any remaining tumor cells, achieving the goal of bone
tumor treatment (Figure 5). Unlike chemo/radiotherapy, Hydration Process A
innovative
scaffolding
incorporating
calcium
sulphate
hemihydrate (CSH) and highly porous calcium Si (MCS) for
BTEwas formed utilizing 3DP and hydration processes, with CSH December 2021 | Volume 9 | Article 754266 Frontiers in Bioengineering and Biotechnology | www.frontiersin.org 9 Development Tendency of 3D-Printed Bioceramic Scaffolds Fang et al. photothermal/magnetothermal therapy has minimal adverse
effects and may selectively and efficiently ablate tumor cells
without
harming
healthy
tissue. Permanent
proteomic
degradation, cellular membranes degradation, and delayed
advanced apoptosis are all caused by high temperatures
generated by functional SCF. As a result, photothermal or
magnetothermal agents can be used locally with functional
SCF
that
have
good
photothermal
or
magnetothermal
performance. Furthermore, these functional BCR SCF have
good
biocompatibility,
encourage
bone
MSC
migration,
attachment, proliferation, and differentiation, and indorse
fresh bone creation in vivo. Some researchers have recently
concentrated on the development of such functional 3DP SCF
for use in bone tumour treatment. the photothermal impact of Fe and the Fenton reaction between
the liberated Fe ions and H2O2 in tumor cells led in synergistic
photothermal and ROS treatments for tumors. In vivo, the
inclusion of CaSiO3 in composite SCF increased the rate of
breakdown, enhanced rBMSC proliferation and differentiation,
and
encouraged
bone
formation. Such
high-compressive-
strength composite SCF can be used as adaptable biomaterials
for the repair of cortical bone defects and the treatment of bone
tumors. Magnetothermal treatment, photothermal therapy, can be
used to destroy tumor cells using functional magnetic SCF. Because of the success of creating superparamagnetic SCF, an
oscillating magnetic field can be used to destroy remaining tumor
cells locally. Zhang et al. created a magnetic scaffold using a GO-
Fe3O4-GO sandwich layer. With only a few Fe3O4 particles on
the surface of the SCF, this sandwich layer had a better
magnetothermal effect (Figure 7). Furthermore, the presence
of GO is advantageous for heat transmission due to its high
thermal conductivity, and for stimulating new bone production
(Zhang et al., 2016). GZ-modified BCR scaffolding with good photothermic upshot
was created by Wu et al. (Ma et al., 2016b). In mice, the unique
photothermal impact of useful SCF may be regulated to efficiently
ablate tumor cells and prevent tumor development. It has been
demonstrated that high temperature produced by functional SCF
significantly reduced tumor cell growth while significantly
promoting tumor cell apoptosis. Hydration Process Furthermore, due to the
bioactive groups and protein absorbance of GO, functional
SCF enhanced the osteogenic differentiation of rabbit bone
MSCs when compared to pure BCR SCF. Patients
may
develop
bacterial
infections
when
using
implanted materials, posing a significant danger. Zhang et al. created silver (Ag)/GO particle-modified b-TCP BCR SCF called
Ag@GO. The inclusion of Ag ion in the SCF resulted in
significant antibacterial activity. GO has a number of beneficial
characteristics, including conductivity, biocompatibility, and
excellent mechanical properties. They focused on its capacity
to boost BMSC angiogenesis and mass formation in this
investigation (Zhang et al., 2017c). CuFeSe2 nanocrystals may develop in situ on the strut surface
of BG SCF using the solvothermal reaction technique, resulting in
tight interactions between CuFeSe2 nanocrystals and BG SCF
(Figure 6). The narrow energy band of CuFeSe2 (0.16 eV) was
attributed to the good photothermal effect of bifunctional SCF,
leading in effective light absorption. The released Ca, Si, P, Fe, Cu,
and Se ions may work synergistically to induce rabbit bone MSC
osteogenesis. A new bifunctional scaffolding containing CuFeSe2
nanocrystals might be used as a therapeutic treatment approach
for tumor-induced bone abnormalities (Dang et al., 2018). CONCLUSION “In conclusion, when it comes to realizing extraordinary
organizational complication, adaptability, and patient-specific
needs, 3DP technology clearly outperforms older approaches. In
terms of content and hierarchical structure, this study detailed
the progress of 3DP-BSCF for presentations ranges from BTE to
bone cancer management (macros, micros, and nanos levels). We focused on the impact of scaffold composition and
hierarchical structure on physical, chemical, and biological
characteristics. 3DP SCF have a hierarchical structure that
can provide anchoring sites for cell expansions, promote cell
spread, and further govern cell network development and tissue
ingrowth. We described certain bi-functional BCR SCF in
particular that
have good photothermal or
antibacterial
performance while also promoting bone tissue regeneration. Such bi-functional BCR SCF provide up new avenues for the
treatment of bone tumors. Surely, the current 3DP technology
has
certain
drawbacks
in
terms
of
additional
needs. Processability
and
regenerability
are
critical
in
tissue
engineering, and factors including mechanical properties,
degradation rate, and scaffold precision must be continually
improved. As a result, 3DP technology’s precision and digital
operating system must be continually enhanced. In the future,
better spatial resolution than the current resolution given by
main 3DP methods will be required to create 3DP ceramic SCF Elements doped BG-ceramic (BGC) SCF demonstrated
photothermal effect and osteogenic differentiation capacity
utilizing the 3DP technique. The photothermal effect followed
this
pattern:
5Cu-BGC
is
superior. The
produced
high
temperature was efficient in killing tumour cells and inhibiting
tumor growth in vivo due to the photothermal impact of 5Cu-
BGC, 5Fe-BGC, and 5Mn-BGC SCF. Furthermore, the ionic
products
produced
by
element-doped
SCF
appeared
to
stimulate the osteogenic differentiation of bone-forming cells,
and 5Fe-BGC and 5Mn-BGC SCF enhanced the affection of
rabbit bone MSCs. The effect of 5Co-BGC SCF on ALP
expression was substantially greater than the effect of other
SCF. Furthermore, as compared to a blank control, element-
doped SCF clearly enhanced VEGF expression in rabbit bone
MSCs. According to the findings, 5Fe-BGC and 5Mn-BGC SCF
might be used in the photothermal therapy of bone tumors and
BTR (Liu et al., 2018). Synergy aimed photothermal and reactive oxygen species
(ROS) treatments outperformed solo photothermal therapy in
terms of tumor therapeutic efficacy. 3DP was used by Wu et al. to
construct CaSiO3-Fe complex SCF (Liu et al., 2018). The complex
SCF had a great tensile grift, which meant they could sustain bone
cortical defects mechanically. CONCLUSION Furthermore, in vitro and in vivo, December 2021 | Volume 9 | Article 754266 Frontiers in Bioengineering and Biotechnology | www.frontiersin.org 10 Development Tendency of 3D-Printed Bioceramic Scaffolds Fang et al. that mimic the natural macros, micros, and nanos structures of
the bone.” live organs, has a lot of potential. The greatest danger, however, is
that the cell structure will be badly disrupted during the printing
process. As a result, the major challenge to address during
material preparation is maintaining the integrity and activity
of the cells before and after printing. It will also be a significant
opportunity and challenge to use biological 3DP technology to
print
certain
bi-functional
BCR
SCF
with
outstanding
photothermal or osteogenic concert. Although current printing
methods may generate SCF with structures that are comparable to
those of many tissues, completely functional tissue engineered
SCF are still a long way off. 3DP tools is an influential inundation
that would persist to propel different areas forward. 3DP
bioactive SCF will be a great fit for bone abnormalities in the
future, and will fulfil clinical individualization criteria. 3DP will
be
increasingly
precise
and
low-cost
as
a
result
of
multidisciplinary
collaboration
and
technical
progress. Furthermore, some of them will be quite small and portable,
and our electronic equipment will be able to manage and
reassemble them”. “To avoid powdered particle aggregation, one option is to
utilize powdered granules with excellent flowability and an
appropriate binder solution. Furthermore, as an emerging
technology, a high-performance 3DP will be costly. To cut
costs, greater technical innovation and multidisciplinary
collaboration are required. 3DP SCF for BTR still require
post-treatments
like
high-temperature
sintering
or
densification. The shrinking of various portions of the SCF
during the sintering process may be uneven, causing the SCF to
fracture
and
become
useless. The
micropore
and
microstructure of 3DP SCF generated by 3DP cannot be
smaller
than
10 lm
due
to
resolution
limitations. Nanomaterials are integrated into 3DP inks or changed on
the
surface
of
3DP-BSCF
to
create
microstructures. Furthermore,
the
sintering
procedure
has
a
significant
impact
on
the
BCR
SCF′
microporosity. During
the
sintering process, the binder will be removed from the SCF,
revealing the micropores. The microporosity of sintered SCF is
determined by the binder concentration. CONCLUSION In the case of bone
tumour therapy, a compelling big animal model is required to
test the bi-functional BCR SCF’ potential to be utilized for both
bone tumor therapy and BTR in the same animal. The
following are thought to be the key issues and obstacles in
the development and implementation of 3DP biological tissue-
engineered SCF in hospital: 1) the difficulties in controlling the
accuracy
of
3DP
bioactive
SCF
based
on
medical
individualization
needs;
2)
the
vast
differences
in
mechanical properties between 3DP-B SCF and the clinical
requirements of implanted materials under various indication
conditions; 3) the difficulty in repairing large bone defects with
3D-printed
bioactive
SCF
under
various
indication
conditions”. AUTHOR CONTRIBUTIONS ZF, JC, JP, and GL collect the data, wrote the article. CZ design the
whole study. FUNDING This work was supported by the Zhejiang Provincial Science and
Technology Projects (No. LGF18H180007 to ZXF), Funds of
Traditional Chinese Medicine Science and Technology Project of
Zhejiang Province (No. 2018ZB131 to ZXF), Zhejiang Provincial
Medicine Health Science and Technology Program (2021KY021
to JHC, 2019KY011, 2020KY017 to CZ), and General Project of
Zhejiang Provincial Department of Science and Technology (B-
2017-J112 to CZ). “Biological 3DP technology is a novel approach that has just
been introduced to the 3DP sector. This new area, which entails
printing tissue-engineered SCF that sustain living cells and even REFERENCES How Useful Is SBF in Predicting In Vivo
Bone
Bioactivity? Biomaterials
27
(15),
2907–2915. doi:10.1016/
j.biomaterials.2006.01.017 Monma, H. (1988). Preparation and Properties of Porous Apatite by the Hydration
Adn Hardening Method. Gypsum Lime 212, 25–28. Kumar, A., Kargozar, S., Baino, F., and Han, S. S. (2019). Additive
Manufacturing
Methods
for
Producing
Hydroxyapatite
and
Hydroxyapatite-Based Composite Scaffolds: A Review. Front. Mater. 6,
313. doi:10.3389/fmats.2019.00313 Ni, S., Lin, K., Chang, J., and Chou, L. (2008). β-CaSiO3/β-Ca3(PO4)2 Composite
Materials for Hard Tissue Repair:In Vitro Studies. J. Biomed. Mater. Res. 85A
(1), 72–82. doi:10.1002/jbm.a.31390 Nowicki, M., Castro, N. J., Rao, R., Plesniak, M., and Zhang, L. G. (2017). Integrating
Three-Dimensional
Printing
and
Nanotechnology
for
Musculoskeletal
Regeneration. Nanotechnology 28 (38), 382001. doi:10.1088/1361-6528/aa8351 Lai, Y., Cao, H., Wang, X., Chen, S., Zhang, M., Wang, N., et al. (2018). Porous
Composite Scaffold Incorporating Osteogenic Phytomolecule Icariin for
Promoting Skeletal Regeneration in Challenging Osteonecrotic Bone in
Rabbits. Biomaterials 153, 1–13. doi:10.1016/j.biomaterials.2017.10.025 Park, J., Bauer, S., von der Mark, K., and Schmuki, P. (2007). Nanosize and Vitality:
TiO2 Nanotube Diameter Directs Cell Fate. Nano Lett. 7 (6), 1686–1691. doi:10.1021/nl070678d Lee, B.-T., Kim, K.-H., Youn, H.-C., and Song, H.-Y. (2007). Functionally Gradient
and Micro-channeled Al2O3?(t-ZrO2)/HAp Composites. J. Am. Ceram. Soc. 90
(2), 629–631. doi:10.1111/j.1551-2916.2006.01394.x Pei, P., Wei, D., Zhu, M., Du, X., and Zhu, Y. (2017). The Effect of Calcium Sulfate
Incorporation on Physiochemical and Biological Properties of 3D-Printed
Mesoporous Calcium Silicate Cement Scaffolds. Microporous Mesoporous
Mater. 241, 11–20. doi:10.1016/j.micromeso.2016.11.031 Li, X., Liu, W., Sun, L., Aifantis, K. E., Yu, B., Fan, Y., et al. (2014). Resin
Composites
Reinforced
by
Nanoscaled
Fibers
or
Tubes
for
Dental
Regeneration. Biomed. Res. Int. 2014, 542958. doi:10.1155/2014/542958 Li, X., and Feng, Q. (2005). Porous Poly-L-Lactic Acid Scaffold Reinforced by
Chitin Fibers. Polym. Bull. 54 (1), 47–55. doi:10.1007/s00289-005-0364-7 Pham, D. T., and Gault, R. S. (1998). A Comparison of Rapid Prototyping
Technologies. Int. J. Machine Tools Manufacture 38 (10), 1257–1287. doi:10.1016/s0890-6955(97)00137-5 Li, X., Sogo, Y., Ito, A., Mutsuzaki, H., Ochiai, N., Kobayashi, T., et al. (2009). The
Optimum Zinc Content in Set Calcium Phosphate Cement for Promoting Bone
Formation In Vivo. Mater. Sci. Eng. C. 29 (3), 969–975. doi:10.1016/
j.msec.2008.08.021 Qi, X., Pei, P., Zhu, M., Du, X., Xin, C., Zhao, S., et al. (2017). Three Dimensional
Printing of Calcium Sulfate and Mesoporous Bioactive Glass Scaffolds for Improving
Bone Regeneration In Vitro and In Vivo. Sci. Rep. 7, 42556. REFERENCES A Bifunctional
Biomaterial with Photothermal Effect for Tumor Therapy and Bone Regeneration. Adv. Funct. Mater. 26 (8), 1197–1208. doi:10.1002/adfm.201504142 Hao, Z., Song, Z., Huang, J., Huang, K., Panetta, A., Gu, Z., et al. (2017). The
Scaffold Microenvironment for Stem Cell Based Bone Tissue Engineering. Biomater. Sci. 5 (8), 1382–1392. doi:10.1039/c7bm00146k Ma, H., Luo, J., Sun, Z., Xia, L., Shi, M., Liu, M., et al. (2016). 3D Printing of Biomaterials
with Mussel-Inspired Nanostructures for Tumor Therapy and Tissue Regeneration. Biomaterials 111, 138–148. doi:10.1016/j.biomaterials.2016.10.005 Hwa, L. C., Rajoo, S., Noor, A. M., Ahmad, N., and Uday, M. B. (2017). Recent
Advances in 3D Printing of Porous Ceramics: A Review. Curr. Opin. Solid State. Mater. Sci. 21 (6), 323–347. doi:10.1016/j.cossms.2017.08.002 Marques, C. F., Olhero, S., Abrantes, J. C. C., Marote, A., Ferreira, S., Vieira, S. I.,
et al. (2017). Biocompatibility and Antimicrobial Activity of Biphasic Calcium
Phosphate Powders Doped with Metal Ions for Regenerative Medicine. Ceramics Int. 43 (17), 15719–15728. doi:10.1016/j.ceramint.2017.08.133 Inagaki, M., and Kameyama, T. (2007). Phase Transformation of Plasma-Sprayed
Hydroxyapatite Coating with Preferred Crystalline Orientation. Biomaterials
28 (19), 2923–2931. doi:10.1016/j.biomaterials.2007.03.008 Mathieu, L., Mueller, T., Bourban, P., Pioletti, D., Muller, R., and Manson, J. (2006). Architecture and Properties of Anisotropic Polymer Composite Scaffolds for
Bone
Tissue
Engineering. Biomaterials
27
(6),
905–916. doi:10.1016/
j.biomaterials.2005.07.015 Inzana, J. A., Olvera, D., Fuller, S. M., Kelly, J. P., Graeve, O. A., Schwarz, E. M.,
et al. (2014). 3D Printing of Composite Calcium Phosphate and Collagen
Scaffolds for Bone Regeneration. Biomaterials 35 (13), 4026–4034. doi:10.1016/
j.biomaterials.2014.01.064 Mestres, G., and Ginebra, M.-P. (2011). Novel Magnesium Phosphate Cements
with High Early Strength and Antibacterial Properties. Acta Biomater. 7 (4),
1853–1861. doi:10.1016/j.actbio.2010.12.008 Jin Woo, L., and Dong-Woo, C. (2015). 3D Printing Technology over a Drug
Delivery for Tissue Engineering. Curr. Pharm. Des. 21 (12), 1606–1617. Mirtchi, A. A., Lemaitre, J., and Terao, N. (1989). Calcium Phosphate Cements:
Study of the β-tricalcium Phosphate - Monocalcium Phosphate System. Biomaterials 10 (7), 475–480. doi:10.1016/0142-9612(89)90089-6 Kentleach, J., Kaigler, D., Wang, Z., Krebsbach, P., and Mooney, D. (2006). Coating
of VEGF-Releasing Scaffolds with Bioactive Glass for Angiogenesis and Bone
Regeneration. Biomaterials
27
(17),
3249–3255. doi:10.1016/
j.biomaterials.2006.01.033 Mondal, S., Haldar, S., Saha, P., and Ghosh, T. K. (2010). Metabolism and Tissue
Distribution of Trace Elements in Broiler Chickens’ Fed Diets Containing
Deficient and Plethoric Levels of Copper, Manganese, and Zinc. Biol. Trace
Elem. Res. 137 (2), 190–205. doi:10.1007/s12011-009-8570-z Kokubo, T., and Takadama, H. (2006). REFERENCES doi:10.1046/j.1365-2818.2000.00647.x Lu, D., Liu, Y., Li, W., Ma, H., Li, T., Ma, X., et al. (2021). Development and
Application of 3D Bioprinted Scaffolds Supporting Induced Pluripotent Stem
Cells. Biomed. Res. Int. 2021, 4910816. doi:10.1155/2021/4910816 Lu, J., Descamps, M., Dejou, J., Koubi, G., Hardouin, P., Lemaitre, J., et al. (2002). Th Bi d
d i
M
h
i
f C l i
Ph
h
Bi
i l i
B Eliaz, N., and Metoki, N. (2017). Calcium Phosphate Bioceramics: a Review of
Their History, Structure, Properties, Coating Technologies and Biomedical
Applications. Materials 10 (4), 334. doi:10.3390/ma10040334 Lu, J., Descamps, M., Dejou, J., Koubi, G., Hardouin, P., Lemaitre, J., et al. (2002). The Biodegradation Mechanism of Calcium Phosphate Biomaterials in Bone. J. Biomed. Mater. Res. 63 (4), 408–412. doi:10.1002/jbm.10259 The Biodegradation Mechanism of Calcium Phosphate Biomaterials in Bone. J. Biomed. Mater. Res. 63 (4), 408–412. doi:10.1002/jbm.10259 j
Luo, Y., Wu, C., Lode, A., and Gelinsky, M. (2013). Hierarchical Mesoporous
Bioactive Glass/alginate Composite Scaffolds Fabricated by Three-Dimensional Feng, C., Zhang, W., Deng, C., Li, G., Chang, J., Zhang, Z., et al. (2017). 3D Printing
of Lotus Root-like Biomimetic Materials for Cell Delivery and Tissue
Regeneration. Adv. Sci. 4 (12), 1700401. doi:10.1002/advs.201700401 Luo, Y., Wu, C., Lode, A., and Gelinsky, M. (2013). Hierarchical Mesoporous
Bioactive Glass/alginate Composite Scaffolds Fabricated by Three-Dimensional
Plotting for Bone Tissue Engineering. Biofabrication 5 (1), 015005. Bioactive Glass/alginate Composite Scaffolds Fabricated by Three Dimensional
Plotting for Bone Tissue Engineering. Biofabrication 5 (1), 015005. García, A., Izquierdo-Barba, I., Colilla, M., de Laorden, C. L., and Vallet-Regí, M. (2011). Preparation of 3-D Scaffolds in the SiO2-P2o5 System with Tailored
Hierarchical
Meso-Macroporosity. Acta
Biomater. 7
(3),
1265–1273. doi:10.1016/j.actbio.2010.10.006 Luo, Y., Wu, C., Lode, A., and Gelinsky, M. (2012). Hierarchical Mesoporous
Bioactive Glass/alginate Composite Scaffolds Fabricated by Three-Dimensional
Plotting for Bone Tissue Engineering. Biofabrication 5 (1), 015005. doi:10.1088/
1758-5082/5/1/015005 Gariboldi, M. I., and Best, S. M. (2015). Effect of Ceramic Scaffold Architectural
Parameters on Biological Response. Front. Bioeng. Biotechnol. 3, 151. doi:10.3389/fbioe.2015.00151 Luo, Y., Zhai, D., Huan, Z., Zhu, H., Xia, L., Chang, J., et al. (2015). Three-Dimensional
Printing of Hollow-Struts-Packed Bioceramic Scaffolds for Bone Regeneration. ACS
Appl. Mater. Inter. 7 (43), 24377–24383. doi:10.1021/acsami.5b08911 Guvendiren, M., Molde, J., Soares, R. M. D., and Kohn, J. (2016). Designing
Biomaterials for 3D Printing. ACS Biomater. Sci. Eng. 2 (10), 1679–1693. doi:10.1021/acsbiomaterials.6b00121 Ma, H., Jiang, C., Zhai, D., Luo, Y., Chen, Y., Lv, F., et al. (2016). REFERENCES Brunello, G., Sivolella, S., Meneghello, R., Ferroni, L., Gardin, C., Piattelli, A., et al. (2016). Powder-based 3D Printing for Bone Tissue Engineering. Biotechnol. Adv. 34 (5), 740–753. doi:10.1016/j.biotechadv.2016.03.009 Armulik, A., Svineng, G., Wennerberg, K., Fässler, R., and Johansson, S. (2000). Expression of Integrin Subunit β1B in Integrin β1-Deficient GD25 Cells Does
Not Interfere with αVβ3 Functions. Exp. Cel Res. 254 (1), 55–63. doi:10.1006/
excr.1999.4722 Butscher, A., Bohner, M., Hofmann, S., Gauckler, L., and Müller, R. (2011). Structural and Material Approaches to Bone Tissue Engineering in Powder-
Based Three-Dimensional Printing. Acta Biomater. 7 (3), 907–920. doi:10.1016/
j.actbio.2010.09.039 Asa’ad, F., Pagni, G., Pilipchuk, S. P., Giannì, A. B., Giannobile, W. V., and
Rasperini, G. (2016). 3D-Printed Scaffolds and Biomaterials: Review of Alveolar
Bone
Augmentation
and
Periodontal
Regeneration
Applications. Int. J. Dentistry 2016, 1239842. doi:10.1155/2016/1239842 Bădilă, A. E., Rădulescu, D. M., Niculescu, A. G., Grumezescu, A. M., Rădulescu,
M., and Rădulescu, A. R. (2021). Recent Advances in the Treatment of Bone
Metastases and Primary Bone Tumors: An Up-To-Date Review. Cancers (Basel)
13 (16), 4229. doi:10.3390/cancers13164229 Cao, H., and Kuboyama, N. (2010). A Biodegradable Porous Composite Scaffold of
PGA/β-TCP for Bone Tissue Engineering. Bone 46 (2), 386–395. doi:10.1016/
j.bone.2009.09.031 Baino, F., and Fiume, E. (2020). 3D Printing of Hierarchical Scaffolds Based on
Mesoporous Bioactive Glasses (MBGs)-Fundamentals and Applications. Materials 13 (7), 1688. doi:10.3390/ma13071688 Mesoporous Bioactive Glasses (MBGs)-Fundamentals and Applications. Materials 13 (7), 1688. doi:10.3390/ma13071688 Bohner, M., van Lenthe, G. H., Grünenfelder, S., Hirsiger, W., Evison, R., and
Müller, R. (2005). Synthesis and Characterization of Porous β-tricalcium
Phosphate
Blocks. Biomaterials
26
(31),
6099–6105. doi:10.1016/
j.biomaterials.2005.03.026 Cha, H.-S., Kim, K.-H., and Ko, Y.-J. (2012). Selecting Optimum Management
Practicesin
Pre-construction
Phase
Considering
Project
Characteristics. J. Construction Eng. Project Management 2 (1), 1–4. doi:10.6106/jcepm.2012.2.1.001 Dang, W., Li, T., Li, B., Ma, H., Zhai, D., Wang, X., et al. (2018). A Bifunctional
Scaffold
with
CuFeSe2
Nanocrystals
for
Tumor
Therapy
and
Bone
Reconstruction. Biomaterials
160,
92–106. doi:10.1016/
j.biomaterials.2017.11.020 Bose, S., Vahabzadeh, S., and Bandyopadhyay, A. (2013). Bone Tissue Engineering
Using
3D
Printing. Mater. Today
16
(12),
496–504. doi:10.1016/
j.mattod.2013.11.017 December 2021 | Volume 9 | Article 754266 Frontiers in Bioengineering and Biotechnology | www.frontiersin.org 11 Development Tendency of 3D-Printed Bioceramic Scaffolds Fang et al. De Aza, Luklinska, Martinez, Anseau, Guitian, and De Aza (2000). Morphological
and Structural Study of Pseudowollastonite Implants in Bone. J. Microsc. 197
(1), 60–67. REFERENCES I., and Reis, R. L. (2010). Vascularization in Bone Tissue Engineering:
Physiology, Current Strategies, Major Hurdles and Future Challenges. Macromol. Biosci. 10 (1), 12–27. doi:10.1002/mabi.200900107 Yves-Christian, H., Jan, W., Wilhelm, M., Konrad, W., and Reinhart, P. (2010). Net
Shaped High Performance Oxide Ceramic Parts by Selective Laser Melting. Phys. Proced. 5, 587–594. doi:10.1016/j.phpro.2010.08.086 Srivastava, A. K., Pyare, R., and Singh, S. P. (2012). In Vitro Bioactivity and
Physical-Mechanical Properties of MnO2 Substituted 45S5 Bioactive Glasses
and Glass-Ceramics. j biomat tissue engng 2 (3), 249–258. doi:10.1166/
jbt.2012.1043 Zadpoor, A. A. (2015). Bone Tissue Regeneration: the Role of Scaffold Geometry. Biomater. Sci. 3 (2), 231–245. doi:10.1039/c4bm00291a j
Sui, B.-D., Hu, C.-H., Liu, A.-Q., Zheng, C.-X., Xuan, K., and Jin, Y. (2019). Stem
Cell-Based Bone Regeneration in Diseased Microenvironments: Challenges and
Solutions. Biomaterials 196, 18–30. doi:10.1016/j.biomaterials.2017.10.046 Zhang, C.-L., Huang, T., Wu, B.-L., He, W.-X., and Liu, D. (2017). Stem Cells in
Cancer
Therapy:
Opportunities
and
Challenges. Oncotarget
8
(43),
75756–75766. doi:10.18632/oncotarget.20798 Zhang, W., Feng, C., Yang, G., Li, G., Ding, X., Wang, S., et al. (2017). 3D-printed
Scaffolds with Synergistic Effect of Hollow-Pipe Structure and Bioactive Ions
for Vascularized Bone Regeneration. Biomaterials 135, 85–95. doi:10.1016/
j.biomaterials.2017.05.005 Suping, H., Baiyun, H., Kechao, Z., and Zhiyou, L. I. (2004). Effects of Coatings on
the Mechanical Properties of Carbon Fiber Reinforced HAP Composites. Mater. Lett. 58 (27), 3582–3585. doi:10.1016/j.matlet.2004.05.086 Suwanprateeb, J., Sanngam, R., Suvannapruk, W., and Panyathanmaporn, T. (2009). Mechanical and In Vitro Performance of Apatite-Wollastonite Glass
Ceramic Reinforced Hydroxyapatite Composite Fabricated by 3D-Printing. J. Mater. Sci. Mater. Med. 20 (6), 1281–1289. doi:10.1007/s10856-009-3697-1 Zhang, Y., Chen, H., Guo, K., Zhang, X., and Eben Li, S. (2017). Electro-hydraulic
Damper for Energy Harvesting Suspension: Modeling, Prototyping and
Experimental
Validation. Appl. Energ. 199,
1–12. doi:10.1016/
j.apenergy.2017.04.085 Tamimi, F., Nihouannen, D. L., Eimar, H., Sheikh, Z., Komarova, S., and Barralet, J. (2012). The Effect of Autoclaving on the Physical and Biological Properties of
Dicalcium Phosphate Dihydrate Bioceramics: Brushite vs. Monetite. Acta
Biomater. 8 (8), 3161–3169. doi:10.1016/j.actbio.2012.04.025 Zhang, Y., Xia, L., Zhai, D., Shi, M., Luo, Y., Feng, C., et al. (2015). Mesoporous
Bioactive
Glass
Nanolayer-Functionalized
3D-Printed
Scaffolds
for
Accelerating
Osteogenesis
and
Angiogenesis. Nanoscale
7
(45),
19207–19221. doi:10.1039/c5nr05421d Tarafder, S., Dernell, W. S., Bandyopadhyay, A., and Bose, S. (2015). SrO- and
MgO-doped Microwave Sintered 3D Printed Tricalcium Phosphate Scaffolds:
Mechanical Properties and In Vivo Osteogenesis in a Rabbit Model. J. Biomed. Mater. Res. 103 (3), 679–690. REFERENCES doi:10.1002/jbm.b.33239 Zhang, Y., Zhai, D., Xu, M., Yao, Q., Chang, J., and Wu, C. (2016). 3D-printed
Bioceramic Scaffolds with a Fe3O4/graphene Oxide Nanocomposite Interface
for Hyperthermia Therapy of Bone Tumor Cells. J. Mater. Chem. B. 4 (17),
2874–2886. doi:10.1039/c6tb00390g Vaezi, M., Seitz, H., and Yang, S. (2013). A Review on 3D Micro-additive
Manufacturing Technologies. Int. J. Adv. Manufacturing Technology 67 (5-
8), 1721–1754. doi:10.1007/s00170-012-4605-2 Zhao, W., Wang, J., Zhai, W., Wang, Z., and Chang, J. (2005). The Self-Setting
Properties and In Vitro Bioactivity of Tricalcium Silicate. Biomaterials 26 (31),
6113–6121. doi:10.1016/j.biomaterials.2005.04.025 Wang, J., Sun, F., Sun, W., Shi, H., Yong, Y., Liu, S., et al. (2014). Busuishengxue Ranules Mediate Their Effects upon Non-severe Aplastic
Anemia
via
Mitogen-Activated
Protein
Kinase/extracellular
Signal-
Regulated Kinase Pathway. J. Traditional Chin. Med. 34 (1), 23–29. doi:10.1016/s0254-6272(14)60049-2 Zhu, M., Li, K., Zhu, Y., Zhang, J., and Ye, X. (2015). 3D-printed Hierarchical Scaffold
for Localized Isoniazid/rifampin Drug Delivery and Osteoarticular Tuberculosis
Therapy. Acta Biomater. 16, 145–155. doi:10.1016/j.actbio.2015.01.034 Zhu, Y., Zhang, K., Zhao, R., Ye, X., Chen, X., Xiao, Z., et al. (2017). Bone
Regeneration with Micro/nano Hybrid-Structured Biphasic Calcium Phosphate
Bioceramics at Segmental Bone Defect and the Induced Immunoregulation of
MSCs. Biomaterials 147, 133–144. doi:10.1016/j.biomaterials.2017.09.018 Wang, Y., Malcolm, D. W., and Benoit, D. S. W. (2017). Controlled and Sustained
Delivery of siRNA/NPs from Hydrogels Expedites Bone Fracture Healing. Biomaterials 139, 127–138. doi:10.1016/j.biomaterials.2017.06.001 Winkel, A., Meszaros, R., Reinsch, S., Müller, R., Travitzky, N., Fey, T., et al. (2012). Sintering of 3D-Printed Glass/HAp Composites. J. Am. Ceram. Soc. 95 (11),
3387–3393. doi:10.1111/j.1551-2916.2012.05368.x Conflict of Interest: The authors declare that the research was conducted in the
absence of any commercial or financial relationships that could be construed as a
potential conflict of interest. Woodard, J. R., Hilldore, A. J., Lan, S. K., Park, C. J., Morgan, A. W., Eurell, J. A. C.,
et
al. (2007). The
Mechanical
Properties
and
Osteoconductivity
of
Hydroxyapatite Bone Scaffolds with Multi-Scale Porosity. Biomaterials 28
(1), 45–54. doi:10.1016/j.biomaterials.2006.08.021 Publisher’s Note: All claims expressed in this article are solely those of the authors
and do not necessarily represent those of their affiliated organizations, or those of
the publisher, the editors, and the reviewers. Any product that may be evaluated in
this article, or claim that may be made by its manufacturer, is not guaranteed or
endorsed by the publisher. Wu, S.-C., Hsu, H.-C., Hsu, S.-K., Wang, W.-H., and Ho, W.-F. (2011). REFERENCES doi:10.1038/srep42556 Liu, W.-J., Tian, K., Jiang, H., Zhang, X.-S., and Yang, G.-X. (2013). Preparation of
Liquid Chemical Feedstocks by Co-pyrolysis of Electronic Waste and Biomass
without
Formation
of
Polybrominated
Dibenzo-P-Dioxins. Bioresour. Technology 128, 1–7. doi:10.1016/j.biortech.2012.10.160 Rainer, A., Giannitelli, S. M., Abbruzzese, F., Traversa, E., Licoccia, S., and
Trombetta, M. (2008). Fabrication of Bioactive Glass-Ceramic Foams
Mimicking
Human
Bone
Portions
for
Regenerative
Medicine. Acta
Biomater. 4 (2), 362–369. doi:10.1016/j.actbio.2007.08.007 Liu, Y., Li, T., Ma, H., Zhai, D., Deng, C., Wang, J., et al. (2018). 3D-printed
Scaffolds with Bioactive Elements-Induced Photothermal Effect for Bone
Tumor Therapy. Acta Biomater. 73, 531–546. doi:10.1016/j.actbio.2018.04.014 Rapacz-Kmita, A., Ślósarczyk, A., and Paszkiewicz, Z. (2006). Mechanical
Properties of HAp-ZrO2 Composites. J. Eur. Ceram. Soc. 26 (8), 1481–1488. doi:10.1016/j.jeurceramsoc.2005.01.059 December 2021 | Volume 9 | Article 754266 Frontiers in Bioengineering and Biotechnology | www.frontiersin.org 12 Development Tendency of 3D-Printed Bioceramic Scaffolds Fang et al. Roohani-Esfahani, S. I., Newman, P., and Zreiqat, H. (2016). Design and
Fabrication of 3D Printed Scaffolds with a Mechanical Strength Comparable
to Cortical Bone to Repair Large Bone Defects. Sci. Rep. 6 (1), 19468. doi:10.1038/srep19468 Xu, S., Lin, K., Wang, Z., Chang, J., Wang, L., Lu, J., et al. (2008). Reconstruction of Calvarial Defect of Rabbits Using Porous Calcium
Silicate
Bioactive
Ceramics. Biomaterials
29
(17),
2588–2596. doi:10.1016/j.biomaterials.2008.03.013 Ryu, H.-S., Hong, K. S., Lee, J.-K., Kim, D. J., Lee, J. H., Chang, B.-S., et al. (2004). Magnesia-doped HA/β-TCP Ceramics and Evaluation of Their Biocompatibility. Biomaterials 25 (3), 393–401. doi:10.1016/s0142-9612(03)00538-6 Yamada, S., Heymann, D., Bouler, J. M., and Daculs, G. (1997). Osteoclastic
Resorption of Calcium Phosphate Ceramics with Different Hydroxyapatite/
β-Tricalcium Phosphate Ratios. Biomaterials 18 (15), 1037–1041. doi:10.1016/
s0142-9612(97)00036-7 Salinas, A. J., Esbrit, P., and Vallet-Regí, M. (2013). A Tissue Engineering Approach
Based on the Use of Bioceramics for Bone Repair. Biomater. Sci. 1 (1), 40–51. doi:10.1039/c2bm00071g Yang, Y., Yang, S., Wang, Y., Yu, Z., Ao, H., Zhang, H., et al. (2016). Anti-infective
Efficacy,
Cytocompatibility
and
Biocompatibility
of
a
3D-Printed
Osteoconductive
Composite
Scaffold
Functionalized
with
Quaternized
Chitosan. Acta Biomater. 46, 112–128. doi:10.1016/j.actbio.2016.09.035 Samavedi, S., Whittington, A. R., and Goldstein, A. S. (2013). Calcium Phosphate
Ceramics in Bone Tissue Engineering: A Review of Properties and Their
Influence on Cell Behavior. Acta Biomater. 9 (9), 8037–8045. doi:10.1016/
j.actbio.2013.06.014 Yun,
H.-s.,
Kim,
S.-e.,
and
Park,
E. K. (2011). Bioactive
Glass-Poly
(ε-Caprolactone) Composite Scaffolds with 3 Dimensionally Hierarchical
Pore
Networks. Mater. Sci. Eng. C. 31
(2),
198–205. doi:10.1016/
j.msec.2010.08.020 Santos, M. REFERENCES Preparation
and Characterization of Four Different Compositions of Calcium Phosphate
Scaffolds for Bone Tissue Engineering. Mater. Characterization 62 (5), 526–534. doi:10.1016/j.matchar.2011.03.014 Xu, M., Li, H., Zhai, D., Chang, J., Chen, S., and Wu, C. (2015). Hierarchically
Porous Nagelschmidtite Bioceramic-Silk Scaffolds for Bone Tissue Engineering. J. Mater. Chem. B. 3 (18), 3799–3809. doi:10.1039/c5tb00435g Copyright © 2021 Fang, Chen, Pan, Liu and Zhao. This is an open-access article
distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the
original author(s) and the copyright owner(s) are credited and that the original
publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with
these terms. Xu, M., Zhang, Y., Zhai, D., Chang, J., and Wu, C. (2013). Mussel-inspired
Bioactive
Ceramics
with
Improved
Bioactivity,
Cell
Proliferation,
Differentiation
and
Bone-Related
Gene
Expression
of
MC3T3
Cells. Biomater. Sci. 1 (9), 933–941. doi:10.1039/c3bm60028a December 2021 | Volume 9 | Article 754266 Frontiers in Bioengineering and Biotechnology | www.frontiersin.org 13
|
https://openalex.org/W4253461400
|
https://www.jmir.org/2021/8/e26786/PDF
|
English
| null |
The Influence of Web-Based Tools on Maternal and Neonatal Outcomes in Pregnant Adolescents or Adolescent Mothers: Mixed Methods Systematic Review (Preprint)
| null | 2,020
|
cc-by
| 12,027
|
Abstract Background: Pregnant adolescent women increasingly seek support during pregnancy and the puerperium through digital
platforms instead of the traditional support system of family, friends, and the community. However, it is uncertain whether digital,
web-based tools are reliable and effective in providing information to the user on a variety of topics such as fetal development,
pregnancy outcomes, delivery, and breastfeeding to improve maternal and infant outcomes. Objective: We aimed to identify web-based tools designed to promote knowledge, attitudes, and skills of pregnant adolescents
or adolescent mothers and determine the efficacy of such web-based tools compared with conventional resources in promoting
good pregnancy and infant outcomes. Methods: A systematic search was conducted using Medline, Scopus, CINAHL, and PsycINFO for articles published from
January 2004 to November 2020 to identify randomized trials and observational studies that evaluated digital, web-based platforms
to deliver resources to pregnant adolescents. All articles written in the author’s languages were included. Two authors independently
reviewed abstracts and full-text articles for inclusion and assessed study quality. Risk of bias in each study was assessed using
appropriate tools recommended by PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-analyses) and the
Joanna Briggs Institute. We adopted a qualitative synthesis and presented the results in a narrative format due to the heterogenous
nature of the studies. Results: Seven articles met the inclusion criteria and were analyzed. The majority of the studies were graded to be of low to
moderate risk for bias. The research methodologies represented were varied, ranging from randomized (n=1) and nonrandomized
controlled trials (n=1) and prospective cohort studies (n=1) to mixed methods studies (n=1) and longitudinal surveys (n=3). Four
studies included active web-based interventions, and 3 described exposure to web-based tools, including the use of social media
and/or other internet content. Web-based tools positively influenced treatment-seeking intentions (intervention 17.1%, control
11.5%, P=.003) and actual treatment-seeking behavior for depression among postpartum adolescents (intervention 14.1%, control
6.5%, P<.001). In contrast, readily available information on the internet may leave adolescents with increased anxiety. The critical
difference lies in information curated by health care professionals specifically to address targeted concerns versus self-acquired
data sourced from various websites. Conclusions: Despite almost universal web use, few studies have used this platform for health promotion and disease prevention. Jania J Y Wu1, MBBS; Nurulhuda Ahmad2, MBBS; Miny Samuel3, MSc, PhD; Susan Logan2,4, MBChB, MD, FRCOG,
FFSRH; Citra N Z Mattar2,4, MBBS, MMED, MRANZCOG, PhD, FAMS 1Yong Loo Lin School of Medicine, National University of Singapore, Singapore, Singapore
2Department of Obstetrics and Gynaecology, National University of Singapore, Singapore, Singapore
3Research Support Unit, Yong Loo Lin School of Medicine, National University of Singapore, Singapore, Singapore
4Department of Obstetrics and Gynaecology, Yong Loo Lin School of Medicine, National University of Singapore, Singapore, Singapore Corresponding Author:
Citra N Z Mattar, MBBS, MMED, MRANZCOG, PhD, FAMS
Department of Obstetrics and Gynaecology
National University of Singapore
1E Kent Ridge Rd
Singapore
Singapore
Phone: 65 67722672
Email: citramattar@nus.edu.sg Review
The Influence of Web-Based Tools on Maternal and Neonatal
Outcomes in Pregnant Adolescents or Adolescent Mothers: Mixed
Methods Systematic Review
Jania J Y Wu1, MBBS; Nurulhuda Ahmad2, MBBS; Miny Samuel3, MSc, PhD; Susan Logan2,4, MBChB, MD, FRCOG,
FFSRH; Citra N Z Mattar2,4, MBBS, MMED, MRANZCOG, PhD, FAMS
1Yong Loo Lin School of Medicine, National University of Singapore, Singapore, Singapore
2Department of Obstetrics and Gynaecology, National University of Singapore, Singapore, Singapore
3Research Support Unit, Yong Loo Lin School of Medicine, National University of Singapore, Singapore, Singapore
4Department of Obstetrics and Gynaecology, Yong Loo Lin School of Medicine, National University of Singapore, Singapore, Singapore
Corresponding Author:
Wu et al
JOURNAL OF MEDICAL INTERNET RESEARCH Wu et al JOURNAL OF MEDICAL INTERNET RESEARCH https://www.jmir.org/2021/8/e26786 J Med Internet Res 2021 | vol. 23 | iss. 8 | e26786 | p. 1
(page number not for citation purposes) Introduction Pregnant adolescents are an especially vulnerable population. Despite the significant decline of adolescent pregnancies in
recent decades [1], the World Health Organization estimates
that 12 million girls aged 15 to 19 years give birth yearly in
developing countries [2]. These adolescents are at increased
risk of prenatal and perinatal complications including gestational
hypertension, preterm delivery, low infant birth weight, and
other neonatal complications [3-5]. Adolescent pregnancies are
more
prevalent
in
socioeconomically
disadvantaged
communities, on a background of disrupted family structures
and limited educational opportunities [6]. Poor pregnancy
outcomes are more frequent among socioeconomically
disadvantaged adolescents, largely due to the complex social
and cultural factors that result in lower or delayed maternal
engagement with health care services [7,8]. While the definition of social media is dynamic and constantly
evolving [20], in general, a social media site is an interactive
online platform that facilitates the exchange of user-generated
content [21,22]. In our study, we defined social media as any
online platform that enables users to exchange content (eg,
Facebook or Instagram), while internet content was defined as
any online platform that does not facilitate content exchange
among users (eg, websites or online reading materials) [21,22]. We defined web-based tools broadly as describing all services
and technologies found on online platforms and consider both
social media and internet content to be subsets of web-based
tools [23]. Despite their convenience and easy access [24], these
tools have limitations. Information found online may not be
verified and may provide pregnant women with inaccurate,
unreliable, or unsupported knowledge [25]. A meta-analysis
evaluating the quality of online health information found that
70% of the included studies concluded that information sources
on the internet were of low quality [26] and often provided
advice with limited or no scientific evidence [27-29]. Specifically
regarding
women’s
health,
inaccurate
celebrity-based advice has been highlighted [30]. Low-quality
pregnancy-related information may be harmful or conflicting
[29] and is often not discussed with health care providers to
clarify misconceptions [24,31], all of which have the potential
to negatively influence pregnancy outcomes. The unregulated
online community can also produce negative experiences for
the naïve user [32]. As the use of social media or internet
resources during pregnancy is a relatively recent phenomenon,
there is an opportunity to explore the association between its
use among at-risk adolescents and perinatal outcomes. The transition from child-free adolescence to motherhood is
daunting. JOURNAL OF MEDICAL INTERNET RESEARCH Wu et al programs. The vastness of the information available on the web limits the ability of health care professionals to monitor or control
sources of information sought by patients. Thus, it is important to create professionally curated platforms to prevent or limit
exposure to potentially misleading or harmful information on the internet while imparting useful knowledge to the user. Trial Registration:
PROSPERO International Prospective Register of Systematic Reviews CRD42020195854;
https://www.crd.york.ac.uk/prospero/display_record.php?RecordID=195854 KEYWORDS pregnancy in adolescence; teenagers; adolescents; pregnancy; postpartum; internet; digital health; digital media; new digital
media; eHealth; social media; social network; communications media pregnancy in adolescence; teenagers; adolescents; pregnancy; postpartum; internet; digital health; digital media; new digital
media; eHealth; social media; social network; communications media including fetal development, delivery, and breastfeeding [16]. As such, information found on the internet has the potential to
influence the mother’s decisions surrounding her pregnancy
care [19]. Abstract Social media interventions or web-based tools have the potential to positively influence both maternal and infant outcomes in
adolescent pregnancy, but there is a need for more well-conducted studies to demonstrate the effectiveness of these support J Med Internet Res 2021 | vol. 23 | iss. 8 | e26786 | p. 1
(page number not for citation purposes) XSL•FO
RenderX XSL•FO
RenderX JOURNAL OF MEDICAL INTERNET RESEARCH Wu et al J Med Internet Res 2021 | vol. 23 | iss. 8 | e26786 | p. 2
(page number not for citation purposes) https://www.jmir.org/2021/8/e26786 Eligibility Criteria All inclusion and exclusion criteria were defined a priori. We
included both quantitative and qualitative studies that explored
the association between the use of web-based tools by pregnant
adolescents or adolescent mothers and maternal and neonatal
outcomes. Studies that defined their population as adolescents
or studied women aged 21 years and younger were included. Social media was defined as online platforms providing avenues
to exchange content with other users (eg, Facebook, Instagram,
Twitter, blogs, vlogs, forums, chatrooms) while internet content
was defined as online platforms that did not provide direct
opportunities to interact with other users (eg, websites, online
reading materials, internet programs). Conventional tools
developed to support adolescent mothers such as brochures,
school-based counseling, community-based counseling, and
group and personal counseling were the comparison of interest. Maternal outcomes measured were physical (nutrition, physical Search Strategy The protocol was registered with the International Prospective
Register
of
Systematic
Reviews
(PROSPERO)
[CRD42020195854]. We followed the Preferred Reporting
Items for Systematic Reviews and Meta-analyses (PRISMA)
guidelines [33] and conducted a systematic search of PubMed,
Scopus, CINAHL, and PsycINFO electronic databases for
articles published from January 2008 to November 2020. We
restricted publications to the last 12 years to ensure studies
promoted updated practices relating to pregnancy care. The
search was initially conducted on May 15, 2020, and updated
on December 5, 2020. The process of updating the search was
guided by methods described by Bramer et al [34]. We also
conducted a secondary search to identify studies published
between 2004 and 2007 as the first concept of Web 2.0, which
is defined as a network platform that spans across all devices,
was introduced in 2004 [35]. No further studies met our
inclusion criteria. Two librarians from the National University
of Singapore Medical Library were consulted on the finalization
of the search strategy. Search terms included “pregnant,”
“adolescents,” “social media,” and “internet,” and the full search
strategy can be found in Multimedia Appendix 1. Selection of Studies The study selection was conducted in two phases. During level
1 screening, two authors (JW, NA) independently screened all
studies retrieved by electronic database searches based on key
terms and resolved discrepancies by discussion with a third
author until a list of studies for level 2 screening was agreed
upon. During level 2 screening, the full texts of studies selected
in level 1 were retrieved and independently reviewed by the
same 2 authors to determine the eligibility of each study;
discrepancies were again resolved by discussion with the third
author until a final list of studies was agreed on. The reasons
for exclusion were coded and recorded systematically. JOURNAL OF MEDICAL INTERNET RESEARCH JOURNAL OF MEDICAL INTERNET RESEARCH Wu et al •
How effective are these web-based tools in promoting good
pregnancy and infant outcomes compared with conventional
resources? activity, breastfeeding practices, birth complications, and risky
behaviors such as smoking and alcohol consumption) and
psychosocial (mental health, depression, anxiety, loneliness and
stress, self-esteem, birth preparedness, and parenting outcomes)
factors. Infant or child outcomes included preterm birth, low
birth weight, sudden infant death syndrome, and obesity. All
outcomes collected met the inclusion criteria. We included
studies with or without a comparison group that were relevant
to answering our research questions and excluded reviews,
abstracts, conference proceedings, letters, editorials, comments,
opinions, and book chapters. We excluded studies that were not
in English, Chinese, Malay, or French, the languages of the
authors. Studies that only examined the use of social media or
internet content with no quantitative or qualitative outcomes
were also excluded. Risk of Bias Assessment We conducted a mixed methods systematic review to assess the
relevant studies using various critical appraisal tools that are
validated and widely used. Qualitative studies [36] and
nonrandomized controlled trials [37] were evaluated using the
Joanna Briggs Institute instrument, which appraises articles as
“included,” “excluded,” or “seek further info” based on 10- and
9-question checklists, respectively. Cohort or case-control
studies were evaluated using the Newcastle-Ottawa Quality
Assessment Scale [38], which rates articles according to
selection, comparability, and exposure categories using a star
system. Randomized controlled trials (RCT) were evaluated
using the Cochrane Risk of Bias 2 describing a “fixed set of
domains of bias, focusing on different aspects of trial design,
conduct, and reporting” [39]. It appraises articles as low risk of
bias, some concerns, and high risk of bias. Studies using mixed
methodologies were evaluated using the Mixed Methods
Appraisal Tool [40] based on a qualitative and quantitative
scoring system. Four authors (JW, NA, SL, CM) independently
and critically appraised each study for quality and potential bias. We resolved discrepancies by discussion until a consensus was
reached. Data Extraction We extracted relevant evidence using a standard proforma
including study design, settings, observational period, sample
size, participant characteristics, description of intervention and
comparison tools, maternal outcomes, infant or child outcomes,
adjusted factors, findings, and limitations. J Med Internet Res 2021 | vol. 23 | iss. 8 | e26786 | p. 3
(page number not for citation purposes) https://www.jmir.org/2021/8/e26786 Introduction Traditionally, those who are pregnant or postpartum
turn to their family, friends, and partners for support [9]. More
recently, community-based and home-visit programs have also
been established to support adolescent mothers. Home-visit
programs may offer better outcomes in adolescents who are
difficult to engage due to close bonds formed [10]. These
programs aim to provide access to information, resources, and
social support in order to maximize coping strategies and
eventual reintegration into society [11,12]. The Teenage
Pregnancy Strategy is an example of a successful,
multicomponent intervention that has reduced adolescent
conceptions and improved outcomes for adolescent mothers by
providing support for mothers targeted at completion of
education and securing appropriate housing [13]. With increasing access to technology, expectant mothers may
seek pregnancy-related information or support from social media
[14,15] or internet-based platforms [16,17] instead of traditional
sources. Due to shifts in contemporary social structure, many
women find themselves geographically and emotionally isolated
from their support system of family and friends [14,15]. Alternative support systems on digital platforms provide
opportunities for like-minded women to exchange experiences
and gain social support. Web-based support systems are
available regardless of time and location and allow for
anonymity of use, reducing stigma and facilitating the discussion
of sensitive topics [18]. Furthermore, the majority of expectant
mothers perceive the internet to be a reliable source of
information and access information on a large variety of topics The aims of this systematic review were to assess the impact
of web-based tools used by pregnant adolescents or adolescent
mothers on maternal and infant outcomes to compare these to
conventional resources and critically appraise the evidence from
relevant quantitative and qualitative studies. The research
questions addressed were as follows: •
What types of available web-based tools are designed to
promote knowledge, attitudes, and skills of pregnant
adolescents or adolescent mothers? •
What types of available web-based tools are designed to
promote knowledge, attitudes, and skills of pregnant
adolescents or adolescent mothers? J Med Internet Res 2021 | vol. 23 | iss. 8 | e26786 | p. 2
(page number not for citation purposes) J Med Internet Res 2021 | vol. 23 | iss. 8 | e26786 | p. 2
(page number not for citation purposes) XSL•FO
RenderX XSL•FO
RenderX Wu et al Synthesis of Results We categorized included studies into 2 groups to answer our
research questions. To explore the variety and effectiveness of
web-based tools in promoting knowledge, we analyzed and
compared all 7 studies [41-47]. To determine their influence on
pregnancy outcomes compared to conventional resources, we
assessed 2 studies [41,42]. The studies analyzed were too
heterogeneous to enable a formal meta-analysis, thus we adopted
a qualitative synthesis and presented the results in a narrative
format. J Med Internet Res 2021 | vol. 23 | iss. 8 | e26786 | p. 3
(page number not for citation purposes) XSL•FO
RenderX XSL•FO
RenderX JOURNAL OF MEDICAL INTERNET RESEARCH Wu et al Wu et al removing the duplicates, 4722 records were eligible for further
screening. After 2 reviewers screened all titles and abstracts,
135 records met the inclusion criteria and were eligible for
further full-text screening. Following the full text review, we
excluded 128 articles based on the predefined exclusion criteria,
and 7 articles were finally included in the systematic review for
analysis (Figure 1). Search Results The search in the various medical databases (Medline [using
PubMed platform], Scopus, CINAHL [using EbscoHost
platform], and PsycINFO) for articles published from January
2008 to November 2020 yielded a total of 6157 records. After Figure 1. PRISMA flow diagram. id
d
d l
t b
th
th
[42 46]
M Figure 1. PRISMA flow diagram. considered an adolescent by the authors [42,46]. Most
participants were defined as adolescents in 6 studies
[41,42,44-47], while 1 study defined participants as first-time
mothers [43]. Participants were currently pregnant (n=1) [44],
already mothers (n=5) [41-43,45,47], or both (n=1) [46]. Various
methods were used to encourage participation: 5 studies
recruited participants from established institutions such as
prenatal clinics or organizations offering services for adolescents
[41,42,44,46,47], while the remaining 2 studies recruited
participants by posting public advertisements or through
personal and professional contacts [43,45]. Two studies included
a control group, which received conventional care [41,42]. https://www.jmir.org/2021/8/e26786 J Med Internet Res 2021 | vol. 23 | iss. 8 | e26786 | p. 4
(page number not for citation purposes) JOURNAL OF MEDICAL INTERNET RESEARCH Table 1. Descriptive characteristics. Key findings
Outcomes evaluated
Control group (if
any)
Intervention/expo-
sure
Participants
Study design
Author,
country
The NMN website is wel
poised for nursing-driven
Maternal:
n=19; usual care:
hospital parenting
n=15; NMNb web-
site: internet-based
Adolescents: 16-21
years (mean 18.3 [SD
1.7] years); 1-week
Quantitative;
RCTa
Hudson
et al [41],
US
•
Mental health
instructions, par-
ent’s own re-
sources
social support intervention
The social support compo
nent was identified as a
key strength with positiv
qualitative comments. education resource,
discussion forum,
direct email contact
with nurses (6
months)
•
Parenting outcomes
postpartum; single,
low-income, African
Americans
•
Birth complications
Infant:
•
Health care use
•
Breastfeeding
The internet intervention
was successful in changin
Maternal:
n=138 (mean
18.2 [SD 1.9]
n=154 (mean 17.9
[SD 2.1] years); in-
Adolescents: 13-21
years; up to 1-year
Quantitative;
matched
Logsdon
et al [42],
US
•
Mental health
years); home visi-
tation program
ternet intervention
website: internet-
based education re-
sources (2 weeks)
postpartum; living in
urban, suburban and ru-
ral counties, mixture of
White (8.6%), Black
(88.0%) and others
(3.4%)
prospective
cohort study
attitudes, perceived con-
trol, intention to seek
treatment, and actually
seeking treatment. The in
tervention effect was equa
in adolescents regardless
f
h
th
li
d b t The NMN website is well
poised for nursing-driven
social support intervention. The social support compo-
nent was identified as a
key strength with positive
qualitative comments. The internet intervention
was successful in changing
attitudes, perceived con-
trol, intention to seek
treatment, and actually
seeking treatment. The in-
tervention effect was equal
in adolescents regardless
of where they lived, but
the impact on changing at-
titudes may be dose depen-
dent. n=7; personal elec-
tronic media use:
First-time mothers;
mean 18-21 [SD 19.5]
Qualitative
Fleming
et al [43],
US
websites, internet
blogs, internet chat
years); 6-8 weeks post-
partum; single (85.7%),
low-income
rooms, online shows
(duration not speci-
fied) This study demonstrated
adolescents’ desire and
need for clear, accurate,
and easily accessible infor-
mation about birthing. Providing credible electron-
ic sources will educate the
mothers and increase their
confidence and birthing
preparedness levels. Adolescents: 13-21
years; up to 1-year
Quantitative;
matched
Logsdon
et al [42],
US
postpartum; living in
urban, suburban and ru-
ral counties, mixture of
White (8.6%), Black
(88.0%) and others
(3.4%)
prospective
cohort study Maternal:
•
Birth preparedness
•
Mental health J Med Internet Res 2021 | vol. 23 | iss. 8 | e26786 | p. 5
(page number not for citation purposes) Study Characteristics Characteristics of the 7 studies are summarized in Table 1. Six
were conducted in the United States [41-44,46,47], and one was
conducted
in
Western
Australia
[45]. The
research
methodologies represented were varied—the most common
study design was qualitative (n=3) [43,45,46], while the
remaining studies were an RCT [41], a non-RCT [47], a
prospective cohort study [42], and a mixed methods study [44]. Study sizes ranged from 7 [43,45] to 292 participants [42]. Despite limiting our inclusion criteria to adolescents, the
participant age range was large, from age 13 years [42] to a
participant aged 22 years, who was included as she was https://www.jmir.org/2021/8/e26786 J Med Internet Res 2021 | vol. 23 | iss. 8 | e26786 | p. 4
(page number not for citation purposes) XSL•FO
RenderX JOURNAL OF MEDICAL INTERNET RESEARCH JOURNAL OF MEDICAL INTERNET RESEARCH Wu et al JOURNAL OF MEDICAL INTERNET RESEARCH p p
Poor diet quality persists
among both adolescent and
Maternal
n=12 (median
29.2 [IQR 23.7,
n=10 (median 17.0
[IQRd 16.4, 17.7]
years);
social media inter-
vention: private
Adolescents: 14-18
years; 12-28 weeks
pregnant; low-income,
mixture of Black
(70%), Hispanic White
(20%), and non-Hispan-
ic White (10%)
Mixed meth-
ods; non-
RCT
Vander
Wyst et
al [44],
US
•
Physical anthropo-
metric data
33.8] years);
adult partici-
pantse
adult low-income pregnant
women. Although social
media-based education
was well received by the
participants, this did not
result in significant
•
Nutrition knowl-
edge
•
Nutrition behavior
Facebook group
with interactive activ-
ities and dissemina-
•
Physical activity
•
Attitudes and be-
liefs on prenatal
changes in dietary intake
and knowledge. tion of health infor-
mation (18 weeks)
health
•
Social support
Infant
•
Birth weight
•
Gestational age
•
Breastfeeding
The use of social network
sites affords adolescent
Maternal:
—
n=7; personal social
network site use:
Adolescents; 16-19
years; 3-17 months
Qualitative
Nolan et
al [45],
WAf
•
Social support
website that enables
users to create pub-
postpartum; single, liv-
ing with parents
mothers access to tangible,
informational, and emotion-
•
Mental health
•
Parenting outcomes
al support. There is a poten-
lic profiles and form
(71.4%), extended rela-
relationships with
tives (14.3%), or part-
ner/friend (14.3%)
tial role for midwives to
use such platforms to pro-
vide additional social sup-
port. other users (duration
not specified) Poor diet quality persists
among both adolescent and
adult low-income pregnant
women. Although social
media-based education
was well received by the
participants, this did not
result in significant
changes in dietary intake
and knowledge. Qualitative
Nolan et
al [45],
WAf The use of social network
sites affords adolescent
mothers access to tangible,
informational, and emotion-
al support. There is a poten-
tial role for midwives to
use such platforms to pro-
vide additional social sup-
port. J Med Internet Res 2021 | vol. 23 | iss. 8 | e26786 | p. 5
(page number not for citation purposes) https://www.jmir.org/2021/8/e26786 https://www.jmir.org/2021/8/e26786 XSL•FO
RenderX JOURNAL OF MEDICAL INTERNET RESEARCH Wu et al JOURNAL OF MEDICAL INTERNET RESEARCH Key findings
Outcomes evaluated
Control group (if
any)
Intervention/expo-
sure
Participants
Study design
Author,
country
The use of technology
among adolescent mothers
living in foster homes is
associated with multiple
social issues. bNMN: New Mothers Network. cNot applicable. dIQR: interquartile range. eControl group (adult participants) is not relevant to answering the research question. fWA: Western Australia. et al [47]. The earlier study in 2013 was conducted as a pilot to
test the prototype of a web-based intervention. Although
included, it was rated as a poor-quality study due to the lack of
a control group, the one-off nature of outcome sampling, lack
of meaningful clinical outcome, and an inadequate description
of statistical analysis used. This was subsequently followed by
a more robust prospective cohort study in 2018 [42]. Risk of Bias Within Studies Table 2 illustrates an overview of the studies’ risk of bias. A
detailed assessment of the relevant studies using various critical
appraisal tools is found in Multimedia Appendix 2, as the criteria
for each study differed by study design. All 7 studies were
included, and 6 were judged to have a low risk of bias [41-46]. Of note, 2 out of 7 studies [42,47] were conducted by Logsdon Table 2. Overview of the studies’ risk of bias. Rating
Quality assessment instrument
Study design
Author
Include; risk of bias: low
Cochrane Risk of Bias 2
Quantitative; RCTa
Hudson et al [41]
Include; selection: ***; comparability:
**; outcome: **; risk of bias: low
Newcastle-Ottawa Quality Assessment
Scale (Cohort studies)
Quantitative; matched prospective
cohort study
Logsdon et al [42]
Include; risk of bias: low
MMATb
Mixed methods; non-RCT
Vander Wyst et al [44]
Include; risk of bias: moderate-high
JBIc (quasi-experimental studies)
Quantitative; non-RCT
Logsdon et al [47]
Include; risk of bias: low
JBI (qualitative research)
Qualitative
Fleming et al [43]
Include; risk of bias: low
JBI (qualitative research)
Qualitative
Nolan et al [45]
Include; risk of bias: low
JBI (qualitative research)
Qualitative
Rueda et al [46] Table 2. Overview of the studies’ risk of bias. aRCT: randomized controlled trial. bMMAT: Mixed Methods Appraisal Tool. cJBI: Joanna Briggs Institute. [41,44], and the remaining 2 studies described purely internet
content [42,47]. The adolescents in the remaining 3 studies
[43,45,46] were exposed to various web-based tools. Of these
studies, the duration of exposure to the web-based tools was
not known for all [43,45,46], and resources accessed by the J Med Internet Res 2021 | vol. 23 | iss. 8 | e26786 | p. 6
(page number not for citation purposes) https://www.jmir.org/2021/8/e26786 JOURNAL OF MEDICAL INTERNET RESEARCH Technology
should be included in vari-
ous models of care to in-
crease understanding be-
tween professionals and
adolescents. Maternal:
•
Relationship with
intimate partners
•
Mental health
Infant:
•
Child protection
—
n=13; personal elec-
tronic media use: so-
cial media websites,
phone apps that facil-
itate communication
between individuals
(duration not speci-
fied)
Adolescents; 14-22
years; currently preg-
nant or mothers; single,
living in residential fos-
ter care home, mixture
of ethnic minorities:
Hispanic (43.5%),
Black (30.4%), Mixed
race (10.9%)
Qualitative
Rueda et
al [46],
US
The testing of a prototype
website for adolescent
mothers with postpartum
depression showed
promising results. Atti-
tudes related to depression
and seeking treatment im-
proved after viewing the
website. Maternal
•
Mental health
—
n=138; internet inter-
vention website: in-
ternet-based educa-
tion resources (sin-
gle class period)
Adolescents; mean 16.8
years; mothers; single,
students of a public
school–based program
for adolescent parents;
mixture of African
American (48.6%),
White (34.1%), and
others (17.3%)
Quantitative;
non-RCT
Logsdon
et al [47],
US The use of technology
among adolescent mothers
living in foster homes is
associated with multiple
social issues. Technology
should be included in vari-
ous models of care to in-
crease understanding be-
tween professionals and
adolescents. Maternal
•
Mental health The testing of a prototype
website for adolescent
mothers with postpartum
depression showed
promising results. Atti-
tudes related to depression
and seeking treatment im-
proved after viewing the
website. aRCT: randomized controlled trial. bNMN: New Mothers Network. cNot applicable. dIQR: interquartile range. eControl group (adult participants) is not relevant to answering the research question. fWA: Western Australia. aRCT: randomized controlled trial. Synthesis of Results The study results are summarized in Tables 3 and 4. Adolescents
in 4 studies [41,42,44,47] received an active intervention. Of
these studies, 1 included adolescent–health care professional
interactions [41], 2 described adolescent-adolescent interactions https://www.jmir.org/2021/8/e26786 XSL•FO
RenderX XSL•FO
RenderX JOURNAL OF MEDICAL INTERNET RESEARCH Wu et al Wu et al available), and general anthropometric data. The outcomes are
discussed as maternal and infant outcomes. We further
categorized maternal outcomes into physical, mental well-being,
parenting outcomes, and others. adolescents include informative [43] and social media websites
[43,45,46]. Various outcomes were collected during and after
the intervention, measured via self-reports, postintervention
surveys, corroboration with medical records, validated tools (if adolescents include informative [43] and social media websites
[43,45,46]. Various outcomes were collected during and after
the intervention, measured via self-reports, postintervention
surveys, corroboration with medical records, validated tools (if Table 3. Synthesis of quantitative results. Table 3. Synthesis of quantitative results. Table 3. Synthesis of quantitative results. Synthesis of Results Non–statistically significant trends
following intervention
Statistically significant outcomes with intervention (P<.05, unless stated otherwise)
Author, country
Increasing:
Assuming α=.10, P<.10;
Hudson et al
[41], US
•
•
Social support
Intervention group had lower self-esteem than control group at 6 months; scale: RSEa
No differences in:
•
Intervention group had higher levels of perceived competence after 6 months; scale:
PPSb
•
Depression symptoms
•
Intervention group had higher parenting satisfaction levels after 6 months; scale:
WPBL-Rc
•
Loneliness
•
Perceived stress
•
ERd use reduced >50% in intervention group compared to control group (35.7% vs
70.6%); data collection: questionnaire
•
Birth complications
•
Intervention group was less likely to exclusively breastfeed compared to control group;
data collection: questionnaire
No differences in:
Logsdon et al
[42], US
•
Intervention group had more positive attitudes toward seeking psychological help
than the control group after 2 weeks; scale: ATSPHe
•
Depression symptoms
•
Intervention group had more positive perceived behavior control than the control
group after 2 weeks; scale: HSDIf
•
Stigma for receiving psychologi-
cal help
•
Intervention group had greater intention to seek treatment for depression than the
control group after 2 weeks; scale: MHIg
•
Intervention group had higher treatment seeking behavior for depression than the
control group after 2 weeks; data collection: questionnaire
No differences in:
Vander Wyst et
al [44], US
•
There was higher sugar intake in both groups after 18 weeks compared to baseline;
data collection: 24-hour diet recall calculated via FPPh
•
Participant anthropometric data
•
There was a lower likelihood of adolescents cooking at home at baseline compared
to adults; data collection: questionnaire
•
Mean caloric consumption
•
Macronutrient distribution of food
•
There was a lower likelihood of adolescents buying their own groceries at baseline
and after 18 weeks compared to adults; data collection: questionnaire
•
Infant birth weight
•
Infant gestational age
•
Adolescents were less knowledgeable in nutrition (eg, identifying fiber rich food,
recommended whole grain consumption, fruit, vegetable and fat intake) compared to
adults at baseline and/or after 18 weeks; data collection method: questionnaire
No differences in:
Logsdon et al
[47], US
•
Adolescents had more positive attitudes toward seeking psychological help postinter-
vention compared to baseline; scale: ATSPH
•
Mental health acceptability
•
Stigma for receiving psychologi-
cal help aRSE: Rosenberg Self-Esteem. bPPS: How I Deal With Problems Regarding Care of My Baby. J Med Internet Res 2021 | vol. 23 | iss. 8 | e26786 | p. 7
(page number not for citation purposes) Synthesis of Results Drawbacks were the absence of adequate privacy controls and negative comments that could potentially threaten
emotional well-being
•
Increased parenting confidence: peer support and positive affirmations significantly increased adolescents’ confidence levels
Rueda et al
[46], US
•
Social media tools provided positive experiences in:
•
Interacting with a potentially intimate partner
•
Maintaining contact and fostering feelings of closeness with their child’s father
•
Social media tools provided negative experiences in:
•
Unwanted sexual solicitations
•
Child protection, as meetings with strangers in offline spaces place both the adolescent and their children at risk
•
Cyber abuse (eg, cyber bullying, stalking) of which adolescents were both victims and perpetuators
•
Adverse emotional side effects fueled by jealousy and mistrust g
Increased parenting confidence: peer support and positive affirmations significantly increased adolescents’ confidence level •
Maintaining contact and fostering feelings of closeness with their child’s father •
Child protection, as meetings with strangers in offline spaces place both the adolescent and their children at ris
•
Cyber abuse (eg, cyber bullying, stalking) of which adolescents were both victims and perpetuators
•
Adverse emotional side effects fueled by jealousy and mistrust p
g
g
p
p
•
Cyber abuse (eg, cyber bullying, stalking) of which adolescents were both victims and perpetuators
•
Adverse emotional side effects fueled by jealousy and mistrust aWA: Western Australia. psychological help (P=.04) were successfully replicated, with
additional data demonstrating that perceived behavior control
(P=.007), intention to seek treatment for depression (P=.003),
and seeking treatment for depression (P<.001) were significantly
higher after 2 weeks in the intervention group compared to the
controls. Data was measured using the Attitude Toward Seeking
Psychological Help scale, Health Self Determination Index
scale, Mental Health Intention scale, and a questionnaire on the
outcomes of seeking treatment, respectively. Fleming et al [43]
conducted qualitative interviews of participants who had
prepared for childbirth through a variety of web-based tools
including websites, blogs, chatrooms, and mass media. The
findings suggested that access to electronic media did not
necessarily prepare adolescents for pregnancy and childbirth
but instead increased anxiety levels. Nolan et al [45] also
conducted qualitative interviews on participants who used social
media and reported that communicating with other adolescent
mothers via social network sites helped reduce loneliness,
parental stress, and anxiety. Synthesis of Results cWPBL-R: What Being the Parent of a Baby is Like–Revised. dER: emergency room. eATSPH: Attitude Toward Seeking Psychological Help. fHSDI: Health Self Determination Index. gMHI: Mental Health Intention. hFPP: Food Processor Program. eATSPH: Attitude Toward Seeking Psychological Help. fHSDI: Health Self Determination Index. gMHI: Mental Health Intention. hFPP: Food Processor Program. J Med Internet Res 2021 | vol. 23 | iss. 8 | e26786 | p. 7
(page number not for citation purposes) J Med Internet Res 2021 | vol. 23 | iss. 8 | e26786 | p. 7
(page number not for citation purposes) https://www.jmir.org/2021/8/e26786 https://www.jmir.org/2021/8/e26786 JOURNAL OF MEDICAL INTERNET RESEARCH Wu et al Table 4. Synthesis of qualitative results. Outcomes
Study, country
Fleming et al
[43], US
•
Increased anxiety due to graphic media, birthing process, potential complications, and neonatal care
•
Birth preparedness: suboptimal birth preparedness due to fragmented, inconsistent, weakly linked, and poorly referenced in-
formation although a small subset of women developed improved or enhanced understanding
•
Social support: platform allowed connection with others and peer support
Vander Wyst et
al [44], US
•
Nutrition behavior: adolescents had improved attitudes toward nutrition with dietary changes (eg, limiting high fat fast food,
increasing vegetable and fruit intake), motivated by time, convenience, and food preferences
•
Physical activity: adolescents had an increased tendency to exercise during pregnancy as they believed it to help with labor
•
Breastfeeding: adolescents tended to be less likely to breastfeed compared to adults
•
Social support: both adolescents and adults had both good and poor sources of social support
Nolan et al [45],
WAa
•
Increased social support and connectedness: participants had unlimited access to relationships, minimizing feelings of exclusion,
and social isolation. They could maintain both old and new friendships. Social network sites provide valuable tangible, emo-
tional, and informational support for adolescent mothers, contributing to mothers’ social capital
•
Parental stress and anxiety: social network sites served as a medium for problem sharing and helped to reduce parental stress
and anxiety. J Med Internet Res 2021 | vol. 23 | iss. 8 | e26786 | p. 8
(page number not for citation purposes) https://www.jmir.org/2021/8/e26786 Maternal Outcomes Maternal mental well-being was explored in 5 studies
[41-43,45,47]. Hudson et al [41] conducted an RCT comparing
participants who were exposed to the New Mothers Network
website with those who received typical parenting instructions
provided by the hospital. The New Mothers Network
intervention provided parenting information through their
electronic library and via communications with other mothers
and nurses. The authors found that self-esteem levels measured
via the Rosenberg Self-Esteem scale were significantly lower
in the intervention group over the 6-month period (P=.04),
although the authors found no clear explanation for this trend. In the 2 studies conducted by Logsdon et al [42,47], the earlier
prospective pilot study [47] explored the efficacy of a website
showcasing pictures and stories of other adolescent mothers’
experiences, county and national mental health resources, and
a frequently asked questions segment. The evidence from this
study suggests that attitudes toward seeking psychological help,
measured via the Attitude Toward Seeking Psychological Help
scale, were significantly higher postintervention (P=.02). A
similar result was observed in their prospective cohort study
[42] of participants who received exposure to the website and
home visits by volunteers as part of a home visitation program. Compared to the earlier pilot study [47], the intervention was
adapted slightly to showcase video vignettes instead of pictures
and stories based on user feedback. Mental health community
resources and the frequently asked questions segment remained. The earlier results of improved attitudes toward seeking Only one study by Vander Wyst et al [44] explored maternal
physical outcomes. In this 18-week longitudinal study,
participants were added into a private Facebook group where
health information (pregnancy fitness, healthy recipes, nutrition,
pregnancy fun facts, and stress management) were disseminated
and interactive group activities were conducted. The authors
compared adolescent and adult mothers and found dietary sugar
distribution among both groups to be significantly increased
after 18 weeks, with a greater increase in the adolescent group
(7.4 [SD 0.2] vs 6.3 [SD 0.1] g/d, P=.005). Adolescent mothers https://www.jmir.org/2021/8/e26786 XSL•FO
RenderX XSL•FO
RenderX JOURNAL OF MEDICAL INTERNET RESEARCH Wu et al their children at risk. While this topic was not discussed with
the adolescent mothers, their dialogue suggests that they are
aware of such risks. Maternal Outcomes This belies an important knowledge and
perception gap of program staff on how adolescent mothers use
technology and social networks, as meeting strangers offline
was either not as common as program staff believed or was
regarded as shameful and not openly discussed by the
adolescents. were less likely to shop for groceries (postintervention; 14.0%
vs 89.0%, P=.002) or cook at home (postintervention; 14.0%
vs 67.0%, P=.054) compared to adult mothers and had
significantly less knowledge regarding nutrition (identifying
fiber rich foods, recommended daily consumption of
wholegrains, fruit and vegetable varieties, and fat) before and
after the intervention. There were no significant differences
concerning anthropometric data, mean caloric consumption, or
macronutrient distribution of food. Principal Findings Based on current available literature, there is much that needs
to be explored concerning the potential benefits and harms of
social media for pregnant and postnatal adolescents. This is
despite most adolescent participants having grown up in the
presence of advanced technology and with extensive social
media use. We have systematically studied the impact of
web-based tools used by adolescents on various maternal and
infant outcomes to address our research questions. We found
that available web-based tools include professionally curated
programs (via websites or social media), readily available
information on the internet, and personal use of various social
media platforms. While the limited evidence shows mixed and conflicting
findings, we observe that web-based tools may be useful in
improving mental health outcomes, positively influencing
treatment-seeking intentions, and actual treatment-seeking
behavior for depression among postpartum adolescents [42,47]. Conversely, readily available information on the internet may
increase anxiety among adolescent recipients [43]. The
difference appears to lie in the source of information, whether
curated by health care professionals specifically to address
common concerns of the target group and presented in a
controlled setting or self-acquired data sourced from various
websites initiated by the adolescent recipients themselves. RCTs
using similar platforms, not limited to adolescent mothers, yield
similar results, with postpartum mothers with depressive
symptoms describing significantly improved parenting
competence and decreased depression severity following social
media interventions compared to unexposed mothers [48-51]. The presence of a professionally moderated program may be a
key differentiating factor in mental health outcomes. Web-based
tools may also be more influential on mental health outcomes
due to the users’ relative anonymity compared to conventional
care. A systematic review found that although online services
did not significantly facilitate mental help-seeking behavior in
youths, many youths regularly use online services and
recommend them to peers as they are easily accessible,
anonymous, and less stigmatizing [52]. Discussion Parenting outcomes were explored in 3 studies [41,43,45]. Hudson et al [41] found that both intervention and control
groups had significantly higher perceived competence (P<.01)
and parenting satisfaction (P<.01) measured by the How I Deal
With Problems Regarding Care of My Baby and evaluation
subscale of the What Being the Parent of a Baby is
Like–Revised, respectively. Fleming et al [43] concluded that
although many young mothers had acquired knowledge on what
to expect during childbirth, much of this was fragmented,
inconsistent, weakly linked, poorly referenced, not always
beneficial, and potentially a greater source of confusion. In
contrast, Nolan et al [45] found that using social network sites
increased young mothers’ confidence levels in parenting roles
and with parenting strategies through positive affirmations,
reassurance by other parents, and the collective sharing of
experiences. Regarding social support to adolescent mothers,
Hudson et al [41] observed a positive trend in social support
levels following intervention supported by the qualitative
comments submitted by participants. Nolan et al [45] described
valuable tangible, emotional, and information support from
social networks. Additionally, Rueda et al [46] explored the
role of technology and social network sites on intimate
relationships among adolescents living in residential foster
homes. They determined these tools to be critical to the
adolescent mothers’ ability to maintain or initiate intimate
relationships with the fathers of their children or with new
partners, facilitating both communication and in person
meetings. J Med Internet Res 2021 | vol. 23 | iss. 8 | e26786 | p. 9
(page number not for citation purposes) https://www.jmir.org/2021/8/e26786 Strengths and Limitations Our systematic review is timely as advanced technology and
social media use are pervasive in modern societies, particularly
among youth. Including both quantitative and qualitative studies
in the review allowed for a broader interpretation of both
statistical outcomes and adolescents’ qualitative feedback. Key limitations of this systematic review are the weaknesses
inherent in the included studies and the lack of research in this
important area, particularly in Asian societies, despite being the
most connected globally. Notwithstanding the extensive internet
and social media use among adolescents, few studies have been
conducted to directly establish the relationship between
technology use in pregnant adolescents or adolescent mothers
and pregnancy outcomes. Across the 7 studies there was great
variability in measurement scales and reported outcomes,
contributing to the heterogeneity of our results, and making it
challenging to draw well-grounded conclusions. For this reason,
we were not able to conduct a meta-analysis and were only able
to provide a descriptive narrative of the studies. These studies
were based in the United States and Western Australia, and
findings may not be directly applicable to more ethnically
diverse populations given the cultural differences in family
traditions, community infrastructure, and child-raising practices
[66]. Babes Pregnancy, a charity based in Singapore [67], has
used social media to support these vulnerable youth, although
the impact of this program is not known. Only 2 studies compared the efficacy of web-based tools and
conventional methods in influencing maternal and neonatal
outcomes in the adolescent population [41,42]. Mental health
was the only common outcome that both studies evaluated, in
which Logsdon et al [42] had a greater focus on depression
compared to Hudson et al [41]. As such, no meaningful
conclusions can be drawn to address our second research
question. A Singaporean study described increased risks of perinatal
complications like anemia and preterm births and reduced
likelihood of regular clinic attendance and sexually transmitted
infection screening in a group of younger, vulnerable,
predominantly Malay parturients with poorer access to prenatal
care [59]. This highlights the importance of increasing
accessibility to prenatal care and making a concerted effort to
improve outreach [59]. Such young women are vulnerable to
poor health literacy [60], closely related to eHealth literacy,
which involves the use of digital technology [61], and are less
likely to gain positive outcomes from internet searches [62]. Infant Outcomes Infant outcomes were studied less. None of the studies
investigated putative associations between the use of web-based
tools and premature birth, low birth weight, infant obesity, or
sudden infant death syndrome. Hudson et al [41] found that
emergency services use for postpartum problems in the first 6
months decreased significantly following intervention compared
to no intervention, as 70.6% of mothers who did not receive
intervention brought their child to the emergency room at least
once compared to 35.7% of mothers who received the
intervention (P=.052). In each group, there was one visit to the
emergency room that was appropriate, with one infant
hospitalized and a mother-infant pair treated for smoke
inhalation. Vander Wyst et al [44] described adolescents as less
likely to breastfeed compared to adults. Hudson et al [41]
observed that adolescents in the intervention group were less
likely to exclusively breastfeed compared to those who received
usual care (P=.06 [assuming α=.10, P<.10]). Rueda et al [46]
reported that residential foster home staff were concerned about
adolescent mothers bringing their children to meetings arranged
over social media with unknown partners, potentially placing Aside from mental health outcomes, there may be other benefits
for adolescent mothers, including fewer emergency room visits
(with increased knowledge and confidence in postpartum care)
[41] and positive, albeit limited, dietary and behavioral changes
[44]. Vulnerable populations like low-income adolescent
mothers do not have ready access to health care professionals
or services, but frequently have questions regarding childcare. These require rapid attention and resolution, resulting in https://www.jmir.org/2021/8/e26786 XSL•FO
RenderX XSL•FO
RenderX JOURNAL OF MEDICAL INTERNET RESEARCH Wu et al Wu et al highlight the effectiveness of web-based tools to reach pregnant
and postpartum adolescents who may be more comfortable
seeking help on online platforms. These tools allow health care
professionals and policy makers to spread valuable
pregnancy-related information to this vulnerable population in
an age and peer acceptable way. Local or regional governments
can potentially harness internet platforms and social media to
drive public health policies. It may prove to be a more efficient
allocation of resources, improving compliance to prenatal
follow-ups and reducing pregnancy-related complications. Aside
from antenatal care, governments can broaden their focus and
cover topics pertaining to general women’s health and,
importantly, contraception use. unnecessary emergency room visits. Despite a lack of theoretical
knowledge, structured interventions can motivate both
adolescents and adults to seek healthier lifestyles. Strengths and Limitations eHealth interventions have proven effective in reaching out to
populations with low levels of literacy [63], enabling health
care professionals to facilitate behavioral change, personalize
management, and improve education surrounding their health
[64]. Aside from designing interventions to reach out to
adolescents, it is also important to educate them and fill in the
skill gap in eHealth literacy so they can access and evaluate
health information accurately [65]. The use of professionally
curated web-based tools incorporating online resources,
communication services with health care professionals and
peers, and social support may hold the key to reduce this
important health inequality gap. However, larger age and
culturally appropriate RCTs are necessary to validate the
efficacy of these nontraditional methods. Infant Outcomes Furthermore,
professional support via web-based tools has demonstrated
positive impact on both infants [53,54] and older mothers
[55-57] in the general prenatal population. Preexisting social media platforms and internet-based tools play
an integral role in an adolescent’s life and are associated with
positive and negative outcomes, particularly with maintaining
beneficial friendships and fostering precarious and potentially
dangerous relationships, respectively [45,46]. Despite some
negative consequences, Sherman et al [32] found online
adolescent pregnancy communities such as online support
groups and discussion forums to be largely supportive and to
serve an important role for those who use them. The authors
noted that adolescents should choose their online community
carefully to reduce additional psychological distress. Social
media platforms may thus fill an unmet need to engage
adolescent pregnant women, given the substantial use of social
media globally [58]. J Med Internet Res 2021 | vol. 23 | iss. 8 | e26786 | p. 10
(page number not for citation purposes) Acknowledgments We would like to thank the medical librarians Ms Wong Suei Nee and Ms Wong Huey Lih Winnifred from the National University
of Singapore for their valuable feedback in the search strategy. Conclusion The vastness of the information available on the web limits the
ability of health care professionals to monitor or control sources
of information sought by patients. It is important to create
professionally curated platforms that patients can be directed
to in order to prevent or limit exposure to potentially misleading
or inaccurate information. Although our study is limited by the
scope of the included studies, it is evident that web-based tools
have the potential to improve outcomes in adolescent
pregnancies. This review highlights the potential for web-based
tools to target this vulnerable population, which is usually
excluded from public health policies. The data presented can
be highly informative to local health care policy makers. Overall,
community-specific research is urgently needed to explore the
potential of social media to guide novel interventions for this
important vulnerable group. Overall, our findings demonstrate a paucity of studies in this
important aspect of managing adolescent pregnancies and J Med Internet Res 2021 | vol. 23 | iss. 8 | e26786 | p. 10
(page number not for citation purposes) J Med Internet Res 2021 | vol. 23 | iss. 8 | e26786 | p. 10
(page number not for citation purposes) https://www.jmir.org/2021/8/e26786 XSL•FO
RenderX XSL•FO
RenderX JOURNAL OF MEDICAL INTERNET RESEARCH Wu et al Authors' Contributions JJYW conceptualized the research, conducted the database search, screened records, performed data extraction, critically appraised
studies, analyzed the data, wrote the first draft of the paper, and reviewed the paper. NA conceptualized the research, screened
articles, critically appraised studies, analyzed the data, and wrote and reviewed the paper. MS wrote and reviewed the paper. SL
and CNZM conceptualized the research, critically appraised studies, and wrote and reviewed the paper. Multimedia Appendix 2 Multimedia Appendix 2
Risk of bias. [DOC File , 79 KB-Multimedia Appendix 2] Risk of bias. Risk of bias. [DOC File , 79 KB-Multimedia Appendix 2] [DOC File , 79 KB-Multimedia Appendix 2] Conflicts of Interest None declared. Multimedia Appendix 1 Search strategy. Search strategy. [DOC File , 27 KB-Multimedia Appendix 1] Search strategy. [DOC File , 27 KB-Multimedia Appendix 1] References Eur J Public Health 2007 Dec;17(6):630-636. [doi: 10.1093/eurpub/ckm014] [Medline: 17387106] 7. Amjad S, MacDonald I, Chambers T, Osornio-Vargas A, Chandra S, Voaklander D, et al. Social determinants of health
and adverse maternal and birth outcomes in adolescent pregnancies: a systematic review and meta-analysis. Paediatr Perinat
Epidemiol 2019 Jan;33(1):88-99. [doi: 10.1111/ppe.12529] [Medline: 30516287] 7. Amjad S, MacDonald I, Chambers T, Osornio-Vargas A, Chandra S, Voaklander D, et al. Social determinants of health
and adverse maternal and birth outcomes in adolescent pregnancies: a systematic review and meta-analysis. Paediatr Perinat
Epidemiol 2019 Jan;33(1):88-99. [doi: 10.1111/ppe.12529] [Medline: 30516287] 8. Kapaya H, Mercer E, Boffey F, Jones G, Mitchell C, Anumba D. Deprivation and poor psychosocial support are key
determinants of late antenatal presentation and poor fetal outcomes: a combined retrospective and prospective study. BMC
Pregnancy Childbirth 2015 Nov 25;15:309 [FREE Full text] [doi: 10.1186/s12884-015-0753-3] [Medline: 26608259] Pregnancy Childbirth 2015 Nov 25;15:309 [FREE Full text] [doi: 10.1186/s12884-015-0753-3] [Medline: 26608259]
9. Bunting L, McAuley C. Research Review: teenage pregnancy and motherhood: the contribution of support. Child Fam Soc
Work 2004 May;9(2):207-215. [doi: 10.1111/j.1365-2206.2004.00328.x] 9. Bunting L, McAuley C. Research Review: teenage pregnancy and motherhood: the contribution of
Work 2004 May;9(2):207-215. [doi: 10.1111/j.1365-2206.2004.00328.x] 10. Gomby DS, Culross PL, Behrman RE. Home visiting: recent program evaluations—analysis and recommendations. Future
Child 1999;9(1):4-26. [Medline: 10414008] 10. Gomby DS, Culross PL, Behrman RE. Home visiting: recent program evaluations—analysis and recommendations. Future
Child 1999;9(1):4-26. [Medline: 10414008] 11. Smithbattle L. Helping teen mothers succeed. J Sch Nurs 2006 Jun;22(3):130-135. [doi: 10.1177/10598405060220030201]
[Medline: 16704281] 11. Smithbattle L. Helping teen mothers succeed. J Sch Nurs 2006 Jun;22(3):130-135. [doi: 10.1177/10598405060220030201]
[Medline: 16704281] 12. Culp AM, Culp RE, Blankemeyer M, Passmark L. Parent education home visitation program: adolescent and nonadolescent
mother comparison after six months of intervention. Infant Ment Health J 1998;19(2):111-123. [doi:
10.1002/(SICI)1097-0355(199822)19:2<111::AID-IMHJ3>3.0.CO;2-R] [Medline: 12294463] 12. Culp AM, Culp RE, Blankemeyer M, Passmark L. Parent education home visitation program: adolescent and nonadolescent
mother comparison after six months of intervention. Infant Ment Health J 1998;19(2):111-123. [doi:
10.1002/(SICI)1097-0355(199822)19:2<111::AID-IMHJ3>3.0.CO;2-R] [Medline: 12294463] mother comparison after six months of intervention. Infant Ment Health J 1998;19(2):111-123. [doi:
10.1002/(SICI)1097-0355(199822)19:2<111::AID-IMHJ3>3.0.CO;2-R] [Medline: 12294463] 13. Skinner SR, Marino JL. England's teenage pregnancy strategy: a hard-won success. Lancet 2016 Aug 06;388(10044):538-540
[FREE Full text] [doi: 10.1016/S0140-6736(16)30589-X] [Medline: 27229191] 13. Skinner SR, Marino JL. England's teenage pregnancy strategy: a hard-won success. Lancet 2016 Aug 06;388(10044):538-540
[FREE Full text] [doi: 10.1016/S0140-6736(16)30589-X] [Medline: 27229191] 14. References 1. Sedgh G, Finer LB, Bankole A, Eilers MA, Singh S. Adolescent pregnancy, birth, and abortion rates across countries: levels
and recent trends. J Adolesc Health 2015 Feb;56(2):223-230. [doi: 10.1016/j.jadohealth.2014.09.007] [Medline: 25620306]
2. Adolescent pregnancy. Geneva: World Health Organization; 2020 Jan 31. URL: https://www.who.int/news-room/fact-sheets/ 1. Sedgh G, Finer LB, Bankole A, Eilers MA, Singh S. Adolescent pregnancy, birth, and abortion rates across countries: levels
and recent trends. J Adolesc Health 2015 Feb;56(2):223-230. [doi: 10.1016/j.jadohealth.2014.09.007] [Medline: 25620306] 1. Sedgh G, Finer LB, Bankole A, Eilers MA, Singh S. Adolescent pregnancy, birth, and abortion rates across countries: levels
and recent trends. J Adolesc Health 2015 Feb;56(2):223-230. [doi: 10.1016/j.jadohealth.2014.09.007] [Medline: 25620306] 2. Adolescent pregnancy. Geneva: World Health Organization; 2020 Jan 31. URL: https://www.who.int/news-room/fact-sheets/
detail/adolescent-pregnancy [accessed 2021-07-26] 3. Cunnington AJ. What's so bad about teenage pregnancy? J Fam Plann Reprod Health Care 2001 Jan;27(1):36-41 [FREE
Full text] [doi: 10.1783/147118901101194877] [Medline: 12457546]
4
A e edo WFD Dini MB Fonseca ESVB A e edo LMRD E angelista CB Complications in adolescent pregnanc : 3. Cunnington AJ. What's so bad about teenage pregnancy? J Fam Plann Reprod Health Care 2001 Jan;27(1):36-41 [FREE
Full text] [doi: 10.1783/147118901101194877] [Medline: 12457546]
4. Azevedo WFD, Diniz MB, Fonseca ESVB, Azevedo LMRD, Evangelista CB. Complications in adolescent pregnancy:
systematic review of the literature. Einstein (Sao Paulo) 2015;13(4):618-626 [FREE Full text] [doi: ] [
] [
]
4. Azevedo WFD, Diniz MB, Fonseca ESVB, Azevedo LMRD, Evangelista CB. Complications in adolescent pregnancy:
systematic review of the literature. Einstein (Sao Paulo) 2015;13(4):618-626 [FREE Full text] [doi:
10.1590/S1679-45082015RW3127] [Medline: 26061075] 5. Chen X, Wen SW, Fleming N, Demissie K, Rhoads GG, Walker M. Teenage pregnancy and adverse birth outcomes: a
large population based retrospective cohort study. Int J Epidemiol 2007 Apr;36(2):368-373. [doi: 10.1093/ije/dyl284]
[Medline: 17213208] 5. Chen X, Wen SW, Fleming N, Demissie K, Rhoads GG, Walker M. Teenage pregnancy and adverse birth outcomes: a
large population based retrospective cohort study. Int J Epidemiol 2007 Apr;36(2):368-373. [doi: 10.1093/ije/dyl284]
[Medline: 17213208] 6. Imamura M, Tucker J, Hannaford P, da Silva MO, Astin M, Wyness L, REPROSTAT 2 group. Factors associated with
teenage pregnancy in the European Union countries: a systematic review. Eur J Public Health 2007 Dec;17(6):630-636. [doi: 10.1093/eurpub/ckm014] [Medline: 17387106] 6. Imamura M, Tucker J, Hannaford P, da Silva MO, Astin M, Wyness L, REPROSTAT 2 group. Factors associated with
teenage pregnancy in the European Union countries: a systematic review. JOURNAL OF MEDICAL INTERNET RESEARCH Wu et al 15. Baker B, Yang I. Social media as social support in pregnancy and the postpartum. Sex Reprod Healthc 2018 Oct;17:31-34. [doi: 10.1016/j.srhc.2018.05.003] [Medline: 30193717] 16. Bjelke M, Martinsson A, Lendahls L, Oscarsson M. Using the Internet as a source of information during pregnancy: a
descriptive cross-sectional study in Sweden. Midwifery 2016 Sep;40:187-191. [doi: 10.1016/j.midw.2016.06.020] [Medline:
27450590] 17. Lupton D, Pedersen S, Thomas GM. Parenting and digital media: from the early web to contem
Compass 2016 Aug 02;10(8):730-743. [doi: 10.1111/soc4.12398] 18. Herman J, Mock K, Blackwell D, Hulsey T. Use of a pregnancy support web site by low-income Afr
J Obstet Gynecol Neonatal Nurs 2005;34(6):713-720. [doi: 10.1177/0884217505282019] [Medline y
19. Lagan BM, Sinclair M, Kernohan WG. Internet use in pregnancy informs women's decision making: a web-based survey. Birth 2010 Jun;37(2):106-115. [doi: 10.1111/j.1523-536X.2010.00390.x] [Medline: 20557533] 20. Aichner T, Grünfelder M, Maurer O, Jegeni D. Twenty-five years of social media: a review of social media applications
and definitions from 1994 to 2019. Cyberpsychol Behav Soc Netw 2021 Apr;24(4):215-222 [FREE Full text] [doi:
10.1089/cyber.2020.0134] [Medline: 33847527] y
21. Aichner T, Jacob F. Measuring the degree of corporate social media use. Int J Market Res 2015 Mar;57(2):257-276. [doi:
10.2501/ijmr-2015-018] y
21. Aichner T, Jacob F. Measuring the degree of corporate social media use. Int J Market Res 2015 Mar;57(2):257-276. [doi:
10.2501/ijmr-2015-018] j
22. Kaplan AM, Haenlein M. Users of the world, unite! The challenges and opportunities of social media. Bus Horizons 2010
Jan;53(1):59-68. [doi: 10.1016/j.bushor.2009.09.003] 23. Boulos MNK, Maramba I, Wheeler S. Wikis, blogs and podcasts: a new generation of Web-based tools for virtual
collaborative clinical practice and education. BMC Med Educ 2006;6:41 [FREE Full text] [doi: 10.1186/1472-6920-6-41]
[Medline: 16911779] 24. Sayakhot P, Carolan-Olah M. Internet use by pregnant women seeking pregnancy-related informatio
BMC Pregnancy Childbirth 2016 Mar 28;16:65 [FREE Full text] [doi: 10.1186/s12884-016-0856-5 24. Sayakhot P, Carolan Olah M. Internet use by pregnant women seeking pregnancy related information: a systematic review. BMC Pregnancy Childbirth 2016 Mar 28;16:65 [FREE Full text] [doi: 10.1186/s12884-016-0856-5] [Medline: 27021727]
25. Narasimhulu DM, Karakash S, Weedon J, Minkoff H. Patterns of internet use by pregnant women, and reliability of
pregnancy-related searches. Matern Child Health J 2016 Dec;20(12):2502-2509. [doi: 10.1007/s10995-016-2075-0] [Medline:
27456311] 25. Narasimhulu DM, Karakash S, Weedon J, Minkoff H. Patterns of internet use by pregnant women, and reliability of
pregnancy-related searches. Matern Child Health J 2016 Dec;20(12):2502-2509. [doi: 10.1007/s10995-016-2075-0] [Medline:
27456311] 26. JOURNAL OF MEDICAL INTERNET RESEARCH PLoS Med 2009 Jul 21;6(7):e1000097 [FREE Full text] [doi: 10.1371/journal.pmed.1000097] [Medline:
19621072] ]
34. Bramer W, Bain P. Updating search strategies for systematic reviews using EndNote. J Med Libr Assoc 2017 ]
34. Bramer W, Bain P. Updating search strategies for systematic reviews using EndNote. J Med Libr Assoc 2017
J l 105(3) 285 289 [FREE F ll t
t] [d i 10 5195/j l 2017 183] [M dli
28670219] 34. Bramer W, Bain P. Updating search strategies for systematic reviews using EndNote. J Med Libr Assoc 2017
Jul;105(3):285-289 [FREE Full text] [doi: 10.5195/jmla.2017.183] [Medline: 28670219] 35. O'Reilly T. What is Web 2.0: design patterns and business models for the next generation of software. 2005 Sep 03. URL:
https://www.oreilly.com/pub/a/web2/archive/what-is-web-20.html [accessed 2021-07-26] 35. O'Reilly T. What is Web 2.0: design patterns and business models for the next generation of software. 2005 Sep 03. URL:
https://www.oreilly.com/pub/a/web2/archive/what-is-web-20.html [accessed 2021-07-26] 36. Lockwood C, Munn Z, Porritt K. Qualitative research synthesis: methodological guidance for systematic reviewers utilizing
meta-aggregation. Int J Evid Based Healthc 2015 Sep;13(3):179-187. [doi: 10.1097/XEB.0000000000000062] [Medline:
26262565] 36. Lockwood C, Munn Z, Porritt K. Qualitative research synthesis: methodological guidance for systematic reviewers utilizing
meta-aggregation. Int J Evid Based Healthc 2015 Sep;13(3):179-187. [doi: 10.1097/XEB.0000000000000062] [Medline:
26262565] 37. Tufanaru C, Munn Z, Aromataris E, Campbell J, Hopp L. Chapter 3: Systematic reviews of effectiveness. In: Aromataris
E, Munn Z, editors. JBI Manual for Evidence Synthesis. Adelaide: Joanna Briggs Institute; 2020. 37. Tufanaru C, Munn Z, Aromataris E, Campbell J, Hopp L. Chapter 3: Systematic reviews of effectiveness. In: Aromataris
E, Munn Z, editors. JBI Manual for Evidence Synthesis. Adelaide: Joanna Briggs Institute; 2020. 38. Wells GA, Shea B, O'Connel D, Peterson J. The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised
studies in meta-analyses. 2009. URL: http://www.ohri.ca/programs/clinical_epidemiology/oxford.asp [accessed 2021-07-26] 38. Wells GA, Shea B, O'Connel D, Peterson J. The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised
studies in meta-analyses. 2009. URL: http://www.ohri.ca/programs/clinical_epidemiology/oxford.asp [accessed 2021-07-26]
39
Sterne JAC Savović J Page MJ Elbers RG Blencowe NS Boutron I et al RoB 2: a revised tool for assessing risk of bias 38. Wells GA, Shea B, O'Connel D, Peterson J. The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised
studies in meta-analyses. 2009. URL: http://www.ohri.ca/programs/clinical_epidemiology/oxford.asp [accessed 2021-07-26]
39. Sterne JAC, Savović J, Page MJ, Elbers RG, Blencowe NS, Boutron I, et al. References Holtz B, Smock A, Reyes-Gastelum D. Connected motherhood: social support for moms and moms-to-be on Facebook. Telemed J E Health 2015 May;21(5):415-421. [doi: 10.1089/tmj.2014.0118] [Medline: 25665177] 14. Holtz B, Smock A, Reyes-Gastelum D. Connected motherhood: social support for moms and moms-to-be on Facebook. Telemed J E Health 2015 May;21(5):415-421. [doi: 10.1089/tmj.2014.0118] [Medline: 25665177] J Med Internet Res 2021 | vol. 23 | iss. 8 | e26786 | p. 11
(page number not for citation purposes) https://www.jmir.org/2021/8/e26786 XSL•FO
RenderX XSL•FO
RenderX JOURNAL OF MEDICAL INTERNET RESEARCH JOURNAL OF MEDICAL INTERNET RESEARCH Eysenbach G, Powell J, Kuss O, Sa E. Empirical studies assessing the quality of health information for consumers on the
world wide web: a systematic review. JAMA 2002;287(20):2691-2700. [doi: 10.1001/jama.287.20.2691] [Medline:
12020305] 27. Chang T, Verma BA, Shull T, Moniz MH, Kohatsu L, Plegue MA, et al. Crowdsourcing and the accuracy of online
information regarding weight gain in pregnancy: a descriptive study. J Med Internet Res 2016 Apr 07;18(4):e81 [FREE
Full text] [doi: 10.2196/jmir.5138] [Medline: 27056465] j
28. Kavlak O, Atan, Güleç D, Oztürk R, Atay N. Pregnant women's use of the internet in relation to their pregnancy in Izmir,
Turkey. Inform Health Soc Care 2012 Dec;37(4):253-263. [doi: 10.3109/17538157.2012.710686] [Medline: 22958087] 28. Kavlak O, Atan, Güleç D, Oztürk R, Atay N. Pregnant women's use of the internet in relation to their pregnancy in Izmir,
Turkey. Inform Health Soc Care 2012 Dec;37(4):253-263. [doi: 10.3109/17538157.2012.710686] [Medline: 22958087]
29. Jesper E. Poor quality online information for pregnant women is a global problem: what can we do about it? BJOG 2015 Turkey. Inform Health Soc Care 2012 Dec;37(4):253-263. [doi: 10.3109/17538157.2012.710686] [Medline: 22958087]
29. Jesper E. Poor quality online information for pregnant women is a global problem: what can we do about it? BJOG 2015
Apr;122(5):740. [doi: 10.1111/1471-0528.13091] [Medline: 25315320] 29. Jesper E. Poor quality online information for pregnant women is a global problem: what can we do about it? BJOG 2015
Apr;122(5):740. [doi: 10.1111/1471-0528.13091] [Medline: 25315320] 30. Wiseman E. Jennifer Gunter: "Women are being told lies about their bodies". The Guardian. 2019 Sep 08. URL: https:/
/www.theguardian.com/lifeandstyle/2019/sep/08/jennifer-gunter-gynaecologist-womens-health-bodies-myths-and-medicine
[accessed 2021-07-26] 31. Larsson M. A descriptive study of the use of the Internet by women seeking pregnancy-related information. Midwifery
2009 Feb;25(1):14-20. [doi: 10.1016/j.midw.2007.01.010] [Medline: 17408822] 32. Sherman LE, Greenfield PM. Forging friendship, soliciting support: a mixed-method examination of message boards for
pregnant teens and teen mothers. Comput Human Behav 2013 Jan;29(1):75-85. [doi: 10.1016/j.chb.2012.07.018] 32. Sherman LE, Greenfield PM. Forging friendship, soliciting support: a mixed-method examination of message boards for
pregnant teens and teen mothers. Comput Human Behav 2013 Jan;29(1):75-85. [doi: 10.1016/j.chb.2012.07.018]
33. Moher D, Liberati A, Tetzlaff J, Altman DG. Preferred reporting items for systematic reviews and meta-analyses: the
PRISMA statement. PLoS Med 2009 Jul 21;6(7):e1000097 [FREE Full text] [doi: 10.1371/journal.pmed.1000097] [Medline:
19621072] 33. Moher D, Liberati A, Tetzlaff J, Altman DG. Preferred reporting items for systematic reviews and meta-analyses: the
PRISMA statement. JOURNAL OF MEDICAL INTERNET RESEARCH RoB 2: a revised tool for assessing risk of bias
in randomised trials. BMJ 2019 Aug 28;366:l4898. [doi: 10.1136/bmj.l4898] [Medline: 31462531] studies in meta analyses. 2009. URL: http://www.ohri.ca/programs/clinical_epidemiology/oxford.asp [accessed 2021 07 26]
39. Sterne JAC, Savović J, Page MJ, Elbers RG, Blencowe NS, Boutron I, et al. RoB 2: a revised tool for assessing risk of bias
in randomised trials. BMJ 2019 Aug 28;366:l4898. [doi: 10.1136/bmj.l4898] [Medline: 31462531] 39. Sterne JAC, Savović J, Page MJ, Elbers RG, Blencowe NS, Boutron I, et al. RoB 2: a revised tool for assessing risk of bias
in randomised trials. BMJ 2019 Aug 28;366:l4898. [doi: 10.1136/bmj.l4898] [Medline: 31462531] J Med Internet Res 2021 | vol. 23 | iss. 8 | e26786 | p. 12
(page number not for citation purposes) https://www.jmir.org/2021/8/e26786 XSL•FO
RenderX XSL•FO
RenderX JOURNAL OF MEDICAL INTERNET RESEARCH Wu et al 40. Pluye P, Robert E, Cargo M, Bartlett G, O'Cathain A, Griffiths F. Proposal: a mixed methods appraisal tool for systematic
mixed studies reviews. 2018. URL: http://mixedmethodsappraisaltoolpublic.pbworks.com/w/file/fetch/127916259/
MMAT_2018_criteria-manual_2018-08-01_ENG.pdf [accessed 2020-11-09] p
41. Hudson DB, Campbell-Grossman C, Hertzog M. Effects of an internet intervention on mothers' psychological, parenting,
and health care utilization outcomes. Issues Compr Pediatr Nurs 2012;35(3-4):176-193. [doi: 10.3109/01460862.2012.734211]
[Medline: 23140413] p
41. Hudson DB, Campbell-Grossman C, Hertzog M. Effects of an internet intervention on mothers' psychological, parenting,
and health care utilization outcomes. Issues Compr Pediatr Nurs 2012;35(3-4):176-193. [doi: 10.3109/01460862.2012.734211]
[Medline: 23140413] [
]
42. Cynthia Logsdon M, Myers J, Rushton J, Gregg JL, Josephson AM, Davis DW, et al. Efficacy of an Internet-based depression
intervention to improve rates of treatment in adolescent mothers. Arch Womens Ment Health 2018 Jun;21(3):273-285
[FREE Full text] [doi: 10.1007/s00737-017-0804-z] [Medline: 29260321] 42. Cynthia Logsdon M, Myers J, Rushton J, Gregg JL, Josephson AM, Davis DW, et al. Efficacy of an Internet-based depression
intervention to improve rates of treatment in adolescent mothers. Arch Womens Ment Health 2018 Jun;21(3):273-285
[FREE Full text] [doi: 10.1007/s00737-017-0804-z] [Medline: 29260321] [
] [
] [
]
43. Fleming SE, Vandermause R, Shaw M. First-time mothers preparing for birthing in an electronic world: internet and mobile
phone technology. J Reprod Infant Psychol 2014 Mar 19;32(3):240-253. [doi: 10.1080/02646838.2014.886104] 43. Fleming SE, Vandermause R, Shaw M. First-time mothers preparing for birthing in an electronic world: internet and mobile
phone technology. J Reprod Infant Psychol 2014 Mar 19;32(3):240-253. [doi: 10.1080/02646838.2014.886104] 44. Vander Wyst KB, Vercelli ME, O'Brien KO, Cooper EM, Pressman EK, Whisner CM. JOURNAL OF MEDICAL INTERNET RESEARCH A social media intervention to
improve nutrition knowledge and behaviors of low income, pregnant adolescents and adult women. PLoS One
2019;14(10):e0223120 [FREE Full text] [doi: 10.1371/journal.pone.0223120] [Medline: 31647852] 45. Nolan S, Hendricks J, Towell A. Social networking sites (SNS): exploring their uses and associated value for adolescent
mothers in Western Australia in terms of social support provision and building social capital. Midwifery 2015
Sep;31(9):912-919. [doi: 10.1016/j.midw.2015.05.002] [Medline: 26026197] p
j
46. Rueda HA, Brown ML, Geiger JM. Technology and dating among pregnant and parenting youth in residential foster care:
a mixed qualitative approach comparing staff and adolescent perspectives. J Adolesc Res 2019 Jul 17;35(4):521-545. [doi:
10.1177/0743558419861087] 47. Logsdon MC, Barone M, Lynch T, Robertson A, Myers J, Morrison D, et al. Testing of a prototype Web based intervention
for adolescent mothers on postpartum depression. Appl Nurs Res 2013 Aug;26(3):143-145. [doi: 10.1016/j.apnr.2013.01.005]
[Medline: 23473677] 48. Boyd RC, Price J, Mogul M, Yates T, Guevara JP. Pilot RCT of a social media parenting intervention for postpartum
mothers with depression symptoms. J Reprod Infant Psychol 2019 Jul;37(3):290-301. [doi: 10.1080/02646838.2018.1556788]
[Medline: 30556428] 49. Forsell E, Bendix M, Holländare F, Szymanska von Schultz B, Nasiell J, Blomdahl-Wetterholm M, et al. Internet delivered
cognitive behavior therapy for antenatal depression: a randomised controlled trial. J Affect Disord 2017 Oct 15;221:56-64
[FREE Full text] [doi: 10.1016/j.jad.2017.06.013] [Medline: 28628768] 50. Haga SM, Drozd F, Lisøy C, Wentzel-Larsen T, Slinning K. Mamma Mia: a randomized controlled trial of an internet-based
intervention for perinatal depression. Psychol Med 2019 Aug;49(11):1850-1858 [FREE Full text] [doi:
10.1017/S0033291718002544] [Medline: 30191779] 51. Yang M, Jia G, Sun S, Ye C, Zhang R, Yu X. Effects of an online mindfulness intervention focusing on attention monitoring
and acceptance in pregnant women: a randomized controlled trial. J Midwifery Womens Health 2019 Jan 29;64(1):68-77. [doi: 10.1111/jmwh.12944] [Medline: 30695166] 52. Kauer SD, Mangan C, Sanci L. Do online mental health services improve help-seeking for young people? A systematic
review. J Med Internet Res 2014;16(3):e66 [FREE Full text] [doi: 10.2196/jmir.3103] [Medline: 24594922] 53. Huang MZ, Kuo S, Avery MD, Chen W, Lin K, Gau M. Evaluating effects of a prenatal web-based breastfeeding education
programme in Taiwan. J Clin Nurs 2007 Aug;16(8):1571-1579 [FREE Full text] [doi: 10.1111/j.1365-2702.2006.01843.x]
[Medline: 17655546] 54. Abuidhail J, Mrayan L, Jaradat D. Evaluating effects of prenatal web-based breastfeeding education for pregnant mothers
in their third trimester of pregnancy: Prospective randomized control trial. Midwifery 2019 Feb;69:143-149. JOURNAL OF MEDICAL INTERNET RESEARCH [doi:
10.1016/j.midw.2018.11.015] [Medline: 30513445] 54. Abuidhail J, Mrayan L, Jaradat D. Evaluating effects of prenatal web-based breastfeeding education for pregnant mothers
in their third trimester of pregnancy: Prospective randomized control trial. Midwifery 2019 Feb;69:143-149. [doi:
10.1016/j.midw.2018.11.015] [Medline: 30513445] j
55. Kuo S, Chen Y, Lin K, Lee T, Hsu C. Evaluating the effects of an Internet education programme on newborn care in Taiwan. J Clin Nurs 2009 Jun;18(11):1592-1601. [doi: 10.1111/j.1365-2702.2008.02732.x] [Medline: 19490296] 55. Kuo S, Chen Y, Lin K, Lee T, Hsu C. Evaluating the effects of an Internet education programme on newborn care in Taiwan. J Clin Nurs 2009 Jun;18(11):1592-1601. [doi: 10.1111/j.1365-2702.2008.02732.x] [Medline: 19490296] 56. Smith K, Lanningham-Foster L, Welch A, Campbell C. Web-based behavioral intervention increases maternal exercise
but does not prevent excessive gestational weight gain in previously sedentary women. J Phys Act Health 2016
Jun;13(6):587-593. [doi: 10.1123/jpah.2015-0219] [Medline: 26594820] 56. Smith K, Lanningham-Foster L, Welch A, Campbell C. Web-based behavioral intervention increases maternal exercise
but does not prevent excessive gestational weight gain in previously sedentary women. J Phys Act Health 2016
Jun;13(6):587-593. [doi: 10.1123/jpah.2015-0219] [Medline: 26594820] jp
57. Bianchi CM, Mariotti F, Lluch A, Journet C, Stehr Y, Beaussier H, et al. Computer-based tailored dietary counselling
improves the nutrient adequacy of the diet of French pregnant women: a randomised controlled trial. Br J Nutr 2019 Dec
18;123(2):220-231. [doi: 10.1017/s0007114519002617] jp
57. Bianchi CM, Mariotti F, Lluch A, Journet C, Stehr Y, Beaussier H, et al. Computer-based tailored dietary counselling
improves the nutrient adequacy of the diet of French pregnant women: a randomised controlled trial. Br J Nutr 2019 Dec
18;123(2):220-231. [doi: 10.1017/s0007114519002617] 58. Twenge JM, Martin GN, Spitzberg BH. Trends in U.S. adolescents’ media use, 1976–2016: the rise of digital media, the
decline of TV, and the (near) demise of print. Psychol Popular Media Culture 2019 Oct;8(4):329-345. [doi:
10.1037/ppm0000203] 58. Twenge JM, Martin GN, Spitzberg BH. Trends in U.S. adolescents’ media use, 1976–2016: the rise of digital media, the
decline of TV, and the (near) demise of print. Psychol Popular Media Culture 2019 Oct;8(4):329-345. [doi:
10.1037/ppm0000203] pp
59. Tung JS, Kwek LK, Nadarajah S, Sulaiman S. Birth outcomes and successive pregnancy rates in ad
specialised antenatal clinic. Ann Acad Med Singap 2018 Jan;47(1):44-47 [FREE Full text] [Medlin 59. Tung JS, Kwek LK, Nadarajah S, Sulaiman S. Birth outcomes and successive pregnancy rates in adolescents attending a
specialised antenatal clinic. JOURNAL OF MEDICAL INTERNET RESEARCH Ann Acad Med Singap 2018 Jan;47(1):44-47 [FREE Full text] [Medline: 29493711]
60. MacLean L. A literature review of health literacy of adolescents during pregnancy. Nurs Womens Health 2020
Dec;24(6):431-439. [doi: 10.1016/j.nwh.2020.09.004] [Medline: 33157069] p
g p
;
( )
[
] [
]
60. MacLean L. A literature review of health literacy of adolescents during pregnancy. Nurs Womens Health 2020
Dec;24(6):431-439. [doi: 10.1016/j.nwh.2020.09.004] [Medline: 33157069] 60. MacLean L. A literature review of health literacy of adolescents during pregnancy. Nurs Womens Health 2020
Dec;24(6):431-439. [doi: 10.1016/j.nwh.2020.09.004] [Medline: 33157069] J Med Internet Res 2021 | vol. 23 | iss. 8 | e26786 | p. 13
(page number not for citation purposes) https://www.jmir.org/2021/8/e26786 XSL•FO
RenderX XSL•FO
RenderX JOURNAL OF MEDICAL INTERNET RESEARCH J Med Internet Res 2021 | vol. 23 | iss. 8 | e26786 | p. 14
(page number not for citation purposes) JOURNAL OF MEDICAL INTERNET RESEARCH Wu et al Wu et al 61. Manganello J, Gerstner G, Pergolino K, Graham Y, Falisi A, Strogatz D. The relationship of health literacy with use of
digital technology for health information: implications for public health practice. J Public Health Manag Pract 2016 Dec
15:380-387. [doi: 10.1097/PHH.0000000000000366] [Medline: 26672402] 61. Manganello J, Gerstner G, Pergolino K, Graham Y, Falisi A, Strogatz D. The relationship of health literacy with use of
digital technology for health information: implications for public health practice. J Public Health Manag Pract 2016 Dec
15:380-387. [doi: 10.1097/PHH.0000000000000366] [Medline: 26672402] 62. Neter E, Brainin E. eHealth literacy: extending the digital divide to the realm of health information. J Med Internet Res
2012 Jan;14(1):e19 [FREE Full text] [doi: 10.2196/jmir.1619] [Medline: 22357448] 64. Lotto M, Strieder AP, Ayala Aguirre PE, Oliveira TM, Andrade Moreira Machado MA, Rios D, et al. Parental-oriented
educational mobile messages to aid in the control of early childhood caries in low socioeconomic children: a randomized
controlled trial. J Dent 2020 Oct;101:103456 [FREE Full text] [doi: 10.1016/j.jdent.2020.103456] [Medline: 32827598] 65. Skopelja EN, Whipple EC, Richwine P. Reaching and teaching teens: adolescent health literacy and the internet. J Consum
Heal Internet 2008 Jun 17;12(2):105-118. [doi: 10.1080/15398280802121406] (
)
]
67. Babes: a helping hand for pregnant teens. URL: https://www.babes.org.sg/ [accessed 2021-07-26] PRISMA: Preferred Reporting Items for Systematic Reviews and Meta-analyses
PROSPERO: International Prospective Register of Systematic Reviews
RCT: randomized controlled trial Please cite as:
Wu JJY, Ahmad N, Samuel M, Logan S, Mattar CNZ
The Influence of Web-Based Tools on Maternal and Neonatal Outcomes in Pregnant Adolescents or Adolescent Mothers: Mixed
Methods Systematic Review
J Med Internet Res 2021;23(8):e26786
URL: https://www.jmir.org/2021/8/e26786
doi: 10.2196/26786
PMID: ©Jania J Y Wu, Nurulhuda Ahmad, Miny Samuel, Susan Logan, Citra N Z Mattar. Originally published in the Journal of Medical
Internet Research (https://www.jmir.org), 26.08.2021. This is an open-access article distributed under the terms of the Creative
Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and
reproduction in any medium, provided the original work, first published in the Journal of Medical Internet Research, is properly
cited. The complete bibliographic information, a link to the original publication on https://www.jmir.org/, as well as this copyright
and license information must be included. J Med Internet Res 2021 | vol. 23 | iss. 8 | e26786 | p. 14
(page number not for citation purposes) https://www.jmir.org/2021/8/e26786 XSL•FO
RenderX
|
https://openalex.org/W4319442913
|
http://umu.diva-portal.org/smash/get/diva2:1735912/FULLTEXT01
|
English
| null |
Negotiating the ambiguity of an (in)authentic working life: a grounded theory study into severe perineal trauma
|
BMC women's health
| 2,023
|
cc-by
| 11,164
|
© The Author(s) 2023. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which
permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the
original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or
other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line
to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory
regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this
licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativeco
mmons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Abstract Background In Sweden, persistent physical and psychological health problems occur in about three in ten women
who sustain severe perineal trauma (SPT) during childbirth. As most Swedish women work outside the home, the
question of if and how SPT-related morbidity influences working life needs exploration. This study aims to qualitatively
explore how women with persistent SPT-related morbidities experience and conceptualise their problems concerning
working life. Further, we theorise the findings by applying Simone de Beauvoir’s feminist framework of immanence
and transcendence as well as authentic and inauthentic life. Methods Ten interviews with women recruited via a Swedish social media community for perineal trauma were
analysed according to Charmaz’s constructivist approach to grounded theory. Results The theoretical model and related core category ‘Negotiating the ambiguity of an (in)authentic working
life’ reflected the women’s negotiations of immanence as ‘the silent covert object’ versus transcendence as the ‘the
resourceful overt subject’. The model also mirrored the conflict of (in)authenticity in working life. An inauthentic work-
ing life was experienced when women were denied their subjectivity at work or constructed themselves as subjects
in denial of their SPT. On the other hand, women who acknowledged their SPT and were constructed as subjects by
‘others’ achieved an authentic working life despite SPT. Conclusions The conflicting gendered process of ‘the silent covert object’ versus ‘the resourceful overt subject’
problematised women’s vulnerability at work. Aspects that enable subjectification and transcendence are essential
for policymakers, employers, healthcare services, and society to eradicate the taboo of SPT and create a working
environment characterised by understanding, support, and flexibility. Further, access to adequate care, sick leave, and
occupational rehabilitation are essential. Such measures support transcendence towards an authentic working life
and, consequently, a more gender-equal working environment that does not deprive women of career opportunities
due to a physical ailment. Keywords Severe perineal trauma, Maternal morbidity, Workability, Grounded theory, Gender theory, Simone de
Beauvoir, Immanence, Transcendence BMC Women’s Health BMC Women’s Health Tjernström et al. BMC Women’s Health (2023) 23:47
https://doi.org/10.1186/s12905-023-02191-9 Open Access Background
Introduction *Correspondence:
Katharina Tjernström
katharina.tjernstrom@umu.se
1 Department of Nursing, Umeå University, Umeå, Sweden
2 Department of Community Medicine and Rehabilitation, Section
of Physiotherapy, Umeå University, Umeå, Sweden *Correspondence:
Katharina Tjernström
katharina.tjernstrom@umu.se
1 Department of Nursing, Umeå University, Umeå, Sweden
2 Department of Community Medicine and Rehabilitation, Section
of Physiotherapy, Umeå University, Umeå, Sweden Severe perineal trauma (SPT), i.e., third and fourth-
degree perineal lacerations affecting the anal sphincter
muscle and the anorectal mucosa [1], relates to short-
and long-term (beyond one year postpartum) physical Rationale In sum, there is a global lack of research on the impact of
long-term SPT-related morbidities on women’s working
life, especially from a gender perspective. As most Swed-
ish women work outside the home [10], the influence of
persistent SPT-related morbidities on working life needs
investigation. To address this knowledge gap, we aim to
qualitatively explore how women with self-reported per-
sistent SPT-related morbidities experience and conceptu-
alise their problems concerning working life. Further, we
theorised the findings by applying a theoretical feminist
perspective. Work‑related consequences of genital problems and SPT Work‑related consequences of genital problems and SPT
Globally, few studies explore how women with SPT, or
other genital problems, experience their situation at work
and whether the difficulties impact their professional
role. Urinary incontinence, the most studied subtype of
pelvic floor dysfunction, negatively impacts work pro-
ductivity in men and women [12]. Experiencing urine
incontinence at work manifests in reduced time man-
agement and diminished focus [13]. Women cope with
urine incontinence at work by using pads, keeping spare
clothing, limiting fluid intake, and frequently visiting the
bathroom [14, 15]. Severe urine incontinence affects con-
centration, the performance of physical activities, self-
confidence, or undisturbed task completion at work [14]. Faecal incontinence may lead to an inability to work part-
or full-time, resulting in feelings of unfulfillment and a
lack of work-related confidence. Employees with faecal
incontinence frequently change pads and struggle with
inaccessible toilets at work [16, 17]. Further, menopau-
sal working women with uterine prolapse report worse
occupational quality of life than mid-aged women with-
out prolapse [18]. However, these studies do not explic-
itly study women with known SPT after childbirth. Theoretical and conceptual framework To explore and theorise women’s ‘situated embodiment’
of SPT in working life, we apply concepts and notions
developed by the existentialist feminist Simone de Beau-
voir [26], see Table 1. The limited research on SPT-related problems in
working life shows that women experience insecurities
regarding incontinence at work [19]. On the other hand,
working may distract their attention from SPT-related
morbidities. Nevertheless, women may in some con-
texts lose their job due to anal incontinence after SPT
[20]. In addition, they may need to change work tasks,
occupations, or workplaces [19, 20], indicating that SPT-
related morbidities may imply extensive professional
consequences for some women. Beauvoir’s well-known expression, ‘One is not born,
but rather becomes, woman’ [26, p. 330], implies that
‘woman’ is a social and gendered construct in a specific
situation and time. According to Beauvoir [26], ‘imma-
nence’ represents the body or nature where the exist-
ence as an ‘object’ is maintained passively, repetitively,
and monotonously. On the other hand, ‘transcendence’
symbolises the mind actively developing life projects
towards freedom as a ‘subject’. Beauvoir sees the human
being as ‘ambiguous’, incorporating both subject and
object. To her, one can never deny one’s facticity, i.e., we
are always a situated embodiment, a lived experience
in a context. An individual can transcend to achieve
an ‘authentic’ life from this immanent situation. Thus,
according to Beauvoir, existing in only immanence or
transcendence can constitute an ‘inauthentic’ life. She
also argues that women are bound to face the conflict © The Author(s) 2023. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which
permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the
original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or
other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line
to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory
regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this
licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativeco
mmons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Tjernström et al. BMC Women’s Health (2023) 23:47 Page 2 of 13 Page 2 of 13 Tjernström et al. BMC Women’s Health (2023) 23:47 Tjernström et al. BMC Women’s Health despite being part of the workforce and contributing to
society. Therefore, a gender perspective is essential to
gain insight into these women’s working experiences. However, to our knowledge, such studies do not exist. A few studies use a feminist approach to explore per-
ineal trauma, but with other perspectives. For example,
Salmon [23] describes women’s experiences of exposure
and vulnerability during suturing, the normalisation of
postpartum pain by health professionals, and the com-
plex emotional impact of perineal trauma, such as fear,
anxiety, anger, frustration, and invisibility. Moreover,
Priddis [24] shows that women with SPT use disem-
bodiment strategies to disconnect themselves from their
broken perineum and related problems to function in
everyday life. Additionally, morbidity after SPT is found
to cause blurring of boundaries between the ‘self’ and
‘other’ [25]. [2–6] and psychological [5, 7, 8] maternal morbidities. In Sweden, women experience more extensive STP-
related morbidity than previously described [5, 8, 9]. Further, about one in three Swedish women report per-
sisting SPT-related problems one year postpartum [9]. 78% of Swedish women work outside the home [10], and
approximately 86% of Swedish one- to five-year-olds are
enrolled in childcare [11]. This raises the question of if
and how SPT influences women at work. Gender perspectives on SPT Gender norms and gendered power relations affect sex-
ual and reproductive health within a social, economic,
and political context [21]. Gender equality, sexual and
reproductive health, and paid employment are part of
the 2030 Agenda for Sustainable Development [22]. The
lack of research on if and how morbidity after SPT influ-
ences workability suggests that women with SPT-related
morbidity are an unrecognised and marginalised group Page 3 of 13 Tjernström et al. BMC Women’s Health (2023) 23:47 Table 1 Definitions of gender theoretical concepts according to Simone de Beauvoir [26]
Concept
Definition
Object
The body or nature
Subject
The mind
Situated embodiment
A lived incorporated experience in a particular situation and context
Immanence
Passive, repetitive, and monotonous existence as an object
Transcendence
Active existence as a subject pursuing developing life projects towards freedom
Ambiguity
Being object and subject at the same time
Authentic life
Embracing both immanence (the body) and transcendence (the mind) simultaneously
Inauthentic life
Denying the existence of either the body (immanence) or the mind (transcendence)
Authentic working life
As defined in the present study: Acknowledging one’s SPT-damaged body in working
life while being constructed as a subject by others relevant to the work context
Inauthentic working life
As defined in the present study: Being denied subjectivity at work by others or con-
structing oneself as a subject in denial of the SPT-damaged body Table 1 Definitions of gender theoretical concepts according to Simone de Beauvoir [26] rding to Simone de Beauvoir [26]
Definition
The body or nature
The mind
A lived incorporated experience in a particular situation and context
Passive, repetitive, and monotonous existence as an object
Active existence as a subject pursuing developing life projects towards freedom
Being object and subject at the same time
Embracing both immanence (the body) and transcendence (the mind) simultaneously
Denying the existence of either the body (immanence) or the mind (transcendence)
As defined in the present study: Acknowledging one’s SPT-damaged body in working
life while being constructed as a subject by others relevant to the work context
As defined in the present study: Being denied subjectivity at work by others or con-
structing oneself as a subject in denial of the SPT-damaged body of freedom (transcendence) or reproduction (imma-
nence) through their bodies and bodily processes, such
as pregnancy and childbirth. These conflicts complicate
transcendence: therefore, a woman’s body becomes an
obstacle to freedom. Study design A constructivist approach to grounded theory (GT),
according to Charmaz [34], inspired this study design. A
GT of a studied topic starts with empirical data and ends
with rendering them in an explanatory theory. Charmaz’s
constructivist approach to GT also seeks to understand
and explore complex social phenomena (such as working
with SPT-related morbidities). Charmaz also stresses that
the researcher is not a neutral observer but a co-partic-
ipant in the study; thus, the findings are not discovered
but constructed by the researcher and the study par-
ticipants. During this construction process, we applied
Simone de Beauvoir’s theoretical framework [26]. Previously, Beauvoir’s ideas have been applied in
research on sexual and reproductive health as well as
working life. Studies on childbirth experience [27],
motherhood [28], labour pain [29], and obstetric vio-
lence [30] problematise the gendered ambiguity embed-
ded in these lived situations. Embodied ambiguity here
embraces both possibilities (I can) and limitations (I
cannot), subjectivity and materiality [29], independence
and relationality [30], as well as stagnation and freedom
[27]. Further, Beauvoir [26] views engagement in paid
work outside the domestic field as a path of transcend-
ence and freedom. Beauvoir [31] also critiques the gen-
der-divided labour market where women are confined
to certain femininely coded occupations with fewer
career options. Veltman [32] states, using a Beauvoirian
lens, that contemporary women are hindered in their
transcendence because they are burdened with double
work duty: employment outside the home and house-
hold chores. Further, Ross [33] analyses the professional
role of nurses stating that the immanent values of ‘help-
ing’ and ‘caring for others’ have been transformed into
transcendent activities by present-day nurses. Gender perspectives on SPT For Beauvoir, women’s bodies are
also the scene for constant body shame tied to sexuality
and reproductive functions such as menstruation, sex-
ual maturation, the postpartum body, or the develop-
ment of breasts in puberty. Hence, ‘becoming a woman’
is a lesson in body shame. Thus, shame controls and
oppresses women’s bodies regarding appearance and
behaviour. Consequently, a woman cannot transcend
due to her eternal status as an object and is therefore
assigned to immanence. of freedom (transcendence) or reproduction (imma-
nence) through their bodies and bodily processes, such
as pregnancy and childbirth. These conflicts complicate
transcendence: therefore, a woman’s body becomes an
obstacle to freedom. For Beauvoir, women’s bodies are
also the scene for constant body shame tied to sexuality
and reproductive functions such as menstruation, sex-
ual maturation, the postpartum body, or the develop-
ment of breasts in puberty. Hence, ‘becoming a woman’
is a lesson in body shame. Thus, shame controls and
oppresses women’s bodies regarding appearance and
behaviour. Consequently, a woman cannot transcend
due to her eternal status as an object and is therefore
assigned to immanence. Still, there is a need to theorise postpartum morbidity
in various contexts. In the following, the Beauvoirian lens
is applied to women with SPT and their working life as
lived situation. Study setting and Swedish context A national sample of Swedish women with self-reported
persistent SPT-related morbidities was recruited using
the closed Swedish Facebook community ‘Förlossn-
ingsskadad? Du är inte ensam!’ (‘Injured at childbirth? You are not alone!’). This social media community
addressing perineal trauma had over 7 600 members
(during recruitment in 2020 to 2021), and only women
with SPT can become members [35].h The employment rate of Swedish women (aged
20–64 years) was 78% in 2021, the highest rate in the
European Union (EU) [36]. The healthcare (76.1%) and
education sectors (71.1%) are dominated by women [37]. Sweden was estimated to have closed 82.3% of its over-
all gender gap by 2021, placing the country in the global
top ten [38]. Women’s labour market outcomes decline Tjernström et al. BMC Women’s Health (2023) 23:47 Tjernström et al. BMC Women’s Health (2023) 23:47 Page 4 of 13 as her problems did not relate to her working life. Final
interviews were performed in February 2021. after childbirth, salary and income decrease, and sick
leave increases compared to men and pre-childbirth. This so-called child penalty effect may be explained by
health status and gender norms [39]. The sickness rate is
higher for Swedish women (26.35%) than men (17.78%)
[36]. Swedish women’s sick leave at five years postpartum
is mainly due to psychiatric diagnoses, such as depres-
sion or anxiety, and musculoskeletal diagnoses [39]. Since
ICD-10 (International Statistical Classification of Dis-
eases and Related Health Problems, 10th Revision) diag-
noses for late complications after perineal trauma were
implemented as late as 2020 in Sweden [40], representa-
tive statistics on diagnoses and related sick leave are una-
vailable. The parental benefit consists of 480 days of paid
leave (80% of annual gross wage) per child up to 12 years
of age, and women are overrepresented in parental leave
(70%) and leave to tend to sick children (60%) [36]. Sampling and procedure Purposive sampling was applied to aim for a heteroge-
neous sample of characteristics and SPT-related mor-
bidities. The inclusion criteria were vaginal deliveries
irrespective of mode, live-born child/children, third or
fourth-degree perineal lacerations from January 2016
until June 2019, and self-reported persistent SPT-related
morbidities. Further inclusion criteria were women
above 18 years of age who understand and speak Swedish
and have finalised the parental leave period (i.e., work-
ing, searching for work, studying or on sickness benefit). Exclusion criteria were stillbirths or death of an infant
within 28 days, ongoing parental full-time leave after the
injury, or subsequent ongoing pregnancy.h i
Interviews were conducted via Zoom® due to the
Covid-19 pandemic and audio-recorded with the wom-
en’s consent. The interviews were carried out by the first
author (KT), except for two interviews early in the data
collection process, which IL or MW also attended. The
interviews lasted 36–112 min with a median of 65 min
and provided comprehensive data. Recordings were tran-
scribed verbatim and validated for accuracy by listening
to all recordings while reviewing the transcripts. The
transcribed data were coded using MAXQDA® [41], a
software for analysing qualitative research data. Data collectionh The researchers developed a semi-structured interview
guide focusing on working life. The construction of the
interview guide was based on literature reviews, the sen-
sitising concept of ‘gender’, and the clinical pre-under-
standing of the authors. KT, IL, and MP are midwives,
and MW is a physiotherapist with expertise in gender
theory. The interview guide was pilot tested by a woman
who met the inclusion criteria except when the injury
occurred and resulted in no significant changes. The indi-
vidual interviews started with a broad, open-ended ques-
tion concerning if and how SPT-related morbidity had
affected working life. Additional questions covered physi-
cal, emotional, or social impact, management of symp-
toms, treatment by superiors or colleagues, sick leave,
and thoughts about the future. Probing questions clari-
fied and explored the concepts more profoundly. Participants Background characteristics are found in Table 2. Par-
ticipants were all cis women, i.e., identifying themselves
as women, and subsequently referred to as ‘the women’
in this paper. The median age was 35.5 years (min.–
max.: 25–42 years) and the mean number of children
was 2 children (min.–max.: 1–4 children). All women
were married or cohabited with their partners. The self-
reported morbidities varied from urine or faecal inconti-
nence, pain, sexual dysfunction, defecation problems, and
prolapse to psychological symptoms, and most women
reported a combination of morbidities. Most women had
partial parental leave (an option within the social security
system to reduce working hours for parents with young
children). Women with no sick leave periods related to
their SPT reported wishing they had been offered the
possibility. More specifically, during the theoretically informed
phase of abduction, several gender theoretical perspec-
tives were discussed to understand the meaning of data
(e.g., intersectionality, crip theory or corporeal femi-
nism). Since the emerging categories and social pro-
cess pointed towards an equivocal, double-sided, and
conflicting situation of passivity/stagnation (body) vs
activity/development (mind) in working life, Simone
de Beauvoir’s notions on immanence and transcend-
ence [26] were adopted (see Table 1). This contributed
to understanding the phenomenon at an abstract level
and aided in further conceptualising and developing a
theoretical model grounded in the emerging data. The
emergent theoretical model was also peer-reviewed and
discussed at research seminars. Negotiating the ambiguity of an (in)authentic working life
When returning to working life, the persisting conse-
quences and traumatised bodies were the women’s shared
experience and mutual departure as professionals. How-
ever, their return to work was multifaceted and discrep-
ant, forming the base for our grounded theory of their
situated embodiment. The theoretical model, inspired by
Beauvoir [26], and the related core category ‘Negotiating
the ambiguity of an (in)authentic working life’ reflected
their negotiations of the conflicting gendered process of
immanence versus transcendence at work (Fig. 1). The
women negotiated work immanence as ‘the silent covert
object’ by objectification and professional stagnation due
to SPT. However, they could also experience work tran-
scendence as ‘the resourceful overt subject’ who gained
professional competency and freedom from restrictions
at work despite SPT.h Results
Participants and the theoretical framework supported the construc-
tion of categories and gave ideas on theoretical con-
structs. Finally, theoretical coding identified how our
categories related to each other. Abduction, memo-
writing, and visual reflections aided this process. Ana-
lytical saturation was reached when the same concepts
kept emerging, and after ten interviews, no new signifi-
cant concepts contributed to our properties. Data analysis The community administrators posted a poster con-
taining brief information about the study and a web
link to the study homepage on November 23rd, 2020. Eligible participants could access further written infor-
mation on the study homepage and contact the respon-
sible researchers for additional information. Women who
wished to participate notified the researchers and were
contacted via telephone to ensure the fulfilment of inclu-
sion criteria. According to the emergent study design,
participants were included collaterally with interviews
and preliminary analysis. A total of 12 participants con-
tacted the research team during the first call. Interviews
contained vivid, detailed descriptions of women’s experi-
ences of working life and provided an extensive amount
of data. Thus, our properties became analytically satu-
rated when approaching ten interviews, providing the
same concepts towards the end of data analysis. There-
fore, there was no need to post another call for partici-
pants on the Facebook community. One woman emailed
her interest to partake but did not respond to our contact
efforts. One participant was excluded from this analysis In line with Charmaz [34], the coding process was ini-
tially inductive, followed by an abductive stage in which
we applied the gender theoretical framework of Simone
de Beauvoir [26]. The analysis started collaterally with
the data collection, and the authors repeatedly discussed
the emerging findings. Memos were written after each
interview and during the analysis. All authors coded one
of the first interviews, and the emerging initial codes
were triangulated. Following Charmaz [34], KT induc-
tively coded the remaining interviews naming lines or
segments of data with predominantly longer in vivo and
in vitro codes close to the data trying not to lose nuances. Verbs and gerunds (i.e., verbs ending with ‘-ing’) in initial
codes aided in exploring inferred meanings and actions
in the data—sparking interest to investigate different
emergent links by constantly comparing data and going
back to the text. After the inductive initial coding process, the codes
were clustered and visualised to identify the first cen-
tral concepts during focused coding. Further, memos Tjernström et al. BMC Women’s Health (2023) 23:47 Tjernström et al. BMC Women’s Health (2023) 23:47 Page 5 of 13 Page 5 of 13 Ethical considerations
Th
h
j The research project was performed in accordance
with the Declaration of Helsinki [42] and approved
by the Swedish Ethical Review Authority (Dnr: 2020-
035410908). Participation was voluntary, and the
women could withdraw without stating a reason. Informed consent was mandatory and was obtained
digitally before each interview. The women could
refrain from answering questions they felt were too
intimate. The project homepage and the informa-
tion letter provided information on available support
options. If any woman experienced negative emotions
after the interview, researchers could assist in referral
to adequate healthcare services. The interpretation of ‘ambiguity’ as the core of the
traumatised professionals’ situated embodiment and pro-
fessional ‘becoming’ refers to the women’s negotiations Table 2 Group-level demographic background data of 10 included participants
Demographic category
Number (n)
Country of origin
Sweden (7); Estonia (1); Syria (1); South Korea (1)
Degree of perineal laceration
Third-degree (7); Fourth-degree (3)
Year of SPT
2016 (4); 2017 (5); 2019 (1)
Education level
University level (7); Higher Secondary (2); Lower secondary (1)
Type of occupation
Physically challenging (5); Sedentary (5)
Type of employment
Permanent (7); Substitute (1); Unemployed (2); Student (2)
Employment rate
Full-time (4); Part-time (6) [20 to 83% of full-time]
Reason for sick leave
Reconstructive surgery (5); Other reasons than SPT (4) Table 2 Group-level demographic background data of 10 included participants Number (n) Demographic category Country of origin
Degree of perineal laceration
Year of SPT
Education level
Type of occupation
Type of employment
Employment rate
Reason for sick leave Tjernström et al. BMC Women’s Health (2023) 23:47 Page 6 of 13 Fig. 1 The theoretical model on women’s negotiations regarding the ambiguity of an (in)authentic working life ‘Internalising shame versus externalising the trauma’ At an individual level, the women negotiated to inter-
nalise the shame of SPT versus externalising the trauma
when facing the ambiguity of an (in)authentic working
life. Embodying ‘a silent covert object’ meant hiding their
condition at work, and shame reinforced this process. Internalised shame implied that the women established
a workplace behaviour characterised by adjustment, iso-
lation, avoidance, or prevention, which interfered with
their workability. The concealing behaviour was inter-
preted as passivity, stagnation, and resignation, corre-
sponding to immanence. Thus, the women became an
object; the women were ashamed of their SPT-related
morbidities, and the workplace situation reinforced this
feeling. After all, it [incontinence] becomes some sort of
shame. It does not fit the norm to be just over 30 and
need a diaper. (Interview 8) Shame restricted them from telling others at the work-
place, asking for help, seeking medical help or sick leave. Some women used parental leave to cover up for dimin-
ished work capacity or came up with excuses for not par-
ticipating in certain activities. Further, they were afraid
that the SPT-related morbidities might damage their pro-
fessional reputation and negatively affect how colleagues
viewed them as professionals. As a result, they internal-
ised others’ gaze and patronising in their self-esteem: In contrast, the gendered process of becoming ‘a
resourceful overt subject’ represents a gendered con-
struct characterised by transcendence at the workplace
despite the sequelae of SPT. Subjectification was por-
trayed by ‘externalising the trauma’ at an individual level,
‘intersubjective corroboration’ at an organisational level,
and ‘institutionalised subjectification’ at an institutional
level. To be understood and constructed as a subject
conveyed the possibility of revealing the SPT, receiv-
ing support or needed aids at the workplace, and being
seen as an individual in need of care and sick leave by
societal institutions. The construct of women with SPT-
related morbidities as subjects was crucial to experienc-
ing transcendence at work. Thus, the women could break
free from restrictions at work due to SPT and embod-
ied ‘a resourceful overt subject’. Although construct-
ing themselves as a working subject, some women faced They [my colleagues] might think: Well, what does
this [SPT] imply? Is this a person who maybe does
not come to work because she must sit on the toilet
at home? model on women’s negotiations regarding the ambiguity of an (in)authentic working life external ‘others’. An inauthentic working life was also
experienced when women constructed themselves as
subjects in denial of their facticity, i.e., a body with SPT. On the other hand, the women who gained access to
their subjectivity through openness, support, and sub-
jectification by ‘others’ achieved an authentic working
life while simultaneously embracing their bodies with
SPT. and efforts to be object and subject simultaneously. Specific individual, organisational, and institutional
aspects played a role in this conflicting gendered process
(Table 3). For instance, the women could be constructed
or construct themselves as ‘transcending subjects’ at the
individual and interpersonal level at work, e.g., by being
open about their SPT and receiving support. Neverthe-
less, they were simultaneously objectified at the insti-
tutional level because healthcare services denied them
adequate help and sick leave. This conflicting gendered
process highlighted the vulnerability of working women
with SPT-related morbidities. The theoretical model should be understood in the
context of gendered societal norms visualised as an
external context consisting of society, workplaces,
healthcare institutions, and national expert bodies,
such as the social security agency or policymakers. Further, inauthenticity at work was reflected in the
denial of subjectivity by objectification via internal or Table 3 Overview of core category, categories, subcategories, and properties
Core category
Negotiating the ambiguity of an (in)authentic working life
Category
Immanence at the workplace—the silent covert
object
Versus
Transcendence at the workplace—the resourceful overt
subject
Subcategory
Internalising shame
Versus
Externalising the trauma
Properties
Hiding the SPT
Shame as a restrictor
Being open about SPT
Being free of shame
Subcategory
Intersubjective tabooing
Versus
Intersubjective corroboration
Properties
The restricting workplace
Tabooing SPT at work
Needing flexibility
Receiving support and
understanding
Subcategory
Institutionalised objectification
Versus
Institutionalised subjectification
Properties
Seeing women as objects
Struggling endlessly
to get sick leave
Finally receiving care
Seeing women as subjects Page 7 of 13 Tjernström et al. BMC Women’s Health (2023) 23:47 Tjernström et al. BMC Women’s Health Immanence and transcendence at the workplace–being
‘the silent covert object’ versus ‘the resourceful overt
subject’h ambiguous situations where the workplace or healthcare
services did not recognise their subjectivity.h The ambiguity of an (in)authentic working life and
being ‘a silent covert object’ and ‘a resourceful overt
subject’ entailed negotiations, i.e., ‘internalising shame’
versus ‘externalising the trauma’ (individual level), ‘inter-
subjective tabooing’ versus ‘intersubjective corroboration’
(organisational level), and ‘institutionalised objectifica-
tion’ versus ‘institutionalised subjectification’ (institu-
tional level). In the following, we present the negotiations
in more detail with citations from the interviews. The (in)authentic working life comprised the negotia-
tion of the ambiguity of being ‘a silent covert object’
and ‘a resourceful overt subject’, which took place at an
individual, organisational, and institutional level. Becoming ‘a silent covert object’ represented a gen-
dered construct characterised by immanence and
objectification at the workplace. The objectification
occurred at different levels; ‘internalising shame’ at an
individual level, ‘intersubjective tabooing’ at an organi-
sational level, and ‘institutionalised objectification’
at an institutional level. Being seen as an object made
women with SPT-related morbidities hide their condi-
tion because SPT was taboo in the workplace. Moreo-
ver, objectified women perceived themselves as ignored
by healthcare services and, thus, received inadequate
care. This lack of care contributed to stagnation at
work because no rehabilitating measures were available
to improve the work situation. Thus, the women were
confined to embodying ‘the silent covert object’. Their
‘bodies’ were problematic in this gendered construct,
and their resources as ‘minds’ were silent. Objectifica-
tion was not a fixed status but ‘ambiguous’ and opened
for negotiation as the women could embody both sub-
ject and object at different levels or in different work
situations. They could define themselves as subjects
but be objectified by others, or vice versa. The women
experienced limited career options because they, for
example, had to quit their jobs, put their studies on
hold, and could not work as intended. Further, the
women described socio-economic limitations. They
used parental social security benefits to compensate
for a default sick leave or reduced work time without
compensation. ‘Internalising shame versus externalising the trauma’ (Interview 3) Some women adjusted their workplace behaviour by
varying work tasks, working fewer hours, or perform-
ing more manageable tasks on a ‘bad day’. Moreover, the
women isolated themselves professionally. This reduced
work travels and colleague-related stress, gave easy access
to a toilet or shower after bowel accidents, and facili-
tated alternating postures to relieve their aching bodies. Page 8 of 13 Tjernström et al. BMC Women’s Health (2023) 23:47 Tjernström et al. BMC Women’s Health (2023) 23:47 Tjernström et al. BMC Women’s Health (2023) 23:47 Avoidance manifested in some women not seeking occu-
pational healthcare and, therefore, denying their trauma. Avoidance manifested in some women not seeking occu-
pational healthcare and, therefore, denying their trauma. The women also avoided exposed situations to hide their
condition from colleagues, such as afternoon meetings
for fear of incontinence after lunch or work travels due to
limited access to toilets. The women described how they
avoided larger groups, heavy lifting, or high-stress levels
because they feared public incontinence. Some women
postponed visits to the bathroom to ensure privacy and
preferred visiting the toilet at home. ‘Intersubjective tabooing versus intersubjective
corroboration’ ‘Intersubjective tabooing versus intersubjective
corroboration’ The women balanced between an intersubjective taboo
where SPT was silenced at work and, on the other hand,
an intersubjective corroboration consisting of under-
standing, support, and flexibility at work. l
Objectification and detention in immanence were done
both ‘by’ the women and ‘to’ the women. Intersubjective
tabooing showed how the workplace objectified women
by treating SPT as a taboo with restricted openness. Some women perceived their workplace as closed and
intolerant, so they remained silent about their situation. I partly do not want them [my colleagues] to notice
that it [toilet visit] takes some time. Moreover, in no
way do I want to push [because my symptoms may
worsen]. And when I am stressed, I may push. So,
I must make the body feel calm to relax because it
[faeces] then comes easier without pushing. (Inter-
view 10) Well, it feels like you try to hide it as much as pos-
sible. It is not like saying, ‘I have diabetes, so if I
collapse, give me a shot’. It does not feel like such a
thing. (Interview 6) Hence, they were met with incomprehension and a lack
of support. Moreover, a restricting workplace made the
women feel that SPT-related problems should only be
mentioned if necessary. Consequently, many women did
not tell their superiors or colleagues about their SPT, and
if they shared their SPT, they did not disclose any details
or only told certain people. The women revealed preventative measures to con-
ceal their SPT, e.g., emptying their bowels before work,
adjusting their nutrition, and wearing incontinence pads. They preferred digital meetings or worked effectively to
cope and cover for a potentially bad period that might
strain them. The women always ensured they had quick
access to a toilet. A sense of security was established by
knowing where the nearest toilet was. Unfortunately,
locating the closest bathroom became an obsession for
some women who internalised the shame attached to
public incontinence. No, not my boss, […] because he would have no sym-
pathies. I have only told him that I do not feel well
and have complications. But I have not told him
why. I was not comfortable telling my boss. (Inter-
view 8) No, not my boss, […] because he would have no sym-
pathies. I have only told him that I do not feel well
and have complications. ‘Intersubjective tabooing versus intersubjective
corroboration’ But I have not told him
why. I was not comfortable telling my boss. (Inter-
view 8) But I am very thorough with knowing where a toilet
is [located]. When I walk down the corridor, I know
which direction to turn. Do I have to go one floor up
or down? Do I have to ride the elevator? Do I have
access to the elevator? (Interview 5) The women experienced SPT as a gendered taboo in
the working environment; being an injured or inconti-
nent working young woman was not the norm. This gen-
dered taboo also functioned as a barrier to being open
about SPT at work. Most women felt uncomfortable tell-
ing men about the injury and did not feel understood by
male co-workers. In contrast, constructing themselves as subjects by
being open about the SPT at work was a prerequisite to
transcendence. It seemed crucial for the women to be
honest about their SPT to become subjects at an indi-
vidual level. Nevertheless, revealing the SPT was a long
process where the women seemed to become more con-
fident in sharing their SPT-related morbidities the longer
they had been dealing with them. The women noted that
being honest about their situation could be therapeutic,
and they thought their revealing process could empower
others to tell. Thus, these women embodied a subject
where shame did not play a significant part. Women con-
ceptualising themselves as subjects also listened to their
needs and revealed positive thinking and hope. It works both ways. He [the boss] is an older man,
nearly 70 years old. Furthermore, he generally has
minimal sympathy for women, and it was no idea to
discuss this with him. So, it was both his personality,
and he was male. I think a man cannot understand
in the same way [as a woman]. (Interview 8) Although SPT was perceived as taboo, some women
felt forced to disclose their injury to get support or
adjustments at work before reconstructive surgery or to
explain any deviant behaviour. In contrast, Beauvoir [26] suggests that a woman is not
solely responsible for her status as an object or subject. Thus, intersubjective corroboration portrayed the wom-
en’s need for specific aids at work to take an active role as
a working subject and functioning at work. The women So, I am frank; I have told them [my colleagues] eve-
rything. ‘Intersubjective tabooing versus intersubjective
corroboration’ I perceive them to be really great and under-
standing. (Interview 7) So, I am frank; I have told them [my colleagues] eve-
rything. I perceive them to be really great and under-
standing. (Interview 7) Tjernström et al. BMC Women’s Health (2023) 23:47 Page 9 of 13 described that flexibility during the workday was essen-
tial to work despite SPT-related morbidities; they needed
to plan their workday, making them feel competent and
able-bodied. covert object’ at work. Lack of care led to internalised
acceptance; the health problems became normality, and
the women were forced to ignore their symptoms to
manage. My work is relatively flexible, so I can slip away for
a little while if I want to. I think it is worse to have
a four-hour shift at the register at the supermarket
because then you are supposed to sit there. (Inter-
view 10) My pain level is so incredibly high that I do not
remember what it was like not to be in pain. So, as
every day goes by, I endure more and more pain. […]
Well, it really hurts, but I have gotten used to it. Eve-
ryday life. (Interview 1) Openness about SPT decreased stress levels and cre-
ated a sense of security and relief, enabling the women
to receive support and understanding. As a result, col-
leagues could take care of specific work tasks. Never-
theless, the women needed genuine interest and trust to
reveal their SPT at work. Consequently, the women could
access understanding, support, and flexibility, enabling
them to embody working subjects. In contrast, acknowledgement as a subject was cru-
cial for transcendence. It was mirrored in the women’s
experiences of subjectification by healthcare providers
and the social security agency, which enabled access to
medical care and sick leave. For example, some women
described how healthcare providers acknowledged them
and saw them as ‘actual patients with real symptoms’
needing treatment and sick leave. I was so lucky to get hold of a physician who went
totally nuts [finding out about my situation]. So,
she has helped me to get help. So that physician, all
credit to her, I have had so much help and support
from her. She did not let go of me. (Interview 8) I remember when we were supposed to go to xxx, we
were supposed to sit in the car for two hours. ‘Intersubjective tabooing versus intersubjective
corroboration’ So, I
said, ‘Well, I decline’. I did not want to sit in that car
for two hours with other people and not know where
to use the bathroom during the car trip. So, I dis-
cussed this with my boss, and I did not have to go;
I could decide if I wanted to join. So, I chose to stay
and work instead. (Interview 5) The extent of their SPT was finally discovered if they
met a healthcare provider who listened to them. Symp-
toms were taken seriously, which was experienced as
a confirmation of not being mentally ill or imagining
problems. Thus, the women appreciated an explanation
of their difficulties and adequate help and support. This
resulted in liberating feelings and validation of their sub-
jectivity, which enabled transcendence in working life. ‘Institutionalised objectification versus institutionalised
subjectification’ At the institutional level, the women faced both objecti-
fication and subjectification by healthcare services or the
social security agency in their negotiations of the ambi-
guity of an (in)authentic working life.hi So, to be taken for real, to be listened to as a woman
and not only reject women’s problems regarding
birth injuries. (Interview 5) The institutionalised objectification manifested in
healthcare providers’ perceptions of women and their
genitals as objects and ignoring their healthcare needs. This lack of care left the women struggling to receive
appropriate help and sick leave if needed. Consequently,
a neglected diagnosis, inadequate support, and denial
of sick leave enforced the women’s immanence at work. Some women experienced that the healthcare profession-
als reduced them to incorporating genitals when asking
for help. Some women noted that SPT-related morbidi-
ties were taboo in healthcare services rendering these
women voiceless because nobody was listening. Health-
care providers made them believe that symptoms were
imagined by calling persistent symptoms ‘normal’ or a
‘psychiatric condition’. Some women only received sick
leave for mental issues, which was perceived as gendered. Such objectification left the women feeling restricted at
work and depressed due to a physical ailment. A denial
of sick leave bound the women to incorporate ‘a silent In summary, working life for women with SPT was
experienced as ambiguous and (in)authentic. The women
negotiated to embody ‘the silent covert object’ and ‘the
resourceful overt subject’, leading to a conflicting gen-
dered process when they returned to work. The women’s
experiences led them to perceive themselves as objects
contained in immanence and subjects transcending any
restrictions at work. Discussion Our conceptualisation of internalised shame indi-
cated that a body with SPT-related morbidities was not
congruent with existing norms of corporal occupational
femininity and therefore silenced at work. Shame [48] is
a lived experience and is consequently embodied. There
must be an audience and ‘gaze’, internal or external, for
shame to occur. Shame can therefore be experienced due
to an internal self-devaluation process where the subject
judges itself through an internalised other, e.g., a ‘male
gaze’ or a ‘biomedical gaze’. These ideas align with our
findings, where shame seemed to urge an internalisa-
tion process of hiding SPT at work. Such concealments
of health conditions at work [49, 50] are documented for
other illnesses such as migraine, fibromyalgia, or multiple
sclerosis. In our study, the women struggled to disclose
SPT at work, negotiating the ambiguity of ‘internalising
shame’ and ‘externalising the trauma’. Similar ambigu-
ity and dilemmas are presented in other studies [51, 52]
where workers struggle with disclosing chronic illnesses
at work. Further, menstruation [53], pregnancy [54], and
breastfeeding [55] are reproductive processes depicted
as taboo in the workplace. Herron-Marx, Williams and
Hicks [45] show that perineal morbidity is a taboo for
women themselves and society. Moreover, Keighley et al. [20] describe the hidden taboo of living with SPT. These
study findings cohere with our conceptualisation of SPT
as an objectifying taboo at the workplace and a demon-
stration of power by shaming working women with devi-
ant bodies. From a Beauvoirian perspective [26], paid work is a
way towards women’s transcendence. Based on our con-
ceptualisation, this calls for embracing one’s SPT factic-
ity at work and being subject and object in symbiotic
ambiguity to achieve an authentic working life. Enhanced
workability also requires a supportive workplace and
organisation. As reflected in our study, findings from
previous studies show that women with SPT may need
to change work tasks, occupations, or workplaces [19,
20]. Further, SPT has been found to reduce work capac-
ity [7], which was also visualised in our results; women
developed an individualised workplace behaviour con-
sisting of adjustment, isolation, avoidance, or preven-
tion with a negative impact on their work capacity. The
women in our study even used parental leave to cover up
their reduced workability. Further, in previous research,
women experience insecurities regarding incontinence
at work [19], similar to our participants, who described
how internalised shame made them hide their condition. Discussion Our generated theoretical model, in which we applied
the ideas of Simone de Beauvoir [26], demonstrated
how the women with persistent SPT-related morbidities
negotiated an (in)authentic working life and faced ambi-
guity regarding being object and subject. This entailed
a negotiation of being objectified and doomed to ‘the
silent overt object’ or subjectively constituted to be a Tjernström et al. BMC Women’s Health (2023) 23:47 Page 10 of 13 Page 10 of 13 Tjernström et al. BMC Women’s Health ‘resourceful overt subject’. The conflicting gendered pro-
cess of either becoming ‘the silent covert object’ or ‘the
resourceful overt subject’ at work highlighted the vulner-
ability of the affected women, tied to inauthentic work-
ing life. In turn, authenticity referred to an integration of
these constructs. adverse effects of SPT on working life. Only one study
by Keighley et al. [20] describes that work may function
as a distraction from SPT-related anal incontinence. Our
theoretical model suggested multiple aspects of tran-
scendence concerning working life. Externalising the
trauma, intersubjective corroboration, and institutional
subjectification were positive inclinations supporting the
working women with SPT in our study. Thus, they could
function at work despite SPT. As mentioned, the research
on SPT-related morbidities at work is limited [19, 20]. Our findings corroborated these few studies but may also
deepen the insights into the occupational situation of
women with SPT. In terms of immanence and/or transcendence, our
model illuminated an ambiguity regarding the situated
embodiment of women with SPT at work. This contrasts
with Beauvoir [26], who understands women’s bodily
processes as solely immanent projects (in a patriarchal
context). Therefore, our study may contribute significant
insight into the complexity of SPT and differing working
conditions. Likewise, postmodern feminist theorists [43]
disregard Cartesian dualism and argue for an indivis-
ible connection between the corporeal and the culturally
constructed body. Our conceptualisation of ambiguity
might also be compared with prior research on childbirth
experience [27] and motherhood [28], concluding that
childbirth and motherhood in terms of immanence/tran-
scendence might be ambiguous experiences. Childbirth
[27] both limits women’s freedom in life (immanence)
and is interpreted as a meaningful and developing pro-
ject for future life (transcendence). Further, motherhood
[28], i.e., pregnancy, birthing, breastfeeding, and raising
children, can be viewed as an ambiguous concept: both
repetitive (immanence) and creative, joyful as well as
empowering (transcendence). Discussion Other studies have also found institutionalised objec-
tification [44–46], where healthcare professionals nor-
malise persistent morbidities or question the legitimacy
of symptoms. The long struggle for adequate healthcare
depicted in our study is also supported by others [25,
44, 47]. Hence, previous research shows almost solely Some women in our findings used partial parental
leave to cover up their impaired workability. This might
misrepresent women’s need for work adjustments or
sick leave absence and underestimate their health status
postpartum. Generally, pregnancy and childbirth cause
no new medical diagnoses requiring sick leave, but sick
leave periods increase after childbirth [39]. Underly-
ing health problems, more common among women
with lower socio-economic status, can be reinforced
after childbirth [39]. Postpartum morbidity can then be
disguised by parental leave. Our study highlighted that Tjernström et al. BMC Women’s Health (2023) 23:47 Tjernström et al. BMC Women’s Health (2023) 23:47 Page 11 of 13 However, the ten included interviews contained vivid,
detailed descriptions providing extensive data, and the
same concepts reappeared towards the end of the analy-
sis. Thus, sufficient information power [57] was obtained,
and the researchers are confident that the constructed
theoretical model was grounded in data that may reso-
nate with the whole picture. In addition, three women
were born outside of Sweden, which adds to resonance
[34] and is a strength as immigrant women are often
excluded from research. Our study also included women
with both sedentary work and physically demanding
occupations, which further strengthens the resonance
of the paper. Nevertheless, future studies should address
the possible differences in experiences depending on the
type of occupation. Our sample included only cis women
in relationships, which may not reflect the experiences
of same-sex parents or single parents. Moreover, most
women had higher education and the majority had third-
degree lacerations. This does not fully reflect the experi-
ences of women with fourth-degree lacerations or those
with lower levels of education. Our results also included
more data on immanence than transcendence. These
aspects impact the paper’s resonance, and future studies
should focus on these groups. childbirth might imply new health problems that nega-
tively influenced the women’s work situation and might
delay professional transcendence and career. In a wider
Beauvoirian sense [26], women with SPT do not only
strive towards transcendence in regard of freedom from
impaired workability due to their SPT but also in regard
of economic freedom after childbirth. Strengths and limitations Charmaz [34] speaks of credibility, originality, resonance,
and usefulness regarding trustworthiness.h The transparent description of our methodology, a
detailed description of our categories, including exten-
sive and vivid descriptions of various experiences, and
interdisciplinary authors’ triangulation enhanced cred-
ibility [34]. Our preunderstanding, a part of credibility
[34], origins from the clinical expertise of midwifery or
physiotherapy and research addressing perineal trauma
and gender studies. We identify as women, feminists, and
mothers. The positive aspects may include a vast clinical
experience of the subject, and some intimate questions
may be easier to discuss with a female researcher. On
the other hand, our preunderstanding of the problems
may bias our perceptions. However, our different pro-
fessional backgrounds and repeated triangulation within
the research group and in seminars may have minimised
any possible preunderstanding bias. In line with Charmaz
[34], we also practised reflexivity by being aware of our
preunderstanding, using it to construct our theoreti-
cal model, and routinely writing memos on analytical
choices. Discussion Working only part-
time or compensating default sick leave with parental
benefit leads to lesser income and pension compared to
their partners; an economical gap that women globally
are already trying to close [22, 38]. This financial loss
places women with SPT back in financial dependence on
their partner; an ‘oppression’ that paid work itself should
have freed them from according to Beauvoir [26]. However, the ten included interviews contained vivid,
detailed descriptions providing extensive data, and the
same concepts reappeared towards the end of the analy-
sis. Thus, sufficient information power [57] was obtained,
and the researchers are confident that the constructed
theoretical model was grounded in data that may reso-
nate with the whole picture. In addition, three women
were born outside of Sweden, which adds to resonance
[34] and is a strength as immigrant women are often
excluded from research. Our study also included women
with both sedentary work and physically demanding
occupations, which further strengthens the resonance
of the paper. Nevertheless, future studies should address
the possible differences in experiences depending on the
type of occupation. Our sample included only cis women
in relationships, which may not reflect the experiences
of same-sex parents or single parents. Moreover, most
women had higher education and the majority had third-
degree lacerations. This does not fully reflect the experi-
ences of women with fourth-degree lacerations or those
with lower levels of education. Our results also included
more data on immanence than transcendence. These
aspects impact the paper’s resonance, and future studies
should focus on these groups. Conclusions We showed that women with persistent SPT-related
morbidities after childbirth experienced an ambiguity
of (in)authentic working life. The women had to con-
stantly negotiate being objectified and doomed to ‘the
silent overt object’ or subjectively constituted to be a
‘resourceful overt subject’. Creating a working environ-
ment characterised by understanding, support, and flex-
ibility is essential to eradicate the taboo of SPT-related
morbidities in working life. Thus, policymakers, employ-
ers, healthcare services, and society should focus on
enabling subjectification and transcendence in work-
ing life. Therefore, further research is needed to develop
an occupational rehabilitation programme focusing on
SPT-related morbidities, which, to our knowledge, does
not exist. Furthermore, healthcare services, employers,
and social security agencies should collaborate to pro-
vide adequate medical help and support so that women
can access social security resources when needed. Such
measures can support transcendence towards an authen-
tic working life and a gender-equal working environment
that does not deprive women of career opportunities due
to a physical ailment. Further, our study has originality [34] because we con-
tribute novel findings and a theory grounded in data to
an area that, to our knowledge, has not been researched
before. Thus, our study conveys new insights and concepts that
are socially and theoretically significant since they may
prove helpful to policymakers, employers, and healthcare
services [34]. Further, the findings, depicted with thick
descriptions, reflect the experiences of our sample and
might be transferable to other women with SPT in similar
working contexts, adding to the usefulness of the paper. When saturation is obtained is a subject of debate;
some suggest about 12 participants [56], while others
argue for up to 20 participants [34]. This paper could have
had a larger sample by repeating the call for participants. Abbreviations
EU
European Union
GT
Grounded theory Funding Open access funding provided by Umeå University. Grants financing this pro-
ject were obtained from The Research Lift and Strategic Research Area Health
Care Science (SFO-V), Umeå University. The funding bodies did not influence
the design of the study or collection, analysis, and interpretation of data or
writing of the manuscript. 11. Skolverket. Barn och personal i förskola - Hösten 2021. 2022. https://www.
skolverket.se/publikationsserier/beskrivande-statistik/2022/barn-och-
personal-i-forskola---hosten-2021. Accessed 30 Nov 2022. 12. Coyne KS, Sexton CC, Irwin DE, Kopp ZS, Kelleher CJ, Milsom I. The
impact of overactive bladder, incontinence and other lower urinary tract
symptoms on quality of life, work productivity, sexuality and emotional
well-being in men and women: results from the EPIC study. BJU Int. 2008;101(11):1388–95. Acknowledgements 8. Lindqvist M, Lindberg I, Nilsson M, Uustal E, Persson M. “Struggling to
settle with a damaged body” – a Swedish qualitative study of women’s
experiences one year after obstetric anal sphincter muscle injury (OASIS)
at childbirth. Sexual Reprod Healthc. 2019;19:36–41. Consent for publicatio
Not applicable. 20. Keighley MR, Perston Y, Bradshaw E, Hayes J, Keighley DM, Webb S. The
social, psychological, emotional morbidity and adjustment techniques for
women with anal incontinence following Obstetric Anal Sphincter Injury:
use of a word picture to identify a hidden syndrome. BMC Pregnancy
Childbirth. 2016;16(1):275. Availability of data and materials The datasets generated and analysed during the current study are not publicly
available due to informed consent and confidentiality but are available from
the corresponding author upon reasonable request. 13. Pierce H, Perry L, Gallagher R, Chiarelli P. Severity of urinary incontinence
and its impact on work productivity among nurses and midwives in
urban Australia. Aust N Z Cont J. 2018;24:7–15. 14. Fultz N, Girts T, Kinchen K, Nygaard I, Pohl G, Sternfeld B. Prevalence, man-
agement and impact of urinary incontinence in the workplace. Occup
Med. 2005;55(7):552–7. Received: 20 March 2022 Accepted: 23 January 2023 22. United Nations. Transforming our world: the 2030 agenda for sustainable
development. 2015. https://sdgs.un.org/2030agenda. Accessed 05 Dec
2022. Abbreviations
EU
E Page 12 of 13 Page 12 of 13 Tjernström et al. BMC Women’s Health (2023) 23:47 ICD-10
International statistical classification of diseases and related health
problems, 10th revision
SFO-V
Strategic Research Area Health Care Science
SPT
Severe perineal trauma
A k
l d
t 6. Andreucci CB, Bussadori JC, Pacagnella RC, Chou D, Filippi V, Say L, et al. Sexual life and dysfunction after maternal morbidity: a systematic review. BMC Pregnancy Childbirth. 2015;15(1):307. 7. Evans E, Falivene C, Briffa K, Thompson J, Henry A. What is the total
impact of an obstetric anal sphincter injury? An Australian retrospective
study. Int Urogynecol J. 2020;31(3):557–66. 7. Evans E, Falivene C, Briffa K, Thompson J, Henry A. What is the total
impact of an obstetric anal sphincter injury? An Australian retrospective
study. Int Urogynecol J. 2020;31(3):557–66. Author contributions KT: conceptualisation; data curation; methodology; investigation; writing—
original draft; writing—review & editing; visualisation. IL: conceptualisation;
methodology; writing—review & editing; visualisation; supervision. MW: con-
ceptualisation; methodology; writing—review & editing; visualisation; supervi-
sion. MP: conceptualisation; data curation; methodology; writing—review &
editing; funding acquisition; visualisation; supervision; project administration. All authors read and approved the final manuscript. 9. The Swedish National Quality Register of Gynecological Surgery. Bristning
vid förlossning grad 3–4 Årsrapport från GynOp-registret avseende
operationer utförda år 2021. 2022. https://www.gynop.se/wp-content/
uploads/2022/05/Arsrapport-GynOp-bristning-grad-3-4-op-2021.pdf. Accessed 16 Oct 2022. 10. European Commission. Employment and unemployment (Labour Force
Survey). 2021. https://ec.europa.eu/eurostat/web/main/data/database. Accessed 20 Nov 2022. Ethics approval and consent to participate 15. Fitzgerald ST, Palmer MH, Berry SJ, Hart K. Urinary incontinence: impact
on working women. AAOHN J. 2000;48(3):112–8. The research project was performed in accordance with the Declaration
of Helsinki [42] and approved by the Swedish Ethical Review Authority
(Dnr: 2020–035410908). Participation was voluntary, and the women could
withdraw without stating a reason. Informed consent was mandatory and
was obtained digitally before each interview. The women could refrain from
answering questions they felt were too intimate. The project homepage and
the information letter provided information on available support options. If
any woman experienced negative emotions after the interview, researchers
could assist in referral to adequate healthcare services. 16. Wilson M. The impact of faecal incontinence on quality of life. Br J Nurs. 2007;16:204–7. 17. Olsson F, Berterö C. Living with faecal incontinence: trying to control the
daily life that is out of control. J Clin Nurs. 2015;24(1–2):141–50. 18. Gazibara T, Rancic B, Radovanovic S, Kurtagic I, Nurkovic S, Kovacevic N,
et al. Climacteric women at work: What lurks behind poor occupational
quality of life? Health Care Women Int. 2018;39(12):1350–65. 19. Beck CT. Effects of fourth-degree perineal lacerations on women’s
physical and mental health. J Obstet Gynecol Neonatal Nurs. 2021;50(2):133–42. Competing interests The authors declare that they have no competing interests. 21. World Health Organization (WHO): Defining sexual health. https://www.
who.int/teams/sexual-and-reproductive-health-and-research/key-areas-
of-work/sexual-health/defining-sexual-health (2022). Accessed 30 Nov
2022. Received: 20 March 2022 Accepted: 23 January 2023 Received: 20 March 2022 Accepted: 23 January 2023 Received: 20 March 2022 Accepted: 23 January 2023 References 1. World Health Organization (WHO). International statistical classification
of diseases and related health problems (ICD) 10th revision. 2019. https://
icd.who.int/browse10/2019/en. Accessed 14 Nov 2022. 2. d’Almeida I. Women’s experiences following obstetric anal sphincter
injury. J Pelvic Obstet Gynaecol Physiother. 2020;127:39–50. 3. Darmody E, Bradshaw C, Atkinson S. Women’s experience of obstetric
anal sphincter injury following childbirth: an integrated review. Mid-
wifery. 2020;91:102820. 4. LaCross A, Groff M, Smaldone A. Obstetric anal sphincter injury and anal
incontinence following vaginal birth: a systematic review and meta-
analysis. J Midwifery Womens Health. 2015;60(1):37–47. 5. Lindqvist M, Persson M, Nilsson M, Uustal E, Lindberg I. “A worse night-
mare than expected” - a Swedish qualitative study of women’s experi-
ences two months after obstetric anal sphincter muscle injury. Midwifery. 2018;61:22–8. 1. World Health Organization (WHO). International statistical classification
of diseases and related health problems (ICD) 10th revision. 2019. https://
icd.who.int/browse10/2019/en. Accessed 14 Nov 2022. 23. Salmon D. A feminist analysis of women’s experiences of perineal trauma
in the immediate post-delivery period. Midwifery. 1999;15(4):247–56. 24. Priddis HS. Autoethnography and severe perineal trauma-an unexpected
journey from disembodiment to embodiment. BMC Womens Health. 2015;15(1):88. 2. d’Almeida I. Women’s experiences following obstetric anal sphincter
injury. J Pelvic Obstet Gynaecol Physiother. 2020;127:39–50. 3. Darmody E, Bradshaw C, Atkinson S. Women’s experience of obstetric
anal sphincter injury following childbirth: an integrated review. Mid-
wifery. 2020;91:102820. 25. Priddis H, Schmied V, Dahlen H. Women’s experiences following severe
perineal trauma: a qualitative study. BMC Womens Health. 2014;14(1):32. 26. De Beauvoir S. The second sex. 1st American edn. New York: Alfred A. Knopf; 1949/2010. y
4. LaCross A, Groff M, Smaldone A. Obstetric anal sphincter injury and anal
incontinence following vaginal birth: a systematic review and meta-
analysis. J Midwifery Womens Health. 2015;60(1):37–47. 27. Söderberg M, Christensson K, Lundgren I. A project for future life-Swedish
women’s thoughts on childbearing lacking experience of giving birth
and parenthood. Int J Qual Stud Health Well-being. 2012;7. 5. Lindqvist M, Persson M, Nilsson M, Uustal E, Lindberg I. “A worse night-
mare than expected” - a Swedish qualitative study of women’s experi-
ences two months after obstetric anal sphincter muscle injury. Midwifery. 2018;61:22–8. 28. Shabot SC. On motherhood as ambiguity and transcendence: reevaluat-
ing motherhood through the Beauvoirian erotic. Comp Cont Philos. 2021:1–13. Page 13 of 13 Page 13 of 13 Tjernström et al. BMC Women’s Health (2023) 23:47 Tjernström et al. References BMC Women’s Health (2023) 23:47 29. Cohen SS. Constructing subjectivity through labour pain: A Beauvoirian
analysis. Eur J Women’s Stud. 2015;24(2):128–42. 52. Kulkarni M. Hiding but hoping to be found: workplace disclosure
dilemmas of individuals with hidden disabilities. Equal Divers Incl Int J. 2022;41(3):491–507. y
30. Shabot SC. We birth with others: towards a Beauvoirian understanding of
obstetric violence. Eur J Women’s Stud. 2020;28(2):213–28. obstetric violence. Eur J Women’s Stud. 2020;28(2):213–28. 53. Sang K, Remnant J, Calvard T, Myhill K. Blood Work: Managing Menstrua-
tion, Menopause and Gynaecological Health Conditions in the Work-
place. Int J Environ Res Public Health. 2021;18(4). 31. Kadoglou T, Sarri K. Women, motherhood and work. Simone Beauvoir
Stud. 2013;29:33–45. 54. Gatrell C. Policy and the pregnant body at work: strategies of secrecy,
silence and supra-performance. Gend Work Organ. 2011;18(2):158–81. 32. Veltman A. The sisyphean torture of housework: Simone de Beauvoir
and inequitable divisions of domestic work in marriage. Hypatia. 2004;19(3):121–43. 55. Gatrell CJ. Secrets and lies: breastfeeding and professional paid work. Soc
Sci Med. 2007;65(2):393–404. 33. Ross BW. Simone de Beauvoir’s idea of transcendence and its applicability
to women who work as nurses. Ann Arbor: New York University; 1982. 56. Boddy CR. Sample size for qualitative research. Qual Market Res. 2016;19(4):426–32. 34. Charmaz K. Constructing grounded theory. 2nd ed. Thousand Oaks, CA:
Sage Publications; 2014. 57. Malterud K, Siersma VD, Guassora AD. Sample size in qualitative
interview studies: guided by information power. Qual Health Res. 2016;26(13):1753–60. 35. Öhman H, Schesny S, Barth L, Löwendahl J. Förlossningsskadad? Du är
inte ensam! . https://www.facebook.com/groups/forlossningsskadad/:
Facebook; 2014. 36. Statistics Sweden. Women and men in Sweden-Facts and Figures. 2022. https://www.scb.se/en/finding-statistics/statistics-by-subject-area/
living-conditions/gender-statistics/gender-statistics/pong/publications/
women-and-men-in-sweden---facts-and-figures-2022/. Accessed 20 Nov
2022. Tjernström et al. BMC Women’s Health (2023) 23:47 Publisher’s Note
S
i
N
i Springer Nature remains neutral with regard to jurisdictional claims in pub-
lished maps and institutional affiliations. Springer Nature remains neutral with regard to jurisdictional claims in pub-
lished maps and institutional affiliations. 37. Statistics Sweden. Arbetskraftsundersökningarna (AKU) 2021. 2022. https://www.scb.se/hitta-statistik/statistik-efter-amne/arbetsmarknad/
arbetskraftsundersokningar/arbetskraftsundersokningarna-aku/pong/
publikationer/arbetskraftsundersokningarna-aku-2021/. Accessed 20 Nov
2022. 38. World Economic Forum. Global gender gap report 2021. 2021. https://
www.weforum.org/reports/global-gender-gap-report-2021. Accessed 20
Nov 2022. 39. Fransson E, Grönqvist E, Iliadis S, Lindahl E. Kvinnors hälsa, sjukfrånvaro
och inkomster efter barnafödande. Vad vet vi om barneffektens orsaker
och vilken roll spelar hälsa och föräldraförsäkringen? Institutet för
arbetsmarknads-och utbildningspolitisk utvärdering (IFAU). 2021. https://
www.ifau.se/Forskning/Publikationer/Rapporter/20212/kvinnors-halsa-
sjukfranvaro-och-inkomster-efter-barnafodande/. Accessed 21 Nov 2022. 40. Bäckenbottenutbildning: Sen behandling av förlossningsskador. https://
backenbottenutbildning.se/index.php/utbildningsmaterial/sen-behan
dling-av-forlossningsskador (2022). Accessed 27 Feb 2022. 41. MAXQDA: Discover MAXQDA’s Features. https://www.maxqda.com/what-
is-maxqda (2022). Accessed 19 Oct 2022. 41. MAXQDA: Discover MAXQDA’s Features. https://www.maxqda.com/what-
is-maxqda (2022). Accessed 19 Oct 2022. q
42. Declaration of Helsinki–Ethical principles for medical research involving
human subjects. 1964. 42. Declaration of Helsinki–Ethical principles for medical research involving
human subjects. 1964. 43. Davis DL, Walker K. Re-discovering the material body in midwifery
through an exploration of theories of embodiment. Midwifery. 2010;26(4):457–62. 43. Davis DL, Walker K. Re-discovering the material body in midwifery
through an exploration of theories of embodiment. Midwifery. 2010;26(4):457–62. 44. Huber M, Tunón K, Lindqvist M. “From hell to healed”-a qualitative study
on women’s experience of recovery, relationships and sexuality after
severe obstetric perineal injury. Sex Reprod Healthc. 2022;33:100736. 44. Huber M, Tunón K, Lindqvist M. “From hell to healed”-a qualitative study
on women’s experience of recovery, relationships and sexuality after
severe obstetric perineal injury. Sex Reprod Healthc. 2022;33:100736. 45. Herron-Marx S, Williams A, Hicks C. A Q methodology study of women’s
experience of enduring postnatal perineal and pelvic floor morbidity. Midwifery. 2007;23(3):322–34. y
46. O’Reilly R, Peters K, Beale B, Jackson D. Women’s experiences of recovery
from childbirth: focus on pelvis problems that extend beyond the puer-
perium. J Clin Nurs. 2009;18(14):2013–9. p
47. Priddis HS, Schmied V, Kettle C, Sneddon A, Dahlen HG. “A patchwork
of services”–caring for women who sustain severe perineal trauma in
New South Wales–from the perspective of women and midwives. BMC
Pregnancy Childbirth. 2014;14:236. •
fast, convenient online submission
•
thorough peer review by experienced researchers in your field
•
rapid publication on acceptance
•
support for research data, including large and complex data types
•
gold Open Access which fosters wider collaboration and increased citations
maximum visibility for your research: over 100M website views per year
•
At BMC, research is always in progress.
Learn more biomedcentral.com/submissions
Ready to submit your research
Ready to submit your research ? Choose BMC and benefit from:
? Choose BMC and benefit from: Publisher’s Note
S
i
N
i •
fast, convenient online submission
•
thorough peer review by experienced researchers in your field
•
rapid publication on acceptance
•
support for research data, including large and complex data types
•
gold Open Access which fosters wider collaboration and increased citations
maximum visibility for your research: over 100M website views per year
•
At BMC, research is always in progress. Learn more biomedcentral.com/submissions
Ready to submit your research
Ready to submit your research ? Choose BMC and benefit from:
? Choose BMC and benefit from: •
fast, convenient online submission
•
thorough peer review by experienced researchers in your field
•
rapid publication on acceptance
•
support for research data, including large and complex data types
•
gold Open Access which fosters wider collaboration and increased citations
maximum visibility for your research: over 100M website views per year
•
At BMC, research is always in progress. Learn more biomedcentral.com/submissions
Ready to submit your research
Ready to submit your research ? Choose BMC and benefit from:
? Choose BMC and benefit from: 48. Dolezal L. The body and shame : phenomenology, feminism, and the
socially shaped body. Lanham, MD: Lexington Books; 2015. 49. Vickers MH. Dark secrets and impression management: workplace
masks of people with multiple sclerosis (MS). Empl Responsib Rights J. 2017;29(4):175–95. 49. Vickers MH. Dark secrets and impression management: workplace
masks of people with multiple sclerosis (MS). Empl Responsib Rights J. 2017;29(4):175–95. 50. Ganesh K, Lazar A. The work of workplace disclosure: invisible chronic
conditions and opportunities for design. Proc ACM Hum Comput Inter-
act. 2021. https://doi.org/10.1145/3449147. 50. Ganesh K, Lazar A. The work of workplace disclosure: invisible chronic
conditions and opportunities for design. Proc ACM Hum Comput Inter-
act. 2021. https://doi.org/10.1145/3449147. 51. Toth KE, Yvon F, Villotti P, Lecomte T, Lachance J-P, Kirsh B, et al. Disclosure
dilemmas: how people with a mental health condition perceive and
manage disclosure at work. Disabil Rehabil. 2021:1–11.
|
https://openalex.org/W3171717359
|
https://figshare.com/ndownloader/files/28401572
|
English
| null |
Elucidating Age and Sex-Dependent Association Between Frontal EEG Asymmetry and Depression: An Application of Multiple Imputation in Functional Regression
|
Journal of the American Statistical Association
| 2,021
|
cc-by
| 11,646
|
Appendix to: Elucidating age and sex-dependent association
between frontal EEG asymmetry and depression: An application of
multiple imputation in functional regression Adam Ciarleglioa, Eva Petkovab,c, Ofer Hareld Adam Ciarleglioa, Eva Petkovab,c, Ofer Hareld a Department of Biostatistics and Bioinformatics, Milken Institute School of Public Health,
George Washington University, Washington, DC
b Department of Population Health, New York University, New York, NY
c Department of Child and Adolescent Psychiatry, New York University, New York, NY
d Department of Statistics, University of Connecticut, Storrs, CT Appendix to: Elucidating age and sex-dependent association
between frontal EEG asymmetry and depression: An application of
multiple imputation in functional regression Appendix to: Elucidating age and sex-dependent association
between frontal EEG asymmetry and depression: An application of
multiple imputation in functional regression 1.1
PFFR for Fitting Functional Response Models (FRMs) Briefly, the PFFR fitting procedure is carried out as follows. First, each coefficient function is
represented by a set of basis functions (e.g., B-splines, cubic-regression splines, thin-plate splines,
etc.). That is, we let βj(t) = B⊺
j (t)bj for j = 0, . . . , p, where Bj(t) = (Bj,1(t), . . . , Bj,Lj(t))⊺is the
vector of known basis functions selected for βj(t) and bj = (bj,1, . . . , bj,Lj)⊺is the corresponding
vector of unknown basis coefficients. Similarly, we let ρk(s, t) = U ⊺
k (s, t)uk for k = 1, . . . , q where
Uk(s, t) = (Uk,1(s, t), . . . , Uk,Lk(s, t))⊺is the vector of known bivariate basis functions (e.g., bivariate
thin-plate splines) selected for ρk(s, t) and uk = (uk,1, . . . , uk,Lk)⊺is the corresponding vector of
unknown basis coefficients. The numbers of basis functions used in the representations (Lj and
Lk) are typically selected to be larger than is assumed necessary to adequately represent a given
function. The integral terms are approximated as Z
Xi,k(s)ρk(s, t)ds ≈
Dk
X
d=1
∆dρ(sd, t)Xi,k(sd) =
( Dk
X
d=1
U ⊺
k (sd, t)Xi,k(sd)∆d
)
uk, where sd are the grid points over which Xi,k is observed and ∆d is the corresponding interval length
between grid points. Substituting the integral approximation and basis representations into model
(1) from the main article gives where sd are the grid points over which Xi,k is observed and ∆d is the corresponding interval length
between grid points. Substituting the integral approximation and basis representations into model
(1) from the main article gives g{µi(t)} = B⊺
0(t)b0 +
p
X
j=1
zi,jB⊺
j (t)bj +
q
X
k=1
( Dk
X
d=1
U ⊺
k (sd, t)Xi,k(sd)∆d
)
uk, and the log-likelihood function is given by l(b0, . . . , bp, u1, . . . , uq) =
Nt
X
g=1
n
X
i=1
log [EF{µi(tg|b0, . . . , bp, u1, . . . , uq), ηt)}] , where {tg; g = 1, . . . , Nt} are the grid points at which Yi is observed and we explicitly write the
mean as a function of the vectors of parameters bj and uk for j = 0, . . . , p and k = 1, . . . , q. 1.1
PFFR for Fitting Functional Response Models (FRMs) As
Lj and Lk are chosen to be large, smoothness in the estimated coefficient functions is achieved by
including penalties on the estimated basis coefficients when maximizing the log-likelihood. Hence
the objective function to be maximized is l(b0, . . . , bp, u1, . . . , uq) −
p
X
j=0
λbjPbj(bj) −
q
X
k=1
λukPuk(uk), where Pbj(bj) and Puk(uk) are known penalty functions and λbj and λuk are the respective non-
negative tuning parameters that control the amount of smoothness. Using quadratic penalties of the
form Pbj(bj) = b⊺
jDbjbj and Puk(uk) = u⊺
kDukuk, where Dbj and Duk are known penalty matrices
allows to use a mixed-effects model framework, with the coefficient functions viewed as random
effects. In this setting, one can use restricted maximum likelihood (REML) to simultaneously
select the tuning parameters and estimate the basis coefficients corresponding to the coefficient
functions. Furthermore, approximate confidence intervals for the basis coefficients can be obtained. The estimated basis coefficients and their approximate intervals can then be substituted into the
basis representations to obtain the estimated coefficient functions and point-wise confidence bands. Full details can be found in Ivanescu et al. (2015). 1.2
PFR for Fitting Scalar Response Models (SRMs) Briefly, PFR is carried out as follows. First the functional covariates are represented using a
truncated Karhunen-Lo´eve decomposition, Xij(t) = PKj
k=1 cijkψjk(t) = ψj(t)cij, where ψjk(t) for
k = 1, . . . , Kj are the first Kj eigenfunctions (functional principal components (FPCs)) of the
smoothed covariance operator corresponding to cov{Xij(s), Xij(t)}, cijk =
R
Xij(t)ψjk(t)dt are
the FPC scores, ψj(t) = (ψj1(t), . . . , ψjKj(t)) is a 1 × Kj vector of the eigenfunctions, and c⊺
ij =
(cij1, . . . , cijKj) is a 1 × Kj vector of FPC scores representing the j-th functional covariate for
the i-th subject. Next, each coefficient function, βj, is represented using a suitable set of basis
functions given by φj(t) = (φj1(t), . . . , φjLj(t)) such that βj(t) = PLj
ℓ=1 bjℓφjℓ(t) = φj(t)bj, where
b⊺
j = (bj1, . . . , bjLj). Using these representations, the right-hand side of the equation (2) from the
main article can be re-expressed as θ0 + ziθ + Pq
j=1 ˜
Xijbj where ˜
Xij = c⊺
ijGj and Gj is a Kj × Lj
dimensional matrix with entry (u, v) given by
R
ψju(t)φjv(t)dt. R
j
j
As with the PFFR method to estimate the FRM given in model (1) from the main article, PFR
uses a penalized log-likelihood objective function with similar penalties on the basis coefficients and
employs a mixed effects model framework to obtain estimates for the basis coefficients. Using this
framework, the basis coefficients are viewed as random effects in a mixed effects model and turning
parameters are estimated via REML. As with PFFR, since PFR employs a mixed effects framework,
it is possible to obtain approximate confidence intervals for the scalar coefficients and approximate
point-wise confidence intervals for the coefficient functions. The reader is referred to Goldsmith
et al. (2011) for the complete details of the PFR estimation procedure. Kj and Lj for j = 1, . . . , q
are tuning parameters that need to be chosen prior to estimation. Selection can be based on a
data-driven approach such as cross-validation, but this approach may be computationally intensive. Alternatively, Goldsmith et al. (2011) note that as long as they are chosen “large enough,” then
their specific values have minor impact on the quality of the estimates. 2 2 2
Additional Simulation Results In the main article, we present results for simulations in which the analysis model is a FRM with
3 scalar predictors. In those simulations, the sample size is n = 350. Here we present additional
results for the same settings and scenarios, but for a smaller sample size of n = 100. In addition,
we include plots for the point-wise coverages and point-wise widths of the 95% confidence intervals
for both n = 350 and n = 100. In the main article, we provide tables with across-the-function mean
point-wise coverage and point-wise width. We hope that the plots included here provide additional
insight into the performance of the different methods for handling missing data in the context of
fitting a FRM. These additional results are provided in Sections 2.2 and 2.3. We also introduce a new simulation study in Section 2.4 where the analysis model is a SRM with
scalar and functional predictors. The complete case, mean imputation, and fregMICE approaches
for handling missing data are applied and evaluated. 2.1
Proportion of Missing Observations Tables 1 and 2 below show mean (sd) proportion of missing observations in the data sets generated
under the simulation settings described in the main article for sample sizes of n = 100 and n = 350
respectively. Mean (SD) proportion of missing observations with n = 100 for Settings 1 and 2. e 1: Mean (SD) proportion of missing observations with n = 100 for Settings 1 and 2
10% Intended
20% Intended
30% Intended
Scenario (a)
Prop. Miss. z2
0.10 (0.00)
0.20 (0.00)
0.30 (0.00)
Scenario (b)
Prop. Miss. z2
0.10 (0.03)
0.20 (0.04)
0.29 (0.05)
Prop. Miss. Y
0.10 (0.03)
0.20 (0.04)
0.31 (0.05)
Prop. Miss. z2 or Y
0.19 (0.04)
0.36 (0.05)
0.51 (0.05) Table 2: Mean (SD) proportion of missing observations with n = 350 for Settings 1 and 2. Mean (SD) proportion of missing observations with n = 350 for Settings 1 and 2. Table 2: Mean (SD) proportion of missing observations with n = 350 for Settings 1 and 2
10% Intended
20% Intended
30% Intended
Scenario (a)
Prop. Miss. z2
0.10 (0.00)
0.20 (0.00)
0.30 (0.00)
Scenario (b)
Prop. Miss. z2
0.10 (0.02)
0.20 (0.02)
0.29 (0.03)
Prop. Miss. Y
0.10 (0.02)
0.20 (0.02)
0.31 (0.03)
Prop. Miss. z2 or Y
0.20 (0.02)
0.36 (0.02)
0.52 (0.03) Note that, for Scenario (a), for either sample size, the missing proportions are exactly 10%,
20%, or 30% when they are intended to take on these values. This is because z2 is set to missing if
the corresponding functional response, Y , has an average value above the 90th percentile (for 10%
missing), 80th percentile (for 20% missing), or 70th percentile (for 30% missing). In Scenario (b), both z2 and/or Y can be missing for an observation. The bottom rows of Tables
1 and 2 show the mean (sd) proportions of incomplete observations in a data set. 2.2.1
Point-wise Mean Plots for n = 350 2.2.1
Point-wise Mean Plots for n = 350 In the main article, we provided point-wise standardized bias plots for each setting and scenario
for n = 350. Figures 1 and 2 show the point-wise means over the 500 simulation runs. 3 10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0
1
2
3
4
0.0
0.3
0.6
0.9
0.5
1.0
1.5
2.0
0.0
0.3
0.6
0.9
t
Method
ANM
CCA
fregMICE Imputation
Mean Imputation
True Function Method
ANM
CCA
fregMICE Imputation
Mean Imputation
True Function 10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0
1
2
3
4
0.0
0.3
0.6
0.9
0.5
1.0
1.5
2.0
0.0
0.3
0.6
0.9
t
Method
ANM
CCA
fregMICE Imputation
Mean Imputation
True Function
10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0
1
2
3
4
0.00
0.25
0.50
0.75
1.00
0.5
1.0
1.5
2.0
−0.2
0.0
0.2
t
Method
ANM
CCA
fregMICE Imputation
Mean Imputation
True Function
1: Point-wise mean curves in Setting 1 Scenario (a) (Top) and Setting 1 Scenario
om). Columns (left to right) correspond to 10, 20, and 30% missing data. Rows (to
) correspond to functional parameters β0, β1, β2, and β3. 10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0
1
2
3
4
0.00
0.25
0.50
0.75
1.00
0.5
1.0
1.5
2.0
−0.2
0.0
0.2
t
Method
ANM
CCA
fregMICE Imputation
Mean Imputation
True Function 10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0
1
2
3
4
0.00
0.25
0.50
0.75
1.00
0.5
1.0
1.5
2.0
−0.2
0.0
0.2
t
Method
ANM
CCA
fregMICE Imputation
Mean Imputation
True Function
Figure 1: Point-wise mean curves in Setting 1 Scenario (a) (Top) and Setting 1 Scenario (b)
(Bottom). Columns (left to right) correspond to 10, 20, and 30% missing data. Rows (top to
bottom) correspond to functional parameters β0, β1, β2, and β3. 2.2.1
Point-wise Mean Plots for n = 350 Method
ANM
CCA
fregMICE Imputation
Mean Imputation
True Function Figure 1: Point-wise mean curves in Setting 1 Scenario (a) (Top) and Setting 1 Scenario (b)
(Bottom). Columns (left to right) correspond to 10, 20, and 30% missing data. Rows (top to
bottom) correspond to functional parameters β0, β1, β2, and β3. 4 10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0.0
0.5
1.0
1.5
2.0
2.5
−1.0
−0.5
0.0
0.5
1.0
0.0
0.2
0.4
0.6
0.8
−0.4
−0.3
−0.2
−0.1
0.0
t
Method
ANM
CCA
fregMICE Imputation
Mean Imputation
True Function Method
ANM
CCA
fregMICE Imputation
Mean Imputation
True Function 10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0.0
0.5
1.0
1.5
2.0
2.5
−1.0
−0.5
0.0
0.5
1.0
0.0
0.2
0.4
0.6
0.8
−0.4
−0.3
−0.2
−0.1
0.0
t
Method
ANM
CCA
fregMICE Imputation
Mean Imputation
True Function
10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0.0
0.5
1.0
1.5
2.0
2.5
−1.0
−0.5
0.0
0.5
1.0
0.0
0.2
0.4
0.6
0.8
−0.4
−0.3
−0.2
−0.1
0.0
t
Method
ANM
CCA
fregMICE Imputation
Mean Imputation
True Function
2: Point-wise mean curves in Setting 2 Scenario (a) (Top) and Setting 2 Scenario
om). Columns (left to right) correspond to 10, 20, and 30% missing data. Rows (to
) correspond to functional parameters β0, β1, β2, and β3. 10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0.0
0.5
1.0
1.5
2.0
2.5
−1.0
−0.5
0.0
0.5
1.0
0.0
0.2
0.4
0.6
0.8
−0.4
−0.3
−0.2
−0.1
0.0
t
Method
ANM
CCA
fregMICE Imputation
Mean Imputation
True Function
Figure 2: Point-wise mean curves in Setting 2 Scenario (a) (Top) and Setting 2 Scenario (b)
(Bottom). Columns (left to right) correspond to 10, 20, and 30% missing data. Rows (top to
bottom) correspond to functional parameters β0, β1, β2, and β3. Point-wise Mean and Point-wise Standardized Bias Plots for n = 100 Figures 3 and 4 show point-wise mean and point-wise standardized bias plots for simulation settings
presented in the main article, but with a smaller sample size of n = 100. (ANM = all no missing;
CCA = complete case analysis; fregMICE = functional regression MICE) Overall, we see that
the performance of the estimates derived from the fregMICE-imputed data, relative the the other
approaches for handling missing data, is similar to that seen in the main article for the larger
sample size of n = 350. 2.2.1
Point-wise Mean Plots for n = 350 10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0.0
0.5
1.0
1.5
2.0
2.5
−1.0
−0.5
0.0
0.5
1.0
0.0
0.2
0.4
0.6
0.8
−0.4
−0.3
−0.2
−0.1
0.0
t
Method
ANM
CCA
fregMICE Imputation
Mean Imputation
True Function
2: Point-wise mean curves in Setting 2 Scenario (a) (Top) and Setting 2 Scenari Method
ANM
CCA
fregMICE Imputation
Mean Imputation
True Function Figure 2: Point-wise mean curves in Setting 2 Scenario (a) (Top) and Setting 2 Scenario (b)
(Bottom). Columns (left to right) correspond to 10, 20, and 30% missing data. Rows (top to
bottom) correspond to functional parameters β0, β1, β2, and β3. 5 2.2.2
Point-wise Mean and Point-wise Standardized Bias Plots for n = 1 2.2.3
Point-wise 95% Confidence Band Coverage and Width Figures 5 and 6 show the coverage probability for the estimated 95% confidence bands at each
point on [0, 10] for Setting 1 Scenarios (a) and (b) with n = 350 and n = 100 respectively, over
the 500 simulation runs for each method of handling the missing data. Figures 7 and 8 show the
corresponding mean point-wise widths for the 95% confidence bands. 6 6 Method
ANM
CCA
fregMICE Imputation
Mean Imputation
True Function 10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0
1
2
3
4
0.0
0.3
0.6
0.9
0.5
1.0
1.5
2.0
0.0
0.3
0.6
0.9
t
Method
ANM
CCA
fregMICE Imputation
Mean Imputation
True Function
10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0.0
0.5
−0.4
−0.2
0.0
0.2
−1.0
−0.5
0.0
0.5
−0.2
0.0
0.2
t
3: (Top) Point-wise mean curves and (Bottom) point-wise standardized bias in Sett
io (a) for n = 100. Columns (left to right) correspond to 10, 20, and 30% missing data. R
o bottom) correspond to functional parameters β0, β1, β2, and β3. Point-wise standar
urve for mean imputation is removed to better compare estimates from ANM, CCA
CE Figure 3: (Top) Point-wise mean curves and (Bottom) point-wise standardized bias in Setting 1
Scenario (a) for n = 100. Columns (left to right) correspond to 10, 20, and 30% missing data. Rows
(top to bottom) correspond to functional parameters β0, β1, β2, and β3. Point-wise standardized
bias curve for mean imputation is removed to better compare estimates from ANM, CCA, and
fregMICE. 2.2.3
Point-wise 95% Confidence Band Coverage and Width 7 7 Method
ANM
CCA
fregMICE Imputation
Mean Imputation
True Function 10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0
1
2
3
4
0.00
0.25
0.50
0.75
1.00
0.5
1.0
1.5
2.0
−0.2
−0.1
0.0
0.1
0.2
0.3
t
Method
ANM
CCA
fregMICE Imputation
Mean Imputation
True Function
10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
−0.3
0.0
0.3
0.6
−0.2
0.0
0.2
−0.50
−0.25
0.00
0.25
0.50
−0.3
−0.2
−0.1
0.0
0.1
t
4: (Top) Point-wise mean curves and (Bottom) point-wise standardized bias in Setti
io (b) for n = 100. Columns (left to right) correspond to 10, 20, and 30% missing data. R
o bottom) correspond to functional parameters β0, β1, β2, and β3. Point-wise standard
urve for mean imputation is removed to better compare estimates from ANM, CCA,
CE 10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
−0.3
0.0
0.3
0.6
−0.2
0.0
0.2
−0.50
−0.25
0.00
0.25
0.50
−0.3
−0.2
−0.1
0.0
0.1
t Figure 4: (Top) Point-wise mean curves and (Bottom) point-wise standardized bias in Setting 1
Scenario (b) for n = 100. Columns (left to right) correspond to 10, 20, and 30% missing data. Rows
(top to bottom) correspond to functional parameters β0, β1, β2, and β3. Point-wise standardized
bias curve for mean imputation is removed to better compare estimates from ANM, CCA, and
fregMICE. 8 8 10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
t
Method
All
Complete Case
fregMICE Imputation
Mean Imputation
10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
t
Method
All
Complete Case
fregMICE Imputation
Mean Imputation
5: Point-wise 95% confidence band coverage for Setting 1 Scenarios (a) (Top) and
om) with n = 350. 2.2.3
Point-wise 95% Confidence Band Coverage and Width Columns (left to right) correspond to 10, 20, and 30% missing d
top to bottom) correspond to functional parameters β0, β1, β2, and β3. 10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
t
Method
All
Complete Case
fregMICE Imputation
Mean Imputation Method
All
Complete Case
fregMICE Imputation
Mean Imputation 10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
t
Method
All
Complete Case
fregMICE Imputation
Mean Imputation
Figure 5: Point-wise 95% confidence band coverage for Setting 1 Scenarios (a) (Top) and (b)
(Bottom) with n = 350. Columns (left to right) correspond to 10, 20, and 30% missing data. Rows (top to bottom) correspond to functional parameters β0, β1, β2, and β3. 10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
t
Method
All
Complete Case
fregMICE Imputation
Mean Imputation Method
All
Complete Case
fregMICE Imputation
Mean Imputation Figure 5: Point-wise 95% confidence band coverage for Setting 1 Scenarios (a) (Top) and (b)
(Bottom) with n = 350. Columns (left to right) correspond to 10, 20, and 30% missing data. Rows (top to bottom) correspond to functional parameters β0, β1, β2, and β3. 2.2.3
Point-wise 95% Confidence Band Coverage and Width 9 10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
t
Method
ANM
CCA
fregMICE Imputation
Mean Imputation
10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
t
Method
ANM
CCA
fregMICE Imputation
Mean Imputation
6: Point-wise 95% confidence band coverage for Setting 1 Scenarios (a) (Top) and
om) with n = 100. Columns (left to right) correspond to 10, 20, and 30% missing d
top to bottom) correspond to functional parameters β0, β1, β2, and β3. 10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
t
Method
ANM
CCA
fregMICE Imputation
Mean Imputation 10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
t
Method
ANM
CCA
fregMICE Imputation
Mean Imputation
Figure 6: Point-wise 95% confidence band coverage for Setting 1 Scenarios (a) (Top) and (b)
(Bottom) with n = 100. Columns (left to right) correspond to 10, 20, and 30% missing data. Rows (top to bottom) correspond to functional parameters β0, β1, β2, and β3. 10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
t
Method
ANM
CCA
fregMICE Imputation
Mean Imputation Method
ANM
CCA
fregMICE Imputation
Mean Imputation Figure 6: Point-wise 95% confidence band coverage for Setting 1 Scenarios (a) (Top) and (b)
(Bottom) with n = 100. Columns (left to right) correspond to 10, 20, and 30% missing data. Rows (top to bottom) correspond to functional parameters β0, β1, β2, and β3. 2.2.3
Point-wise 95% Confidence Band Coverage and Width 10 10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0.3
0.4
0.5
0.20
0.25
0.30
0.35
0.40
0.45
0.10
0.15
0.20
0.25
0.09
0.12
0.15
0.18
0.21
t
Method
All
Complete Case
fregMICE Imputation
Mean Imputation
10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0.2
0.3
0.4
0.5
0.2
0.3
0.4
0.5
0.10
0.15
0.20
0.25
0.10
0.15
0.20
0.25
t
Method
All
Complete Case
fregMICE Imputation
Mean Imputation
7:
Point-wise 95% confidence band width for Setting 1 Scenarios (a) (Top) and
om) with n = 350. Columns (left to right) correspond to 10, 20, and 30% missing da
top to bottom) correspond to functional parameters β0, β1, β2, and β3. 10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0.3
0.4
0.5
0.20
0.25
0.30
0.35
0.40
0.45
0.10
0.15
0.20
0.25
0.09
0.12
0.15
0.18
0.21
t
Method
All
Complete Case
fregMICE Imputation
Mean Imputation Method
All
Complete Case
fregMICE Imputation
Mean Imputation t
10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0.2
0.3
0.4
0.5
0.2
0.3
0.4
0.5
0.10
0.15
0.20
0.25
0.10
0.15
0.20
0.25
t
Method
All
Complete Case
fregMICE Imputation
Mean Imputation
Figure 7:
Point-wise 95% confidence band width for Setting 1 Scenarios (a) (Top) and (b)
(Bottom) with n = 350. Columns (left to right) correspond to 10, 20, and 30% missing data. Rows (top to bottom) correspond to functional parameters β0, β1, β2, and β3. 10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0.2
0.3
0.4
0.5
0.2
0.3
0.4
0.5
0.10
0.15
0.20
0.25
0.10
0.15
0.20
0.25
t
Method
All
Complete Case
fregMICE Imputation
Mean Imputation Method
All
Complete Case
fregMICE Imputation
Mean Imputation Figure 7:
Point-wise 95% confidence band width for Setting 1 Scenarios (a) (Top) and (b)
(Bottom) with n = 350. Columns (left to right) correspond to 10, 20, and 30% missing data. Rows (top to bottom) correspond to functional parameters β0, β1, β2, and β3. .1
Point-wise Mean and Point-wise Standardized Bias Plots for n = 100 Figures 9 and 10 show point-wise mean and point-wise standardized bias plots for simulation
settings presented in the main article, but with a smaller sample size of n = 100. Overall, we see
that the performance of the estimates derived from the fregMICE-imputed data, relative the the
other approaches for handling missing data, is similar to that seen in the main article for the larger
sample size of n = 350. 2.2.3
Point-wise 95% Confidence Band Coverage and Width 11 10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0.4
0.5
0.6
0.7
0.8
0.3
0.4
0.5
0.6
0.7
0.20
0.25
0.30
0.35
0.40
0.45
0.15
0.20
0.25
0.30
0.35
t
Method
ANM
CCA
fregMICE Imputation
Mean Imputation
10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0.4
0.6
0.8
0.4
0.6
0.8
0.2
0.3
0.4
0.2
0.3
0.4
t
Method
ANM
CCA
fregMICE Imputation
Mean Imputation
8:
Point-wise 95% confidence band width for Setting 1 Scenarios (a) (Top) and
om) with n = 100. Columns (left to right) correspond to 10, 20, and 30% missing da
top to bottom) correspond to functional parameters β0, β1, β2, and β3. 10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0.4
0.5
0.6
0.7
0.8
0.3
0.4
0.5
0.6
0.7
0.20
0.25
0.30
0.35
0.40
0.45
0.15
0.20
0.25
0.30
0.35
t
Method
ANM
CCA
fregMICE Imputation
Mean Imputation 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
t
10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0.4
0.6
0.8
0.4
0.6
0.8
0.2
0.3
0.4
0.2
0.3
0.4
t
Method
ANM
CCA
fregMICE Imputation
Mean Imputation
Figure 8:
Point-wise 95% confidence band width for Setting 1 Scenarios (a) (Top) and (b)
(Bottom) with n = 100. Columns (left to right) correspond to 10, 20, and 30% missing data. Rows (top to bottom) correspond to functional parameters β0, β1, β2, and β3. 10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0.4
0.6
0.8
0.4
0.6
0.8
0.2
0.3
0.4
0.2
0.3
0.4
t
Method
ANM
CCA
fregMICE Imputation
Mean Imputation Figure 8:
Point-wise 95% confidence band width for Setting 1 Scenarios (a) (Top) and (b)
(Bottom) with n = 100. Columns (left to right) correspond to 10, 20, and 30% missing data. Rows (top to bottom) correspond to functional parameters β0, β1, β2, and β3. 12 2.3.2
Point-wise 95% Confidence Band Coverage and Width Figures 11 and 12 show the coverage probability for the estimated 95% confidence bands at each
point on [0, 10] for Setting 2 Scenarios (a) and (b) with n = 350 and n = 100 respectively, over
the 500 simulation runs for each method of handling the missing data. Figures 13 and 14 show the
corresponding mean point-wise widths for the 95% confidence bands. 13 Method
ANM
CCA
fregMICE Imputation
Mean Imputation
True Function 10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0.0
0.5
1.0
1.5
2.0
2.5
−1.0
−0.5
0.0
0.5
1.0
0.0
0.2
0.4
0.6
0.8
−0.4
−0.3
−0.2
−0.1
0.0
t
Method
ANM
CCA
fregMICE Imputation
Mean Imputation
True Function
10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0
1
2
−0.1
0.0
0.1
−2
−1
0
0
1
2
t
9: (Top) Point-wise mean curves and (Bottom) point-wise standardized bias in Sett
io (a) for n = 100. Columns (left to right) correspond to 10, 20, and 30% missing data. o bottom) correspond to functional parameters β0, β1, β2, and β3. Point-wise standar
urve for mean imputation is removed to better compare estimates from ANM, CCA
ICE. 14 β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0.0
0.2
0.4
0.6
−0.4
−0.3
−0.2
−0.1
0.0
t
10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0
1
2
−0.1
0.0
0.1
−2
−1
0
0
1
2
t Figure 9: (Top) Point-wise mean curves and (Bottom) point-wise standardized bias in Setting 2
Scenario (a) for n = 100. Columns (left to right) correspond to 10, 20, and 30% missing data. Rows
(top to bottom) correspond to functional parameters β0, β1, β2, and β3. 2.3.2
Point-wise 95% Confidence Band Coverage and Width Point-wise standardized
bias curve for mean imputation is removed to better compare estimates from ANM, CCA, and
f egMICE 15 10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
t
Method
All
Complete Case
fregMICE Imputation
Mean Imputation
10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
t
Method
All
Complete Case
fregMICE Imputation
Mean Imputation
11: Point-wise 95% confidence band coverage for Setting 2 Scenarios (a) (Top) and
om) with n = 350. Columns (left to right) correspond to 10, 20, and 30% missing d
top to bottom) correspond to functional parameters β0, β1, β2, and β3. 10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
t
Method
All
Complete Case
fregMICE Imputation
Mean Imputation Method
All
Complete Case
fregMICE Imputation
Mean Imputation 10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
t
Method
All
Complete Case
fregMICE Imputation
Mean Imputation
Figure 11: Point-wise 95% confidence band coverage for Setting 2 Scenarios (a) (Top) and (b)
(Bottom) with n = 350. Columns (left to right) correspond to 10, 20, and 30% missing data. Rows (top to bottom) correspond to functional parameters β0, β1, β2, and β3. 10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
t
Method
All
Complete Case
fregMICE Imputation
Mean Imputation Method
All
Complete Case
fregMICE Imputation
Mean Imputation Figure 11: Point-wise 95% confidence band coverage for Setting 2 Scenarios (a) (Top) and (b)
(Bottom) with n = 350. 2.3.2
Point-wise 95% Confidence Band Coverage and Width Point-wise standardized
bias curve for mean imputation is removed to better compare estimates from ANM, CCA, and
fregMICE 14 Method
ANM
CCA
fregMICE Imputation
Mean Imputation
True Function 10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0.0
0.5
1.0
1.5
2.0
2.5
−1.0
−0.5
0.0
0.5
1.0
0.0
0.2
0.4
0.6
0.8
−0.4
−0.3
−0.2
−0.1
0.0
t
Method
ANM
CCA
fregMICE Imputation
Mean Imputation
True Function
10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0
1
2
3
−2
0
2
−3
−2
−1
0
1
−1
0
1
2
3
4
t
10: (Top) Point-wise mean curves and (Bottom) point-wise standardized bias in Sett
io (b) for n = 100. Columns (left to right) correspond to 10, 20, and 30% missing data. R
o bottom) correspond to functional parameters β0, β1, β2, and β3. Point-wise standar
urve for mean imputation is removed to better compare estimates from ANM, CCA
CE. 15 β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0.0
−0.4
−0.3
−0.2
−0.1
0.0
t
10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0
1
2
3
−2
0
2
−3
−2
−1
0
1
−1
0
1
2
3
4
t Figure 10: (Top) Point-wise mean curves and (Bottom) point-wise standardized bias in Setting 2
Scenario (b) for n = 100. Columns (left to right) correspond to 10, 20, and 30% missing data. Rows
(top to bottom) correspond to functional parameters β0, β1, β2, and β3. 2.3.2
Point-wise 95% Confidence Band Coverage and Width Columns (left to right) correspond to 10, 20, and 30% missing data. Rows (top to bottom) correspond to functional parameters β0, β1, β2, and β3. 16 10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
t
Method
ANM
CCA
fregMICE Imputation
Mean Imputation
10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
t
Method
ANM
CCA
fregMICE Imputation
Mean Imputation
12: Point-wise 95% confidence band coverage for Setting 2 Scenarios (a) (Top) and
om) with n = 100. Columns (left to right) correspond to 10, 20, and 30% missing d
top to bottom) correspond to functional parameters β0, β1, β2, and β3. 10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
t
Method
ANM
CCA
fregMICE Imputation
Mean Imputation Method
ANM
CCA
fregMICE Imputation
Mean Imputation Method
ANM
CCA
fregMICE Imputation
Mean Imputation 10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
t
Method
ANM
CCA
fregMICE Imputation
Mean Imputation
Figure 12: Point-wise 95% confidence band coverage for Setting 2 Scenarios (a) (Top) and (b)
(Bottom) with n = 100. Columns (left to right) correspond to 10, 20, and 30% missing data. Rows (top to bottom) correspond to functional parameters β0, β1, β2, and β3. 2.3.2
Point-wise 95% Confidence Band Coverage and Width 10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
t
Method
ANM
CCA
fregMICE Imputation
Mean Imputation Method
ANM
CCA
fregMICE Imputation
Mean Imputation Figure 12: Point-wise 95% confidence band coverage for Setting 2 Scenarios (a) (Top) and (b)
(Bottom) with n = 100. Columns (left to right) correspond to 10, 20, and 30% missing data. Rows (top to bottom) correspond to functional parameters β0, β1, β2, and β3. 17 10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0.3
0.4
0.5
0.3
0.4
0.5
0.15
0.20
0.25
0.30
0.15
0.20
0.25
0.30
t
Method
All
Complete Case
fregMICE Imputation
Mean Imputation 10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0.3
0.4
0.5
0.3
0.4
0.5
0.15
0.20
0.25
0.30
0.15
0.20
0.25
0.30
t
Method
All
Complete Case
fregMICE Imputation
Mean Imputation
10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0.3
0.4
0.5
0.6
0.2
0.3
0.4
0.5
0.6
0.15
0.20
0.25
0.30
0.10
0.15
0.20
0.25
0.30
0.35
t
Method
All
Complete Case
fregMICE Imputation
Mean Imputation
13:
Point-wise 95% confidence band width for Setting 1 Scenarios (a) (Top) and
om) with n = 350. Columns (left to right) correspond to 10, 20, and 30% missing da
top to bottom) correspond to functional parameters β0, β1, β2, and β3. Method
All
Complete Case
fregMICE Imputation
Mean Imputation t
10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0.3
0.4
0.5
0.6
0.2
0.3
0.4
0.5
0.6
0.15
0.20
0.25
0.30
0.10
0.15
0.20
0.25
0.30
0.35
t
Method
All
Complete Case
fregMICE Imputation
Mean Imputation
Figure 13:
Point-wise 95% confidence band width for Setting 1 Scenarios (a) (Top) and (b)
(Bottom) with n = 350. Columns (left to right) correspond to 10, 20, and 30% missing data. Rows (top to bottom) correspond to functional parameters β0, β1, β2, and β3. 2.3.2
Point-wise 95% Confidence Band Coverage and Width 10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0.3
0.4
0.5
0.6
0.2
0.3
0.4
0.5
0.6
0.15
0.20
0.25
0.30
0.10
0.15
0.20
0.25
0.30
0.35
t
Method
All
Complete Case
fregMICE Imputation
Mean Imputation Method
All
Complete Case
fregMICE Imputation
Mean Imputation Figure 13:
Point-wise 95% confidence band width for Setting 1 Scenarios (a) (Top) and (b)
(Bottom) with n = 350. Columns (left to right) correspond to 10, 20, and 30% missing data. Rows (top to bottom) correspond to functional parameters β0, β1, β2, and β3. 18 10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0.5
0.6
0.7
0.8
0.9
0.4
0.5
0.6
0.7
0.8
0.9
0.3
0.4
0.5
0.2
0.3
0.4
0.5
t
Method
ANM
CCA
fregMICE Imputation
Mean Imputation 10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0.5
0.6
0.7
0.8
0.9
0.4
0.5
0.6
0.7
0.8
0.9
0.3
0.4
0.5
0.2
0.3
0.4
0.5
t
Method
ANM
CCA
fregMICE Imputation
Mean Imputation
10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0.4
0.6
0.8
1.0
0.3
0.5
0.7
0.9
0.2
0.3
0.4
0.5
0.2
0.3
0.4
0.5
t
Method
ANM
CCA
fregMICE Imputation
Mean Imputation
14:
Point-wise 95% confidence band width for Setting 1 Scenarios (a) (Top) and
om) with n = 100. Columns (left to right) correspond to 10, 20, and 30% missing da
top to bottom) correspond to functional parameters β0, β1, β2, and β3. Method
ANM
CCA
fregMICE Imputation
Mean Imputation β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0.2
0.3
0.4
0.5
t
10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0.4
0.6
0.8
1.0
0.3
0.5
0.7
0.9
0.2
0.3
0.4
0.5
0.2
0.3
0.4
0.5
t
Method
ANM
CCA
fregMICE Imputation
Mean Imputation
Figure 14:
Point-wise 95% confidence band width for Setting 1 Scenarios (a) (Top) and (b)
(Bottom) with n = 100. Columns (left to right) correspond to 10, 20, and 30% missing data. 2.3.2
Point-wise 95% Confidence Band Coverage and Width Rows (top to bottom) correspond to functional parameters β0, β1, β2, and β3. 10
20
30
β0
β1
β2
β3
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0.4
0.6
0.8
1.0
0.3
0.5
0.7
0.9
0.2
0.3
0.4
0.5
0.2
0.3
0.4
0.5
t
Method
ANM
CCA
fregMICE Imputation
Mean Imputation Figure 14:
Point-wise 95% confidence band width for Setting 1 Scenarios (a) (Top) and (b)
(Bottom) with n = 100. Columns (left to right) correspond to 10, 20, and 30% missing data. Rows (top to bottom) correspond to functional parameters β0, β1, β2, and β3. 19 2.4
Simulation Study with a Scalar Response Here we consider a simulation study where the analysis model has a scalar response with scalar
and functional predictors. As in the simulation study presented in the main article, we compare
the complete case, mean imputation, and fregMICE approaches. 2.4.2
Procedures Compared and Performance Measures For each simulation setting, we fit the correctly specified model for the scalar response, y, using
PFR, implemented by the pfr function from the R package refund (Goldsmith et al., 2018). For
the PFR fitting procedure, the functional predictor was represented using functional principle
components (FPCs) by smoothed covariance (Yao et al., 2003) where the number of FPCs was
selected to be the minimum number of components explaining at least 99% of the variance in the
functional observations. The coefficient function, β1, was represented using a basis set of 30 thin-
plate regression splines and the fitting procedure penalized the magnitude of the second derivative. As a benchmark, we used all of the data, prior to imposing missingess on any of the variables to
fit model (1). In the results, we refer to this as “all no missing” (ANM). For mean imputation, we
filled in any missing X functions with the point-wise mean function of the observed X functions. For complete case analysis (CCA), the analysis model was fit on observations with complete data
only. For our fregMICE procedure, the imputation model for the missing X functions was given by
Xi(t) = γ0(t)+γ1(t)zi +γ2(t)yi +ε(t) where the error term at each point is assumed to be from the
N(0, σ2
ε) distribution with constant error variance, σ2
ε. We used the pffr function from the refund
package (Goldsmith et al., 2018) to fit this model in each iteration of the fregMICE procedure
using 20 cubic B-splines to represent each coefficient function and we penalized the magnitude of
the second derivative. We ran the fregMICE procedure for 20 iterations and constructed 5 imputed
data sets. We fit model (1) on each imputed data set and used the extension of Rubin’s rules
described in Section 4.2 of the article to pool estimates from the 5 data sets. ¯ We computed the mean squared error (MSE) and standardized bias, given by (ˆθ1 −θ1)/sd(ˆθ1),
for estimates of the scalar coefficient θ1. For the standardized bias computation, ¯ˆθ1 is the average
of the 500 estimates of θ1 and sd(ˆθ1) is the Monte-Carlo standard deviation of these estimates. These values are given in Table 4. Mean coverage and width of the 95% confidence intervals for θ1
are given in Table 5. As in the main article, we provide point-wise mean and point-wise standardized bias curves for
the estimates of the β1 coefficient function from each method. 2.4.2
Procedures Compared and Performance Measures These plots are displayed in Figures
16 and 17. Point-wise coverage and width of the 95% confidence bands for β1 are displayed in
Figures 18 and 19. Across-the-function mean coverage and width of the 95% confidence bands for
β1 are given in Table 6. 2.4.1
Data Generation 10% Intended
20% Intended
30% Intended
MCAR
n = 100
0.10 (0.03)
0.20 (0.04)
0.27 (0.04)
n = 350
0.10 (0.02)
0.20 (0.02)
0.27 (0.02)
MAR
n = 100
0.11 (0.03)
0.20 (0.04)
0.30 (0.04)
n = 350
0.10 (0.02)
0.20 (0.02)
0.30 (0.03) 2.4.3
Results Point and Interval Estimates for θ1 2.4.1
Data Generation For this simulation, each complete observation consists of a scalar response, y, a scalar predictor, z,
and a functional predictor, X. For each observation (i = 1, . . . , n), we generated a scalar predictor
zi ∼N(0, 1) and a functional predictor, Xi, observed on a grid {tg =
g
10 : g = 0, 1, . . . , 100} such
that Xi(tg) = ui1 + ui2tg + ui3 20
√
2πe−(tg−3)2/2 + P10
k=1
vik1sin
2πk
10 tg
+ vik2cos
2πk
10 tg
where
ui1 ∼U(0, 5), ui2 ∼N(1, 0.04), ui3 ∼N(0, 1), vik1, vik2 ∼N
0, 1
k2
for k = 1, . . . , 10. Figure 15
displays a set of 50 simulated predictor functions. We generated the scalar response according to
the model, yi = θ0 + θ1zi +
Z
Xi(t)β1(t) + εi,
(1) (1) where θ0 = 0, θ1 = 1, β1(t) = sin
πt
5
, and εi ∼N(0, 0.5). where θ0 = 0, θ1 = 1, β1(t) = sin
πt
5
, and εi ∼N(0, 0.5). We considered two settings: one in which the functional predictor is MCAR and a second in
which it is MAR. In both settings, yi and zi are always observed and the probability of Xi being
observed is given by logit{P(RXi = 1)} = ψ0 + ψ1yi. In the MCAR setting, we have ψ1 = 0 and
ψ0 = log
0.9
0.1
, log
0.8
0.2
, and log
0.7
0.3
to achieve about 10%, 20%, and 30% missingness respectively. In the MAR setting, we have ψ1 = −6 and ψ0 = 6.2, 1, and −3 to achieve about 10%, 20%, and 30%
missingness respectively. For each setting we considered two sample sizes: n = 350 and n = 100
and generated 500 data sets for each combination of setting, missingness, and sample size. The
average proportion of missing data for each combination is given in Table 3. −10
0
10
20
0.0
2.5
5.0
7.5
10.0
t
Figure 15: 50 simulated functional predictors (X) with three highlighted observations. −10
0
10
20
0.0
2.5
5.0
7.5
10.0
t Figure 15: 50 simulated functional predictors (X) with three highlighted observations. 20 Table 3: Mean (SD) proportion of missing X functions. Point and Interval Estimates for θ1 Table 4 shows that the MSE and standardized bias for estimating θ1 for the fregMICE and
CCA estimates are comparable across the different amounts of missingness and sample size in the
MCAR setting. Mean imputation yields higher MSE, but similar standardized biases. In the MAR
setting, the fregMICE estimates for θ1 have smaller magnitude MSE and standardized bias under
all combinations of missingness and sample size in comparison to both mean imputation and CCA. 21 Table 4: MSE (SD) and standardized bias for θ1 estimates from scalar response simulations. MSE
(SD) values are ×10−4. )
10%
20%
30%
MCAR
MSE (SD)
Std.Bias
MSE (SD)
Std.Bias
MSE (SD)
Std.Bias
n = 100
ANM
5.42 (7.82)
-0.02
5.42 (7.82)
-0.02
5.42 (7.82)
-0.02
Mean
8.91 (12.35)
0.05
10.87 (15.14)
0.03
13.49 (20.07)
-0.01
CCA
6.25 (9.43)
-0.02
7.14 (10.63)
-0.03
8.07 (11.76)
-0.03
fregMICE
6.29 (9.40)
-0.01
6.98 (10.55)
-0.03
7.85 (10.75)
-0.06
n = 350
ANM
1.43 (2.05)
0.03
1.43 (2.05)
0.03
1.43 (2.05)
0.03
Mean
2.33 (3.30)
0.00
3.27 (4.43)
-0.01
3.72 (5.20)
-0.02
CCA
1.64 (2.36)
0.02
1.88 (2.73)
0.01
2.05 (3.05)
0.01
fregMICE
1.63 (2.35)
0.01
1.92 (2.82)
0.00
2.06 (2.98)
0.02
MAR
n = 100
ANM
5.42 (7.82)
-0.02
5.42 (7.82)
-0.02
5.42 (7.82)
-0.02
Mean
45.13 (50.99)
1.39
61.48 (64.80)
1.59
67.65 (71.32)
1.51
CCA
8.94 (12.38)
-0.63
13.89 (18.29)
-0.93
21.17 (25.27)
-1.17
fregMICE
6.64 (9.25)
-0.05
7.28 (9.90)
-0.09
9.08 (13.24)
-0.13
n = 350
ANM
1.43 (2.05)
0.03
1.43 (2.05)
0.03
1.43 (2.05)
0.03
Mean
31.97 (21.95)
2.64
46.95 (28.93)
2.97
51.44 (32.71)
2.81
CCA
3.43 (3.97)
-1.10
6.97 (6.39)
-1.70
12.54 (10.21)
-2.16
fregMICE
1.68 (2.30)
0.04
2.01 (2.73)
0.02
2.47 (3.33)
-0.02
Table 5: Coverages (SD) and widths (SD) of 95% confidence intervals for θ1. able 5: Coverages (SD) and widths (SD) of 95% confidence intervals for θ1. 10%
20%
30%
MCAR
Cov. (SD)
Width (SD)
Cov. (SD)
Width (SD)
Cov. Point and Interval Estimates for θ1 (SD)
Width (SD)
n = 100
ANM
0.95 (0.21)
0.29 (0.03)
0.95 (0.21)
0.29 (0.03)
0.95 (0.21)
0.29 (0.03)
Mean
0.94 (0.23)
0.36 (0.05)
0.95 (0.23)
0.42 (0.06)
0.95 (0.23)
0.46 (0.06)
CCA
0.94 (0.24)
0.30 (0.04)
0.94 (0.24)
0.32 (0.04)
0.94 (0.24)
0.34 (0.04)
fregMICE
0.94 (0.24)
0.30 (0.04)
0.94 (0.24)
0.31 (0.04)
0.93 (0.26)
0.33 (0.05)
n = 350
ANM
0.96 (0.20)
0.15 (0.01)
0.96 (0.20)
0.15 (0.01)
0.96 (0.20)
0.15 (0.01)
Mean
0.96 (0.20)
0.19 (0.01)
0.94 (0.23)
0.22 (0.02)
0.95 (0.21)
0.24 (0.02)
CCA
0.95 (0.22)
0.16 (0.01)
0.94 (0.24)
0.17 (0.01)
0.94 (0.23)
0.18 (0.01)
fregMICE
0.95 (0.21)
0.16 (0.01)
0.94 (0.23)
0.17 (0.02)
0.93 (0.25)
0.17 (0.02)
MAR
n = 100
ANM
0.95 (0.21)
0.29 (0.03)
0.95 (0.21)
0.29 (0.03)
0.95 (0.21)
0.29 (0.03)
Mean
0.71 (0.45)
0.49 (0.07)
0.69 (0.46)
0.56 (0.07)
0.71 (0.45)
0.61 (0.07)
CCA
0.90 (0.30)
0.31 (0.04)
0.84 (0.37)
0.33 (0.04)
0.78 (0.41)
0.36 (0.05)
fregMICE
0.95 (0.23)
0.32 (0.04)
0.95 (0.22)
0.34 (0.06)
0.93 (0.26)
0.37 (0.07)
n = 350
ANM
0.96 (0.20)
0.15 (0.01)
0.96 (0.20)
0.15 (0.01)
0.96 (0.20)
0.15 (0.01)
Mean
0.25 (0.43)
0.25 (0.02)
0.19 (0.39)
0.29 (0.02)
0.21 (0.41)
0.32 (0.02)
CCA
0.84 (0.37)
0.16 (0.01)
0.62 (0.49)
0.17 (0.01)
0.45 (0.50)
0.19 (0.01)
fregMICE
0.95 (0.21)
0.16 (0.01)
0.95 (0.21)
0.18 (0.02)
0.95 (0.23)
0.20 (0.03) Table 5 shows that the 95% confidence intervals from the fregMICE imputed data achieve or
nearly achieve the nominal coverage in both the MCAR and MAR settings for both sample sizes. Intervals based on mean imputed data and CCA perform similarly in the MCAR setting, but show
considerably worse coverage in the MAR setting for each combination of sample size and missing-
ness. The widths of the 95% confidence intervals from the fregMICE imputed data tend to be the
same as or slightly larger than those for CCA with both being close to the ANM benchmark widths
or slightly wider in both the MCAR and MAR settings. Point and Interval Estimates for β1 Figure 16 shows that, in the MCAR setting the point-wise means and point-wise standardized
biases for the β1 function are similar for the fregMICE imputation and CCA approaches with freg-
MICE showing slightly less bias for greater amounts of missing data. Estimates derived from mean
imputed data perform especially poorly for the smaller sample size of n = 100 and performance 22 deteriorates with greater amounts of missing data. Figure 17 shows that, in the MAR setting,
the fregMICE imputation approach yields estimates for β1 that perform nearly identically to the
ANM benchmark estimates while the estimates derived from mean imputed data and CCA show
considerable amounts of bias. The bias increases with smaller sample size and with greater amounts
of missing data. For the smaller sample size of n = 100 in the MCAR setting, Figure 18 shows that the the point-
wise coverage for the 95% confidence intervals derived from the fregMICE imputed data is similar
to or better than that of the ANM benchmark and typically better than the coverage provided by
intervals based on mean imputed data or CCA. The point-wise width for the intervals derived from
the fregMICE imputed data sets increases with greater amounts of missing data. The coverage is
poor for the intervals, regardless of the method used, since the small sample size tended to result
in estimates of β1 that were linear (i.e., the PFR fitting procedure placed a large penalty on the
second derivative of the coefficient estimate). The larger widths of the fregMICE-based intervals
for larger amounts of missing data explain why the coverage of these intervals is better than than
the coverage of intervals from the ANM benchmark as the amount of missing data increases. For
the larger sample size of n = 350 in the MCAR setting, Figure 18 shows that point-wise coverage is
typically greater than 95% for fregMICE-based intervals and those based on CCA. Intervals based
on mean imputed data perform poorly with respect to coverage as the amount of missing data
increases. As in the cases with the smaller sample size, the widths of the fregMICE-based intervals
increase for larger amounts of missing data. In the MAR setting, Figure 19 shows similar results, with respect to point-wise 95% confidence
interval coverage and width, compared to those in the MCAR setting discussed above. Point and Interval Estimates for β1 The main
difference is that we see a more pronounced widening of the fregMICE-based intervals as the amount
of missing data increases in both the small and large sample settings. Table 6 provides additional
summary measures that support the findings in Figures 18 and 19. Table 6:
Across-the-function mean (SD) point-wise 95% confidence band coverage (pwCov) of β1
and width (pwWidth) for the MCAR and MAR settings. Point and Interval Estimates for β1 10%
20%
30%
MCAR
pwCov (SD)
pwWidth (SD)
pwCov (SD)
pwWidth (SD)
pwCov (SD)
pwWidth (SD)
n = 100
ANM
0.82 (0.32)
0.95 (0.35)
0.82 (0.32)
0.95 (0.35)
0.82 (0.32)
0.95 (0.35)
Mean
0.65 (0.36)
0.91 (0.49)
0.53 (0.34)
0.84 (0.54)
0.49 (0.30)
0.80 (0.53)
CCA
0.78 (0.34)
0.95 (0.38)
0.75 (0.35)
0.94 (0.41)
0.72 (0.36)
0.95 (0.43)
fregMICE
0.83 (0.29)
0.92 (0.30)
0.85 (0.27)
1.02 (0.36)
0.86 (0.25)
1.13 (0.52)
n = 350
ANM
0.99 (0.04)
0.70 (0.07)
0.99 (0.04)
0.70 (0.07)
0.99 (0.04)
0.70 (0.07)
Mean
0.99 (0.08)
0.80 (0.08)
0.96 (0.18)
0.89 (0.16)
0.88 (0.29)
0.91 (0.27)
CCA
0.98 (0.06)
0.72 (0.08)
0.98 (0.06)
0.75 (0.09)
0.97 (0.09)
0.77 (0.10)
fregMICE
0.98 (0.06)
0.69 (0.07)
0.98 (0.06)
0.74 (0.11)
0.97 (0.07)
0.79 (0.19)
MAR
n = 100
ANM
0.82 (0.32)
0.95 (0.35)
0.82 (0.32)
0.95 (0.35)
0.82 (0.32)
0.95 (0.35)
Mean
0.48 (0.30)
0.77 (0.48)
0.45 (0.23)
0.75 (0.42)
0.48 (0.19)
0.79 (0.37)
CCA
0.78 (0.34)
0.93 (0.36)
0.73 (0.36)
0.91 (0.40)
0.68 (0.37)
0.90 (0.43)
fregMICE
0.83 (0.28)
0.96 (0.32)
0.86 (0.25)
1.11 (0.41)
0.87 (0.23)
1.32 (0.64)
n = 350
ANM
0.99 (0.04)
0.70 (0.07)
0.99 (0.04)
0.70 (0.07)
0.99 (0.04)
0.70 (0.07)
Mean
0.92 (0.23)
0.92 (0.21)
0.74 (0.36)
0.89 (0.38)
0.54 (0.36)
0.76 (0.45)
CCA
0.98 (0.07)
0.70 (0.09)
0.97 (0.08)
0.72 (0.09)
0.95 (0.11)
0.73 (0.10)
fregMICE
0.98 (0.06)
0.71 (0.10)
0.98 (0.06)
0.80 (0.18)
0.98 (0.06)
0.93 (0.28) 23 Method
ANM
CCA
fregMICE Imputation
Mean Imputation
True Function 10
20
30
n = 100
n = 350
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
−1.5
−1.0
−0.5
0.0
0.5
1.0
1.5
−1.0
−0.5
0.0
0.5
1.0
t
Method
ANM
CCA
fregMICE Imputation
Mean Imputation
True Function
10
20
30
n = 100
n = 350
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
−3
−2
−1
0
1
2
−1.5
−1.0
−0.5
0.0
0.5
1.0
t
igure 16: (Top) Point-wise mean curves and (Bottom) point-wise standardized bias for β1 in
MCAR setting. Columns (left to right) correspond to 10, 20, and 30% missing data. Rows corre-
pond to sample size n = 100 or n = 350. 30 30 Figure 16: (Top) Point-wise mean curves and (Bottom) point-wise standardized bias for β1 in
MCAR setting. Point and Interval Estimates for β1 Columns (left to right) correspond to 10, 20, and 30% missing data. Rows corre-
spond to sample size n = 100 or n = 350. 24 10
20
30
n = 100
n = 350
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
−1
0
1
−1.0
−0.5
0.0
0.5
1.0
t
Method
ANM
CCA
fregMICE Imputation
Mean Imputation
True Function
10
20
30
n = 100
n = 350
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
−8
−4
0
4
−2
0
2
t
Figure 17: (Top) Point-wise mean curves and (Bottom) point-wise standardized bias for β1 in
MAR setting. Columns (left to right) correspond to 10, 20, and 30% missing data. Rows correspond
to sample size n = 100 or n = 350. Method
ANM
CCA
fregMICE Imputation
Mean Imputation
True Function 10
20
30
n = 100
n = 350
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
−8
−4
0
4
−2
0
2
t 30 Figure 17: (Top) Point-wise mean curves and (Bottom) point-wise standardized bias for β1 in
MAR setting. Columns (left to right) correspond to 10, 20, and 30% missing data. Rows correspond
to sample size n = 100 or n = 350. 25 10
20
30
n = 100
n = 350
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
t
Method
ANM
CCA
fregMICE Imputation
Mean Imputation
10
20
30
n = 100
n = 350
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
1
2
3
0.5
1.0
1.5
2.0
t
Figure 18: (Top) Point-wise 95% confidence band coverage of β1 in MCAR setting and (Bottom)
point-wise 95% confidence band width. Columns (left to right) correspond to 10, 20, and 30%
missing data. Rows correspond to sample size n = 100 or n = 350. Method
ANM
CCA
fregMICE Imputation
Mean Imputation 20 30 Figure 18: (Top) Point-wise 95% confidence band coverage of β1 in MCAR setting and (Bottom)
point-wise 95% confidence band width. Columns (left to right) correspond to 10, 20, and 30%
missing data. Rows correspond to sample size n = 100 or n = 350. 3
Additional Results from EMBARC EEG Application 3.1
fregMICE Computational Considerations Point and Interval Estimates for β1 26 10
20
30
n = 100
n = 350
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
t
Method
ANM
CCA
fregMICE Imputation
Mean Imputation
10
20
30
n = 100
n = 350
0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0 0.0
2.5
5.0
7.5
10.0
1
2
3
0.5
1.0
1.5
2.0
2.5
t
Figure 19: (Top) Point-wise 95% confidence band coverage of β1 in MAR setting and (Bottom)
point-wise 95% confidence band width. Columns (left to right) correspond to 10, 20, and 30%
missing data. Rows correspond to sample size n = 100 or n = 350. Method
ANM
CCA
fregMICE Imputation
Mean Imputation 20 30 0.0
2.5
5.0
7.5
10.0 Figure 19: (Top) Point-wise 95% confidence band coverage of β1 in MAR setting and (Bottom)
point-wise 95% confidence band width. Columns (left to right) correspond to 10, 20, and 30%
missing data. Rows correspond to sample size n = 100 or n = 350. 27 3.2
Functional Convergence Plots Figures 20 and 21 show plots of the means from the imputed frontal asymmetry (FA) curves at
each iteration of the fregMICE algorithm for each of the 20 streams in the health control (HC) and
major depressive disorder (MDD) subsets, respectively. Note that we have scaled the horizontal
axis to be between 0 and 1. If we look across the 20 panels for each subset, we see overlap in the
streams and no discernible patterns to suggest convergence issues with the imputed FA values. 11
12
13
14
15
16
17
18
19
20
1
2
3
4
5
6
7
8
9
10
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
−0.2
0.0
0.2
−0.2
0.0
0.2
5
10
15
20
Iteration
Number
Figure 20: Functional convergence plots. Point-wise mean of the imputed function values for HC
subjects. Each panel corresponds to one imputation stream. Dark colors correspond to later
iterations. Figure 20: Functional convergence plots. Point-wise mean of the imputed function values for HC
subjects. Each panel corresponds to one imputation stream. Dark colors correspond to later
iterations. 3.1
fregMICE Computational Considerations We conducted the EMBARC data analysis using R version 4.0.0 on an iMAC with a 4.2GHz Intel
Core i7 processor. To run one iteration of the fregMICE procedure on the subset of healthy control
data (n = 40) it took 9.14 seconds, on average (sd = 2.52 seconds) based on 10 runs. To run one
iteration of the fregMICE procedure on the subset of MDD data (n = 295) it took 47.69 seconds,
on average (sd = 6.78 seconds) based on 10 runs. 28 3.3
Functional Strip Plots Figures 22 and 23 show strip plots for the FA curves from the HC and MDD subsets respectively. In each figure, there is one panel for each of the 20 imputed data sets. In each of the numbered
panels, we show the imputed FA values. The panels titled “Observed” show the observed FA curves
for the sample. These plots show that the imputed FA functions tend to fall within the range of
values of the observed FA functions and have similar characteristics. alues of the observed FA functions and have similar characteristics. Observed
11
12
13
14
15
16
17
18
19
20
Observed
1
2
3
4
5
6
7
8
9
10
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
−1.0
−0.5
0.0
0.5
1.0
−1.0
−0.5
0.0
0.5
1.0
Figure 22: Functional strip plot for HC subset. Observed
11
12
13
14
15
16
17
18
19
20
Observed
1
2
3
4
5
6
7
8
9
10
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
−1.0
−0.5
0.0
0.5
−1.0
−0.5
0.0
0.5
Figure 23: Functional strip plot for MDD subset. Observed
11
12
13
14
15
16
17
18
19
20
Observed
1
2
3
4
5
6
7
8
9
10
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
−1.0
−0.5
0.0
0.5
1.0
−1.0
−0.5
0.0
0.5
1.0
Figure 22: Functional strip plot for HC subset. Figure 22: Functional strip plot for HC subset. 3.2
Functional Convergence Plots 11
12
13
14
15
16
17
18
19
20
1
2
3
4
5
6
7
8
9
10
0
5
0
5
0
0
5
0
5
0
0
5
0
5
0
0
5
0
5
0
0
5
0
5
0
0
5
0
5
0
0
5
0
5
0
0
5
0
5
0
0
5
0
5
0
0
5
0
5
0
−0.10
−0.05
0.00
0.05
0.10
0.15
−0.10
−0.05
0.00
0.05
0.10
0.15
5
10
15
20
Iteration
Number 11
12
13
14
15
16
17
18
19
20
1
2
3
4
5
6
7
8
9
10
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
−0.10
−0.05
0.00
0.05
0.10
0.15
−0.10
−0.05
0.00
0.05
0.10
0.15
5
10
15
20
Iteration
Number
Figure 21: Functional convergence plots. Point-wise mean of the imputed function values for HC
subjects. Each panel corresponds to one imputation stream. Dark colors correspond to later
iterations. Figure 21: Functional convergence plots. Point-wise mean of the imputed function values for HC
subjects. Each panel corresponds to one imputation stream. Dark colors correspond to later
iterations. 29 3.3
Functional Strip Plots Observed
11
12
13
14
15
16
17
18
19
20
Observed
1
2
3
4
5
6
7
8
9
10
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
−1.0
−0.5
0.0
0.5
−1.0
−0.5
0.0
0.5
Figure 23: Functional strip plot for MDD subset. Figure 23: Functional strip plot for MDD subset. 30 3.4
Scalar Convergence and Strip Plots Figure 24 shows diagnostic plots of the mean value of the imputed WASIV scores at each iter-
ation of the fregMICE algorithm for each of the 20 streams (various line colors) in the HC and
MDD subsets. We see a fair amount of mixing in the streams and no discernible patterns that
would suggest convergence issues. Strip plots showing the imputed and observed values for the the
WASIV scores from the 20 imputed data sets are provided in Figure 25. The plots reveal several
instances when the imputed WASIV scores were higher than the maximum possible score of 80. These scores were set to 80 prior to fitting the analysis model. HC
MDD
1
2
3
4
5
6
7
8
9
10 11 12 13 14 15 16 17 18 19 20
1
2
3
4
5
6
7
8
9
10 11 12 13 14 15 16 17 18 19 20
60
63
66
60
65
70
75
Iteration
Figure 24: Scalar convergence plots: Mean of the imputed WASIV values for HC and MDD subjects. The different colors correspond to the different streams. Figure 24: Scalar convergence plots: Mean of the imputed WASIV values for HC and MDD subjects. The different colors correspond to the different streams. HC
MDD
Observed
1 2 3 4 5 6 7 8 9 1011121314151617181920
Observed
1 2 3 4 5 6 7 8 9 1011121314151617181920
20
40
60
80
100
Imputation Number
Figure 25: Strip plots of WASIV values. Horizontal axis shows imputation number (m). Vertical
axis shows WASIV value. Observed values are blue and imputed values are red. Figure 25: Strip plots of WASIV values. Horizontal axis shows imputation number (m). Vertical
axis shows WASIV value. Observed values are blue and imputed values are red. 31 References Goldsmith, J., Bobb, J., Crainiceanu, C., and Reich, D. (2011). Penalized functional regression. Journal of Computational and Graphical Statistics 20:830–851. Goldsmith, J., Scheipl, F., Huang, L., Wrobel, J., Gellar, J., Harezlak, J., McLean, M., Swihart,
B., Xiao, L., Crainiceanu, C., and Reiss, P. (2018). refund: Regression with Functional Data. R
package version 0.1-17. Ivanescu, A. E., Staicu, A.-M., Scheipl, F., and Greven, S. (2015). Penalized function-on-function
regression. Computational Statistics 30:539–568. Yao, F., M¨uller, H., Clifford, A., Dueker, S., Follett, J., Lin, Y., Buchholz, B., and Vogel, J. (2003). Shrinkage estimation for functional principal component scores with application to the
population. Biometrics 59:676 – 685. 32 32
|
https://openalex.org/W4293213366
|
https://www.researchsquare.com/article/rs-1412884/latest.pdf
|
English
| null |
Image-based disease classification in grape leaves using convolutional capsule network
|
Soft computing
| 2,022
|
cc-by
| 8,591
|
Abstract: Crop protection is the prime hindrance for food security. The plant diseases destroy the overall
quality and quantity of the agricultural products. Grape is an important fruit and major source
of vitamin C nutrients. The automatic decision making system plays a paramount role in
agricultural informatics. This paper aims to detect the diseases in grape leaves using
convolutional capsule networks. The capsule network is a promising neural network in the field
of deep learning. This network uses a group of neurons as capsules and effectively represents
spatial information of features. The novelty of the proposed work relies on the addition of
convolutional layers before the primary caps layer which indirectly decrease the number of
capsules and speed up the dynamic routing process. The proposed method is experimented with
augmented and non-augmented datasets. It effectively detects the diseases of grape leaves with
the accuracy of 99.12%. The performance of the method is compared with state-of-the art deep
learning methods and produces reliable results. Research Article Keywords: disease detection, convolutional capsule network, grapes, agriculture
Posted Date: March 9th, 2022
DOI: https://doi.org/10.21203/rs.3.rs-1412884/v1
License: This work is licensed under a Creative Commons Attribution 4.0 International License. Read Full License License: This work is licensed under a Creative Commons Attribution 4.0 International License. R
d F ll Li Image based Disease Classification in Grape Leaves using Convolutional
Capsule Network
1*A.Diana Andrushia, 2T.Mary Neebha, 3A.Trephena Patricia, 4S.Umadevi, 5N.Anand, 6Atul
Varshney
1andrushia@gmail.com, 2neebha08@gmail.com, 3 trephyece@gmail.com, 4umadevi.s@vit.ac.in,
5davids1612@gmail.com, 6atul.varshney@gkv.ac.in
1,2,5Karunya Institute of Technology and Sciences, Coimbatore, Tamil Nadu, India
3 Panimalar Engineering College, Chennai, Tamil Nadu, India
4 Centre for Nano electronics and VLSI Design, SENSE, VIT University, Chennai campus,
Tamil Nadu, India
6Gurukula Kangri University, Haridwar, Uttarakhand, India 1. Introduction: In agriculture, plant diseases have a vital distress because they reduce the crop quality
constantly and also reduce the production. The properties of plant diseases range from minor
symptoms to serious damage and thereby reduce the agricultural economy drastically [Savary,
S et al., 2012]. In order to avoid the losses, different methods have been developed to detect
the disease. The methods that are associated with molecular biology give correct identification
of disease agents. But, these methods are not available to the farmers directly. It requires a lot
of money and domain knowledge to adhere to. With the above reasons, a lot of research has
been carried out to derive the methodologies which will be accurate enough and to be accessible by the farmers. The decision making technologies are used to solve the problems of precision
agriculture [Arsenovic, M et al., 2019]. In India grapes (scientific name vitis vinifera) are one of the common fruits. India is holding
18th position for the production of grapes. The Indian states of Tamil Nadu, Andhra Pradesh,
Karnataka and Maharashtra are the major producing states of grapes. The grape plant grows
well with the temperature of 15oC-40oC. The leaves are the paramount part of any plant. It has
a long span of time in comparison with the buds and flowers. It shows the entire characteristics
of plants. The leaves of the grape plants are commonly affected by black rot, black esca and
leaf blight diseases. All these diseases affect the grape yield and produce loss to the cultivators
[Andrushia et al., 2019, 2020]. Early detection and identification of diseases help the farmers
to reduce the losses. Hence, it is essential to find the diseases by automatic systems at the earlier
stages. The latest technologies are adopted to enhance the decision making process of precision
agriculture [Gebbers, R et al., 2010]. The larger amount of data has been collected from real
time and various artificial intelligence techniques are used to give optimal decisions, which
lead to the cost reduction. Still, the field of decision making systems in agriculture informatics
is in infancy and open for improvements. It has been reviewed that various machine learning
algorithms are used for this purpose. Decision trees, logistic regressions, k-nearest neighbors
(KNN), support vector machine (SVM) and extreme learning machine (ELM) are the few
machine learning techniques used for this purpose. 1. Introduction: The easy accessibility of cameras and the
tremendous growth of internet facilities make the automatic detection system a viable one. Due
to the cheap availability of gadgets, the automatic detection system is one of the less complex
tasks. The recent boom in the deep learning methods also adopted for the automatic process in
agriculture. Owing to the advancement in computer memory and hardware setup can lead to
extra solutions. These deep learning methods can be used to solve complex tasks in a reasonable
time. This paper proposes a convolutional capsule network for disease detection and
classification of grape leaves. It improves the detection accuracy and analyzes the
improvements of other deep learning models. The influence of capsule dimensions on the
detection accuracy is explored. By increasing convolution layers before the primary caps
layers, the proposed model provides accurate detection of diseases. The other parts of the paper is organized as follows: Section 2 elaborates the literature survey
on plant disease detection. Section 3 briefly introduces the convolutional capsule network. Section 4 details about the experimental implementation and result analysis. The final section
concludes the findings and limitations of the research work. 2. Motivation and related work: Timely diagnosis of plant diseases is the prime factor to control the plant loss. If it is performed
by human, then it needs lot of time, chances of getting error and costly. The automatic
equipment and methods are the prime solution to monitor the crop fields [Ampatzidis et al.,
2017]. Recent years, the researchers worked on the automatic techniques to detect and classify
the diseases in plants. This section elaborates the computer vision methods to detect the disease
in plant leaves. Qin F et al., [2016] investigated a leaf disease detection by supervised classification algorithm. Initially, k-median and fuzzy C-means clustering methods are used to obtain lesion images. After the lesion segmentation, texture, color and shape features are extracted. The optimal
features are extracted from a set of 129 features. The optimal features are trained by the
machine learning methods of logistic regression, linear discriminant analysis, navie bayes and
regression tree and their corresponding diseases are detected. The detection accuracy of the
method depends on the adopted features and it is consuming much time. Waghmare, H et al., (2016) investigated a system for grape plant leaf disease detection. Initially, the background of the leaf images are removed. The segmentation technique is applied
to segment the diseased part of the leaves. The fractal based texture features are extracted. The
extracted features are given into multiclass SVM. The proposed method classifies the grape
disease of downy mildew, black rot and powdery mildew. The classification accuracy of 89%
is obtained through MSVM. It varies, if the training and testing samples are increased. The size
of the dataset is less and the performance metric is varied in accordance with the size of the
dataset. The need for finding automatic disease detection methods are seriously evolving due to the
accomplishment of machine learning techniques. The traditional machine learning algorithms
manually extract the features from the images and feed those features into a detection
algorithm. It is a time consuming process and less robust. These manual feature extraction steps make the machine learning method less versatile. But, deep learning models consist of many
processing layers and fetch the features directly from the image. Recent advancement in artificial neural network architecture made a foundation of hybrid deep
learning models. Different techniques are adopted to build the deep learning based algorithm. 2. Motivation and related work: Auto encoders, sparse coding, restricted Boltzmann machines and customized convolution
neural networks (CNN) are some of the widely used architectures. Among these models CNN
is mostly used for computer vision applications [Guo Y et al., 2016]. It has outstanding success
in the agriculture field also. LeNet, AlexNet, ResNet, GoogleNet, Visual Geometry Group
(VGG) etc., are some of the examples of CNN based models. Few studies reported the use of
CNN based plant disease classification [Chen, J et al., 2019, Lu Y et al., 2017, Wang G et al.,
2017, Liu et al., 2017, Ma et al., 2018, Polder, G et al., 2019, Fuentes, A et al., 2018] and yielded
promising results. Table 1 highlights the plant disease classification methods for different
plants. Table 1 Deep learning based plant disease classification References
Species
Classes
Dataset
Architecture
Accuracy
Liu B et al.,
(2018)
Apple
4
collected
Modified
AlexNet
97.62
Amara, J., et al.,
(2017)
Banana
3
PlantVillage
Modified LeNet
92.88
Fuentes, A. et
al., (2017)
Tomato
9
Collected
R-FCN
85.98
Liu B et al.,
(2020)
Grape
7
Collected
Inspection
97.22
Zilvan, V et al.,
(2019)
Corn
5
PlantVillage
Autoencoder
87.09
Marino, S et al.,
(2019)
Potato
8
Collected
CNN
-
Ma et al., (2018)
cucumber
4
PlantVillage
DCNN
93.4%
Lu et al., (2017)
rice
-
collected
DCNN
95.48%
Huang, Z et al.,
(2020)
Grape
5
Collected
MobileNet and
improved
AlexNet
97% Few researchers have used CNN based models for feature extraction and the extracted features
are given as inputs into machine learning algorithms for further classification. Athiwara Kun,
B et al., (2015) extracted the features from CNN and fed the inputs into machine learning
approaches of SVM and random forest. The performance of the model is improved in
comparison with the general CNN method. Kerkech, M. et al., [2018] investigated disease
detection in grapes using CNN model. The color space and vegetation indices are combined
with CNN LeNet-5 to classify the diseases and obtain 95.8% of detection accuracy. This
method is limited by the number of expert labeled input. Gandhi, R et al., 2018 explored CNN for plant disease detection. The deep architectures of
Inception V3 and MobileNet are used to detect the plant diseases with accuracy of 88.3% and
92%. Initial, pre-processing steps are adopted to remove the noises. Deep convolutional
generative adversarial network (DCGAN) is used for augmentation steps in order to increase
the number of images. The model is deployed in mobile applications. Even though the images
are augmented, the detection accuracy of the model is less. Arnal Barbedo, J.G (2019)
presented plant disease identification for ten different diseases. The authors increased the data
samples of each disease category by gathering 60% of images under controlled conditions and
40% under real field conditions. Initially, the individual lesions and spots are segmented. After
the removal of backgrounds, the pre-trained GoogleNet CNN architecture is applied to classify
the diseases. The detection accuracies are varied in accordance with the plant species. The
usage of spot segmentation and background removal greatly impact on the detection accuracies. From the literature, it is understood that the deep learning models are effective to detect
diseases in the plant leaves. Table 1 Deep learning based plant disease classification CNNs are highly used networks which have complex layers and
are used to extract the features such as rotation, translation and scale [Sezer, A et al., 2019]. The prime requirement of any CNN based architecture is massive training data which is not
feasible for individual plant based image analysis. Another major shortcoming is loss of
information due to local details such as position and posture during image analysis. Hence,
CNNs are not a good choice for extracting diverse information such as orientation, scale and
semantic class which are required for image reconstruction. Recently, Hinton team [Sabour, S
et al., 2017] introduced a new deep learning architecture capsule network, which overcomes the disadvantages of traditional deep learning techniques. It is a novel building block that
represents spatial relationships of features effectively. It is mostly applied in image recognition
and classification tasks. Even though the capsule network is applied to the different research
fields such as tumor classification [Afshar, P et al, 2018], disease classification [Sezer, A et
al., 2019, Afshar, P et al., 2020, Verma, S et al., 2020] , drug detection [Wang, Y et al., 2020],
object detection [Kumar, A.D, 2019], hyper spectral image classification [Deng, F et al., 2018,
Yin, J et al., 2019, M. E. Paoletti et al., 2019], it is still in infancy. Actually, a capsule network
uses a set of neurons as a capsule. The capsules are vectors to denote internal properties that
can be utilized to learn the relationship between various features. So, the potential variant can
be effectively inferred by the model with lesser training inputs. Few studies only used the capsule network in plant disease analysis, because it is one of the
recent techniques in the field of computer vision. The limitations in the field of plant diseases
detection are listed below: reduction in trainable parameters to design an efficient network,
development of an efficient model to detect the plant diseases for a large dataset. These
limitations are addressed in the proposed method. The novelty of the work is as follows: the
proposed method combines CNN and the capsule network to obtain superior results in
comparison with the state-of-the-art methods. The convolutional layers are added before the
primary caps layer which indirectly decrease the number of capsules and speed up the dynamic
routing process. Table 1 Deep learning based plant disease classification The influence factors of the proposed model are dimension of capsules and routing number of
training phases which are analyzed towards classification results. It can give valuable guidance
for subsequent research on plant disease classification using a capsule network. 3.1 Dataset In this study, the grape leaves dataset is collected from the farms which are cultivating grape
plants. The images are captured in the grape fields that are located in Madampatti, Coimbatore
district, Tamil Nadu, India. In addition with this, the images from plant village dataset [Hughes
DP et al., 2015] also used for the proposed experiment. The healthy grape leaf images and
diseased leaf images are taken. The considered disease categories are grape black rot, grape
esca and grape leaf blight. Figure 2 shows the images under each disease category. Figure 2: Images of grape leaf diseases classes (a) black rot (b) black esca (c) leaf blight
(d) No disease Figure 2: Images of grape leaf diseases classes (a) black rot (b) black esca (c) leaf blight
(d) No disease 3. Materials and Methods The complete process of proposed grape leaf disease detection using convolutional capsule
networks is highlighted in figure 1. The key steps which are involved in the process are image
collection, image augmentation, dataset partitioning, convolutional capsule network, diseased
classification and performance analysis. Figure 1: Flow chart of proposed grape leaf disease detection Figure 1: Flow chart of proposed grape leaf disease detection 3.2 Capsule Networks Capsule network has a deep learning architecture. It handles affine transformation well [Deng,
F et al., 2018]. The capsules of the capsule network are designed to represent the output in
terms of vectors. The capsules consist of a set of neurons that can learn the entities of an image
such as pose, size and orientation. Dynamic routing mechanism is followed to route the data
from one layer to another layer. The lower layer capsules predict the response of higher level
capsules. The higher level capsules are triggered only when the lower level predictions agree. Consider ‘i’ is the lower level capsule and its output is 𝜇𝑖. The prediction of higher level capsule
‘j’ is given by 𝜇̃𝑗|𝑖= 𝑊𝑖𝑗𝜇𝑖 (1) (1) 𝜇̃𝑗|𝑖= 𝑊𝑖𝑗𝜇𝑖 Where 𝑊𝑖𝑗 is the weight matrix. It is learnt by back propagation technique. Each and every
capsule try to predict the response of the higher level capsules. If the prediction of the capsule
imitates the actual response of higher level capsules, then the coupling coefficients between
the capsules will increase. The coupling coefficient can be calculated by following equation: 𝐶𝑖𝑗=
𝑒𝑏𝑖𝑗
∑𝑒𝑏𝑖𝑗
𝑘
(2) (2) Where 𝑏𝑖𝑗 is the log prior probability. It is given as zero initially which highlights whether
lower level capsule ‘i’ is coupled with higher level capsule ‘j’. By using equation 3 the input
vector of higher level capsule j is calculated. 𝑆𝑗= ∑𝐶𝑖𝑗
𝑖
𝜇̃𝑗|𝑖 (3) 𝑆𝑗= ∑𝐶𝑖𝑗
𝑖
𝜇̃𝑗|𝑖 (3) (3) The length of the output vector shows the probability of existence. The nonlinear squash
function is used as ant activation function. It shrinks long vectors close to one and chokes short
vectors to almost zero. It is given by, The length of the output vector shows the probability of existence. The nonlinear squash
function is used as ant activation function. It shrinks long vectors close to one and chokes short
vectors to almost zero. It is given by, 𝑉𝑗= 𝑆𝑞𝑢𝑎𝑠ℎ (𝑆𝑗) =
‖𝑆𝑗‖
2
1 + ‖𝑆𝑗‖
2
𝑆𝑗
‖𝑆𝑗‖ (4) (4) The input and output vector of jth capsule is represented by 𝑆𝑗 and 𝑉𝑗. The equation 1 to 4 is
used for the routing procedure to find 𝑉𝑗. The routing number can decide the number of
iterations. The output length of lower capsules encodes the existence probability of their
entities [Zhang, W et al., 2019]. 3.2 Capsule Networks The vector directions encode different properties of the entities
such as orientation, size and posture. So, the capsules learn the spatial relationship between
entities within the input image. The margin loss is used to detect if the entities of the particular
class are present and can be obtained by equation 5. ‘k’ is the last layer of the capsule and 𝑙𝑘
is the loss function. 𝑙𝑘= 𝑇𝑘max (0, 𝑚+ −‖𝑉𝑘‖)2 + 𝛽(1 −𝑇𝑘) max(0, ‖𝑉𝑘‖ −𝑚−)2 (5) (5) If 𝑇𝑘 is one then ‘k’ is present. At the initial training process, the parameters of 𝑚+, 𝑚− and 𝛽
are indicated. The sum of the losses of all output capsules of the output layer is known as total
loss. 3.3 Convolutional Capsule Network The proposed method uses convolutional capsule networks for the disease classification of
grape leaves. The architecture of the convolutional capsule network of the proposed method is
shown in figure 3. The input image of size 128 × 28 is used for the classification. The
architecture consists of 5 convolution layers, 1 primary layer, 1 digit layer and 3 fully connected
layers. In order to generate an effective feature map, more convolutional layers are added. Initial layer consists of a 5×5 size of 16 convolutional kernels with 1 stride. 2×2 max-pooling
with 2 strides is performed after the first layer. The second layer consists of a 5×5 size of 32
convolutional kernels with 1 stride. The third layer consists of a 5×5 size of 64 convolutional
kernels with 1 stride. 2×2 max-pooling with 2 strides is performed after the second and third
layer. The fourth layer consists of a 9×9 size of 128 convolutional kernels with 1 stride. Primary
capsule layer is the fifth layer which consists of 32 capsules. Each capsule is applied with a
9×9 size of convolutional kernel with stride 1. The digit caps layer consists of 16 dimensional
capsules with four classes. Figure 3: The convolutional capsule network for grape disease classification
3.3.1 Pre-processing and data augmentation Figure 3: The convolutional capsule network for grape disease classification Figure 3: The convolutional capsule network for grape disease classification 3.3.1 Pre-processing and data augmentation The input size of the proposed model is 128 × 128. The size of the images in the input dataset
are not the same. So, all the images are resized into 128 × 128. Inadequate inputs cause
overfitting problems in the deep neural network. In order to avoid overfitting problems, the input images are augmented. Rotation, scaling transformation, gamma correction, flipping and
color augmentation techniques are used to generate the augmented dataset. The augmented
images are randomly shuffled and grouped into training, testing and validation dataset. Figure
4 shows the example of augmented images. Table 2 highlights the number of grape leaf images
under each category. Table 2: Details of image augmentation
Types
Number of images
Without data
augmentation
With data augmentation
Grape black rot
2800
7000
Grape black esca
2500
7000
Grape leaf blight
2000
6000
No disease
4000
8000
Total
11300
28000
Figure 4: Data augmentation results (A) input image (B)-(H) augmented images
2 Optimization of hyper parameter Table 2: Details of image augmentation
Types
Number of images
Without data
augmentation
With data augmentation
Grape black rot
2800
7000
Grape black esca
2500
7000
Grape leaf blight
2000
6000
No disease
4000
8000
Total
11300
28000 Table 2: Details of image augmentation
Types
Number of images
Without data
augmentation
With data augmentation
Grape black rot
2800
7000
Grape black esca
2500
7000
Grape leaf blight
2000
6000
No disease
4000
8000
Total
11300
28000 Table 2: Details of image augmentation Figure 4: Data augmentation results (A) input image (B)-(H) augmented images Figure 4: Data augmentation results (A) input image (B)-(H) augmented images Figure 4: Data augmentation results (A) input image (B)-(H) augmented images
.3.2 Optimization of hyper-parameter 3.3.2 Optimization of hyper-parameter The hyper-parameter optimization of convolutional capsule networks depends on the number
of kernels in the convolutional layer, dimensions of primary caps and digit caps layers. In
addition, routing number is one of the vital parameters to find coupling coefficients. It is tested
from 1 to 3 with an increment of 1. The performance of the architecture is tested for each
parameter setting. Based on the training dataset ten-fold cross validation is adopted. After
achieving highest accuracy, the best set of parameters are chosen and applied for testing the
dataset. In order to reduce overfitting, early stopping is adopted. Whenever the error in the validation set is less than the previous iteration, the training is stopped immediately and the
optimal hyper-parameter is derived. Table 3: Hyper-parameter of convolutional capsule network
3.3.3 Training of Convolutional Capsule Network
Hyper-parameter
setting
Activation
ReLU
Maximum Epoch
150
Optimizer
Adam
Batch size
50
Routing Number
1-3
Learning rate
0.0001
Number of kernels in Conv1 layer
16
Number of kernels in Conv2 layer
32
Number of kernels in Conv3 layer
64
Number of kernels in Conv4 layer
128
Number of kernels in Primary Caps layer
256 Table 3: Hyper-parameter of convolutional capsule network During the training process, the weights of convolutional neural networks are randomly
initialized using normal distribution with 0.01 standard deviation. Rectified linear unit
activation function is used in all layers of the network. Batch normalization is used to reduce
the internal covariate shift. The input of each layer is normalized to the standard Gaussian
distribution [Ioffe, S et al., 2015]. The Adam optimizer of stochastic gradient descent algorithm
is utilized as the optimizer. This algorithm learns sparse data and fine tunes the learning rate
for every parameter setting. In comparison with non-adaptive methods, it has a high
convergence rate. Let c1,c2,c3,c4,p5 are the parameters of four convolutional layers and the primary caps layer. The convolutional operations are represented with conv. The output from the convolutional
layers and primary caps are denoted by O_conv1, O_conv2, O_conv3, O_conv4 and O_prim. Feature extraction process is achieved through convolutional layers and primary caps layer. The dynamic routing is used to generate digit capsules. The stochastic gradient is used to find
the network parameters. The detail algorithm for training convolutional capsule network is
given below: Algorithm Input:
Feature vector (F)
No of training epoch (E)
No of dynamic routing iteration (R)
Output:
Capsule length (L)
for n=1 to E do
O_conv1 ← conv (F,c1); O_conv2 ← conv (c1,c2); O_conv3 ← conv (c2,c3);
O_conv4 ← conv (c3,c4); O_prim ← conv (c1,p5);
µ←encapule(O_prim)
for capsule ‘i’ in primary caps layer
compute 𝜇̃𝑗|𝑖←𝑊𝑖𝑗𝜇𝑖 (Equation 1) for m=1 to R do for capsule ‘i’ in primary caps layer for capsule ‘i’ in primary caps layer for capsule ‘i’ in primary caps layer and capsule ‘j’ in digit caps layer (Equation 5) (Equation 5) Input: Feature vector (F) for capsule ‘i’ in primary caps layer and capsule ‘j’ in digit caps layer 𝑏𝑖𝑗←0 𝑏𝑖𝑗←0 for m=1 to R do
for capsule ‘i’ in primary caps layer
𝑐𝑖←softmax(𝑏𝑖) (Equation 2)
for capsule ‘j’ in digit caps layer
compute 𝑆𝑗←∑𝐶𝑖𝑗
𝑖
𝜇̃𝑗|𝑖 (Equation 3)
for capsule ‘j’ in digit caps layer
𝑉𝑗←𝑆𝑞𝑢𝑎𝑠ℎ (𝑆𝑗) (Equation 4)
for capsule ‘i’ in primary caps layer and capsule ‘j’ in digit caps layer 3.4 Performance Assessment K fold is the efficient cross-validation method which is used for larger dataset. 10-fold cross
validation is used to evaluate the performance of the proposed method. The dataset is randomly
separated into ten subset. Of the 10 subsets, 9 subset is used for training and the remaining
dataset is used for testing the network. The cross validation procedure is repeated 10 times. With each of the ten subsets used exactly once as the validation set. The performance of the
method is obtained by taking an average of all ten runs. With each of the ten subsets used exactly once as the validation set. The performance of the
method is obtained by taking an average of all ten runs. The performance of the proposed method is assessed through precision, recall, f-measure and
accuracy. Precision is used to quantify the number of positive observations to the total
observation. Recall is used to quantify the number of positive observations made out of all
positive samples. F1 measure or F1 score is the weighted average of precision and recall which
replicate the total number of observations that are correctly classified by the network. The
equations of the performance metrics are given below, 𝑃𝑟𝑒𝑐𝑖𝑠𝑖𝑜𝑛 (𝑃) =
𝑇𝑃
𝑇𝑃+ 𝐹𝑃 (6)
𝑅𝑒𝑐𝑎𝑙𝑙 (𝑅) =
𝑇𝑃
𝑇𝑃+ 𝐹𝑁 (7) 𝑃𝑟𝑒𝑐𝑖𝑠𝑖𝑜𝑛 (𝑃) =
𝑇𝑃
𝑇𝑃+ 𝐹𝑃 (6)
𝑅𝑒𝑐𝑎𝑙𝑙 (𝑅) =
𝑇𝑃
𝑇𝑃+ 𝐹𝑁 (7)
𝐹1 𝑀𝑒𝑎𝑠𝑢𝑟𝑒 (𝐹1) =
2𝑇𝑃
2𝑇𝑃+ 𝐹𝑃+ 𝐹𝑁 (8)
𝐴𝑐𝑐𝑢𝑟𝑎𝑐𝑦 (𝐴) =
𝐹𝑃+ 𝐹𝑁
𝑇𝑃+ 𝑇𝑁+ 𝐹𝑃+ 𝐹𝑁 (9)
4. Experimental results and discussion (6) (7) 𝐹1 𝑀𝑒𝑎𝑠𝑢𝑟𝑒 (𝐹1) =
2𝑇𝑃
2𝑇𝑃+ 𝐹𝑃+ 𝐹𝑁 (8) (8) 𝐴𝑐𝑐𝑢𝑟𝑎𝑐𝑦 (𝐴) =
𝐹𝑃+ 𝐹𝑁
𝑇𝑃+ 𝑇𝑁+ 𝐹𝑃+ 𝐹𝑁 (9) (9) 4. Experimental results and discussion The proposed method for convolutional capsule network based grape leaves disease detection
and classification is implemented in the system having 2.90 GHZ processor with 12 GB RAM
and 4 GB GPU card NVIDIA GTX 1050. All the calculations are carried out in the aforesaid
system. Epochs are varied in the range of 30-150 with respect to the testing accuracy. The
learning rate is varied from 0.01to 0.0001 Different batch sizes are tried for the network in
order to fit in the GPU. The batch size varies from 5 to 50. In order to cater the computer
memory, it is fixed as 50. Different volumes of training parameters are considered with respect
to the dataset in the training process. 85% of the dataset is used for training and 15% of dataset
is used for testing. The optimal set of hyper-parameters are listed in table 3. The detailed
quantitative and qualitative analysis is carried out for augmented and non-augmented datasets. 4.1Performance Analysis All the images in the dataset are resized into 128 × 128. The collection of image dataset for
diseased grape leaves is a time consuming process. The image augmentation steps are used to
provide sufficient images for training and testing of the model. The proposed convolutional
capsule network for grape disease detection and classification is experimented for two datasets
(with and without augmentation). It is trained with optimal hyper-parameters. The performance of the model is evaluated for 10 fold cross validation. In the training process, epochs play vital
role in order to check the performance of the model. The proposed method has been tried with
different epochs which are starting from 30. The model has converged after 150 epochs. Hence,
the training accuracy and loss of the model is obtained up-to 150 epochs. Table 4: 10-fold cross validation results of augmented dataset
Folds
P %
R %
F1 %
A %
Fold 1
99.23
97.21
97.43
97.23
Fold 2
97.12
99.17
98.25
98.21
Fold 3
98.37
99.20
99.27
99.25
Fold 4
98.44
99.92
99.72
99.81
Fold 5
98.53
99.98
99.81
99.66
Fold 6
98.45
99.23
99.92
99.86
Fold 7
99.11
99.23
98.63
98.65
Fold 8
98.23
99.21
99.16
99.12
Fold 9
98.45
98.36
99.42
99.34
Fold 10
98.13
99.13
99.83
99.76
Mean
98.41
99.06
99.10
99.12 The performance metrics of the proposed method for augmented dataset and non-augmented
dataset are highlighted in table 4 and 5. The detailed results of 10-fold cross validation is
shown. The experimental results highlight that the classification accuracy is high for
augmented dataset. 99.12% of classification accuracy is obtained by convolutional capsule
network for augmented dataset. Recall is one of the important performance metrics which is
used to minimize the false negatives. In the proposed method, the recall value is as high as
99.06%. Adding convolutional layers in the capsule network and the parameters which are
employed in the network results in accurate detection of grape leaf diseases. Table 5: 10-fold cross validation results of non-augmented dataset
Folds
P %
R %
F1 %
A %
Fold 1
98.13
94.10
94.92
95.11
Fold 2
93.12
85.31
89.04
90.10
Fold 3
83.84
83.23
83.67
88.45 Table 5: 10-fold cross validation results of non-augmented dataset Fold 4
91.32
90.10
89.75
89.94
Fold 5
95.13
87.71
91.01
91.12
Fold 6
98.11
93.13
95.52
95.72
Fold 7
82.98
87.92
84.23
85.61
Fold 8
97.01
92.33
94.51
94.61
Fold 9
83.46
83.22
83.13
88.61
Fold 10
87.89
86.96
85.96
90.32
Mean
91.01
88.51
89.23
92.13
Figure 5: Training and loss graphs of non-augmented dataset (left) and augmented
dataset (right) Fold 4
91.32
90.10
89.75
89.94
Fold 5
95.13
87.71
91.01
91.12
Fold 6
98.11
93.13
95.52
95.72
Fold 7
82.98
87.92
84.23
85.61
Fold 8
97.01
92.33
94.51
94.61
Fold 9
83.46
83.22
83.13
88.61
Fold 10
87.89
86.96
85.96
90.32
Mean
91.01
88.51
89.23
92.13 Figure 5: Training and loss graphs of non-augmented dataset (left) and augmented
dataset (right) Figure 5: Training and loss graphs of non-augmented dataset (left) and augmented
dataset (right) Figure 5 shows the graphs of performance metrics under training and validation phases. The
epochs play a paramount role in the learning process of deep learning models. The proposed
model of convolutional capsule network for grape leaves disease classification achieved
99.12% of validation accuracy and 0.091loss. It is for augmented dataset. For non-augmented
dataset the validation accuracy is 92.13% and loss is 0.132. Around the 120th epoch the high
accuracy is obtained but it is not consistent. It converged after 150 epochs. The proposed model
is trained for augmented dataset and non-augmented dataset in which the augmented datasets
give higher performances. 4.2.1 Capsule Dimensions The capsules are an important component of the convolutional capsule network. It consists of
many neurons. The capsules in the primary caps layer are the lower level capsules. It learns
from convolutions layers and it can represent small entities of the input image. The capsules in
the digit caps layers are higher dimensions which represent the complex entities of the input. Hence, the dimensions of the capsule play a major role in the classification tasks. If the
dimensions of the capsule are low then the representation ability of the capsule is less. In
addition, if the dimensions of the capsule is high then the redundant information will occur. Both cases provide negative results in the classification. The set of values for capsule
dimensions [(4,8), (6,12), (8,16), (10,20)] are used to evaluate the performance of the model. The other parameters are given as per table 3. The effect of capsule dimensions over two
datasets are given in figure 6. D1 is an augmented dataset. D2 is a non-augmented dataset. The
classification accuracy is high for the capsule dimensions of (8,16). The performance of the
model gradually increases, as the dimension increases. When the dimensionality is too high
then it causes dimensional disaster and results in performance degradation. Figure 6: Effect of capsule dimensions Figure 6: Effect of capsule dimensions 4.2.1 Routing Number The routing number is the prime factor to detect whether a capsule network can obtain accurate
coupling coefficients. Hence, it is necessary to determine the optimal routing number. In this
experiment, the routing number is checked for 1,2 and 3. The network used other parameters
which are listed in table 3. Figure 7 shows the effect of routing number with respect to the
accuracy. By increasing the routing number the accuracy is increased initially. Afterwards it decreased. If the routing number is high, then the training time of the network is also high. Hence, the optimal routing number 2 is chosen in this experiment. Figure 7: Effect of routing number Figure 7: Effect of routing number 4.3 Time complexity In the proposed work, a convolutional capsule network is used to detect the grape leaves
diseases. The network consists of convolutional layers before the primary caps layer. It
decreases the number of primary capsules and speeds up the dynamic routing process. As the
network uses dynamic routing over multiple iterations, the number of operations is higher, but
it also makes the entire network converge quickly. Hence, the design complexity of the network
also reduced. The time complexity is based on the required number of multiplications in the
capsule layer and fully connected layers. The lower layer capsule ‘i’ having the dimensions of
d1 and the higher layer capsule ‘j’ having dimensions of d2. By considering ‘k’ capsules in
lower layers and ‘m’ capsules in the higher layer, the total number of operations is m×d2×k×d1. Each prediction involves d1 × d2 multiplications. According to equation 3 each capsule is
calculated as a weighted average over predictions. d2 multiplications involved to weight each
prediction (𝜇̃𝑗|𝑖)with coupling coefficients (𝐶𝑖𝑗). Each routing process includes d2×k ×m
multiplications. The capsule networks have less number of layers in comparison with
convolutional neural networks. The proposed method took 1.82 seconds to process each input
in the NVIDIA GTX 1050 computer. 4.4 Comparative Analysis The performance of the proposed method is compared with five state-of-the-art deep learning
methods. All the methods have adopted a deep learning network for grape leaves disease
detection. Recent research works which are undergone for grape leaves disease detection methods only selected for the comparison. Table 6 shows the comparative deep learning
methods. CNN based architectures are used mostly to detect the diseases in grape leaves. Most Kerkech,
M. et al., [2018] investigated disease detection using CNN model. The color space and
vegetation indices are combined with CNN LeNet-5 to classify the diseases and obtained
95.8% of detection accuracy. The real time datasets are used in this method. Xie, X et al.,
[2020] experimented the disease detection of grape leaves by faster R-CNN model. The pre-
trained models of Inception v1, Inception ResNet v2 blocks are used to derive faster DR-
IACNN model. The implementation results highlight that faster IACNN method produces
81.1% of average precision. The four common diseases of grape leaves are classified. The real
time grape leaf disease detection dataset (GLDD) is used. The improved convolution neural
network (DICNN) is used by Liu B et al., [2020] to detect the common diseases of grape leaves. 97.22 % of detection accuracy is obtained by DICNN model. It is higher than the GoogleNet
and ResNet-34 model. Rao U.S et al (2021) explored disease detection for mango leaves and
grape leaves. The pre-trained model with AlexNet is used to detect the diseases. The plant
village datasets and real time datasets are used. 99.03% of detection accuracy is obtained. Huang et al (2020) explored transfer learning based disease detection for grape leaves. The real
time self-acquired dataset is used for the study. The pre-trained models of VGG16, MobileNet
and AlexNet was used. MobileNet and improved AlexNet produced 97% of classification
accuracy. Ji M et al (2020) used multiple CNN to extract the complementary features from the
input. Inception V3 and ResNet50 based CNN architecture is used to detect the diseases. 98.57% testing accuracy is obtained by the model. The grapevine yellow disease is analysed
through various CNN models [Cruz, A et al., 2019] in which ResNet-50 produced better
performance. Table 6: Comparative analysis with state-of-the-art methods
References
Algorithm
Selected plant
Datasets
Performance
metric
Liu B et al.,
[2020]
DICNN
Grape
Real
time
images
Accuracy
97.22%
Kerkech, M. 4.4 Comparative Analysis et
al., [2018]
CNN-LetNet
Grape
Real
time
images
Accuracy 95.8%
Xie, X et al.,
[2020]
DR-IACNN
Grape
Real
time
images
Precision:
81.1%
Rao U.S et al
(2021)
Transfer
Learning
with
AlexNet
Grape
Plant
village+real
time dataset
Accuracy
99.03% Huang, Z et al.,
(2020)
MobileNet and
improved
AlexNet
Grape
real time dataset Accuracy
97%
Ji
M
et
al.,(2020)
Multiple CNN
Grape
Plant Village
98.57%
Proposed
Convolutional
Capsule
Network
Grape
Plant
village+real
time dataset
Accuracy
99.12% 4.5 Discussions The automated systems with deep learning techniques produce reliable results in most of the
pattern recognition problems. Among these deep learning techniques, CNN is largely used by
the researchers for image classification tasks. Even though CNN is used extensively, there are
few disadvantages. It fails to set relationships among spatial relationships such as posture, size,
orientation etc. The prime reason is that the input is processed through subsampling and max
pooling techniques. The first capsule networks were proposed by Hinton (Sabour et al., 2017). This network
received attention from the researchers because of its performance. The remarkable
performance of the network is based on the dynamic routing algorithm. The capsule is a set of
neurons that are in the form of a vector. These vectors represent the input information in terms
of spatial orientation, magnitude etc., The capsules are routed via dynamic routing from one
layer to another layer. It will catch and hold many fine information than conventional deep
learning models. Nevertheless, the small changes in the input vector components are often
neglected by deep learning techniques. The capsule networks adopt vector neurons to have a
better performance than the other deep neural models. The proposed method combines CNN and the capsule network to obtain superior results in
comparison with the state-of-the-art methods. The novelty of the proposed work depends on
the additional convolutional layers which are added before the primary caps layer. These
layers indirectly decrease the number of capsules and speed up the dynamic routing process. The influence factors of the proposed model are dimension of capsules and routing number of
training phases which are analyzed towards classification results. It can give valuable guidance
for subsequent research on plant disease classification using a capsule network. The major advantages of proposed method are listed below: In this work, the convolutional capsule network is used to detect the diseases of grape leaves. Unlike CNN, the diseases from grape leaves are detected from a small number of layers. Four
convolutional layers and primary capsule layers are used. The CNN model several
convolutional layers are used to detect the diseases of the same task [Geetharamani, G et al.,
2019, Ghoury, S et al., 2019]. Less number of layers make the network to be less complex. In
order to increase the reliability of the automatic system, a large number of images are used for
comparison with previous studies. 4.5 Discussions By decreasing the size of the input image, the information
in the image is lost. In the proposed method, even though the input image is given with the size
of 128 x 128, reliable classification is achieved. The CNN based models used the input images
with size higher than 128 x 128 [Xie et al.,2020]. The major limitations of the research work are listed below: When processing large images,
the convolutional capsule network requires a lot of hardware resources. There is no standard
rule for finding the number of primary caps. It is discovered artificially by the convolution
mode of the convolution layer. Ethical approval This article does not contain any studies with human participants or animals performed by any of the
authors. 5. Conclusions: The automatic plant disease detection system is a universal detector which detects the
abnormalities caused by fungal and bacterial deformities. In this work, a convolutional capsule
network is used to detect the diseases on grape leaves. CNN is immensely used in several
agricultural applications. The major shortcomings of CNNs are addressed by the capsule
network. In the proposed convolutional capsule network, convolutional layers are added before
the primary caps layer which indirectly decrease the number of capsules and speed up the
dynamic routing process. The influence factors of the proposed model are dimension of
capsules and routing number of training phases which are analyzed towards classification
results. The experimental results are evaluated under augmented and non-augmented datasets. The proposed model yields 99.12% of classification accuracy. The performance of the
proposed method is analyzed through the performance metrics of precision, recall, f1 score and
accuracy. In future, the research work can be extended by adding additional disease classes of grape
leaves. The above work can be extended by developing an automatic drone which can compute
the overall health of fields and take immediate actions. It will empower the farmers and much
useful to decrease the losses of fields. Data Availability The datasets generated and analysed during the current study are available from the corresponding
author on reasonable request. Conflict of interest Authors declare that they have not conflict of interest. Funding This study was not funded by any other organization. References: Gebbers,
R.,
&
Adamchuk,
V. I. (2010). Precision
agriculture
and
food
security. Science, 327(5967), 828-831. Savary, S., Ficke, A., Aubertot, J. N., & Hollier, C. (2012). Crop losses due to diseases and their
implications for global food production losses and food security. Food security, 4(4), 519-537. Andrushia, A. D., & Patricia, A. T. (2020). Artificial bee colony optimization (ABC) for grape
leaves disease detection. Evolving Systems, 11(1), 105-117. Andrushia, A. D., & Patricia, A. T. (2020). Artificial bee colony optimization (ABC) for grape
leaves disease detection. Evolving Systems, 11(1), 105-117. Andrushia, A. D., & Patricia, A. T. (2019). Artificial bee colony based feature selection for
automatic skin disease identification of mango fruit. In Nature Inspired Optimization Techniques
for Image Processing Applications (pp. 215-233). Springer, Cham Krishnaswamy Rangarajan, A., & Purushothaman, R. (2020). Disease classification in eggplant
using pre-trained VGG16 and MSVM. Scientific reports, 10(1), 1-11. Arsenovic, M., Karanovic, M., Sladojevic, S., Anderla, A., & Stefanovic, D. (2019). Solving current
limitations of deep learning based approaches for plant disease detection. Symmetry, 11(7), 939. Ampatzidis, Y., De Bellis, L., Luvisi, A., (2017). iPathology: robotic applications and management
of plants and plant diseases. Sustainability 9 (6), 1010. https://doi.org/ 10.3390/su9061010 Guo Y, Liu Y, Oerlemans A, Lao S, Wu S, Lew SL (2016) Deep learning for visual understanding:
a review. Neurocomputing 187:27 Polder, G., Blok, P. M., De Villiers, H. A., Van der Wolf, J. M., & Kamp, J. (2019). Potato virus
detection in seed potatoes using deep learning on hyperspectral images. Frontiers in plant
science, 10, 209. Fuentes, A. F., Yoon, S., Lee, J., and Park, D. S. (2018). High-performance deep neural network-
based tomato plant diseases and pests diagnosis system with refinement filter bank. Front. Plant
Sci. 9:1162. doi: 10.3389/fpls.2018.01162 Fuentes, A. F., Yoon, S., Lee, J., and Park, D. S. (2018). High-performance deep neural network-
based tomato plant diseases and pests diagnosis system with refinement filter bank. Front. Plant
Sci. 9:1162. doi: 10.3389/fpls.2018.01162 Fuentes, A. F., Yoon, S., Lee, J., and Park, D. S. (2018). High-performance deep neural network-
based tomato plant diseases and pests diagnosis system with refinement filter bank. Front. Plant
Sci. 9:1162. doi: 10.3389/fpls.2018.01162 Waghmare, H., Kokare, R., & Dandawate, Y. (2016 February). Detection and classification of
diseases of grape plant using opposite colour local binary pattern feature and machine learning for
automated decision support system. References: In 2016 3rd international conference on signal processing and
integrated networks (SPIN) (pp. 513-518). IEEE Waghmare, H., Kokare, R., & Dandawate, Y. (2016 February). Detection and classification of
diseases of grape plant using opposite colour local binary pattern feature and machine learning for
automated decision support system. In 2016 3rd international conference on signal processing and
integrated networks (SPIN) (pp. 513-518). IEEE Qin F, Liu D, Sun B, Ruan L, Ma Z, Wang H (2016) Identification of Alfalfa leaf diseases using
image recognition technology. PLoS ONE 11(12):e0168274 Chen, J., Liu, Q., & Gao, L. (2019). Visual tea leaf disease recognition using a convolutional neural
network model. Symmetry, 11(3), 343. Afshar, P., Heidarian, S., Naderkhani, F., Oikonomou, A., Plataniotis, K. N., & Mohammadi, A. (2020). Covid-caps: A capsule network-based framework for identification of covid-19 cases from
x-ray images. Pattern Recognition Letters, 138, 638-643. Lu, Y., Yi, S., Zeng, N., Liu, Y., & Zhang, Y. (2017). Identification of rice diseases using deep
convolutional neural networks. Neurocomputing, 267, 378-384. Wang, G., Sun, Y., & Wang, J. (2017). Automatic image-based plant disease severity estimation
using deep learning. Computational intelligence and neuroscience, 2017. Hughes DP, Salathé M (2015) An open access repository of images on plant health to enable the
development of mobile disease diagnostics. CoRR abs/1511.08060. Liu, B., Zhang, Y., He, D. & Li, Y. (2017) Identification of apple leaf diseases based on deep
convolutional neural networks, Symmetry. https://doi.org/10.3390/sym10010011 Ma, J., Du, K., Zheng, F., Zhang, L., Gong, Z., & Sun, Z. (2018). A recognition method for
cucumber diseases using leaf symptom images based on deep convolutional neural
network. Computers and electronics in agriculture, 154, 18-24. Arnal Barbedo, J.G. (2019) Plant disease identification from individual lesions and spots using deep
learning. Biosyst. Eng. 2180, 96–107. Athiwaratkun, B., & Kang, K. (2015). Feature representation in convolutional neural
networks. arXiv preprint arXiv:1507.02313. Gandhi, R., Nimbalkar, S., Yelamanchili, N., & Ponkshe, S. (2018, May). Plant disease detection
using CNNs and GANs as an augmentative approach. In 2018 IEEE International Conference on
Innovative Research and Development (ICIRD) (pp. 1-5). IEEE. Kerkech, M., Hafiane, A., & Canals, R. (2018). Deep leaning approach with colorimetric spaces
and vegetation indices for vine diseases detection in UAV images. Computers and electronics in
agriculture, 155, 237-243. Sabour, S., Frosst, N., & Hinton, G. E. (2017). Dynamic routing between capsules. Advances in
neural information processing systems, 30. Afshar, P., Mohammadi, A., & Plataniotis, K. References: N. (2018, October). Brain tumor type classification
via capsule networks. In 2018 25th IEEE international conference on image processing (ICIP) (pp. 3129-3133). IEEE. Deng, F., Pu, S., Chen, X., Shi, Y., Yuan, T., & Pu, S. (2018). Hyperspectral image classification
with capsule network using limited training samples. Sensors, 18(9), 3153. Yin, J., Li, S., Zhu, H., & Luo, X. (2019). Hyperspectral image classification using CapsNet with
well-initialized shallow layers. IEEE Geoscience and Remote Sensing Letters, 16(7), 1095-1099. Paoletti, M. E., Haut, J. M., Fernandez-Beltran, R., Plaza, J., Plaza, A., Li, J., & Pla, F. (2018). Capsule networks for hyperspectral image classification. IEEE Transactions on Geoscience and
Remote Sensing, 57(4), 2145-2160. Wang, Y., Huang, L., Jiang, S., Wang, Y., Zou, J., Fu, H., & Yang, S. (2020). Capsule networks
showed excellent performance in the classification of hERG blockers/non blockers. Frontiers in
pharmacology, 10, 1631. Kumar, A. D. (2018). Novel deep learning model for traffic sign detection using capsule
networks. arXiv preprint arXiv:1805.04424. Sezer, A., & Sezer, H. B. (2019). Capsule network-based classification of rotator cuff pathologies
from MRI. Computers & Electrical Engineering, 80, 106480. Zhang, W., Tang, P., & Zhao, L. (2019). Remote sensing image scene classification using CNN-
CapsNet. Remote Sensing, 11(5), 494. Ioffe, S., and Szegedy, C. (2015). Batch normalization: accelerating deep network training by
reducing internal covariate shift. arXiv preprint arXiv:1502.03167. doi: arXiv:1502.03167v3 Xie, X., Ma, Y., Liu, B., He, J., Li, S., & Wang, H. (2020). A deep-learning-based real-time detector
for grape leaf diseases using improved convolutional neural networks. Frontiers in plant
science, 11, 751. Liu, B., Ding, Z., Tian, L., He, D., Li, S., & Wang, H. (2020). Grape leaf disease identification
using improved deep convolutional neural networks. Frontiers in Plant Science, 11, 1082. Rao, U. S., Swathi, R., Sanjana, V., Arpitha, L., Chandrasekhar, K., & Naik, P. K. (2021). Deep Rao, U. S., Swathi, R., Sanjana, V., Arpitha, L., Chandrasekhar, K., & Naik, P. K. (2021). Deep
Learning Precision Farming: Grapes and Mango Leaf Disease Detection by Transfer
Learning. Global Transitions Proceedings, 2(2), 535-544. Learning Precision Farming: Grapes and Mango Leaf Disease Detection by Transfer
Learning. Global Transitions Proceedings, 2(2), 535-544. Huang, Z., Qin, A., Lu, J., Menon, A., & Gao, J. (2020, November). Grape Leaf Disease Detection
and Classification Using Machine Learning. References: In 2020 International Conferences on Internet of
Things (iThings) and IEEE Green Computing and Communications (GreenCom) and IEEE Cyber,
Physical and Social Computing (CPSCom) and IEEE Smart Data (SmartData) and IEEE Congress
on Cybermatics (Cybermatics) (pp. 870-877). IEEE. Ji, M., Zhang, L., & Wu, Q. (2020). Automatic grape leaf diseases identification via UnitedModel
based on multiple convolutional neural networks. Information Processing in Agriculture, 7(3), 418-
426. Ghoury, S., Sungur, C., & Durdu, A. (2019). Real-time diseases detection of grape and grape leaves
using faster r-cnn and ssd mobilenet architectures. In International Conference on Advanced
Technologies, Computer Engineering and Science (ICATCES 2019). Geetharamani G. Arun Pandian J (2019) Identification of plant leaf diseases using a nine-layer
deep convolutional neural network Computers and Electrical Engineering 76, 323–338 Verma, S., Chug, A., & Singh, A. P. (2020). Exploring capsule networks for disease classification
in plants. Journal of Statistics and Management Systems, 23(2), 307-315. Cruz, A., Ampatzidis, Y., Pierro, R., Materazzi, A., Panattoni, A., De Bellis, L., & Luvisi, A. (2019). Detection of grapevine yellows symptoms in Vitis vinifera L. with artificial
intelligence. Computers and electronics in agriculture, 157, 63-76. Marino, S., Beauseroy, P. and Smolarz, A., 2019, February. Deep Learning-based Method for
Classifying
and
Localizing
Potato
Blemishes. In ICPRAM (pp. 107-117),
https://dx.doi.org/10.5220/0007350101070117. Ma, J., Du, K., Zheng, F., Zhang, L., Gong, Z., & Sun, Z. (2018). A recognition method for
cucumber diseases using leaf symptom images based on deep convolutional neural
network. Computers and electronics in agriculture, 154, 18-24. Lu Y, Yi S, Zeng N, Liu Y, Zhang Y, 2017. Identification of rice disease using deep convolutional
neural networks. Neuro Comput 267: 378-384. https://doi.org/10.1016/j. neucom.2017.06.023
Amara, J., Bouaziz, B., & Algergawy, A. (2017). A deep learning-based approach for banana leaf Ma, J., Du, K., Zheng, F., Zhang, L., Gong, Z., & Sun, Z. (2018). A recognition method for
cucumber diseases using leaf symptom images based on deep convolutional neural
network. Computers and electronics in agriculture, 154, 18-24. cucumber diseases using leaf symptom images based on deep convolutional neural
network. Computers and electronics in agriculture, 154, 18-24. Lu Y, Yi S, Zeng N, Liu Y, Zhang Y, 2017. Identification of rice disease using deep convolutional
neural networks. Neuro Comput 267: 378-384. https://doi.org/10.1016/j. neucom.2017.06.023
Amara, J., Bouaziz, B., & Algergawy, A. (2017). A deep learning-based approach for banana leaf
diseases classification. Datenbanksysteme für Business, Technologie und Web (BTW 2017)-
Workshopband. References: u
,
S,
e g N,
u
,
a g
, 0
. de t
cat o o
ce d sease us g deep co vo ut o a
neural networks. Neuro Comput 267: 378-384. https://doi.org/10.1016/j. neucom.2017.06.023
Amara, J., Bouaziz, B., & Algergawy, A. (2017). A deep learning-based approach for banana leaf
diseases classification. Datenbanksysteme für Business, Technologie und Web (BTW 2017)-
Workshopband. Fuentes, A.; Yoon, S.; Kim, S.; Park, D. A (2017) Robust deep-learning-based detector for real-
time tomato plant diseases and pests recognition. Sensors 17, 2022. Fuentes, A.; Yoon, S.; Kim, S.; Park, D. A (2017) Robust deep-learning-based detector for real-
time tomato plant diseases and pests recognition. Sensors 17, 2022. Liu, B.; Zhang, Y.; He, D.; Li, Y. (2018 ) Identification of apple leaf diseases based on deep
convolutional neural networks. Symmetry , 10, 11. Liu, B.; Zhang, Y.; He, D.; Li, Y. (2018 ) Identification of apple leaf diseases based on deep
convolutional neural networks. Symmetry , 10, 11. Zilvan, V., Ramdan, A., Suryawati, E., Kusumo, R. B. S., Krisnandi, D., & Pardede, H. F. (2019). Denoising convolutional variational autoencoders-based feature learning for automatic detection
of plant diseases. In 2019 3rd International Conference on Informatics and Computational
Sciences (ICICoS) (pp. 1-6). IEEE, https://doi.org/10.1109/ICICoS48119.2019.8982494.
|
W3144147576.txt
|
https://www.researchsquare.com/article/rs-64010/v1.pdf?c=1604108156000
|
en
|
Interferon-based therapy did not attenuate hepatitis C-associated rheumatic disorder risk: A nationwide population-based cohort study
|
Research Square (Research Square)
| 2,020
|
cc-by
| 5,686
|
Interferon-based therapy did not attenuate hepatitis
C-associated rheumatic disorder risk: A nationwide
population-based cohort study
Jur-Shan Cheng
Chang Gung Memorial Hospital
Jing-Hong Hu
Chang Gung Memorial Hospital
Yu-sheng Lin
Chang Gung Memorial Hospital
Ming-Shyan Lin
Chang Gung Memorial Hospital
Hsin-Ping Ku
Chang Gung Memorial Hospital
Rong-Nan Chien
Chang Gung Memorial Hospital
Ming-Ling Chang ( mlchang8210@gmail.com )
Chang Gung Memorial Hospital https://orcid.org/0000-0003-4902-4401
Original Research
Keywords: HCV, rheumatic, interferon, TNHIRD, mortality
Posted Date: August 27th, 2020
DOI: https://doi.org/10.21203/rs.3.rs-64010/v1
License: This work is licensed under a Creative Commons Attribution 4.0 International License.
Read Full License
Page 1/19
Abstract
Background
Whether hepatitis C virus (HCV) infection increases the risk of rheumatic disorders and whether the
associated risk is reversed by anti-HCV therapy remain elusive. We aimed to investigate these topics.
Methods
A nationwide population-based cohort study of Taiwan National Health Insurance Research Database
(TNHIRD) was conducted.
Results
Of 19,298,735 subjects, 3 TNHIRD cohorts (1:4:4, propensity score-matched), including HCV-treated
(6,919 HCV-infected subjects with interferon and ribavirin therapy > 6 months), HCV-untreated (n=27,676)
and HCV-uninfected (n=27,676) cohorts, were enrolled and had been followed since 2003 to 2015. HCVuninfected cohort had the lowest 11-year cumulative incidence of rheumatic disorders [9.535%; 95%
confidence interval (CI): 8.416~10.734%] (p<0.0001), while HCV-treated (14.95%; 12.417~17.704%) and
HCV-untreated (14.999%; 13.585~16.479%) cohorts showed no difference in the cumulative incidences of
rheumatic disorders (p=0.8316). Multivariate analyses showed, HCV infection [hazard ratio (HR):1.671;
95% CI HR: 1.562-1.788, p<0.0001], female sex (1.67; 1.565-1.782, p<0.0001), age> 49 years (1.168; 1.0881.253, p<0.0001), Charlson comorbidity index > 1 (1.156; 1.074-1.245, p=0.0001), liver cirrhosis (0.795;
0.674-0.937, p=0.0062), chronic obstruction pulmonary disease (1.244; 1.134-1.364, p<0.0001), end stage
renal disease (0.732; 0.549-0.974, p=0.0326), diabetes mellitus (0.91; 0.835-0.993, p=0.0333) and
dyslipidemia (1.205; 1.108-1.31, p<0.0001) were associated with incident rheumatic disorders. Among the
3 cohorts, the untreated cohort had the highest cumulative incidence of overall mortality (29.163%;
27.218-31.133%, p<0.0001), while the treated (13.662%; 11.389-16.140%) and un-infected (9.99%; 8.54811.559%) cohorts had indifferent mortalities (p=0.1796).
Conclusions
HCV infection, baseline demographics and comorbidities were associated with the risks of rheumatic
disorders. Although HCV-associated risk of rheumatic disorders might not be reversed by interferon-based
anti-HCV therapy, which reduced the overall mortality in HCV-infected patients.
Introduction
Hepatitis C virus (HCV) is a human pathogen responsible for acute and chronic liver disease that infects
an estimated 150 million individuals worldwide [1]. In addition to hepatic complications including
cirrhosis and hepatocellular carcinoma, HCV may cause many extrahepatic complications such as
diabetes mellitus (DM), hypolipidemia, cardiovascular events [1], and rheumatic disorders [2]. HCV is both
hepatotropic and lymphotropic [3]. HCV lymphotropism represents the most important step in the
Page 2/19
pathogenesis of virus-related immunological diseases [4], especially rheumatic disorders. Rheumatologic
extrahepatic manifestations are observed in 2–38% of HCV-infected patients [5], and this variability is
attributed to the various geographic region and design of the studies [6–8]. Moreover, HCV antibodies
were found in 18.5% among patients admitted to the rheumatology ward [9], being higher than the
estimated global prevalence (2.2–2.8%) of HCV infection [10]. HISPAMEC Registry showed, the systemic
autoimmune diseases most associated with chronic HCV infection were Sjogren syndrome (SS),
rheumatoid arthritis (RA) and systemic lupus erythematosus (SLE) [11]. Specifically, the co-prevalence of
HCV and SS ranged from 49% [12] to 80% [13], HCV infection was found in 13% of a large series of
Spanish patients with SS [14], and sicca symptoms were reported in 11% of French HCV patients [15].
HCV infection was also associated with increased RA risks [16–17], and the pooled prevalence of RA was
4.5% (0.6%-25.7%) of chronic HCV-infected patients in East Asia [2]. Moreover, the prevalence of HCV
infection among SLE patients was found up be 10% [18].
The combination of pegylated interferon (Peg-IFN) and ribavirin has provided a “cure” for a considerable
proportion of patients with chronic hepatitis C infection (CHC), particularly in patients with a favorable
interferon λ 3 (IFNL3) genotype [1]. These cure rates were further improved by replacing interferon-based
therapy with potent, direct-acting antiviral agents (DAAs) [1], and the sustained virological response rate
(SVR) to DAA in HCV-infected patients is approaching 100% [19]. However, some HCV-associated
complications such as cardiometabiolic and oncogenic events cannot be reversed, even after viral
clearance [1, 20–21]. Whether the HCV-associated risk of rheumatic disorders can be attenuated after the
completion of anti-HCV therapy thus is still a crucial issue of public health in the era of DAA to eradicate
HCV infection but remains elusive.
Accordingly, we conducted a nationwide population-based cohort study in Taiwan, a country where HCV
infection is rampant [22]. The impacts of HCV infection and anti-HCV therapy on the risk of rheumatic
disorders were investigated by comparing the cumulative incidences of rheumatic disorders and of the
overall mortalities among HCV-infected subjects with and without anti-HCV therapy and the subjects
without HCV infection, based on data from Taiwan National Health Insurance Research Database
(TNHIRD). This database provides medical information of the nationwide population, which comprises
26,573,661 individuals.
Methods
TNHIRD samples and measurements
This population-based retrospective cohort study used nation-level data, including the National Health
Insurance (NHI) administrative database, the Cancer Registry Database, and the Death Registry Database
of Taiwan. The mandatory, single-payer NHI program provides comprehensive coverage including
ambulatory care, hospital services, laboratory tests, and prescription drugs. Over 99% of the population is
enrolled in the program and approximately 90% of the healthcare organizations are contracted with NHI
Administration. Given that Taiwan is a hyperendemic area for hepatitis B virus (HBV) infection, which is
Page 3/19
highly oncogenic, causes many hepatic complications and prominently biases the phenotype of HCV
infection [23], the subjects diagnosed with HBV infection in the observation period (2003–2015), or with
any baseline rheumatic disorders including RA [International Classification of Disease, Ninth. Revision,
Clinical Modification (ICD-9-CM) code (714)], ankylosing spondylitis [ICD-9-CM code (720)], psoriatic
arthopathy [ICD-9-CM code (696.0)], sicca syndrome (also called SS) [ICD-9-CM code ( 710.2), systemic
sclerosis [ICD-9-CM code ( 710.1)], SLE [ICD-9-CM code (710.0)], Behcet’s syndrome [ICD-9-CM code
(136.1)], Raynaud’s syndrome [ICD-9-CM code (443.0)], polyarteritis nodosa and allied conditions [ICD-9CM code (446)], and psoriasis [ICD-9-CM code (696.X)] or mortality occurred prior to 6 months after
completing anti-HCV treatment (the baseline), when is the time to ensure therapeutic response, were
excluded.
HCV-treated cohort included subjects who had HCV RNA test and received ribavirin and pegylated
interferon (Peg-IFN) in 2003–2015. Their first HCV test was assumed to be the index date of diagnosis.
The baseline for the HCV-treated cohort was the date of 6 months after completing the combination
therapy. Untreated HCV-infected patients were those who had HCV test (HCV antibody or HCV RNA test)
(their first HCV test was the index date), were diagnosed with HCV [The International Classification of
Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) codes: 070.41, 070.44, 070.51, 070.54, 070.70,
070.71, V02.62], were prescribed hepatoprotective agents (silymarin, liver hydrolysate, choline bitartrate,
or ursodeoxycholic acid), but did not receive any anti-HCV therapy (ribavirin or peg-interferon). HCVuninfected individuals were who did not have HCV diagnosis or HCV tests, and received no
hepatoprotective agents or anti-HCV therapy, and they were classified as being HCV-uninfected. The HCVtreated cohort was matched with untreated HCV-infected patients (HCV-untreated cohort) and with HCVuninfected individuals (HCV-uninfected cohort) through a propensity score matching method indicating
the probability of receiving the combination therapy, was estimated by using a logistic model. The
covariates in the model included sex (male, female), age (20–39, 40–49, 50–59, ≥ 60), NHI registration
location (city, township, rural area), Charlson Comorbidity Index (CCI) score (0, 1, ≥ 2), and year of the
index date (2003–2006, 2007–2009, 2010–2012). This method was used to assure that the HCV-treated
cohort and the selected counterparts were comparable in observed characteristics. The baselines for the
HCV-untreated and HCV-uninfected cohorts were assigned according to the period from the index date to
the baseline of their matched counterparts of the HCV-treated cohort, and subjects with rheumatic
disorder or mortality occurred before the baselines were not selected. The index date of the HCVuninfected individuals was the date of one of their physician visits randomly selected from their claims
database. The matching process for the 3 cohorts was shown in Supplementary Fig. 1.
Outcomes were defined as the development of rheumatic disorders as mentioned above. Subjects were
followed until the date of the event, death, or the end of follow-up (December 31, 2015), whichever came
first. Date of death were adopted from the Death Registry database. For the HCV-treated group, only the
rheumatic disorder or mortality occurred 6 months after the complement of anti-HCV therapy (the
baseline) were recorded.
Statistical analysis
Page 4/19
All statistical analyses were performed using the Statistical Analysis System ((SAS version 9.4, SAS
Institute Inc., Cary, NC, USA) software. Continuous variables were analyzed using a Student’s t-test or
Analysis of variance, as appropriate, and categorical variables were analyzed using a chi-square test or
Fisher’s exact test, as appropriate. Cumulative incidences of outcomes were estimated and compared by
using the modified Kaplan-Meier method and the Gray method with death being a competing risk event.
Sub-distribution hazards models for competing risks, an extension of Cox proportional hazards models
taking competing mortality into consideration, were used to estimate adjusted hazard ratio of developing
rheumatic disorders, adjusting for age, sex, NHI registration location, the CCI score, year of the index date,
and comorbid liver cirrhosis, chronic obstructive pulmonary disease (COPD), end stage renal disease
(ESRD), DM, hypertension, dyslipidemia, cardiovascular events (including percutaneous coronary
intervention, coronary artery bypass graft, myocardial infarction, heart failure, cardiogenic shock, and
peripheral vascular disease) and stroke. Statistical significance was defined at the 5% level
Results
Baseline characteristics
From a total of 19,298,735 individuals between January 1, 2003, and December 31, 2015, 11,223,475
patients without HBV infection and baseline rheumatic disorders were identified; 104,281 patients with
HCV infection and 11,119,194 patients without HCV infection were eligible for the study. Of all, 3 cohorts
including HCV-treated (n = 6,919), HCV-untreated (n = 27,676) and HCV-uninfected (n = 27,676) cohorts
were enrolled (Fig. 1). The 3 cohorts were matched with the propensity scores (1:4:4), did not differ in
demographic factors, residency, CCI score and index year, which were the covariates in the models to
calculate propensity scores, although baseline comorbidities were not similar (Table 1). Compared with
HCV-untreated cohorts, the HCV-treated cohort had higher rates of baseline cirrhosis, comparable rates of
COPD, but lower rates of other comorbidities. Compared with the HCV-uninfected cohort, the HCV-treated
cohort had higher rates of most comorbidities including cirrhosis, comparable rates of DM and
cardiovascular events, but lower rates of dyslipidemia and stroke. Compared with the HCV-uninfected
cohort, the HCV-untreated cohort had higher rates of all baseline comorbidities except stroke. To lineate
the HCV-associated complications, we compared the baseline factors between the HCV-infected cohort,
which was a combination of the HCV-treated and HCV-untreated cohorts, and HCV-uninfected cohort. The
HCV-infected cohort had higher rates of all baseline comorbidities except indifferent rates of dyslipidemia
and lower rates of stroke, than the HCV-uninfected cohort (Supplementary Table 1).
Page 5/19
Table 1
Baseline characteristics of the 3 HCV cohorts of TNHIRD.
(1)
(2)
(3)
p values
Treated
Untreated
Uninfected
(1)-(2)
(1)-(3)
(2)-(3)
6,919
27,676
27,676
Male
3,832,
(55.38)
15,328,
(55.38)
15,328,
(55.38)
1
1
1
Female
3,087,
(44.62)
12,348,
(44.62)
12,348,
(44.62)
20–39
1,312,
(18.96)
5,247,
(18.96)
5,248,
(18.96)
1
1
1
40–49
1,811,
(26.17)
7,243,
(26.17)
7,244,
(26.17)
50–59
2,443,
(35.31)
9,774,
(35.32)
9,772,
(35.31)
≥ 60
1,353,
(19.55)
5,412,
(19.55)
5,412,
(19.55)
city
1,482,
(21.42)
5,928,
(21.42)
5,928,
(21.42)
1
1
1
township
2,174,
(31.42)
8,696,
(31.42)
8,696,
(31.42)
rural area
3,263,
(47.16)
13,052,
(47.16)
13,052,
(47.16)
0
3,443,
(49.76)
13,774,
(49.77)
13,772,
(49.76)
0.9999
1
0.9998
1
2,138,
(30.90)
8,550,
(30.89)
8,552,
(30.90)
>=2
1,338,
(19.34)
5,352,
(19.34)
5,352,
(19.34)
N
Gender, N, (%)
Age range (yrs), N,
(%)
Area, N, (%)
CCI score, N, (%)
TNHIRD: Taiwan National Health Insurance Research Database; HCV: hepatitis C virus; CCI: Charlson
Comorbidity Index; COPD: Chronic obstructive pulmonary disease; ESRD: end stage renal disease; DM:
diabetes mellitus.
Page 6/19
(1)
(2)
(3)
p values
2003–2006
3,601,
(52.05)
14,404,
(52.05)
14,404,
(52.05)
0.9997
1
0.9992
2007–2009
2,274,
(32.87)
9,099,
(32.88)
9,096,
(32.87)
2010–2012
1,044,
(15.09)
4,173,
(15.08)
4,176,
(15.09)
Liver cirrhosis
695,
(10.04)
1,685, (6.09)
9, (0.03)
<
0.0001
<
0.0001
<
0.0001
COPD
775, (11.2)
3,160,
(11.42)
2548, (9.21)
0.6114
<
0.0001
<
0.0001
ESRD
47, (0.68)
722, (2.61)
81, (0.29)
<
0.0001
<
0.0001
<
0.0001
DM
1,320,
(19.08)
6,166,
(22.28)
5,004,
(18.08)
<
0.0001
0.0549
<
0.0001
Hypertension
2,011,
(29.06)
9,485,
(34.27)
7,422,
(26.82)
<
0.0001
0.0002
<
0.0001
Dyslipidemia
815,
(11.78)
5,268,
(19.03)
4,815, (17.4)
<
0.0001
<
0.0001
<
0.0001
Cardiovascular
events
165, (2.38)
1059, (3.83)
685, (2.48)
<
0.0001
0.6642
<
0.0001
Stroke
227, (3.28)
1369, (4.95)
1407, (5.08)
<
0.0001
<
0.0001
0.4593
index_yr, N, (%)
Baseline factor, N,
(%)
TNHIRD: Taiwan National Health Insurance Research Database; HCV: hepatitis C virus; CCI: Charlson
Comorbidity Index; COPD: Chronic obstructive pulmonary disease; ESRD: end stage renal disease; DM:
diabetes mellitus.
Cumulative incidences and associated factors of rheumatic
disorders
The HCV-treated, untreated, and uninfected cohorts were followed up until 2015 or death, with the longest
observation of 11 years. Rheumatic disorders occurred cumulatively at 11 years in 14.95%, 14.999%, and
9.535% of the HCV-treated, untreated, and uninfected cohorts, respectively (Fig. 2, Table 2). The HCVuninfected cohort had the lowest cumulative incidence of rheumatic disorders among the 3 cohorts.
However, no difference of cumulative incidences of rheumatic diseases was identified between the HCVtreated and HCV-untreated cohorts. The multivariate analysis of the 3 cohorts showed, compared with the
Page 7/19
HCV-uninfected cohort, both the HCV-treated and HCV-untreated cohorts had higher hazard ratios (HRs)
to develop rheumatic disease. In addition, female sex, baseline age ≥ 49 years, CCI score ≥ 1, baseline
COPD and dyslipidemia were associated with increased HRs of rheumatic diseases, while baseline liver
cirrhosis, ESRD and DM were associated with decreased HRs of rheumatic diseases (Supplementary
Fig. 2). Given that HCV-treated and HCV-untreated cohorts yielded similar HRs to develop rheumatic
disorders, we thus combined HCV-treated and HCV-untreated cohorts to form HCV-infected cohort as
mentioned above and compared the HCV-infected cohort with the HCV-uninfected cohort to view the
impact of HCV infection on the development of rheumatic disorders. In addition to sex, age, CCI score,
baseline COPD, dyslipidemia, cirrhosis, ESRD and DM, HCV infection was significantly associated with
the development of rheumatic disorders, with a HR of 1.671 (Fig. 3).
Table 2
Comparison of the cumulative incidences of rheumatic disorders among (1) HCV-treated, (2) HCVuntreated and (3) HCV-uninfected cohorts.
Rheumatic
disorders
(1) Treated
(2)
Untreated
(3)
Uninfected
Number
6,919
27,676
27,676
Follow-up
(years),
mean ± SD
4.61+/-1.90
4.62+/-1.07
4.89+/-1.96
Event
number, N
(%)
503 (7.27)
2140 (7.73)
1310 (4.73)
Competing
mortality, N
(%)
281 (4.06)
3,478
(12.57)
1316 (4.10)
Cumulative
incidence, %
(95% CI)
14.95
(12.417–
17.704)
14.999
(13.585–
16.479)
9.535
(8.416–
10.734)
p values
(1)(2)
(3)
(1)-(2)
(1)-(3)
(2)-(3)
<
0.0001
0.8316
<
0.0001
<
0.0001
CI: confidence interval.
Cumulative incidences of mortality
Of the 3 cohorts, the HCV-untreated cohort had the highest cumulative incidence (29.163%) of overall
mortality at 11 years (p < 0.0001). The HCV-treated and HCV-uninfected cohorts yielded indifferent
mortality rates (p = 0.1796) (Table 3).
Page 8/19
Table 3
Comparison of the cumulative incidences of overall mortality among (1) HCV-treated, (2) HCV-untreated
and (3) HCV-uninfected cohorts.
Overall
mortality
(1) Treated
(2)
Untreated
(3)
Uninfected
Number
6,919
27,676
27,676
Follow-up
(years), mean
± SD
4.82
+/-1.84
4.86
+/-2.03
5.03+/-1.91
Event number,
N (%)
304 (4.39)
3,669
(13.26)
1,170
(4.23)
Cumulative
incidence, %
(95% CI)
13.662
(11.389–
16.140)
29.163
(27.218–
31.133)
9.99
(8.548–
11.559)
p values
(1)(2)
(3)
(1)-(2)
(1)-(3)
(2)-(3)
<
0.0001
<
0.0001
0.1796
<
0.0001
CI: confidence interval.
Discussion
The most compelling results of the current study are as follows: (1) The HCV-uninfected cohort had the
lowest cumulative incidence of rheumatic disorders among the 3 cohorts, while indifferent cumulative
incidences were identified between the HCV-treated and HCV-untreated cohorts. (2) HCV infection, female
gender, baseline age ≥ 49 years, CCI score ≥ 1, baseline COPD and dyslipidemia were associated with
increased HRs of rheumatic disorders, while baseline liver cirrhosis, ESRD and DM were associated with
decreased HRs. (3) The HCV-untreated cohort had the highest cumulative incidence of overall mortality at
11 years, while HCV-treated and HCV-uninfected cohorts yielded indifferent mortality rates.
The higher rate of baseline cirrhosis in the HCV-treated than the HCV-untreated cohorts of TNHIRD was
coincided with the fact that only patients with significant fibrosis were reimbursed with anti-HCV therapy
[24], and the other different baseline variables between these 2 cohorts highlight the idea that patients
with comorbidities were ineligible for the interferon-based therapy and had been excluded for anti-HCV
therapy. The different rates in baseline variables between HCV-infected and HCV-uninfected cohorts were
consistent with the phenomenon that HCV infection elicits many cardiometabolic events and
hypolipidemia [1]. Therefore, the baseline comparisons of the 3 cohorts supported the reliability of the
data based on TNHIRD.
The facts that the HCV-uninfected cohort had the lowest cumulative incidence of rheumatic disorders,
and HCV infection increased the HR of developing rheumatic disorders based on multivariate analyses
endorsed the concept that HCV infection might cause rheumatic disorders, despite some studies did not
support the participation of HCV infection in the pathogenesis of RA [25–27]. However, given that the HRs
Page 9/19
in developing rheumatic disorders between the HCV-treated and HCV-untreated cohorts were indifferent,
the HCV-associated risk of rheumatic disorders might not be attenuated by interferon-based anti-HCV
therapy. In particular, cryoglobulinemic vasculitis represents the prototype of HCV-related rheumatic
disorders [28], long-term mixed cryoglobulinemia after SVR is common since cryoglobulin-generating B
lymphocytes might have reached an HCV-independent autonomous phase before viral clearance [29].
HCV-associated rheumatic disorder therefore might persist despite viral clearance. Moreover, whether
interferon-based therapy reduces the risk of RA had remained conflicting [30–31], and interferon-based
anti-HCV therapy may work as a "trigger" for RA [32–33] or SLE [34] had been showed in some case
reports. Although treatment with interferon-alpha may lead to substantial clinical improvement of HCVrelated arthritis even without a complete biochemical or virological response [35], autoimmune disorders
indeed occur in 4–19% of patients receiving interferon-based anti-HCV therapy and the associated
symptoms developed between 2 weeks and 7 years after initiation of therapy [36]. The interferon-based
anti-HCV therapy thus has been contraindicated for many rheumatologic autoimmune/inflammatory
disorders based on the concern of triggering rheumatic disorders. New oral interferon-free combinations
of various DAAs offer an opportunity for HCV-infected patients with rheumatic disorders to be cured with
a short treatment duration and a low risk of side effects [37]. However, SVR following DAA might lead to
immune reconstitution as tuberculosis reactivation had been reported [38]. Whether DAA therapy precisely
attenuates the risks of HCV-associated rheumatic disorder without introducing other harm as mentioned
above [38] demands further investigation.
On the other hand, that female sex and baseline age ≥ 49 years are positively associated with the
increased HRs of rheumatic disorders is consistent with the fact that female sex and old age had been
identified as risk factors for RA [39]. CCI score ≥ 1 and baseline COPD were associated with increased
HRs of rheumatic disorders coincides that comorbidities including respiratory disease were more
common in patients with RA at diagnosis than controls [40]. Patients with rheumatic disorders have
increased prevalence of metabolic syndrome including dylipidemia [41] and acute myocardial infarction
risk increased by 38% [42] in RA patients might explained why dyslipidemia were associated with
increased HR of rheumatic diseases. Of note, that baseline liver cirrhosis, ESRD and DM are associated
with reduced HRs of rheumatic disorders is a novel finding. Interestingly, the connections with cirrhosis
are variable among different rheumatic disorders. For example, the overall incidences of cirrhosis were
reported to be lower in the RA cohort than in the non-RA cohort [43–44], while patients with psoriasis were
found to have increased risk of cirrhosis than patients without psoriasis [44].With regard to ESRD and
DM, in contrast to their negative associations with the rheumatic disorder risks, chronic kidney disease is
a common complication of rheumatic disorders [45], patients undergoing hemodialysis therapy may
develop serious rheumatic complications [46], newly diagnosed RA patients are at higher risk of DM [47]
and the prevalence of DM is higher in patients with psoriatic arthritis compared with the general
population [48]. That rheumatic disorders might be mistaken as ESRD- or DM-related complications in
patients with ESRD and DM potentially explains the paradox.
Among the 3 cohorts, the HCV-untreated cohort yielded the highest overall mortality, which might be
caused by other HCV-associated events such as cirrhosis, HCC or cardiometabolic events [1] other than
Page 10/19
rheumatic disorder-associated complications, since HCV-treated and HCV-uninfected cohorts had
indifferent mortalities although the latter obviously had lower risk of rheumatic disorders. This
phenomenon indicates the importance to prescribe anti-HCV therapy in HCV-infected patients in
decreasing overall mortality, regardless of the risk for rheumatic disorders.
There are limitations recognized in the current study. First, because linking the results from TNHIRD to the
laboratory results of individual patients was forbidden for privacy protection, the correlation of SVR with
rheumatic disorders could not be identified. However, we are confident of the antiviral efficacy in the HCVtreated cohort since interferon-based therapy for HCV infection generally achieves an SVR rate up to 90%
in Taiwan [49], where a favorable genetic variation in IFNL3 is prevalent [49]. Second, as mentioned
above, interferon-based therapy might elicit rheumatic disorders in SVR patients [32–34, 36] and blunt the
impact of viral clearance in attenuating rheumatic disorder risks. Third, because most of the rheumatic
diseases accounted for the minority of the whole population and our preliminary statistical tests did not
show any significance for any individual rheumatic disease, we thus had put all rheumatic diseases
together as rheumatic disorders to yield the maximal statistical power. Some specific rheumatic diseases
might have different connections with HCV infection or anti-HCV therapy. Anyhow, that SVR did not
reduce the incidences of SLE and RA in CHC patients [50] supported our observation. Future prospective
studies in other independent large cohorts with identifiable SVR following DAA therapy, subgroup
analyses for specific rheumatic diseases and sophisticated molecular investigations are required to
elucidate the fundamental mechanisms underlying the findings described here.
Conclusions
Taken together, HCV infection, female sex, baseline age ≥ 49 years, and other comorbidities were
associated with risks of rheumatic disorders. Although interferon-based therapy did not attenuate the
rheumatic disorder risk, it indeed decreased the overall mortality of HCV-infected patients. These findings
may merit further study for preventing or treating rheumatic disorders in HCV-infected patients.
Abbreviations
HCV: hepatitis C virus; TNHIRD:Taiwan National Health Insurance Research Database; CI:confidence
interval; HR:hazard ratio; DM:diabetes mellitus; SS:Sjogren syndrome; RA:rheumatoid arthritis;
SLE:systemic lupus erythematosus; Peg-IFN:pegylated interferon; CHC:hepatitis C infection;
IFNL3:interferon λ 3; DAAs:direct-acting antivirals; SVR:sustained virological response rate; NHI:National
Health Insurance; HBV:hepatitis B virus; ICD-9-CM:International Classification of Disease, Ninth. Revision,
Clinical Modification; SAS:Statistical Analysis System; COPD:chronic obstructive pulmonary disease;
ESRD:end stage renal disease
Declarations
Ethics approval and consent to participate
Page 11/19
The study protocol conformed to the ethical guidelines of the 1975 Declaration of Helsinki and was
approved by the local Institutional Review Board. The need for consent was waived because the nationallevel data used in this study were de-identified by encrypting personal identification information.
Consent for publication
Not applicable.
Availability of data and materials
The datasets used and/or analysed during the current study are available from the corresponding author
on reasonable request.
Competing interests
The authors declare that they have no competing interests.
Funding
This study was supported by grants from the Chang Gung Medical Research Program (CMRPG3I0412
and CMRPG3K0721) and the National Science Council (MOST 108-2314-B-182-051-, MOST 109-2314-B182-024- and MOST 109-2629-B-182-002-). The funders had no role in study design, data collection and
analysis, decision to publish, or preparation of the manuscript. The opinions expressed in this paper are
those of the authors and do not necessarily represent those of Chang Gung Medical Hospital and
National Science Council, Taiwan.
Authors' contributions
JSC: study implementation and manuscript drafting
JHH, YSL, MSL, and RNC: data collection and manuscript writing.
HPK: statistical analysis and manuscript writing.
MLC: study design and implementation, manuscript drafting, and critical revision of the manuscript for
important intellectual content.
All authors read and approved the final manuscript.
Acknowledgements
The authors thank Ms. Shu-Chun Chen, Ms. Chia-Hui Tsai, Mr. Chun-Kai Liang and Mr. Shuen-Shian Shiau
from the Liver Research Center, Chang Gung Memorial Hospital, Taiwan, for their assistance with data
mining.
Page 12/19
References
1. Chang ML. Metabolic alterations and hepatitis C: From bench to bedside. World J Gastroenterol.
2016;22(4):1461–76.
2. Younossi ZM, Henry L, Ong P. J, et al. Systematic review with meta-analysis: extrahepatic
manifestations in chronic hepatitis C virus-infected patients in East Asia. Aliment Pharmacol Ther.
2019;49(6):644–53.
3. Sebastiani M, Giuggioli D, Colaci M, Fallahi P, Gragnani L, Antonelli A, Zignego AL. Ferri C.HCVRelated Rheumatic Manifestations and Therapeutic Strategies. Curr Drug Targets. 2017;18(7):803–
10.
4. Calvaruso V. Craxì A.Immunological alterations in hepatitis C virus infection. World J Gastroenterol.
2013 Dec;21(47):8916–23. 19(.
5. Sayiner ZA, Haque U, Malik MU, Gurakar A. Hepatitis C virus infection and its rheumatologic
implications. Gastroenterol Hepatol (N Y). 2014 May;10(5):287–93.
6. Cacoub P, Renou C, Rosenthal E, et al. Extrahepatic manifestations associated with hepatitis C virus
infection. A prospective multicenter study of 321 patients. The GERMIVIC Groupe d’Etude et de
Recherche en Medecine Interne et Maladies Infectieuses sur le Virus de l’ Hepatite C. Med (Baltim).
2000;79(1):47–56.
7. Mohammed RH, ElMakhzangy HI, Gamal A, et al. Prevalence of rheumato-logic manifestations of
chronic hepatitis C virus infection among Egyptians. Clin Rheumatol. 2010;29(12):1373–80.
8. Palazzi C, Olivieri I, Cacciatore P, Pennese E, D’Amico E. Difficulties in the differential diagnosis
between primitive rheumatic diseases and hepatitis C virus-related disorders. Clin Exp Rheumatol.
2005;23(1):2–6.
9. El Garf A, El Zorkany B, Gheith R, Sheba H, Abdel Moneim G. El Garf K.Prevalence and clinical
presentations of hepatitis C virus among patients admitted to the rheumatology ward. Rheumatol Int.
2012 Sep;32(9):2691–5.
10. Mohd Hanafiah K, Groeger J, Flaxman AD, Wiersma ST. Global epidemiology of hepatitis C virus
infection: new estimates of age-specific antibody to HCV seroprevalence. Hepatology.
2013;57:1333–42.
11. Ramos-Casals M, Muñoz S, Medina F, et al. Systemic autoimmune diseases in patients with hepatitis
C virus infection: characterization of 1020 cases (The HISPAMEC Registry). J Rheumatol.
2009;36(7):1442–8.
12. Pawlotsky JM, Ben Yahia M, Andre C, Voisin MC, Intrator L, Roudot-Thoraval F, Deforges L, Duvoux C,
Zafrani ES, Duval J. Immunological disorders in C virus chronic active hepatitis: a prospective casecontrol study.Hepatology. 1994 Apr; 19(4):841–8.
13. Loustaud-Ratti V, Riche A, Liozon E, et al. Prevalence and characteristics of Sjögren’s syndrome or
Sicca syndrome in chronic hepatitis C virus infection: a prospective study. J Rheumatol.
2001;28(10):2245–51.
Page 13/19
14. Brito-Zerón P, Gheitasi H, Retamozo S, Bové A, Londoño M, Sánchez-Tapias JM, Caballero M, Kostov
B, Forns X, Kaveri SV. Ramos-Casals M.How hepatitis C virus modifies the immunological profile of
Sjögren syndrome: analysis of 783 patients. Arthritis Res Ther. 2015 Sep;10:17:250.
15. Cacoub P, Poynard T, Ghillani P, Charlotte F, Olivi M, Piette JC, Opolon P. Extrahepatic manifestations
of chronic hepatitis C. MULTIVIRC Group. Multidepartment Virus C. Arthritis Rheum. 1999;42:2204–
12.
16. Su FH, Wu CS, Sung FC, et al. Chronic hepatitis C virus infection is associated with the development
of rheumatoid arthritis: a nationwide population-based study in Taiwan. [PLoS One.
2014;9(11):e113579.
17. Kudaeva FM, Speechley MR, Pope JE. A systematic review of viral exposures as a risk for rheumatoid
arthritis. Semin Arthritis Rheum. 2019;48(4):587–96.
18. Mercado U, Avendaño-Reyes M, Araiza-Casillas R, Díaz-Molina R. Prevalance of antibodies against
hepatitis C and B viruses in patients with systemic lupus erythematosus [in Spanish]. Rev
Gastroenterol Mex. 2005;70(4):399–401.
19. Vermehren J, Park JS, Jacobson I, et al. Challenges and perspectives of direct antivirals for the
treatment of hepatitis C virus infection. J Hepatol. 2018;69:1178–87.
20. Toyoda H, Kumada T, Tada T, et al. Risk factors of hepatocellular carcinoma development in noncirrhotic patients with sustained virologic response for chronic hepatitis C virus infection. J
Gastroenterol Hepatol. 2015;30:1183–9.
21. Kalaitzakis E, Gunnarsdottir SA, Josefsson A, et al. Increased risk for malignant neoplasms among
patients with cirrhosis. Clin Gastroenterol Hepatol. 2011;9:168–74.
22. Hu JH, Chen MY, Yeh CT, et al. Sexual Dimorphic Metabolic Alterations in Hepatitis C Virus-infected
Patients: A Community-Based Study in a Hepatitis B/Hepatitis C Virus Hyperendemic Area. Med
(Baltim). 2016;95:e3546.
23. Chang ML, Lin YJ, Chang CJ, et al. Occult and Overt HBV Co-Infections Independently Predict
Postoperative Prognosis in HCV-Associated Hepatocellular Carcinoma. PLoS One. 2013;8(1):e64891.
24. Chen CW, Cheng JS, Chen TD, Le PH, Ku HP, Chang ML. The irreversible HCV-associated risk of
gastric cancer following interferon-based therapy: a joint study of hospital-based cases and
nationwide population-based cohorts. Therap Adv Gastroenterol. 2019;12:1756284819855732.
25. Yılmaz N, Karadağ Ö, Kimyon G, Yazıcı A, Yılmaz S, Kalyoncu U, Kaşifoğlu T, Temiz H, Baysal B,
Tözün N. Prevalence of hepatitis B and C infections in rheumatoid arthritis and ankylosing
spondylitis: A multicenter countrywide study. Eur J Rheumatol. 2014 Jun;1(2):51–4.
26. Maillefert JF, Muller G, Falgarone G, Bour JB, Ratovohery D, Dougados M, Tavernier C, Breban M.
Prevalence of hepatitis C virus infection in patients with rheumatoid arthritis. Ann Rheum Dis. 2002
Jul;61(7):635–7.
27. Hsu FC, Starkebaum G, Boyko EJ, Dominitz JA. Prevalence of rheumatoid arthritis and hepatitis C in
those age 60 and older in a US population based study. J Rheumatol. 2003 Mar;30(3):455–8.
Page 14/19
28. Sebastiani M, Giuggioli D, Colaci M, et al. HCV-Related Rheumatic Manifestations and Therapeutic
Strategies. Curr Drug Targets. 2017;18(7):803–10.
29. Cheng YT, Cheng JS, Lin CH, Chen TH, Lee KC, Chang ML. Rheumatoid factor and immunoglobulin M
mark hepatitis C-associated mixed cryoglobulinaemia: an 8-year prospective study. Clin Microbiol
Infect. 2020;26(3):366–72.
30. Tung CH, Lai NS, Li CY, Tsai SJ, Chen YC, Chen YC. Risk of rheumatoid arthritis in patients with
hepatitis C virus infection receiving interferon-based therapy: a retrospective cohort study using the
Taiwanese national claims database. BMJ Open. 2018 Jul 23;8(7):e021747.
31. Rossi C, Jeong D, Wong S, McKee G, Ahmed Butt Z, Buxton J, Wong J, Darvishian M, Bartlett S, Samji
H, Yu A, Binka M, Alvarez M, Adu PA, Tyndall M, Krajden M, Zafar Janjua N. BC Hepatitis Testers
Cohort Team.Sustained Virologic Response from Interferon-based Hepatitis C regimes IS associated
with Reduced risk of Extrahepatic Manifestations. J Hepatol. 2019 Aug 6. pii: S0168-8278(19)304623.
32. Izumi Y, Komori A, Yasunaga Y, et al. Rheumatoid arthritis following a treatment with IFNalpha/ribavirin against HCV infection. Intern Med. 2011;50(9):1065–8.
33. Pittau E, Bogliolo A, Tinti A, Mela Q, Ibba G, Salis G, Perpignano G. Development of arthritis and
hypothyroidism during alpha-interferon therapy for chronic hepatitis C. Clin Exp Rheumatol. 1997
Jul-Aug;15(4):415–9.
34. Niewold TB, Swedler WI. Systemic lupus erythematosus arising during interferon-alpha therapy for
cryoglobulinemic vasculitis associated with hepatitis C. Clin Rheumatol. 2005 Apr;24(2):178–81.
35. Zuckerman E, Keren D, Rozenbaum M, et al. Hepatitis C virus-related arthritis: characteristics and
response to therapy with interferon alpha. Clin Exp Rheumatol. 2000;18(5):579–84.
36. Wilson LE, Widman D, Dikman SH, Gorevic PD. Autoimmune disease complicating antiviral therapy
for hepatitis C virus infection. Semin Arthritis Rheum. 2002 Dec;32(3):163–73.
37. Cacoub P, Comarmond C, Desbois AC, Saadoun D. Rheumatologic Manifestations of Hepatitis C
Virus Infection. Clin Liver Dis. 2017;21(3):455–64.
38. Kida T, Umemura A, Kaneshita S, et al. Effectiveness and safety of chronic hepatitis C treatment with
direct-acting antivirals in patients with rheumatic diseases: A case-series. Mod Rheumatol. 2019;1–7.
39. Safiri S, Kolahi AA, Hoy D, et al. Global, regional and national burden of rheumatoid arthritis 1990–
2017: a systematic analysis of the Global Burden of Disease study 2017. Ann Rheum Dis.
2019;78(11):1463–71.
40. Nikiphorou E, de Lusignan S, Mallen C, et al. Prognostic value of comorbidity indices and lung
diseases in early rheumatoid arthritis: a UK population-based study. Rheumatology.
2020;59(6):1296–305.
41. Medina G, Vera-Lastra O, Peralta-Amaro AL, et al. Metabolic syndrome, autoimmunity and rheumatic
diseases. Pharmacol Res. 2018;133:277–88.
42. Chung WS, Lin CL, Peng CL, et al. Rheumatoid arthritis and risk of acute myocardial infarction–a
nationwide retrospective cohort study. Int J Cardiol. 2013;168(5):4750–4.
Page 15/19
43. Hsu CS, Lang HC, Huang KY, Chao YC, Chen CL. Risks of hepatocellular carcinoma and cirrhosisassociated complications in patients with rheumatoid arthritis: a 10-year population-based cohort
study in Taiwan. Hepatol Int. 2018;12(6):531–43.
44. Tung CH, Lai NS, Lu MC, Lee CC. Liver cirrhosis in selected autoimmune diseases: a nationwide
cohort study in Taiwan. Rheumatol Int. 2016;36(2):199–205.
45. Ciszek M, Kisiel B, Czerwinski J, et al. Kidney Transplant Recipients With Rheumatic Diseases:
Epidemiological Data From the Polish Transplant Registries 1998–2015. Transplant Proc.
2018;50(6):1654–1657.
46. Akasbi N, Houssaini TS, Tahiri L, et al. Rheumatic complications of long term treatment with
hemodialysis. Rheumatol Int. 2012;32(5):1161–3.
47. Emamifar A, Levin K, Jensen Hansen IM. Patients with newly diagnosed rheumatoid arthritis are at
increased risk of diabetes mellitus: an observational cohort study. Patients with newly diagnosed
rheumatoid arthritis are at increased risk of diabetes mellitus: an observational cohort study. Acta
Reumatol Port. 2017;42(4):310–7.
48. Eder L, Chandran V, Cook R, Gladman DD. The Risk of Developing Diabetes Mellitus in Patients with
Psoriatic Arthritis: A Cohort Study. J Rheumatol. 2017;44(3):286–91.
49. Yu ML, Dai CY, Huang JF, et al. A randomised study of peginterferon and ribavirin for 16 versus 24
weeks in patients with genotype 2 chronic hepatitis C. Gut. 2007;56:553–9.
50. Hsu WF, Chen CY, Tseng KC, et al. Sustained virological response to hepatitis C therapy does not
decrease the incidence of systemic lupus erythematosus or rheumatoid arthritis. Sci Rep.
2020;10:5372.
Figures
Page 16/19
Figure 1
Flow chart of TNHIRD study subjects selection. TNHIRD: Taiwan National Health Insurance Research
Database; HCV: hepatitis C virus; Peg-IFN: pegylated interferon; PS: propensity score.
Page 17/19
Figure 2
Cumulative incidences of rheumatic disorders among the 3 TNHIRD cohorts including HCV-treated, HCVuntreated and HCV-uninfected cohorts. TNHIRD: Taiwan National Health Insurance Research Database;
HCV: hepatitis C virus.
Page 18/19
Figure 3
Forrest plot of factors associated with incident rheumatic disorders in the 2 TNHIRD cohorts: HCVpositive (untreated) and HCV-negative (combination of treated and uninfected) cohorts. TNHIRD: Taiwan
National Health Insurance Research Database; HR: hazards ratio; LCL: lower confidence interval limit;
HCL: higher confidence interval limit; HCV: hepatitis C virus; CCI: Charlson Comorbidity Index score; COPD:
Chronic obstructive pulmonary disease; ESRD: end stage renal disease; DM: diabetes mellitus.
Supplementary Files
This is a list of supplementary files associated with this preprint. Click to download.
SupplementaryFigure2.tif
SupplementaryFigure1061320.tif
SupplementaryTables081020.doc
Page 19/19
|
|
https://openalex.org/W2782788273
|
http://www.scielo.br/pdf/dn/v11n4/1980-5764-dn-11-04-0434.pdf
|
Portuguese
| null |
Efficacy of cognitive stimulation therapy for older adults with vascular dementia
|
Dementia & Neuropsychologia
| 2,017
|
cc-by
| 5,943
|
Original Article Original Article Dement Neuropsychol 2017 December;11(4):434-441 DOI: 10.1590/1980-57642016dn11-040014 EFICÁCIA DA TERAPIA DE ESTIMULAÇÃO COGNITIVA PARA IDOSOS COM DEMÊNCIA VASCULAR RESUMO. Introdução: A terapia de estimulação cognitiva (TSC) é uma intervenção psicossocial baseada em evidências
para pessoas com demência leve a moderada resultante de vários fatores etiológicos. Objetivo: O objetivo do presente
estudo foi avaliar a eficácia do programa TSC, adaptação italiana -TSC-IT-, com indivíduos com demência vascular
(DV). Métodos: Os idosos com DV leve a moderada (N = 35) foram atribuídos a um dos dois programas: um grupo (N =
21) frequentou as 14 sessões do programa TSC-IT, enquanto o outro, grupo de controle ativo (N = 14) participaram de
atividades alternativas. Os seguintes domínios foram examinados: funcionamento cognitivo, qualidade de vida, humor,
comportamento, atividades funcionais da vida diária. Resultados: Em comparação com os controles ativos, o grupo
TSC-IT mostrou uma maior melhora no funcionamento cognitivo geral após a intervenção (ou seja, um aumento nos
escores de Mini-exame do estado mental e uma diminuição na Escala de avaliação da doença de Alzheimer – subescala
cognitiva). Uma tendência para uma melhoria também foi identificada na memória de curto prazo/trabalho – a tarefa
de extensão de dígitos atrasados – e a qualidade de vida percebida (Quality of Life - escala da doença de Alzheimer). Não houve diferenças significativas entre os dois grupos nos outros domínios considerados. Conclusão: Os resultados
atuais suportam a eficácia do TSC em pessoas com demência vascular. Palavras-chave: demência vascular, terapia de estimulação cognitiva (TSC), funcionamento cognitivo geral, qualidad
de vida. Efficacy of cognitive stimulation therapy
for older adults with vascular dementia Federica Piras1, Elena Carbone2, Silvia Faggian3, Elisa Salvalaio3, Simona Gardini4, Erika Borella2 Federica Piras1, Elena Carbone2, Silvia Faggian3, Elisa Salvalaio3, Simona Gardini4, Erika Borella2 ABSTRACT. Background: Cognitive stimulation therapy (CST) is an evidence-based psychosocial intervention for people
with mild-to-moderate dementia due to various etiological factors. Objective: The aim of the present study was to
assess the efficacy of the CST program, Italian adaptation -CST-IT-, in individuals who have vascular dementia (VaD). Methods: Older adults with mild-to-moderate VaD (N = 35) were assigned to one of two programs: one group (N = 21)
attended the 14 sessions of the CST-IT program, while the other, active control group (N = 14) took part in alternative
activities. The following domains were examined: cognitive functioning, quality of life, mood, behavior, functional activities
of daily living. Results: Compared with the active controls, the CST-IT group showed a greater improvement in general
cognitive functioning after the intervention (i.e. score increase on the Mini-Mental State Examination and decrease on the
Alzheimer’s Disease Assessment Scale – Cognitive subscale). A trend towards improvement was also identified in short-
term/working memory – the backward digit span task- and perceived quality of life (Quality of Life – Alzheimer’s Disease
scale). No significant differences emerged between the two groups for the other domains considered. Conclusion: The
present results support the efficacy of CST in people with vascular dementia. Key words: vascular dementia, cognitive stimulation therapy (CST), general cognitive functioning, quality of life. This study was conducted at the Department of General Psychology, University of Padova, Via Venezia 8 Padova 35131, Italy.
1IRCSS Santa Lucia Foundation – Department of Clinical and Behavioral Neurology, Neuropsychiatry Laboratory Rome, Lazio, Italy.2 Department of General Psychol
ogy, University of Padova, Veneto, Italy. 3Centro Servizi Anni Sereni Scorzè (VE), Veneto, Italy. 4University of Parma, Italy.
Erika Borella. Department of General Psychology, University of Padova – Via Venezia 8 Padova – 35131 Italy. E-mail: erika.borella@unipd.it
Disclosure: The authors report no conflits of interest.
Received August 23, 2017. Accepted in final form October 31, 2017. This study was conducted at the Department of General Psychology, University of Padova, Via Venezia 8 Padova 35131, Italy. INTRODUCTION ments). Given the dubious results achieved to
date with pharmacological therapies,2,3 inter
est in psychosocial approaches to dementia
has increased considerably in recent years. Some psychosocial treatments focusing on
cognitive stimulation have already proved D D
ementia is a major global public health
challenge for our generation.1 Options
for the management of dementia include
pharmacological treatments and psychosocial
interventions (non-pharmacological treat 434
434
Cognitive stimulation therapy in vascular dementia
Piras et Dement Neuropsychol 2017 December;11(4):434-441 effective.4 To our knowledge, cognitive stimulation
therapy (CST) is the most widely used psychosocial
(non-pharmacological) treatment for people with mild-
to-moderate dementia,5 and has been studied in various
countries.6-8 A recent review9 on the use of the CST pro
gram found moderate evidence of its efficacy in sustain
ing general cognitive functioning and promoting quality
of life for people with mild-to-moderate dementia. The
CST protocol, and its Italian adaptation (CST-IT),7 is a
person-centered practice oriented towards the “person
hood” of the individual suffering from dementia.10,11 The
therapy is based on implicit learning, stimulating lan
guage and executive functioning with activities focusing
on orientation, reminiscence, new ideas, thoughts, and
associations to promote continuity between treatment
sessions. The CST is designed to create an environment
in which individuals have fun, learn, and strengthen
their abilities and their social relationships with other
members of a group and with the operators, preserving
their social and cognitive skills for as long as possible.12,13 notype. Indeed, its response to cognitive interventions,
as opposed to the irreversibility of the impairments and
loss of independence associated with AD, may be cru
cially important in clarifying the clinical phenotype of
VaD. The present study thus sought to ascertain the effi
cacy of CST-IT7,18 in individuals with VaD. INTRODUCTION Taking the bottom-up approach proposed by Aguirre
et al.,13 we conducted a single-blind, multicenter, ran
domized controlled trial to assess the potential benefits
of the CST-IT in the following domains: (i) cognition,
using traditional measures of general cognitive function
ing such as the Mini-Mental State Examination and the
Alzheimer’s Disease Assessment Scale - Cognitive sub
scale, and considering more specific cognitive functions,
such as short-term/working memory (measured with
the backward digit span task), and language abilities
(measured with the narrative language test); (ii) qual
ity of life, using the Quality of Life – Alzheimer’s Dis
ease scale; (iii) mood and behavior, examined with the
Cornell scale for depression in dementia, the social and
emotional loneliness scale, and the Neuropsychiatric
Inventory; and (iv) everyday life functioning, assessed
with the Disability Assessment for Dementia tool. All of the studies adopting the CST protocol to date
have included individuals with dementia of various eti
ologies, i.e. Alzheimer’s disease (AD), vascular dementia
(VaD), Parkinson’s disease, atypical/mixed forms, and
other (Lewy body dementia, mixed type dementia, Kor
sakov’s disease), or unspecified dementia. It therefore
remains to be seen whether the intervention’s efficacy
is dependent on the type of dementia considered. In line with previous studies, we only expected to
find improvements in the CST-IT group for traditional
measures of general cognitive functioning and per
ceived quality of life.9 Changes in performance in the
other dimensions were examined mainly to shed light
on whether, and to what extent, the CST program could
promote improvements in specific cognitive domains,
such as communication skills, which are probably tar
geted during the intervention.7,19,20 The diagnostic criteria for VaD have been dissociated
from those of AD14 only recently, and the term vascu
lar cognitive impairment (VCI) has been introduced15
to better reflect the full range of cognitive alterations
resulting from vascular factors. VaD (the most severe
form of VCI) is nonetheless, the second-most-common
form of dementia, and is associated with considerable
morbidity and mortality rates. The construct of the
vascular contribution to cognitive impairment and
dementia encompasses the whole spectrum of cogni
tive disorders associated with all forms of vascular brain
injury - and not only stroke, as was previously believed.16
Central to the mechanism behind the disease is the cru
cial role of cerebral blood vessels in keeping the brain
healthy. INTRODUCTION New experimental findings have revealed a
previously unrecognized functional and pathogenic syn
ergy between neurons, glia, and vascular cells.17 Despite
advances in our understanding, however, there are still
no approved pharmacological treatments for VCI and
VaD, and the development of alternative compensatory
interventions remains a challenge.3 The heterogene
ity of the cognitive disorders attributable to vascular
causes points to the need to better define the VaD phe 43
Piras et al. Cognitive stimulation therapy in vascular dementia METHODS
i i Participants. The sample was selected from Italian resi
dential homes for the elderly. Individuals were referred
by clinicians based on the NINDS-AIREN (National
Institute of Neurological Disorders and Stroke - Associa
tion Internationale pour la Recherche et l’Enseignement
en Neurosciences) criteria,21 which encompass: 1)
documented dementia (a decline in memory and in at
least two other domains of intellectual ability, resulting
in impaired activities of daily living); 2) evidence of
cerebrovascular disorder (focal neurological signs, gait
abnormalities, mood changes, psychomotor slowing,
and extrapyramidal signs) in clinical history, on clinical
examination, or on brain imaging (single strokes are
eligible if the other criteria apply); and 3) a “reason
able relation” between dementia and cerebrovascular
disorder (no early onset, a gradually worsening deficit
in memory or other cognitive functions in the absence 4
Piras et al. Cognitive stimulation therapy in vascular dementia 435 Dement Neuropsychol 2017 December;11(4):434-441 of focal lesions on neuroradiological scans). Eligi
bility for the CST was based on the criteria proposed
by Spector et al.,5 but restricted to individuals with: a
diagnosis of VaD in the mild to moderate range, i.e. a
score of at least 14 on the Mini-Mental State Examina
tion (MMSE);22 a satisfactory ability to understand and
communicate; no learning disability and/or current
physical illness or impairment reported in the person’s
clinical documents that might affect their participa
tion; no severe behavioral or psychological symptoms
of dementia that might affect their participation; no
diagnosis of comorbid psychiatric disorders; and no
ongoing anticholinesterase treatment. then repeating them in the reverse order. Each level con
sists of two sequences of the same length. The test ends
when the respondent fails to repeat both sequences. The
dependent variable was the highest level of difficulty
reached. Language. The narrative language test25 examines
textual competence and discourse information content,
and assesses discourse abilities as a function of the effec
tive communication of information. Participants have
to first describe a single figure (the picture “Picnic,”)26
and then sets of figures (two cartoon sequences).27 The
descriptions are recorded, transcribed verbatim, and
segmented using correct information unit analysis,27
followed by a quantitative textual analysis.28 The depen
dent variable was the sum of the items correctly and
accurately reported. Thirty-five eligible individuals were identified and
allocated to a group that was administered the Italian
version of the CST (CST-IT) (N = 21), or to an active
control group (N = 14). Mood Depression. The Cornell scale30 consists of 19 items
that assesses signs and symptoms of major depres
sion in individuals with dementia. Each item is rated
for severity on a scale from 0 (absent) to 2 (severe). The dependent variable was the sum of scores for the
19 items. Total scores below 6 indicate the absence of
significant depressive symptoms, while those above 10
indicate probable major depression, and those above 18
indicate definite major depression. METHODS
i i The two groups did not differ in
terms of age, gender, or years of education (Table 1). The
study was performed in accordance with the Declaration
of Helsinki (1964 and its later amendments). Quality of life
Q
l The Quality of Life – Alzheimer’s Disease scale (QoL-
AD)29 is a self-report questionnaire comprising 13
items, each rated on a 4-point scale from 1 (poor) to 4
(excellent). The dependent variable was the sum of all
items, where higher scores indicate better quality of life. Study design. The study was designed as a single-blind,
multicenter, randomized controlled study of CST-IT for
dementia. Activities of daily living h The Disability Assessment for Dementia (DAD)33,34
covers basic, instrumental and leisure activities in
10 areas, from personal hygiene to managing money
and medicines. The items in each area assess the indi
vidual’s ability regarding three dimensions: initiation
(the ability to decide and/or start an action), planning/
organization (problem-solving and decision-making), Behaviorh The Neuropsychiatric Inventory (NPI)32 assesses 10
behavioral disturbances in dementia patients. The total
score (frequency × severity), taken as a dependent vari
able, ranges from 1 to 12, where higher scores corre
spond to more frequent and more severe problems. Cognitive measures General cognitive status. The MMSE22 comprises items
that test temporal and spatial orientation, immediate
and delayed verbal memory, language, attention and
praxis. The dependent variable was the sum of the
scores (max. 30), corrected for age and education. The Alzheimer’s Disease Assessment Scale -Cognitive
subscale (ADAS-Cog)23 consists of 11 tasks that assess
orientation, memory, language, praxis, attention, and
other cognitive abilities. The dependent variable was the
sum of the scores (max. 70), with higher scores indicat
ing worse functioning. Loneliness. The 6-item social and emotional loneli
ness scale31 assesses emotional loneliness (items 1, 3,
and 4), and social loneliness (items 2, 5, and 6). Each
item can be scored from 1 (absolutely true) to 5 (abso
lutely not true). The dependent variable was the sum of
scores for the 6 items. Lower scores on the social scale
correspond to lower levels of social loneliness, whereas Memory. Short-term/working memory. The backward
digit span task24 entails participants listening to an
increasingly long sequence of numbers (from 2 to 8) and 436
Cognitive stimulation therapy in vascular dementia
Piras et al
Table 1. Characteristics of participants at baseline assessment by group. CST-IT group
N = 21 (15 females)
Active control group
N = 14 (13 females)
Between-group comparisons
M
SD
M
SD
Age
83.81
10.93
85.43
5.18
t = –0.51
p = 0.61
Education (years)
5.90
3.22
4.64
1.69
t = 1.34
p = 0.18
CST-IT: Cognitive stimulation therapy – Italian adaptation. Table 1. Characteristics of participants at baseline assessment by group. CST-IT: Cognitive stimulation therapy – Italian adaptation. Dement Neuropsychol 2017 December;11(4):434-441 lower scores on the emotional scale indicate higher lev
els of emotional loneliness. and actual performance (the ability to complete an
action). The scoring options are: 1 (ability to perform
the activity without help); 0 (inability to perform the
activity); or N/A (activities never performed before
the onset of the disease, or not performed in the past
2 weeks). The dependent variable was the total score
obtained by adding the rating for each question (‘yes’ =
1 point; ‘no’ = 0 points) expressed as a percentage. The
total number of items rated N/A were subtracted from
the denominator. Treatment procedure All participants in both groups attended 18 sessions: the
first two and last two were pre- and post-test individual
sessions; the other 14 were group sessions during which
the treatment group took part in the CST-IT program,
while the active control group was involved in alterna
tive educational activities (Table 2). RESULTS CST-IT intervention.7 The program consists of 14
structured group sessions, to be delivered twice a week
for 7 weeks in small groups (7-8 people), homogeneous
in terms of cognitive abilities; two trained operators act
as facilitators.5 Each session is organized as follows: (i)
Introduction: (10 min) with personalized welcome, dis
cussion on selecting the group’s name and a theme song,
discussion on the day, month, year, weather, time, and
the name and address of the center, using a whiteboard,
discussion regarding current news, and refreshments;
(ii) Main cognitive stimulation activities: (25 min),
adapted to the cognitive abilities of the participants, and
divided into level A (more difficult) and level B (easier)
(Table 2); (iii) Conclusion (10 min): thanking everyone
for attending and contributing, singing the theme song,
reminding everyone the day and time of the next ses
sion, and its content, and saying goodbye. Preliminary analyses to disclose baseline differences
within each group were conducted on participants’ pre-
test performance for each measure of interest. The t-test
results revealed no significant differences between the
two groups at baseline. The descriptive statistics and
results by group are presented in Table 3. To assess the gains derived from the treatment, uni
variate ANCOVAs were run for each measure of interest
using the post-test scores as the dependent variables,
the pre-test scores as covariates, and group (CST-IT vs
active control) as a between-subjects factor. The descrip
tive statistics for each measure of interest by group and
by session are given in Table 3, and the results of the
ANCOVAs are summarized in Table 4. In general, the analyses showed a main effect of the
covariate for all the measures of interest, except the
backward digit span task (Table 4). Akin to the CST-IT group, the active control group
engaged in educational activities (Table 2) for 14 ses
sions, again in small groups. The main effect of group was significant for the two
measures of general cognitive functioning, the MMSE Table 3. Descriptive statistics (means and standard deviations), and Student’s t-test results at baseline, for all measures of interest by group (CST-IT vs
active controls). Table 2. Activities conducted in the CST-IT and active control groups. Table 2. Activities conducted in the CST-IT and active control groups. • Pre-test session 1: Mini-Mental State Examination, Alzheimer’s Disease Assessment Scale - Cognitive subscale, Cornell
scale, Disability Assessment for Dementia, Backward digit span task. • Pre-test session 2: Neuropsychiatric Inventory, Narrative language test, Quality of Life - Alzheimer’s Disease scale
CST-IT group
Main themes/activities
Active control group
Main themes/activities
1
Physical games
Reading stories/newspapers, discussion
2
Sounds
Creative activities: painting, coloring, decorating, cooking. 3
My life
Reading stories/newspapers, discussion
4
Food
Creative activities: painting, coloring, decorating, cooking. 5
Current affairs
Reading stories/newspapers, discussion
6
Faces and places
Creative activities: painting, coloring, decorating, cooking. 7
Word associations
Reading stories/newspapers, discussion
8
Being creative
Creative activities: painting, coloring, decorating, cooking. 9
Categorizing objects
Reading stories/newspapers, discussion
10
Orientation
Creative activities: painting, coloring, decorating, cooking. 11
Using money
Reading stories/newspapers, discussion
12
Number games
Creative activities: painting, coloring, decorating, cooking. 13
Word games
Reading stories/newspapers, discussion
14
Team games
Creative activities: painting, coloring, decorating, cooking. • Post-test session 1: Mini-Mental State Examination, Alzheimer’s Disease Assessment Scale - Cognitive subscale,
Cornell scale, Disability Assessment for Dementia, Backward digit span task. • Post-test session 2: Neuropsychiatric Inventory, Narrative language test, Quality of Life - Alzheimer’s Disease scale. CST-IT: Cognitive stimulation therapy – Italian adaptation in which some of the original materials that did not fit with the Italian culture were changed, e.g. popular
English songs were replaced with popular Italian songs, traditional English food with traditional Italian food, etc., without changing the structure and activities of
the original program. • Pre-test session 1: Mini-Mental State Examination, Alzheimer’s Disease Assessment Scale - Cognitive
scale, Disability Assessment for Dementia, Backward digit span task. 437
Piras et al. Cognitive stimulation therapy in vascular dementia 437
Piras et al. Cognitive stimulation therapy in vascular dementia Dement Neuropsychol 2017 December;11(4):434-441 n therapy in vascular dementia Piras et al. RESULTS CST-IT group
Pre-test
Active controls
Pre-test
CST-IT group vs
Active controls
CST-IT group
Post-test
Active controls
Post-test
M
SD
M
SD
t (33)
p
M
SD
M
SD
MMSE
20.29
3.42
19.03
4.65
0.92
0.36
21.72
4.45
17.79
3.47
ADAS-Cog
31.85
10.25
31.22
12.14
0.16
0.86
28.99
9.36
32.29
9.75
Backward digit span task
2.88
0.62
2.25
0.97
1.97
0.08
3.19
1.33
2.17
0.39
Narrative language test
9.90
4.56
13.69
7.19
–1.86
0.10
12.35
5.28
12.77
6.34
Quality of life – Alzheimer’s Disease
Scale, people with dementia
25.05
9.78
28.43
7.82
–1.08
0.27
27.35
9.41
28.00
6.87
Quality of life – Alzheimer’s Disease
Scale, caregivers
28.94
7.42
26.92
6.01
0.79
0.44
29.89
7.01
27.75
6.20
Social and emotional loneliness scale
Total score
16.71
3.84
16.93
5.08
–0.14
0.88
18.33
5.09
18.64
4.89
Social and emotional loneliness scale:
Social loneliness
8.52
2.58
8.21
3.02
0.32
0.74
8.90
3.21
8.14
3.23
Social and emotional loneliness scale:
Emotional loneliness
8.19
2.18
8.71
2.84
–0.62
0.54
9.43
2.54
10.50
1.95
Cornell scale
5.90
5.45
4.14
2.54
1.13
0.27
4.19
4.33
3.50
4.01
NPI – total score
9.71
12.81
4.43
4.64
1.47
0.15
7.86
9.24
4.00
3.92
NPI – distress score
4.50
4.89
1.92
2.72
1.72
0.09
3.85
4.68
2.08
2.57
Disability Assessment for Dementia
42.33
26.45
38.44
21.61
0.36
0.72
43.10
22.63
41.92
18.98
CST-IT: Cognitive stimulation therapy – Italian adaptation; MMSE: Mini-Mental State Examination; ADAS-Cog: Alzheimer’s Disease Assessment Scale – Cognitive subscale; NPI: Neuropsychiatric
Inventory. cs (means and standard deviations), and Student’s t-test results at baseline, for all measures of interest by group (CST-IT vs tistics (means and standard deviations), and Student’s t-test results at baseline, for all measures of interest by group (CST-IT 438
Cognitive stimulation therapy in vascular dementia Piras et al. Dement Neuropsychol 2017 December;11(4):434-441 Table 4. Results of one-way ANCOVAs. RESULTS Main effects
F
(1,35)
η2
p
MMSE
covariate
40.59
0.56***
group
9.39
0.23**
CST-IT group > Active controls
ADAS-Cog
covariate
98.33
0.75***
group
5.20
0.14*
CST-IT group < Active controls
Backward digit span task
covariate
2.07
0.08
group
3.54
0.12^
CST group > Active controls
Narrative language test
covariate
27.05
0.47***
group
1.91
0.06
Quality of life - Alzheimer’s Disease Scale, people with dementia
covariate
107.01
0.77***
group
3.22
0.09^^
CST group > Active controls
Quality of life - Alzheimer’s Disease Scale, caregivers
covariate
122.55
0.82***
group
<1
Social and emotional loneliness scale: total score
covariate
68.54
0.68***
group
<1
Social and emotional loneliness scale: social loneliness
covariate
48.59
0.60***
group
<1
Social and emotional loneliness scale: emotional loneliness
covariate
37.98
0.54***
group
1.62
0.04
Cornell scale
covariate
32.97
0.51***
group
<1
NPI – total score
covariate
53.54
0.62***
group
4.98
0.63
NPI – distress score
covariate
118.09
0.80***
group
<1
Disability Assessment for Dementia
covariate
78.82
0.81***
group
<1
* p<0.05; ** p<0.01; ***p<0.001. ^ trend towards significance (p = 0.07); ^^ trend towards significance (p = 0.08). MMSE: Mini-Mental State Examination; ADAS-Cog: Alzheimer’s Disease As
sessment Scale – Cognitive subscale; NPI: Neuropsychiatric Inventory. Since the goal of evidence-based medicine is to pro
duce reliable and relevant data for the purpose of mak
ing clinical decisions for a particular patient, and given
the limited currently available evidence on the efficacy
of cognitive training and cognitive rehabilitation inter
ventions for people with mild AD or VaD, well-designed
single-blind RCTs to test the effectiveness of psychoso
cial non-pharmacological compensatory interventions
are crucial.37 and the ADAS-Cog, with the CST-IT group outperform
ing the active controls. A trend towards an improvement
for the CST-IT group, compared with the active controls,
emerged for short-term/working memory, as measured
by the backward digit span (p = 0.07), and for quality
of life, as measured by the QoL-AD, in individuals with
dementia only (p = 0.08). For all other measures, the two
groups did not differ significantly (Table 4). 43
Piras et al. Cognitive stimulation therapy in vascular dementia DISCUSSION The present study assessed the efficacy of the CST-IT
program in individuals with VaD in terms of improv
ing their cognition, including both general and specific
cognitive outcome measures (i.e. short-term memory
and language), and quality of life, mood, behavior and After controlling for vascular risk factors (e.g. hyper
tension, hypercholesterolemia), the effects of pharma
cological treatments for people with VaD have so far
proved to be modest.35,36 4
Piras et al. Cognitive stimulation therapy in vascular dementia 439 Dement Neuropsychol 2017 December;11(4):434-441 other forms of dementia43, the measure adopted here
may not have been sensitive enough to capture changes
in our participants’ communication skills. everyday life functioning. In fact, it has been suggested
that the endpoints developed for studies on AD are not
necessarily applicable to studies on VaD-related cogni
tive impairments. The primary outcome variables in VaD
trials should be multidimensional and include cognition,
global functioning, activities of daily living, and behav
ioral symptoms.38 In addition, given the specificity of
the cognitive profile in VaD,39 it is important to examine
the contribution of executive processes to cognition in
individuals with this disorder.38 No significant results emerged for the mood or
behavior domains. It is noteworthy to mark that studies
on the efficacy of CST have generally produced less con
sistent evidence regarding improvements in these areas
as compared with findings concerning cognition.3,9 As
expected, and in line with previous reviews,3,9 no changes
were identified in participants’ proficiency in activities
of daily living as measured by the DAD, after complet
ing the CST. Although nobody has suggested amending
measures of mood and behavioral changes when con
sidering individuals with VaD, it has been claimed that
special care should be taken in assessing this group on
activities of daily living because comorbidities (e.g. focal
neurological signs and symptoms) resulting from the
vascular disease itself might confound the issue.38 More
over, the potential particular impact of VaD on execu
tive functions needed in everyday life activities was
taken into account (because the tool used here distin
guishes between planning, sequencing, and completing
the tasks), but functional assessments developed spe
cifically for VCI44 might prove more effective in detect
ing treatment-induced changes in functional abilities. DISCUSSION In line with previous studies,9 our results showed
that the CST-IT group’s general cognitive functioning
improved, as measured by both the MMSE and the
ADAS-Cog, immediately after completing the treat
ment, while functioning in the active control group did
not improve. There was also a trend towards an improve
ment in the CST-IT group’s perceived quality of life, not
seen in the active controls. These findings support the
efficacy of CST in sustaining the global cognitive func
tioning of people with dementia, also in the particular
case of VaD. A trend towards a significant improvement was also
identified in the short-term/working memory domain,
as measured by the backward digit span task, but not
in the language domain. The task used in our study
demands cognitive control abilities to actively main
tain and manipulate memory representations,40 so the
former result (the effect on active short-term memory)
should be considered particularly informative regarding
the efficacy of treatment in VaD. Unlike the predomi
nant memory dysfunction in AD, the cognitive impair
ment that occurs in VCI and VaD is characterized by
attention/concentration and executive deficits36,41 and
the improvement seen in our participants (which might
have reached statistical significance in a larger sample)
on measures requiring executive attentional control sug
gests that the CST may target cognitive abilities spe
cifically associated with a vascular neuropsychological
impairment. When the contribution of vascular disease
was not partialed out, and the program’s efficacy was
tested in a sample with dementia of different etiolo
gies, the benefits of CST for short-term/working mem
ory were not so clear. On the other hand, some studies
found gains in the language domain after administering
the CST,17,19,42 an outcome due mainly to the structure
of the activities presented during the sessions, which
specifically stimulate language skills. The dependent
variable considered here was the number of correctly
reported items in a referential narrative task, however,
as VaD patients are not impaired on the semantic and
pragmatic levels of language processing as in AD and While the present results support the efficacy of
CST in individuals with vascular dementia, a number
of limitations of this study have to be acknowledged. First, the relatively small sample size makes it difficult
to generalize on the strength of our findings. 440
Cognitive stimulation therapy in vascular dementia Piras et al. REFERENCES 23. Rosen WG, Mohs RC, Davis KL. A new rating scale for Alzheimer’s
disease. Am J Psychiatry 1984;141:1356-64. 1. Alzheimer’s disease International, World Alzheimer Report 2015: The
global impact of dementia. 2015. Retrieved from https://www.alz.co.uk/
research/world-report-2015 1. Alzheimer’s disease International, World Alzheimer Report 2015: The
global impact of dementia. 2015. Retrieved from https://www.alz.co.uk/
research/world-report-2015 24. De Beni R, Borella E, Carretti B, Marigo C, Nava LA. BAC. In Portfolio
per la Valutazione del Benessere e Delle Abilità Cognitive nell’età Adulta e
Avanzata [The Assesment of Well-Being and Cognitive Abilities in Adult
hood and Aging]. Firenze, IT: Giunti OS; 2008. 2. Galimberti D, Scarpini E. Progress in Alzheimer’s disease. J Neurol. 2012;259(2):201-211. 3. Iadecola C. The Pathobiology of Vascular Dementia. Neuron 2013;80:
844-66. 25. Carlomagno S, Vorano L, Razzano C, Marini A. Assessing production
and comprehension of narratives in brain-damaged adults with commu
nicative deficits. Rivista di Psicologia Clinica 2013;1:76-96. 4. Woods B, Aguirre E, Spector AE, Orrell M. Cognitive stimulation to
improve cognitive functioning in people with dementia. Cochrane Data
base Syst Rev. 2012;15(2):CD005562. nicative deficits. Rivista di Psicologia Clinica 2013;1:76-96. 26. Kertesz A. The Western Aphasia Battery. Grune & Stratton: New York;
1982. 5. Spector A, Thorgrimsen L, Woods B, Orrell M. Making a difference: an
evidence-based group programme to offer Cognitive Stimulation Therapy
(CST) to people with dementia. Wandsworth, UK: Hawker Publications;
2006. 27. Nicholas LE, Brookshire RH. A system for quantifying the informativeness
and efficiency of the connected speech of adults with aphasia. J Speech
Lang Hear Res. 1993;36:338-50. 6. Apóstolo JLA, Cardoso DFB, Rosa A I, Paúl C. The effect of cognitive
stimulation on nursing home elders: A randomized controlled trial. J Nurs
Scholarsh. 2014;46(3):157-66. 28. Marini A, Carlomagno S. Analisi del Discorso e Patologia del Linguaggio. Springer Press; 1994. 29. Logsdon RG, Gibbons LE, McCurry SM, Teri L. Quality of life in Alzheim
er’s disease: patient and caregiver reports. J Mental Health Aging 1999;
5:21-32. 7. Capotosto E, Belacchi C, Gardini S, Faggian S, Piras, F, Mantoan V,
et al. Cognitive Stimulation Therapy in the Italian context: Its efficacy in
cognitive and non-cognitive measures in older adults with dementia. Int
J Geriatr Psychiatry 2017;32(3):331-40. 30. Alexopoulos GS, Abrams PC, Young RC, Shamoian CA. Cornell Scale
for depression in dementia. Biol Psychiatry 1988;23(3):271-84. 31. De Jong J, Van Tilburg T. A 6-item scale for overall emotional, and
social loneliness. Confirmatory tests on survey data. REFERENCES Res Aging 2006;
28(5):582-98. 8. Yamanaka K, Kawano Y, Noguchi D, Nakaaki S, Watanabe N, Amano T,
et al. Effects of cognitive stimulation therapy - Japanese version (CST-J)
for people with dementia: a single-blind, controlled clinical trial. Aging
Ment Health 2013;17(5):579-86. 32. Cummings JL, Mega M, Gray K, Rosenberg-Thompson S, Carusi DA,
Gornbein J. The Neuropsychiatric Inventory. Comprehensive assess
ment of psychopathology in dementia. Neurology 1994;44(12):2308-14. 9. Lobbia A, Carbone E, Faggian S, Gardini S, Piras F, Spector A, Borella
E. The efficacy of Cognitive Stimulation Therapy (CST) for people with
mild to moderate dementia: a review. Under review. 33. Gélinas I, Gauthier L, McIntyre MC, Gauthier S. Development of a func
tional measure for persons with Alzheimer disease: The disability assess
ment for dementia. Am J Occup Ther. 1999;53:471-81. 10. Kitwood T. Dementia Reconsidered: The Person Comes First. Buck
ingam, UK: Open University Press; 1997. ingam, UK: Open University Press; 1997. 11. Woods B. The person in dementia care. Generations 1999;23:35-9. 34. De Vreese LP, Caffarra P, Savarè R, Cerutti R, Franceschi M, Grossi
E. Functional disability in early Alzheimer’s disease – a validation study
of the Italian version of the disability assessment for dementia scale. Dement Geriatr Cogn Dis. 2008;25:186-94. 11. Woods B. The person in dementia care. Ge 12. Aguirre E, Hoare Z, Streater A, Spector A, Woods B, Hoe J, et al. Cogni
tive stimulation therapy (CST) for people with dementia - Who benefits
most? Int J Geriatr Psychiatry 2013;28(3):284-90. 13. Aguirre E, Spector A, Orrell M. Guidelines for adapting cognitive stimu
lation therapy to other cultures. Clinical Interv Aging. 2014;9:1003-7. 35. Baskys A, Hou AC. Vascular dementia: pharmacological treatment
approaches and perspectives. Clin Interv Aging. 2005;2:327-35. 36. Moorhouse P, Rockwood K. Vascular cognitive impairment: current
concepts and clinical developments. Lancet Neurol. 2008;7:246-55. 14. Román GC, Tatemichi TK, Erkinjuntti T, Cummings JL, Masdeu JC,
Garcia JH. Vascular dementia: diagnostic criteria for research studies. Report of the NINDS-AIREN International Workshop. Neurology
1993;43:250-60. 37. Olazarán J, Reisberg B, Clare L, Cruz I, Peña-Casanova J, del Ser T, et
al. Nonpharmacological Therapies in Alzheimer’s Disease: A Systematic
Review of Efficacy. Dement Geriatr Cogn Disord. 2010;30:161-78. ;
15. Hachinski VC, Bowler JV. Vascular dementia. Neurology 1993;43: 2159-61. Review of Efficacy. Dement Geriatr Cogn Disord. 2010;30:161 38. O’Brien JT, Erkinjuntti T, Reisberg B, Roman G, Sawada T, Pantoni L, et
al. Vascular cognitive impairment. Lancet Neurol. 2003; 2:89-98. 16. DISCUSSION Consider
ing the variable clinical course of VCI, the potential for
symptoms to improve and the non-linear progression of
VaD, a follow-up and a larger sample size may be needed
to conduct meaningful clinical trials. It has also been
suggested38 that disease staging should be applied to
VaD populations receiving treatment in order to address
the whole range of severity of this condition, and to
prove the effectiveness of a given treatment in multiple
domains of the related disability. The present study is
nonetheless the first to have examined the impact of
this evidence-based CST intervention on this particular
form of dementia. In conclusion, our results support the utility of CST,
and the importance of promoting its application and dif
fusion as an evidence-based intervention for sustaining
general cognitive functioning (at least) in individuals
with the vascular type of dementia. Since the main treat
ments currently available for VCI are merely preventive,
the knowledge emerging from the study should orient
future controlled trials on more homogeneous VCI sub
types, possibly stratified by clinical features, course of 440
Cognitive stimulation therapy in vascular dementia Piras et al. Dement Neuropsychol 2017 December;11(4):434-441 the disease, underlying pathophysiological conditions,
and magnetic resonance imaging measures. Faggian, designed the study and collected part of the
data. Elisa Salvalaio, designed the study and collected
part of the data. Simona Gardini, designed the study
and collected part of the data. Erika Borella, designed
the study, supervised data collection carried out the
statistical analyses, and wrote the paper. Author contribution. Federica Piras, assisted in designing
the study and wrote the paper. Elena Carbone, carried
out the statistical analyses, and wrote the paper. Silvia REFERENCES Hachinski VC, Lassen NA, Marshall J. Multi-infarct dementia. A cause of
mental deterioration in the elderly. Lancet 1974;2:207-10. 17. Iadecola C. The overlap between neurodegenerative and vascular factors
in the pathogenesis of dementia. Acta Neuropathol. 2010;120:287-96. 17. Iadecola C. The overlap between neurodegenerative and vas 39. Erkinjuntti T, Inzitari D, Pantoni L, Wallin A, Scheltens P, Rockwood K, et
al. Limitations of clinical criteria for the diagnosis of vascular dementia
in clinical trials. Is a focus on subcortical vascular dementia a solution? Ann N Y Acad Sci. 2000;903:262-72. 18. Gardini S, Pradelli S, Borella E. Intervento di stimolazione cognitiva effi
cace per le persone con demenza. Franco Angeli: Italy; 2015. 40. Engle RW, Tuholski SW, Laughlin JE, Conway AR. Working memory,
short-term memory, and general fluid intelligence: a latent-variable
approach. J Exp Psychol Gen. 1999;128:309-31. 19. Hall L, Orrell M, Stott J, Spector A. Cognitive stimulation therapy
(CST): Neuropsychological mechanisms of change. Int Psychogeriatr. 2013;25(03):479-89. 41. Lee AY. Vascular dementia. Chonnam Med J. 2001;47:66-71 20. Spector A, Orrell M, Woods B. Cognitive Stimulation Therapy (CST):
Effects on different areas of cognitive function for people with dementia. Int J Geriatr Psychiatry 2010; 25(12):1253-8. 42. Spector A, Thorgrimsen L, Woods B, Royan L, Davies S, Butterworth
M, Orrell M. Efficacy of an evidence-based cognitive stimulation therapy
programme for people with dementia. Randomised controlled trial. Br J
Psychiatry 2003;183:248-54. 21. Román GC, Tatemichi TK, Erkinjuntti T, Cummings JL, Masdeu JC,
Garcia JH, et al. Vascular dementia: diagnostic criteria for research
studies. Report of the NINDS-AIREN International Workshop. Neurology
1993;43:250-60. 43. Karantzoulis S, Galvin JE. Distinguishing Alzheimer’s disease from other
major forms of dementia. Expert Rev Neurother. 2011;11:1579-91. 22. Folstein ME, Folstein SE, McHugh PR. “Mini-Mental State”: A practical
method for grading the cognitive state of patients for the clinician. J
Psychiat. Res. 1975;12:189-98. 44. Pantoni L, Inzitari D, Wallin A. The matter of white matter: clinical and patho
physiological aspects of white matter disease related to cognitive decline
and vascular dementia. Academic Pharmaceutical Productions; 2000. 441
Piras et al. Cognitive stimulation therapy in vascular dementia 441
Piras et al. Cognitive stimulation therapy in vascular dementia
|
W3006652496.txt
| null |
en
|
Preferences for Sun Protection With a Self-Monitoring App: Protocol of a Discrete Choice Experiment Study
|
JMIR research protocols
| 2,020
|
cc-by
| 7,905
|
Zurich Open Repository and
Archive
University of Zurich
University Library
Strickhofstrasse 39
CH-8057 Zurich
www.zora.uzh.ch
Year: 2020
Preferences for sun protection with a self-monitoring app: protocol of a discrete
choice experiment study
Nittas, Vasileios ; Mütsch, Margot ; Puhan, Milo Alan
Abstract: BACKGROUND: The incidence of sun-exposure-related skin conditions, such as melanoma, is a gradually increasing and largely preventable public health problem. Simultaneously, the availability of mobile apps
that enable the self-monitoring of health behavior and outcomes is ever increasing. Inevitably, recent years have
seen an emerging volume of electronic patient-generated health data (PGHD), as well as their targeted application
across primary prevention areas, including sun protection and skin health. Despite their preventive potential, the
actual impact of these apps relies on the engagement of health care consumers, who are primarily responsible for
recording, sharing, and using their PGHD. Exploring preferences is a key step toward facilitating consumer engagement and ultimately realizing their potential. OBJECTIVE: This paper describes an ongoing research project
that aims to elicit the preferences of health care consumers for sun protection via app-based self-monitoring.
METHODS: A discrete choice experiment (DCE) will be conducted to explore how healthy consumers choose
between two alternative preventive self-monitoring apps. DCE development and attribute selection were built
on extensive qualitative work, consisting of the secondary use of a previously conducted scoping review, a rapid
review of reviews, 13 expert interviews, and 12 health care consumer interviews, the results of which are reported in this paper. Following D-optimality criteria, a fractional factorial survey design was generated. The
final DCE will be administered in the waiting room of a travel clinic, targeting a sample of 200 participants.
Choice data will be analyzed with conditional logit and multinomial logit models, accounting for individual participant characteristics. RESULTS: An ethics approval was waived by the Ethics Committee Zurich. The study
started in September 2019 and estimated data collection and completion is set for January 2020. Five two-level
attributes have been selected for inclusion in the DCE, addressing (1) data generation methods, (2) privacy control, (3) data sharing with general practitioner, (4) reminder timing, and (5) costs. Data synthesis, analysis, and
reporting are planned for January and February 2020. Results are expected to be submitted for publication by
February 2020. CONCLUSIONS: Our results will target technology developers, health care providers, and policy
makers, potentially offering some guidance on how to design or use sun-protection-focused self-monitoring apps
in ways that are responsive to consumer preferences. Preferences are ultimately linked to engagement and motivation, which are key elements for the uptake and success of digital health. Our findings will inform the design
of person-centered apps, while also inspiring future preference-eliciting research in the field of emerging and
complex eHealth services. INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID): PRR1-10.2196/16087.
DOI: https://doi.org/10.2196/16087
Posted at the Zurich Open Repository and Archive, University of Zurich
ZORA URL: https://www.zora.uzh.ch/191073
Journal Article
Published Version
The following work is licensed under a Creative Commons: Attribution 4.0 International (CC BY 4.0) License.
Originally published at:
Nittas, Vasileios; Mütsch, Margot; Puhan, Milo Alan (2020). Preferences for sun protection with a self-monitoring
app: protocol of a discrete choice experiment study. JMIR research protocols, 9(2):e16087.
DOI: https://doi.org/10.2196/16087
2
JMIR RESEARCH PROTOCOLS
Nittas et al
Protocol
Preferences for Sun Protection With a Self-Monitoring App:
Protocol of a Discrete Choice Experiment Study
Vasileios Nittas, BSc, MSc; Margot Mütsch, MPH, MSc, PhD; Milo Alan Puhan, MD, PhD
Epidemiology, Biostatistics and Prevention Institute, University of Zurich, Zurich, Switzerland
Corresponding Author:
Vasileios Nittas, BSc, MSc
Epidemiology, Biostatistics and Prevention Institute
University of Zurich
Hirschengraben 84
Zurich, 8001
Switzerland
Phone: 41 44 63 44946
Email: vasileios.nittas@uzh.ch
Abstract
Background: The incidence of sun-exposure-related skin conditions, such as melanoma, is a gradually increasing and largely
preventable public health problem. Simultaneously, the availability of mobile apps that enable the self-monitoring of health
behavior and outcomes is ever increasing. Inevitably, recent years have seen an emerging volume of electronic patient-generated
health data (PGHD), as well as their targeted application across primary prevention areas, including sun protection and skin
health. Despite their preventive potential, the actual impact of these apps relies on the engagement of health care consumers, who
are primarily responsible for recording, sharing, and using their PGHD. Exploring preferences is a key step toward facilitating
consumer engagement and ultimately realizing their potential.
Objective: This paper describes an ongoing research project that aims to elicit the preferences of health care consumers for sun
protection via app-based self-monitoring.
Methods: A discrete choice experiment (DCE) will be conducted to explore how healthy consumers choose between two
alternative preventive self-monitoring apps. DCE development and attribute selection were built on extensive qualitative work,
consisting of the secondary use of a previously conducted scoping review, a rapid review of reviews, 13 expert interviews, and
12 health care consumer interviews, the results of which are reported in this paper. Following D-optimality criteria, a fractional
factorial survey design was generated. The final DCE will be administered in the waiting room of a travel clinic, targeting a
sample of 200 participants. Choice data will be analyzed with conditional logit and multinomial logit models, accounting for
individual participant characteristics.
Results: An ethics approval was waived by the Ethics Committee Zurich. The study started in September 2019 and estimated
data collection and completion is set for January 2020. Five two-level attributes have been selected for inclusion in the DCE,
addressing (1) data generation methods, (2) privacy control, (3) data sharing with general practitioner, (4) reminder timing, and
(5) costs. Data synthesis, analysis, and reporting are planned for January and February 2020. Results are expected to be submitted
for publication by February 2020.
Conclusions: Our results will target technology developers, health care providers, and policy makers, potentially offering some
guidance on how to design or use sun-protection-focused self-monitoring apps in ways that are responsive to consumer preferences.
Preferences are ultimately linked to engagement and motivation, which are key elements for the uptake and success of digital
health. Our findings will inform the design of person-centered apps, while also inspiring future preference-eliciting research in
the field of emerging and complex eHealth services.
International Registered Report Identifier (IRRID): PRR1-10.2196/16087
(JMIR Res Protoc 2020;9(2):e16087) doi: 10.2196/16087
KEYWORDS
preventive medicine; mobile health; health informatics; health economics; patient preferences; discrete choice experiment
http://www.researchprotocols.org/2020/2/e16087/
XSL• FO
RenderX
JMIR Res Protoc 2020 | vol. 9 | iss. 2 | e16087 | p. 1
(page number not for citation purposes)
JMIR RESEARCH PROTOCOLS
Introduction
Background
As the mHealth market rapidly expands, digitally
self-monitoring our health and well-being is easier than ever
before. Inevitably, the volume of available electronic
patient-generated health data (PGHD) grows exponentially.
Defined as nonclinical health information, generated and
controlled by consumers, patients, and their designees, PGHD
are widely used across public health domains to facilitate
primary prevention and strengthen health promotion [1-4].
Mobile phones and wearables come as fully functional
measurement devices, accompanied by an abundance of apps
that collect PGHD and provide prevention-relevant feedback
[2,5]. Many of those apps are capable of capturing physical and
contextual signals, as well as communicating risks and
supporting behavior change [6]. With such an unprecedented
number of self-monitoring apps comes an equally unprecedented
need to understand how these should be designed and utilized
for successful primary prevention.
Mobile Self-Monitoring for Sun Protection
Serious skin conditions related to sun exposure, such as
melanoma, are on the rise. Melanoma—one form of skin
cancer—is a potentially fatal malignancy of the skin arising
from atypical melanocytes, primarily affecting young and
middle-aged population groups [7,8]. While disease onset
depends on multiple factors (eg, family history and genetics),
exposure to UV light (eg, sun and indoor tanning) is considered
a primary risk factor [7]. The global incidence of melanoma
indicates upward trends, with most rapid increases recorded in
western and Caucasian populations [7]. While the
epidemiological trends of melanoma indicate a very present and
most likely growing public health problem, targeted behavioral
change in relation to sun protection can mitigate much of its
burden [7]. With increasing popularity of mobile self-monitoring
across prevention areas, including weight loss, physical activity,
nutrition, smoking, alcohol consumption, and mental health,
the use of PGHD is gradually gaining popularity in sun
protection [1-4,9-13]. Mobile apps are designed to monitor
behavior (eg, sunbathing intensity, use of sunscreen, and use
of protective clothing), as well as environmental exposure (eg,
UV-light intensity), and to combine that with behavior change
techniques, such as tailored messages, sensitive reminders,
motivational feedback, gamification, and education [11-13].
Acknowledging the need for person-sensitive and personalized
primary prevention, the emergence of mobile self-monitoring
is a unique opportunity and resource in reducing sun-related
skin conditions, such as melanoma and other skin cancers.
An Emphasis on Health Care Consumer Preferences
A prerequisite of digital and mobile self-monitoring is the
motivation of consumers to engage with technology. This is
driven by individual, technical, social, and environmental
factors, such as personal motivation, appropriate use, long-term
engagement, and satisfaction [14,15]. While existing theories
identify overall drivers of motivations of technology
http://www.researchprotocols.org/2020/2/e16087/
XSL• FO
RenderX
Nittas et al
engagement, consumer preferences regarding concrete
characteristics of technology have been less explored [16,17].
When it comes to one’s health and well-being, as well as the
prevention of malaise (ie, discomfort), focusing on health care
consumer preferences is a central component of person-centered
care. Person-centeredness requires a full focus on the needs,
values, and desires of individuals, as well as their environments
and social contexts [18,19]. Evidence suggests that
person-centeredness can enhance satisfaction with, and
acceptance of, health services while ensuring engagement and
adherence [20]. Understanding preferences and their predictors
is key to developing acceptable self-monitoring technologies.
This study outlines the methodology and preparatory qualitative
results of a discrete choice experiment (DCE) that aims to elicit
consumer preferences for facilitating sun protection with
self-monitoring apps. Our findings target health app providers,
practitioners, citizens, and policy makers, aiming to guide a
more preference-sensitive development and use of
self-monitoring apps.
Aims
Our study aims to assess the relative importance of modifiable
elements of self-monitoring apps that focus on sun protection.
We have identified the following study objectives:
1.
2.
3.
Identify and explore which elements of self-monitoring
apps are deemed important by health care experts and health
care consumers (qualitative results).
Among those preidentified elements, elicit the relative
importance of health care consumer preferences (DCE
results).
Determine whether those preferences vary across participant
characteristics, including age, gender, education, health app
attitudes, and perceived health (DCE results).
Methods
Overview of Approach
A DCE is a robust survey-based methodology that enables the
elicitation of consumer preferences [21]. Rooted in
psychometrics and based on strong theoretical grounds, DCEs
have been widely used in economic research and are rapidly
gaining popularity within health care [21-23]. The technique’s
core assumption suggests that any good or service consists of
distinguishable characteristics, also known as attributes, from
which consumers derive utility [21,24]. Each attribute can take
alternative forms, often described as levels. The derived utility
varies with changing levels of these attributes. Individual choices
among alternatives of these characteristics are assumed to
indicate a person’s preferences, underlying values, and perceived
service utility [21,24]. Developing a DCE and selecting
appropriate attributes requires a range of preparatory qualitative
steps [25]. We conducted literature reviews and interviews with
health care consumers and experts, the results of which are
detailed in this study [22]. Figure 1 provides an overview of the
study’s methodological steps.
JMIR Res Protoc 2020 | vol. 9 | iss. 2 | e16087 | p. 2
(page number not for citation purposes)
JMIR RESEARCH PROTOCOLS
Nittas et al
Figure 1. Overview of the study’s methodology. DCE: discrete choice experiment.
Rationale for Using a Discrete Choice Experiment
Both mobile self-monitoring and primary prevention rely on
engaged health care consumers. Understanding their preferences
is, therefore, key to the successful development and use of
self-monitoring technology for preventive purposes. DCEs
enable the identification of those relative preferences by asking
consumers to choose between at least two versions (ie,
scenarios) of a good or service, each consisting of different
bundles of attribute levels [21,26]. Respondents are requested
to make repeated choices, which provides enough information
to statistically elicit those elements that are perceived to yield
the highest utility [26].
In having to choose one scenario over another, thus, being
requested to make trade-off choices, DCEs provide strong
indices of preferences and are gradually gaining popularity in
eHealth research [21]. For example, Cranen and colleagues used
the methodology to elicit preferences of chronically ill patients
regarding telerehabilitation, exploring attributes such as
physician communication modes, feedback provision, and the
use of digital monitoring tools [27]. Similarly, Kaambwa and
colleagues applied a DCE to investigate the telehealth
preferences of the elderly, identifying an inclination toward
http://www.researchprotocols.org/2020/2/e16087/
XSL• FO
RenderX
comprehensive and inexpensive eHealth services that target
those who face constrained access to traditional care [28]. Using
DCEs will enable us to identify which attributes of
prevention-focused self-monitoring apps are considered
important, as well as how consumer preferences are distributed
across them [26]. DCEs require a thorough and well-conducted
qualitative basis, which enables the selection of correct and
appropriate attributes. The qualitative work and its results are
presented in the following paragraphs.
Discrete Choice Experiment Scenario
Each DCE is framed around a hypothetical scenario that should
be relevant to the targeted topic and specific enough to allow
participants to make their choices accurately. For this DCE,
each participant will be asked to imagine a mobile prevention
app that targets sun protection and skin health by collecting
information on the duration and intensity of sun exposure,
followed by educational reminders on when and how to protect.
Discrete Choice Experiment Development: Methods
of Qualitative Preparatory Work
Prior to developing the DCE, we completed a thorough
three-step qualitative study, using existing literature and
JMIR Res Protoc 2020 | vol. 9 | iss. 2 | e16087 | p. 3
(page number not for citation purposes)
JMIR RESEARCH PROTOCOLS
stakeholder input to identify and select key attributes. We (1)
used the output of a previous scoping review, (2) conducted a
rapid review of systematic reviews on the use of electronic
PGHD for primary prevention, (3) conducted 13 semistructured
expert interviews, and (4) conducted 12 health care consumer
interviews [29]. The literature reviews and expert interviews
were merely meant to provide a preliminary basis of potential
attributes and, therefore, had a broader scope on electronic
self-monitoring for primary prevention. The health care
consumer interviews were framed around sun protection and
skin health, allowing us to identify attributes that are context
specific.
Literature Reviews
Both reviews aimed at mapping current evidence on the use of
electronic PGHD for prevention and health promotion, as well
as associated barriers and facilitators. The previously conducted
scoping review entailed searches in seven databases,
complemented by multiple additional and grey literature
searches, yielding 183 eligible primary studies [30]. The rapid
review was conducted in two databases—PubMed and the
Cochrane Database of Systematic Reviews—and was limited
to systematic reviews, yielding 13 eligible studies. Data
extraction was based on predefined templates and analysis was
thematic, with raw data being thematically clustered and
mapped. Multimedia Appendix 1 provides the rapid review’s
Preferred Reporting Items for Systematic Reviews and
Meta-Analyses (PRISMA) flowchart, inclusion criteria,
information on data extraction, and a list of all included studies
[31].
Expert Interviews
The 13 semistructured expert interviews were conducted
between March 12 and April 4, 2019, either face-to-face or via
Skype. They aimed to expand on and validate the list of
attributes identified in the literature. Our expert selection
criterion entailed that after completion of the interviews, each
of the following areas should be the expertise of at least one
interviewee: eHealth research, self-monitoring, digital
prevention, data science, eHealth and data ethics, primary
prevention, clinical practice, and citizen science. The number
of interviews was not prespecified but continued until saturation
was reached. Interviews were guided by semistructured
questions and a list of fixed topics that had to be addressed.
Informed by the literature reviews, those topics included the
following: (1) barriers and facilitators of digital self-monitoring
for primary prevention, (2) the technical aspects of these barriers
and facilitators, and (3) the broader components of
self-monitoring-based primary prevention interventions, such
as the use of behavior change techniques. In addition, experts
were provided with a list of 22 attributes identified in the
reviews and asked to comment on them, mention potentially
missing ones, and expand on those perceived as highly
important. Attributes were categorized according to the three
above-mentioned (1-3) or newly emerging themes. All experts
provided verbal consent for researchers to audiotape, transcribe,
and analyze the interviews. Recordings were deleted after
transcription, without linkages to any personal information.
http://www.researchprotocols.org/2020/2/e16087/
XSL• FO
RenderX
Nittas et al
Health Care Consumer Interviews
The 12 semistructured consumer interviews were conducted in
Zurich, Switzerland, between May 15 and May 24, 2019. They
aimed to capture which attributes of self-monitoring apps for
sun protection are perceived as most relevant by health care
consumers. Eligibility required a minimum age of 18 years and
no chronic conditions. Participants were recruited at the
University of Zurich Travel Clinic and selected purposively to
ensure age and gender balance. Interviews were guided by
semistructured questions and a list of fixed topics that had to
be addressed, including (1) barriers and facilitators of
self-monitoring for sun protection and skin health promotion
and (2) all attributes that were identified in the reviews and
expert interviews. In the interview’s first part, participants were
asked to discuss what would encourage or discourage them to
electronically collect their health data for sun protection
purposes. The second part consisted of a Likert-scale rating of
a list of attributes that were identified by the literature reviews
and expert interviews. Each interview required approximately
20 minutes and all participants provided prior written informed
consent, including a confirmation of all eligibility criteria. All
contact and personal identification information required for
recruitment and invitation was deleted immediately after
completion of the interviews and replaced by unique ID
numbers. Multimedia Appendix 2 provides the interview
schedule, participant demographics (eg, age and sex), and
selected interview quotes.
Analysis of Expert and Health Care Consumer
Interviews
All interviews were transcribed and analyzed with MAXQDA,
version 18.2.0 (VERBI Software) [32]. Our analysis followed
a hybrid approach of inductive and deductive coding [33]. Initial
deductive coding was based on the above-mentioned fixed topics
that guided the interviews, followed by an inductive, data-driven
generation of new codes and their connection to subthemes and
overarching themes. To ensure that our codes were
understandable and complete, a random sample of three
interviews was provided to an external coder who was instructed
to use our code system and analyze the interviews
independently. The coded interviews were compared and
inconsistencies discussed and resolved. This led to a wording
change for two codes, as well as the merging of two codes that
were not distinguishable.
Discrete Choice Experiment Development: Results of
Qualitative Preparatory Work
Combined, the literature reviews and 13 expert interviews
yielded a list of attributes that were categorized into six groups,
including (1) effort and support, (2) trust and control, (3) data
sharing, (4) technology and design, (5) prevention-related
content, and (6) incentives and disincentives. A more detailed
account of these categories is provided in Multimedia Appendix
3.
The 12 health care consumer interviews revealed six overall
attributes that were perceived as important to the use of
self-monitoring apps for sun protection; these included (1) costs,
(2) privacy and trust, (3) added value, (4) time and effort, (5)
JMIR Res Protoc 2020 | vol. 9 | iss. 2 | e16087 | p. 4
(page number not for citation purposes)
JMIR RESEARCH PROTOCOLS
user-friendliness, and (6) incentives. A more detailed account
is provided in Multimedia Appendix 3.
Selection of Attributes and Levels
The final list of all possible attributes, resulting from a synthesis
of review and interview findings, was reviewed by the research
team and one consulted health care consumer. The selection
was based on three overall criteria, as outlined by Bridges and
colleagues, including (1) research question relevance, (2)
decision context relevance, and (3) interrelations between
attributes. The group’s choice was additionally guided by the
Nittas et al
importance of attributes, primarily by health care consumers,
as well as by whether each attribute was realistic and could be
defined for a DCE. We considered a realistic attribute to be
compliant with current legal and policy regulations as well as
with existing technology and primary prevention services.
Consensus was reached on the following five attributes: (1)
method of data generation, (2) privacy control, (3) data sharing
with the general practitioner, (4) reminder timing, and (5) costs.
The process and flow of attribute selection are provided in
Figure 2.
Figure 2. Attribute selection process. DCE: discrete choice experiment.
We continued with assigning each attribute to realistic, relatable,
and understandable levels. As suggested by Bridges and
colleagues, we avoided ambiguous wording and tried to keep
the number of levels to a minimum [22]. All attributes, levels,
and descriptions were formulated in German and by native
speakers. The wording has been reviewed by a native German
speaker with a nonhealth background, ensuring ease of
understanding and clarity. All included attributes and assigned
levels are listed in Table 1.
We will additionally collect information on age, gender, highest
attained education, attitudes toward health apps, and perceived
health, as all of those factors have been previously associated
http://www.researchprotocols.org/2020/2/e16087/
XSL• FO
RenderX
with eHealth usage [34-37]. Perceived health will be measured
with a widely used single-item measure, asking participants to
rate their current general health on a 5-point scale from very
good to poor [38,39]. Attitude will be captured with one
reworded item on perceived usefulness, which was derived from
previous research and asks participants to indicate their
agreement on whether health apps are useful in promoting their
health; they will rate their agreement on a scale from totally
agree to totally disagree [40,41]. We chose single-item
questions to keep that part of the questionnaire short and simple,
considering that the actual DCE will require higher cognitive
and time resources.
JMIR Res Protoc 2020 | vol. 9 | iss. 2 | e16087 | p. 5
(page number not for citation purposes)
JMIR RESEARCH PROTOCOLS
Nittas et al
Table 1. Identified attributes, attribute levels, and prior assumptions.
Attributes
Descriptions
Attribute levels
Prior assumptions
1. Data generation
method
How would you like your data (a) no manual entry
to be collected?
(b) manual entry once a day
It is expected that most will prefer no manual
data entry, as that is linked to lower effort.
However, those who are more privacy-concerned might tend toward manual data entry.
2. Privacy control
If your data are being shared
with third-party commercial
entities, how would you like
to control when and with
whom your data are shared?
A clear a priori expectation is difficult to be
formulated. Those who are concerned about
constant push messages will likely prefer (a),
and those who are more privacy-concerned
will most likely choose (b).
(a) I will only receive information on potential data sharing with third parties once and
will be asked to provide informed consent
once
(b) I will be informed and provide consent
whenever my data are provided to third parties
3. Data sharing with
general practitioner
Would you like to share the
(a) yes
data collected with your gen- (b) no
eral practitioner, to be discussed at your next visit?
A clear a priori expectation is difficult to
formulate. We expect that those who have a
trusting general practitioner relationship and
lower perceived health will prefer (a).
4. Reminder timing
How would you prefer the
times and frequency of your
reminders to be set?
It is expected that most will prefer setting the
time and frequency of reminders themselves,
to avoid nuisance.
5. Costs
(a) I set the time and frequency of my reminders myself
(b) the app sets the times and frequency of
reminders automatically, based on my data
Are there any costs associated (a) free
with downloading the app
(b) one-time payment of 3 Swiss Francs
and, if yes, how high are
these?
Experimental Design and Choice Sets Selection
The combination of included attributes and levels, as shown in
Table 1, results in a full factorial design of 25=32 possible
distinct choice sets [42]. Limited time and cognitive capacities
deem such a large survey design unrealistic. In our context, the
time factor is particularly constraining, as questionnaires will
have to be answered in the waiting room, often during short
preconsultation windows. Although full factorial designs hold
desirable features, such as perfect orthogonality and balance, it
is common practice to use only subsets of those, known as
fractional factorial designs. Thus, we developed a fractional
factorial design, following D-optimality criteria and using R,
version 3.5.3 (The R Foundation), the open-source software for
statistical computing [43]. The 32 choice sets were reduced to
a fractional factorial sample of eight. The quality of responses
will be assessed through the inclusion of one additional choice
set that will be identical to a previous one. In line with the axiom
of completeness, participants that provide consistent answers
are expected to choose the same alternative twice [22,44]. We
will calculate percentages of inconsistent responses and assess
their distribution across individual participant characteristics,
as excluding participants is not recommended [22,45]. The final
questionnaire version will include eight original and one
repeated-choice set, yielding a total of nine choice sets.
Discrete Choice Experiment Piloting and Validity
The survey was piloted face-to-face with 8 participants recruited
from the University of Zurich Travel Clinic, ensuring
understandability and feasibility. Participants provided written
informed consent and received written detailed information on
the study’s purpose, the research question, and all attributes.
Participants were asked to complete the questionnaire in a
think-aloud manner. The DCE’s face validity was tested through
http://www.researchprotocols.org/2020/2/e16087/
XSL• FO
RenderX
It is expected that most will prefer a free app.
However, the cost might be accepted if combined with other desired attributes, such as
low effort and high privacy.
a discussion on the survey’s understandability and relevance,
as well as on perceived ease of answering. Participants were
asked to provide input on the survey’s wording, content, and
design, including the provided background material, pictures,
pictographs, and choice of colors. Particular attention was given
to the perceived relevance, formulation, and understandability
of the sun protection scenario. Time to completion was measured
to ensure that the survey is feasible within a given time frame.
We additionally assessed whether overall results are in line with
our hypotheses (see Table 1) [46]. The pilot was followed by
subsequent DCE adjustments, ensuring that the final
questionnaire is easy to comprehend and complete.
Participant Sampling, Recruitment, and Survey
Administration
Estimating an adequate DCE sample size is lacking scientific
consensus and remains a largely complex decision. Sample size
decisions ultimately depend on multiple factors, such as task
complexity, available resources, the sample’s composition, and
the target statistical precision of findings [22,47,48]. Although
parametric approaches have been proposed, when it comes to
identifying minimum sample sizes for specific hypotheses
testing, they are considered unsuitable [48-50]. That leaves
many researchers to use rule-of-thumb-based estimations [50].
Examples of those range from an overall sample of 100-300
participants to a minimum of 20 participants per choice set
[47,50]. Carefully considering available time and financial
resources, we utilized the rule of thumb proposed by Johnson
and Orme, which depends on the number of choice tasks,
alternatives, and analysis cells [51,52]. Aiming for a large
enough sample size to identify the main effects and interactions,
we will target a sample of 200 participants.
JMIR Res Protoc 2020 | vol. 9 | iss. 2 | e16087 | p. 6
(page number not for citation purposes)
JMIR RESEARCH PROTOCOLS
Participants will be recruited in the waiting room of the
University of Zurich walk-in Travel Clinic. The clinic constitutes
a hub for pretravel preventive consultation, as well as general
preventive services, including vaccinations. The approximately
20,000 annual consultations render the travel clinic an ideal
recruitment site. During recruitment days, everyone entering
the clinic will be informed about the possibility to participate.
Interested volunteers will be shortly briefed by a team member
on the study’s topic, purpose, and methodology; all this
information will be additionally provided in written form. The
completion of a DCE often poses high cognitive demands, for
which the first pages will solely serve the purpose of preparing
participants to answer the questionnaire. These first pages
include information on (1) the study aims, (2) the study topic
and key concepts, (3) detailed descriptions of each attribute,
complemented by pictographs, and (4) instructions on how a
DCE is filled out. Participants will have to confirm that all
eligibility criteria are fulfilled, including a minimum age of 18
years, no chronic conditions, and mobile phone ownership. As
the study does not include a follow-up session or any identifiable
personal information, a signed participant informed consent
form is not required. The survey will be administered in paper
form and completed in a quiet room within the clinic. During
recruitment and survey administration, a member of the staff
will be present to answer questions and resolve uncertainties.
We expect a survey completion time of about 10-15 minutes.
Analysis Plan
The DCE’s main end points are the individual preferences of
our participants, defined via the chosen attributes and their
levels. Those will be assessed using a conditional logit model.
This allows for the estimation of the relative importance of each
attribute over the remaining ones, using the retrieved mean
preference weights, as given by model coefficients [53]. To
achieve this, our analysis will assess changes in weights within
attributes—when changing from level (a) to level (b)—and the
relative sizes of those across attributes [53]. The conditional
logit model treats our findings as a function of the choice
alternatives. It was developed by McFadden in 1973 and has
been proven to be in accordance with the random utility theory,
dividing a respondent’s utility into a systematic and a random
element [53,54]. We will use the R package support.CEs to
convert our dataset to a form that is suitable for analysis [55].
We will additionally explore our data with mixed multinomial
logit (MMNL) models, treating our findings as a function of
choice alternatives and individual participant characteristics.
Expecting some preference heterogeneity, we chose an MMNL
model over a multinomial logit model, as the addition of the
error term can adjust for unobserved heterogeneity and adds to
the generalization of results [51]. In contrast to conditional logit
modeling, mixed logit models provide preference-weight
estimates and standard deviations of those, based on the
assumption of an underlying distribution of preference weights
and, thus, capturing preference heterogeneity among participants
[53].
We will additionally assess the amount of missing data. If the
percentage is considerable, we will use multiple imputation
using chained equations, using the R package mice, assessing
http://www.researchprotocols.org/2020/2/e16087/
XSL• FO
RenderX
Nittas et al
the impact of missingness on our findings. Goodness of fit will
be assessed by looking at the distribution of the residuals and
by calculating McFadden’s pseudo R-squared [53]. Exploring
variation, our analysis will test several individual characteristics
for inclusion in our models, including age, gender, education,
health app attitudes, and perceived health.
Results
An ethics approval was requested by the Ethics Committee
Zurich and was waived since this study does not fall under the
Swiss human research law, which only applies to clinical studies
that involve a certain level of risk, as well as the collection of
sensitive health data. The study began in September 2019, and
estimated data collection completion is set for January 2020.
Data synthesis, analysis, and reporting are planned for January
and February 2020. Results are expected to be submitted for
publication by February 2020.
Discussion
To the best of our knowledge, this is the first DCE that will
explore health care consumer preferences for sun protection
with self-monitoring apps. Our results will target technology
developers, health care providers, and policy makers, potentially
offering some guidance on how to design or use self-monitoring
apps in ways that are responsive to consumer preferences and,
thus, likely to maximize their engagement. Following good
practice, our methodology is based on extensive and carefully
designed qualitative work, ensuring that all included attributes
are relevant and relatable. Nonetheless, the inclusion of all
potentially relevant attributes is practically impossible in a DCE,
as its feasibility depends on the required cognitive workload,
as imposed by the number and nature of selected attributes.
These should ideally be practical and limited to the most
essential ones. This requires a careful and reasoned reduction
process to allow for a small and feasible number of included
attributes. Inevitably, this process includes trade-offs and the
exclusion of attributes that might be relevant for a considerable
proportion of the target population.
While DCEs constitute a robust and well-accepted approach for
preference exploration, their focus on a limited number of
variables inherently limits their capacity to capture broader
factors that influence preferences toward sun-protection-focused
self-monitoring apps. To fully understand the topic, our findings
need to be followed up by qualitative and mixed-method
research that will focus on understanding the individual and
contextual factors contributing to certain preferences. Finally,
as data collection will occur at the University of Zurich Travel
Clinic and is subject to certain participant inclusion criteria, the
data may not be fully generalizable to the entire Swiss or
European population.
Despite these limitations, the attributes we have identified cover
a considerable range of self-monitoring app characteristics,
which are modifiable and, thus, adjustable to health care
consumer preferences. Data collection effort, privacy, the flow
of information, the sensitivity of push messages, and costs are
all topics that are well-discussed in the literature and perceived
JMIR Res Protoc 2020 | vol. 9 | iss. 2 | e16087 | p. 7
(page number not for citation purposes)
JMIR RESEARCH PROTOCOLS
as key by health care consumers and experts, which signals their
potential to enhance the impact of self-monitoring apps.
Preferences are ultimately linked to engagement and motivation,
which are key elements for the uptake and success of any digital
Nittas et al
health approach. Ultimately, our work and findings will inform
the design of person-centered self-monitoring apps for sun
protection, while also inspiring future preference-eliciting
research in the field of emerging and complex eHealth services.
Acknowledgments
The authors thank all consulted experts, including Dr Andrea Farnham, Dr Andrea Horn, Prof Dr Barbara Stiglbauer, Dr Daniela
Gunz, Prof Dr Gerhard Schwabe, Mr Mathis Brauchbar, Mr Matthias Heuberger, Prof Dr Mathias Allemand, Mr Matthias von
Entreß-Fürsteneck, Dr Philipp Ackermann, Prof Dr Rudolf Marcel Füchslin, Mrs Sibylle Brunner, and Dr Julia Amann, for their
valuable thematic contributions. We would also like to thank the staff from the University of Zurich Travel Clinic for their constant
support throughout the recruitment and interview process. We would also like to thank Dr Julia Braun for her statistical support
during the design phase of our study.
Authors' Contributions
VN contributed to the conceptualization of the study and wrote and edited the manuscript. MM contributed to the conceptualization
of the study, supervised the entire process, and edited the manuscript. MAP contributed to the conceptualization of the study,
supervised the entire process, and edited the manuscript. The first author’s salary is funded by the Béatrice Ederer-Weber
Fellowship.
Conflicts of Interest
None declared.
Multimedia Appendix 1
Rapid review: Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flowchart, inclusion criteria,
information on data extraction, and list of included studies.
[DOCX File , 42 KB-Multimedia Appendix 1]
Multimedia Appendix 2
Health care consumer interviews: interview schedule, participant demographics, and sample quotes.
[DOCX File , 20 KB-Multimedia Appendix 2]
Multimedia Appendix 3
Attribute categories resulting from qualitative work.
[DOCX File , 14 KB-Multimedia Appendix 3]
References
1.
2.
3.
4.
5.
6.
7.
8.
Shapiro M, Johnston D, Wald J, Mon D. Patient-Generated Health Data. Washington, DC: Office of the National Coordinator
for Health Information Technology; 2012 Apr. URL: https://www.itheum.com/resources/PGHD.pdf [accessed 2019-12-30]
Semper HM, Povey R, Clark-Carter D. A systematic review of the effectiveness of smartphone applications that encourage
dietary self-regulatory strategies for weight loss in overweight and obese adults. Obes Rev 2016 Sep;17(9):895-906. [doi:
10.1111/obr.12428] [Medline: 27192162]
Ludwig K, Arthur R, Sculthorpe N, Fountain H, Buchan DS. Text messaging interventions for improvement in physical
activity and sedentary behavior in youth: Systematic review. JMIR Mhealth Uhealth 2018 Sep 17;6(9):e10799 [FREE Full
text] [doi: 10.2196/10799] [Medline: 30224335]
Lentferink AJ, Oldenhuis HK, de Groot M, Polstra L, Velthuijsen H, van Gemert-Pijnen JE. Key components in eHealth
interventions combining self-tracking and persuasive eCoaching to promote a healthier lifestyle: A scoping review. J Med
Internet Res 2017 Aug 01;19(8):e277 [FREE Full text] [doi: 10.2196/jmir.7288] [Medline: 28765103]
Cheatham SW, Stull KR, Fantigrassi M, Motel I. The efficacy of wearable activity tracking technology as part of a weight
loss program: A systematic review. J Sports Med Phys Fitness 2018 Apr;58(4):534-548. [doi:
10.23736/S0022-4707.17.07437-0] [Medline: 28488834]
Higgins JP. Smartphone applications for patients' health and fitness. Am J Med 2016 Jan;129(1):11-19. [doi:
10.1016/j.amjmed.2015.05.038] [Medline: 26091764]
Rastrelli M, Tropea S, Rossi CR, Alaibac M. Melanoma: Epidemiology, risk factors, pathogenesis, diagnosis and
classification. In Vivo 2014;28(6):1005-1011. [Medline: 25398793]
World Health Organization. 2016. International Statistical Classification of Diseases and Related Health Problems 10th
Revision URL: https://icd.who.int/browse10/2016/en [accessed 2019-12-30]
http://www.researchprotocols.org/2020/2/e16087/
XSL• FO
RenderX
JMIR Res Protoc 2020 | vol. 9 | iss. 2 | e16087 | p. 8
(page number not for citation purposes)
JMIR RESEARCH PROTOCOLS
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
Luxton DD, McCann RA, Bush NE, Mishkind MC, Reger GM. mHealth for mental health: Integrating smartphone technology
in behavioral healthcare. Prof Psychol Res Pr 2011 Dec;42(6):505-512. [doi: 10.1037/a0024485]
Kao C, Liebovitz DM. Consumer mobile health apps: Current state, barriers, and future directions. PM R 2017
May;9(5S):S106-S115. [doi: 10.1016/j.pmrj.2017.02.018] [Medline: 28527495]
Dunstone K, Conway C. There is an app for that! Communicating UV via the SunSmart app. In: Proceedings of the NIWA
UV Workshop. 2014 Presented at: NIWA UV Workshop; April 15-17, 2014; Auckland, New Zealand URL: https://niwa.
co.nz/sites/niwa.co.nz/files/Dunstone%20Conway_UV%20Workshop_2014_SunSmart%20app.pdf
Darlow S, Heckman C. Results from a tailored SMS and behavior-tracking pilot study on sun-safe behaviors in young
women. Health Educ Behav 2017 Dec;44(6):937-944. [doi: 10.1177/1090198117699507] [Medline: 28363246]
Rodrigues AM, Sniehotta FF, Birch-Machin MA, Olivier P, Araújo-Soares V. Systematic and iterative development of a
smartphone app to promote sun-protection among holidaymakers: Design of a prototype and results of usability and
acceptability testing. JMIR Res Protoc 2017 Jun 12;6(6):e112 [FREE Full text] [doi: 10.2196/resprot.7172] [Medline:
28606892]
Holden RJ, Karsh BT. A theoretical model of health information technology usage behaviour with implications for patient
safety. Behav Inf Technol 2009 Jan;28(1):21-38 [FREE Full text] [doi: 10.1080/01449290601138245]
Or CK, Karsh B. A systematic review of patient acceptance of consumer health information technology. J Am Med Inform
Assoc 2009;16(4):550-560 [FREE Full text] [doi: 10.1197/jamia.M2888] [Medline: 19390112]
Buchwald A, Letner A, Urbach N, von Entreß-Fürsteneck M. Towards explaining the willingness to disclose personal
self-tracking data to service providers. In: Proceedings of the 25th European Conference on Information Systems. 2017
Presented at: 25th European Conference on Information Systems; June 5-10, 2017; Guimarães, Portugal.
Pfeiffer J, von Entress-Fuersteneck M, Urbach N, Buchwald A. Quantify-me: Consumer acceptance of wearable self-tracking
devices. In: Proceedings of the 24th European Conference on Information Systems. 2016 Presented at: 24th European
Conference on Information Systems; June 12-15, 2016; Istanbul, Turkey.
Epstein RM, Street RL. The values and value of patient-centered care. Ann Fam Med 2011;9(2):100-103 [FREE Full text]
[doi: 10.1370/afm.1239] [Medline: 21403134]
Carey M, Noble N, Mansfield E, Waller A, Henskens F, Sanson-Fisher R. The role of eHealth in optimizing preventive
care in the primary care setting. J Med Internet Res 2015 May 22;17(5):e126 [FREE Full text] [doi: 10.2196/jmir.3817]
[Medline: 26001983]
Rathert C, Wyrwich MD, Boren SA. Patient-centered care and outcomes: A systematic review of the literature. Med Care
Res Rev 2013 Aug;70(4):351-379. [doi: 10.1177/1077558712465774] [Medline: 23169897]
Ryan M, Farrar S. Using conjoint analysis to elicit preferences for health care. BMJ 2000 Jun 03;320(7248):1530-1533
[FREE Full text] [doi: 10.1136/bmj.320.7248.1530] [Medline: 10834905]
Bridges JF, Hauber AB, Marshall D, Lloyd A, Prosser LA, Regier DA, et al. Conjoint analysis applications in health--a
checklist: A report of the ISPOR Good Research Practices for Conjoint Analysis Task Force. Value Health 2011
Jun;14(4):403-413 [FREE Full text] [doi: 10.1016/j.jval.2010.11.013] [Medline: 21669364]
Clark MD, Determann D, Petrou S, Moro D, de Bekker-Grob EW. Discrete choice experiments in health economics: A
review of the literature. Pharmacoeconomics 2014 Sep;32(9):883-902. [doi: 10.1007/s40273-014-0170-x] [Medline:
25005924]
Weymann D, Veenstra DL, Jarvik GP, Regier DA. Patient preferences for massively parallel sequencing genetic testing of
colorectal cancer risk: A discrete choice experiment. Eur J Hum Genet 2018 Sep;26(9):1257-1265 [FREE Full text] [doi:
10.1038/s41431-018-0161-z] [Medline: 29802320]
Coast J, Horrocks S. Developing attributes and levels for discrete choice experiments using qualitative methods. J Health
Serv Res Policy 2007 Jan;12(1):25-30. [doi: 10.1258/135581907779497602] [Medline: 17244394]
Ryan M, Bate A, Eastmond CJ, Ludbrook A. Use of discrete choice experiments to elicit preferences. Qual Health Care
2001 Sep;10 Suppl 1:i55-i60 [FREE Full text] [doi: 10.1136/qhc.0100055] [Medline: 11533440]
Cranen K, Groothuis-Oudshoorn CG, Vollenbroek-Hutten MM, IJzerman MJ. Toward patient-centered telerehabilitation
design: Understanding chronic pain patients' preferences for Web-based exercise telerehabilitation using a discrete choice
experiment. J Med Internet Res 2017 Jan 20;19(1):e26 [FREE Full text] [doi: 10.2196/jmir.5951] [Medline: 28108429]
Kaambwa B, Ratcliffe J, Shulver W, Killington M, Taylor A, Crotty M, et al. Investigating the preferences of older people
for telehealth as a new model of health care service delivery: A discrete choice experiment. J Telemed Telecare 2017
Mar;23(2):301-313. [doi: 10.1177/1357633X16637725] [Medline: 26985004]
Nittas V, Mütsch M, Ehrler F, Puhan MA. Electronic patient-generated health data to facilitate prevention and health
promotion: A scoping review protocol. BMJ Open 2018 Aug 10;8(8):e021245 [FREE Full text] [doi:
10.1136/bmjopen-2017-021245] [Medline: 30099392]
Nittas V, Lun P, Ehrler F, Puhan MA, Mütsch M. Electronic patient-generated health data to facilitate disease prevention
and health promotion: Scoping review. J Med Internet Res 2019 Oct 14;21(10):e13320 [FREE Full text] [doi: 10.2196/13320]
[Medline: 31613225]
http://www.researchprotocols.org/2020/2/e16087/
XSL• FO
RenderX
Nittas et al
JMIR Res Protoc 2020 | vol. 9 | iss. 2 | e16087 | p. 9
(page number not for citation purposes)
JMIR RESEARCH PROTOCOLS
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
53.
54.
55.
Moher D, Shamseer L, Clarke M, Ghersi D, Liberati A, Petticrew M, PRISMA-P Group. Preferred reporting items for
systematic review and meta-analysis protocols (PRISMA-P) 2015 statement. Syst Rev 2015 Jan 01;4:1 [FREE Full text]
[doi: 10.1186/2046-4053-4-1] [Medline: 25554246]
MAXQDA. URL: https://www.maxqda.com [accessed 2019-05-15]
Fereday J, Muir-Cochrane E. Demonstrating rigor using thematic analysis: A hybrid approach of inductive and deductive
coding and theme development. Int J Qual Methods 2016 Nov 29;5(1):80-92 [FREE Full text] [doi:
10.1177/160940690600500107]
Kontos E, Blake KD, Chou WS, Prestin A. Predictors of eHealth usage: Insights on the digital divide from the Health
Information National Trends Survey 2012. J Med Internet Res 2014 Jul 16;16(7):e172 [FREE Full text] [doi:
10.2196/jmir.3117] [Medline: 25048379]
Carroll JK, Moorhead A, Bond R, LeBlanc WG, Petrella RJ, Fiscella K. Who uses mobile phone health apps and does use
matter? A secondary data analytics approach. J Med Internet Res 2017 Apr 19;19(4):e125 [FREE Full text] [doi:
10.2196/jmir.5604] [Medline: 28428170]
Borycki EM, Kushniruk AW, Kuziemsky CE, Nøhr C, editors. Context Sensitive Health Informatics: Many Places, Many
Users, Many Contexts, Many Uses. Amsterdam, the Netherlands: IOS Press; 2015.
Bender MS, Choi J, Arai S, Paul SM, Gonzalez P, Fukuoka Y. Digital technology ownership, usage, and factors predicting
downloading health apps among Caucasian, Flipino, Korean, and Latino Americans: The Digital Link to Health Survey.
JMIR Mhealth Uhealth 2014 Oct 22;2(4):e43 [FREE Full text] [doi: 10.2196/mhealth.3710] [Medline: 25339246]
Bowling A. Just one question: If one question works, why ask several? J Epidemiol Community Health 2005
May;59(5):342-345 [FREE Full text] [doi: 10.1136/jech.2004.021204] [Medline: 15831678]
Eriksson I, Undén AL, Elofsson S. Self-rated health. Comparisons between three different measures. Results from a
population study. Int J Epidemiol 2001 Apr;30(2):326-333. [doi: 10.1093/ije/30.2.326] [Medline: 11369738]
Cho J. The impact of post-adoption beliefs on the continued use of health apps. Int J Med Inform 2016 Mar;87:75-83. [doi:
10.1016/j.ijmedinf.2015.12.016] [Medline: 26806714]
Davis FD, Bagozzi RP, Warshaw PR. User acceptance of computer technology: A comparison of two theoretical models.
Manage Sci 1989 Aug;35(8):982-1003. [doi: 10.1287/mnsc.35.8.982]
Hensher DA, Rose JM, Greene WH. Applied Choice Analysis. A Primer. Cambridge, UK: Cambridge University Press;
2005.
Aizaki H, Nishimura K. Design and analysis of choice experiments using R: A brief introduction. Agric Inf Res
2008;17(2):86-94 [FREE Full text] [doi: 10.3173/air.17.86]
Rezaei A, Patterson Z. Detecting Non-Transitive, Inconsistent Responses in Discrete Choice Experiments. Quebec City,
QC: Interuniversity Research Centre on Enterprise Networks, Logistics and Transportation (CIRRELT); 2015 Jul. URL:
https://www.cirrelt.ca/DocumentsTravail/CIRRELT-2015-30.pdf [accessed 2019-12-30]
Janssen EM, Hauber AB, Bridges JFP. Conducting a discrete-choice experiment study following recommendations for
good research practices: An application for eliciting patient preferences for diabetes treatments. Value Health 2018
Jan;21(1):59-68 [FREE Full text] [doi: 10.1016/j.jval.2017.07.001] [Medline: 29304942]
Ryan M, Major K, Skåtun D. Using discrete choice experiments to go beyond clinical outcomes when evaluating clinical
practice. J Eval Clin Pract 2005 Aug;11(4):328-338. [doi: 10.1111/j.1365-2753.2005.00539.x] [Medline: 16011645]
Lancsar E, Louviere J. Conducting discrete choice experiments to inform healthcare decision making: A user's guide.
Pharmacoeconomics 2008;26(8):661-677. [doi: 10.2165/00019053-200826080-00004] [Medline: 18620460]
Louviere JJ, Hensher DA, Swait JD. Stated Choice Methods: Analysis and Applications. Cambridge, UK: Cambridge
University Press; 2000.
Rose JM, Bliemer MCJ. Sample size requirements for stated choice experiments. Transportation 2013 Feb 1;40(5):1021-1041.
[doi: 10.1007/s11116-013-9451-z]
de Bekker-Grob EW, Donkers B, Jonker MF, Stolk EA. Sample size requirements for discrete-choice experiments in
healthcare: A practical guide. Patient 2015 Oct;8(5):373-384 [FREE Full text] [doi: 10.1007/s40271-015-0118-z] [Medline:
25726010]
Orme BK. Getting Started with Conjoint Analysis: Strategies for Product Design and Pricing Research. Madison, WI:
Research Publishers; 2010.
Johnson R, Orme B. Getting the Most from CBC. Sequim, WA: Sawtooth Software; 2003. URL: https://www.
sawtoothsoftware.com/downloadPDF.php?file=cbcmost.pdf [accessed 2019-12-30]
Hauber AB, González JM, Groothuis-Oudshoorn CG, Prior T, Marshall DA, Cunningham C, et al. Statistical methods for
the analysis of discrete choice experiments: A report of the ISPOR Conjoint Analysis Good Research Practices Task Force.
Value Health 2016 Jun;19(4):300-315 [FREE Full text] [doi: 10.1016/j.jval.2016.04.004] [Medline: 27325321]
McFadden D. Conditional logit analysis of qualitative choice behavior. In: Zarembka P, editor. Frontiers in Econometrics.
New York, NY: Academic Press; 1973:105-142.
Aizaki H. Basic functions for supporting an implementation of choice experiments in R. J Stat Softw 2012 Sep;50(Code
Snippet 2):1-24 [FREE Full text] [doi: 10.18637/jss.v050.c02]
http://www.researchprotocols.org/2020/2/e16087/
XSL• FO
RenderX
Nittas et al
JMIR Res Protoc 2020 | vol. 9 | iss. 2 | e16087 | p. 10
(page number not for citation purposes)
JMIR RESEARCH PROTOCOLS
Nittas et al
Abbreviations
DCE: discrete choice experiment
MMNL: mixed multinomial logit
PGHD: patient-generated health data
PRISMA: Preferred Reporting Items for Systematic Reviews and Meta-Analyses
Edited by G Eysenbach; submitted 02.09.19; peer-reviewed by M Falahee, R Steffen; comments to author 23.10.19; revised version
received 15.11.19; accepted 26.11.19; published 08.02.20
Please cite as:
Nittas V, Mütsch M, Puhan MA
Preferences for Sun Protection With a Self-Monitoring App: Protocol of a Discrete Choice Experiment Study
JMIR Res Protoc 2020;9(2):e16087
URL: http://www.researchprotocols.org/2020/2/e16087/
doi: 10.2196/16087
PMID:
©Vasileios Nittas, Margot Mütsch, Milo Alan Puhan. Originally published in JMIR Research Protocols
(http://www.researchprotocols.org), 09.02.2020. This is an open-access article distributed under the terms of the Creative Commons
Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction
in any medium, provided the original work, first published in JMIR Research Protocols, is properly cited. The complete bibliographic
information, a link to the original publication on http://www.researchprotocols.org, as well as this copyright and license information
must be included.
http://www.researchprotocols.org/2020/2/e16087/
XSL• FO
RenderX
JMIR Res Protoc 2020 | vol. 9 | iss. 2 | e16087 | p. 11
(page number not for citation purposes)
|
|
https://openalex.org/W2949107990
|
https://zenodo.org/records/3257886/files/BISS_article_37276.pdf
|
English
| null |
Big Data Knowledge of Major Lineages of Life and Priorities for Genomic Research
|
Biodiversity Information Science and Standards
| 2,019
|
cc-by
| 603
|
© Coddington J et al. This is an open access article distributed under the terms of the Creative Commons Attribution License (CC
BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are
credited. Corresponding author: Jonathan Coddington (coddington@si.edu) Corresponding author: Jonathan Coddington (coddington@si.edu) Received: 13 Jun 2019 | Published: 19 Jun 2019 Received: 13 Jun 2019 | Published: 19 Jun 2019
Citation: Coddington J, Barker K, Droege G, Seberg O (2019) Big Data Knowledge of Major Lineages of Life and
Priorities for Genomic Research. Biodiversity Information Science and Standards 3: e37276. h
//d i
/10 3897/bi
3 37276 Jonathan Coddington , Katharine Barker , Gabriele Droege , Ole Seberg
‡
‡
§
| ‡ National Museum of Natural History, Smithsonian Institution, Washington, D.C., United States of America
§ Botanic Garden and Botanical Museum Berlin, Berlin, Germany
| Natural History Museum Denmark, Copenhagen, Denmark Biodiversity Information Science and Standards 3: e37276 Coddington J et al Coddington J et al 2 2 important questions at deeper clade levels: What is it? Where is it? What do we know
about it? What do we know about its genome? Approximately 4,500 eukaryotic genomes
have been sequenced. The converse of what we know is what we do not know, another
meaning of “dark taxa.” We can use the distribution and density of big data at deeper clade
levels (families, genera) to quantitatively analyze “dark taxa” and therefore
to strategically optimize knowledge and preservation of biodiversity at a global scale. Technicalities of the quantitative prioritization scheme are debatable, but some initial,
simple scoring systems can help to prioritize lineages for collection and genetic research
so as to most efficiently illuminate regions in the tree of life that that are neither
preserved, imaged, geo-located, studied, nor known genomically. This analysis presents
criteria and goals for collaborating to build a global genomic collection to maximize efficient
acquisition of biodiversity genomic knowledge, and identifies the most valuable and highest
priority taxa for genomic research. Presenting author Jonathan A. Coddington Keywords Families of Life, barcode, GBIF, EOL, NCBI, BOLD, GGBN, phylogeny, phylogenetic
diversity, genome Abstract Genomic science is revolutionizing and accelerating biodiversity research. For collections-
based institutions to continue to lead and support biodiversity research, they must adapt to
this new reality. Simultaneously, “big data” is accumulating so rapidly that we have
unprecedented capacity to plan strategically to use genomics to advance basic and applied
science on multiple fronts. For example, seven “big data” sources have the following
numbers of records (2018 data): Global Biodiversity Information Facility (GBIF), ~1B;
Biodiversity Heritage Library (BHL), ~3.6M; National Center for Biotechnology Information
(NCBI), ~220M; Open Tree of Life (OToL), 1.9M; Barcode of Life Data System (BOLD),
~6.3M; Encyclopedia of Life (EOL), ~99K; Global Genome Biodiversity Network (GGBN),
~2M. Collectively, they offer more than 1.2B records on biodiversity. At the scale of species
(~2M described, multiple millions undescribed), these data are still too sparse to permit
comprehensive conclusions. At the scale of families (i.e. deeper clades of life), the situation
is far more promising: about 9,911 families are known, and relatively few are discovered
each year. This suggests that at the family rank (and above), our knowledge of life on Earth
is reasonably complete. Approximately 160,000 valid and accepted genera exist, but
certainly many new genera await discovery and description. Genomics is the fastest way to
group species into more inclusive lineages such as genera and families, and is certainly
faster than traditional alpha taxonomy. Synergistically, these “big data” answer four Families of Life, barcode, GBIF, EOL, NCBI, BOLD, GGBN, phylogeny, phylogenetic
diversity, genome Presented at Biodiversity_Next 2019
|
https://openalex.org/W4366826904
|
https://www.mdpi.com/2076-3417/13/9/5268/pdf?version=1682238483
|
English
| null |
A Time-Scale Varying Finite Difference Method for Analyzing the Influence of Rainfall and Water Level on the Stability of a Bank Slope
|
Applied sciences
| 2,023
|
cc-by
| 7,471
|
Article
A Time-Scale Varying Finite Difference Method for Analyzing
the Influence of Rainfall and Water Level on the Stability of a
Bank Slope Guorong Zhang 1,2, Guangyin Lu 1, Chengzhi Xia 1,3, Dongxin Bai 1 and Taoying Liu 1,* 1
School of Resource and Safety Engineering, Central South University, Changsha 410083, China
2
Zhejiang Jining Expressway Co., Ltd., Anji 313300, China
3
School of Civil Engineering, Tongji University, Shanghai 200092, China
*
Correspondence: taoying@csu.edu.cn Abstract: A time-scale varying finite difference method (TSFDM) was proposed to analyze the
influence of rainfall infiltration and water level fluctuation on the stability of a bank slope in the
Yunnan Pulang tailings pond. The stability of the slope obtained by the TSFDM and the limit
equilibrium method (LEM) were compared. The result indicates that the TSFDM can obtain reliable
spatiotemporal variation features of geotechnical parameters of the bank slope under different
conditions. The water level fluctuation has a greater impact on the stability than rainfall infiltration. In addition, the safety factor decreases as the shear strain rate increases on the weathering interface. The safety factors under all conditions are smaller than one, so corresponding measures should be
taken. The research provides references for preventing disasters caused by bank slope failure. Keywords: tailings pond bank slope; water level; rainfall; finite difference method Many methods have been proposed to evaluate the stability of bank slope, such as limit
equilibrium method (LEM) and limit difference method (LDM). A stability analysis based
on LEM cannot constrain the evolution characteristics of the slope stress and strain during
slope failure [1] (Chen, 2014). A time-varying analysis method was proposed to analyze
the temporal and spatial changes of the mechanical properties of engineer objects [2]. This
method focuses on the change of pore water pressure over time, but pays little attention to
the change of safety factors. Zhang et al. [3] established a time-varying analysis model for
slope stability on the basis of Geostuido, and replaced the traditional independent safety
factors with the safety factor curve, but using the safety factor alone can hardly assess
the accuracy of the results. Jiang et al. [4] improved the FISH function in the FLAC3D
software, and studied the three-dimensional evolution features of the seepage field of the
unsaturated slope under rainfall infiltration, but little on the slope mechanism. At present,
many secondary developments have been added to FLAC3D software, and they concentrate
on establishing constitutive models of rock-soil mass [5–7], but rarely consider the time
series analysis of slope safety factors. These related studies either use LEM to explore the
time-varying characteristics of slopes, or use the finite difference method to calculate the
safety factor of the slope at a single time. However, the finite difference method is rarely
used to analyze the stability in a long time span and with a small time interval, especially
for tailing pond bank slopes. Citation: Zhang, G.; Lu, G.; Xia, C.;
Bai, D.; Liu, T. A Time-Scale Varying
Finite Difference Method for
Analyzing the Influence of Rainfall
and Water Level on the Stability of a
Bank Slope. Appl. Sci. 2023, 13, 5268. https://doi.org/10.3390/app13095268
Academic Editor: Daniel Dias
Received: 10 March 2023
Revised: 18 April 2023
Accepted: 19 April 2023
Published: 23 April 2023 Citation: Zhang, G.; Lu, G.; Xia, C.;
Bai, D.; Liu, T. A Time-Scale Varying
Finite Difference Method for
Analyzing the Influence of Rainfall
and Water Level on the Stability of a
Bank Slope. Appl. Sci. 2023, 13, 5268. https://doi.org/10.3390/app13095268 Received: 10 March 2023
Revised: 18 April 2023
Accepted: 19 April 2023
Published: 23 April 2023 Copyright:
© 2023 by the authors. Licensee MDPI, Basel, Switzerland. applied
sciences applied
sciences applied
sciences 2.3. Strength Reduction Method The strength reduction method assumes that the slope failure shows as the shear
failure of rock mass if the external load remains the same. So the geotechnical parameters
are adjusted as follows: ci = c0
Ki
, ϕi = arctantan ϕ0
Ki
(5) (5) where c0 and ϕ0 are the initial cohesion and internal friction angle of rock mass, respectively;
ci and ϕi are those with the reduction coefficient Ki. The implementation of strength
reduction method varies over software, and its implementation process in TSFDM can be
found in Song et al. [10]. 2.2. Principle of TSFDM In the TSFDM, Darcy’s law is used for analyzing the changes of the seepage field. Formula (2) describes the relationship among saturation, fluid pressure and fluid volume
change [9]: 1
M
∂p
∂t + n
s
∂s
∂t = 1
s
∂ζ
∂t
(2) (2) where M is the Biot modulus; n is the porosity of the porous medium; t is time; ζ represents
the fluid volume change in the porous media per unit volume. From Formula (2), we
can derive the formula describing the pore water pressure of saturated soil, by setting the
saturation degree as 1.0 (FLAC3D, 2012), which is: where M is the Biot modulus; n is the porosity of the porous medium; t is time; ζ represents
the fluid volume change in the porous media per unit volume. From Formula (2), we
can derive the formula describing the pore water pressure of saturated soil, by setting the
saturation degree as 1.0 (FLAC3D, 2012), which is: 1
M
∂p
∂t = 1
s
∂ζ
∂t
(3) (3) For calculating the saturation degree of the unsaturated soil, the pore water pressure
on the node is set as 0, so the node saturation degree is For calculating the saturation degree of the unsaturated soil, the pore water pressure
on the node is set as 0, so the node saturation degree is n
s
∂s
∂t = 1
s
∂ζ
∂t
(4) (4) As Formulas (3) and (4) show, obtaining the relationship between saturation degree
and negative pressure is important for calculating the permeability of unsaturated soil. The
saturation degree can be used to obtain the permeability coefficient in the unsaturated zone. In the calculation, we adjusted the unit permeability coefficient to analyze the seepage
process in the unsaturated zone. 2.1. Shear Strength Theory of Unsaturated Soil This paper adopted the shear strength formula based on the dual stress variable
strength theory from [8], as follows: τf = c′ + (σ −ua) tan φ′ + (ua −uw) tan φD
(1) (1) τf is the soil shear strength; c′ and φ′ are the effective cohesion and effective friction
angle, respectively; uw is the void water pressure; ua is the void air pressure; and φD is the
matrix suction internal friction angle. When the soil becomes saturated, the pore water pressure uw and pore gas pressure
us are similar, that is, uw −us tends to be 0. In such case, the matrix suction term in
Formula (1) is equal to 0, and the shear strength formula of unsaturated soil is the same as
that of saturated soil. This article is an open access article
distributed
under
the
terms
and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/). g p
p
In this study, a time-varying analysis method (TSFDM) is proposed for analyzing
slope stability based on the finite difference method. We use Python to improve the FISH
function in the TSFDM. The improved FISH function can analyze the temporal change
features of the unsaturated slope seepage and numerically simulate the slope under rainfall,
water level changes and a combination of the two conditions. https://www.mdpi.com/journal/applsci Appl. Sci. 2023, 13, 5268. https://doi.org/10.3390/app13095268 Appl. Sci. 2023, 13, 5268 2 of 13 2. Method 2.1. Shear Strength Theory of Unsaturated Soil 2.4. Improvement of FISH Function (1) Set the fluid tensile strength threshold, which allows the nodal flow outflow and
f
i
d
i
h
l
l
i forms a negative pressure zone during the seepage calculation. (2) The pore water pressure changes with the hydraulic boundary. Use the built-in
variable zone.pp (xcz) to obtain the unit negative pore water pressure, and then calculate
the unit saturation degree by Formula (7). Distribute one constitutive model and the initial
shear strength parameters to each unit. (3) Substitute the obtained unit saturation to Formula (8) to calculate the unit relative
to permeability coefficient, and then arrive at the corrected unit permeability coefficient by
Formula (9). Meanwhile, adjust the unit shear strength parameters according to the pore
water pressure. Figure 1 shows the flowchart using the FISH function to calculate the unit permeability
coefficient and shear strength parameters of the unsaturated soil. Figure 1 shows the flowchart using the FISH function to calculate the unit permeability
coefficient and shear strength parameters of the unsaturated soil. 2.4. Improvement of FISH Function For pores with negative pore water pressure (matrix suction) above the infiltration
line, the TSFDM automatically sets their pore water pressure as 0. Then the permeability Appl. Sci. 2023, 13, 5268 3 of 13 coefficient is a constant. However, this is unreasonable, as the permeability coefficient of
the rock-soil mass in the slope is not a constant, but changes with the saturation degree and
porosity. The permeability coefficient can be expressed by the Van Genuchten model that
describes the soil water characteristic curve and the permeability of unsaturated slope [11]. coefficient is a constant. However, this is unreasonable, as the permeability coefficient of
the rock-soil mass in the slope is not a constant, but changes with the saturation degree and
porosity. The permeability coefficient can be expressed by the Van Genuchten model that
describes the soil water characteristic curve and the permeability of unsaturated slope [11]. θ = θr +
θs −θr
[1 + (p/a)n]m
(6)
s = sr +
1 −sr
[1 + (p/a)n]m
(7)
kr = S1/2[1 −(1 −S1/m)
m]
2
(8)
k = kwkr = kwS1/2[1 −(1 −S1/m)
m]
2
(9) (6) (8) k = kwkr = kwS1/2[1 −(1 −S1/m)
m]
2
(9) k = kwkr = kwS1/2[1 −(1 −S1/m)
m]
2 (9) where θ, θr and θs are the corrected, residual and saturated volumetric water content,
respectively; s represents the corrected saturation degree, sr is the residual saturation
degree, a, n and m are the fitting parameters. Formula (8) expresses the relationship
between the relative permeability coefficient and saturation degree, from which we can
derive Formula (9). kw, kr and k are the saturated, relative and corrected permeability
coefficients, respectively. The following analysis is based on Formulas (6)–(9). p
y
g
y
The shear strength parameters of saturated and unsaturated soil are quite different,
so we improved the FISH function to analyze the shear strength changes in the saturated
and unsaturated slope separately. The function can adjust the shear strength parameters
automatically according to the pore water pressure [12]. The specific improvements are as follows: The specific improvements are as follows: The specific improvements are as follows: (1) Set the fluid tensile strength threshold, which allows the nodal flow outflow and
forms a negative pressure zone during the seepage calculation. 2.5. Development of the FISH Program In the time-varying analysis of the tailings pond slope stability, when setting the
boundary conditions we should consider their changes over time. According to the im-
proved FISH function, the hydraulic boundary conditions change over time, and the
TSFDM generates a seepage field file every two steps [13]. If the time interval is small and
the time span is large, the generated seepage field files are abundant. These seepage field
files are input to the stress field module for analysis. The obtained stress-strain field is then
adopted in strength reduction to calculate the safety factor with Formula (5). To obtain the
safety factor for each period, the seepage field file should be restored repeatedly, but the
original FISH function cannot restore the seepage field file by batch, so we improved the
FISH function by Python programming. Appl. Sci. 2023, 13, 5268
2023, 13, x FOR PEER REV 4 of 13 e 1. Flowchart for calculating the unit permeability coefficient and shear strength parameters
unsaturated slope using the FISH function
Figure 1. Flowchart for calculating the unit permeability coefficient and shear strength parameters of
the unsaturated slope using the FISH function. . Flowchart for calculating the unit permeability coefficient and shear strength parameters
t
t d l
i
th FISH f
ti
Figure 1. Flowchart for calculating the unit permeability coefficient and shear strength parameters of
the unsaturated slope using the FISH function. nsaturated slope using the FISH function. We used Python to realize the slope stability analysis in TSFDM as follows (Figure 2) elopment of the FISH Program
the time-varying analysis of the tailings pond slope stability, when setting the
ry conditions we should consider their changes over time According to the
(1) Use the improved FISH function to obtain the seepage field changes in each unit
time under the condition of the pond water level change and rainfall, and store it in a
specific location. ary conditions we should consider their changes over time. According to the
ed FISH function, the hydraulic boundary conditions change over time, and the
M generates a seepage field file every two steps [13]. If the time interval is small and
e span is large the generated seepage field files are abundant These seepage field
p
(2) Traverse all seepage field files (.f3sav) in a specific location on Python, which are
called into the ITASC module in the Ipython console. 2.5. Development of the FISH Program Restore all seepage field files in
batches and input the pore water pressure field. e span is large, the generated seepage field files are abundant. These seepage field
e input to the stress field module for analysis. The obtained stress-strain field is
dopted in strength reduction to calculate the safety factor with Formula (5). To
the safety factor for each period the seepage field file should be restored repeatedly
p
p
p
(3) Set the mechanical boundary. Use the FDSRM to calculate the safety factor corre-
sponding to the seepage field at each unit time. Finally, the safety factor curve is used to
analyze the slope stability of the tailing pond. e original FISH function cannot resto
ed the FISH fu
tio
by Pytho
o
2.6. Establish a Finite Difference Model ed the FISH function by Python programming. e used Python to realize the slope stability analysis in TSFDM as follows (Figure
Use the improved FISH function to obtain the seepage field changes in each unit
nder the condition of the pond water level change and rainfall, and store it in a
location. Traverse all seepage field files (.f3sav) in a specific location on Python, which are
nto the ITASC module in the Ipython console. Restore all seepage field files in
and input the pore water pressure field. S t th
h
i
l b
d
U
th
FDSRM t
l
l t
th
f t
f
t
This study analyzed the bank slope of a tailing pond in Yunnan Pulang mine. The
slope is 160 m high, 368 m wide, and 10 m thick (Figure 3a). We established a numerical
simulation model for the slope (Figure 3b). The model is composed of hexahedral mesh
elements, with 8867 nodes and 11,018 units. The slope has 3 stratum layers (I: Gravel soil
(GS), II: Strongly weathered carbonaceous slate (SWCS), and III: Weathered carbonaceous
slate (WCS). The geological parameters from survey data are shown in Table 1. The
Mohr Coulomb criterion was adopted as the material failure criterion. The bank slope
stability was calculated by the TSFDM and the Morganston-Price based LEM method,
separately [14]. 5 of 13
5 of 14 Appl. Sci. 2023, 13, 5268
Appl. Sci. 2023, 13, x FOR Figure 2. Flowchart for analyzing the bank slope stability. Figure 2. Flowchart for analyzing the bank slope stability. Figure 2. Flowchart for analyzing the bank slope stability. 2.6. Establish a Finite Difference Model
This study analyzed the bank slope of a tailing pond in Yunnan P
slope is 160 m high, 368 m wide, and 10 m thick (Figure 3a). We establis
simulation model for the slope (Figure 3b). The model is composed of h
elements, with 8867 nodes and 11,018 units. The slope has 3 stratum laye
(GS), Ⅱ: Strongly weathered carbonaceous slate (SWCS), and Ⅲ: Weather
Figure 2. Flowchart for analyzing the bank slope stability. Table 1. Geological parameters of the pond bank slope. ithology
Status
Elastic Modulus
(MPa)
Poisson’s
Ratio
Unit Weight
(KN/m3)
Cohesion
(KPa)
Int
GS
Natural
261.6
0.4
20.5
15
Saturated
241.6
0.41
21
10
SWCS
Natural
2644.9
0.38
22.4
93.6 Figure 2. Flowchart for analyzing the bank slope stability. Figure 2. e original FISH function cannot resto
ed the FISH fu
tio
by Pytho
o
2.6. Establish a Finite Difference Model Flowchart for analyzing the bank slope stability. Figure 2. Flowchart for analyzing the bank slope stability. Figure 2. Flowchart for analyzing the bank slope stability. 2.6. Establish a Finite Difference Model
Table 1. Geological parameters of the pond bank slope. E a i
a
i i e
iffe e
e
o e
This study analyzed the bank slope of a tailing pond in Yunnan Pulang mine. The
slope is 160 m high, 368 m wide, and 10 m thick (Figure 3a). We established a numerical
simulation model for the slope (Figure 3b). The model is composed of hexahedral mesh
elements, with 8867 nodes and 11,018 units. The slope has 3 stratum layers (Ⅰ: Gravel soil
(GS), Ⅱ: Strongly weathered carbonaceous slate (SWCS), and Ⅲ: Weathered carbonaceous
slate (WCS). The geological parameters from survey data are shown in Table 1. The Mohr
Coulomb criterion was adopted as the material failure criterion. The bank slope stability
was calculated by the TSFDM and the Morganston-Price based LEM method, separately
[14]
g
p
p
p
Strata
Lithology
Status
Elastic Modulus
(MPa)
Poisson’s
Ratio
Unit Weight
(KN/m3)
Cohesion
(KPa)
Internal Friction
Angle (◦)
I
GS
Natural
261.6
0.4
20.5
15
32
Saturated
241.6
0.41
21
10
30
II
SWCS
Natural
2644.9
0.38
22.4
93.6
33.3
Saturated
975.8
0.40
23.5
77.4
28.9
III
WCS
Saturated
3786.1
0.36
27
100
30
Natural
5561
0.35
26.5
120.0
35.0 Appl. Sci. 2023, 13, 5268
and B 6 of 13 (a)
(b)
Figure 3. Actual photos and numerical model of the trailing pond slope. (a) Actual photos. (b)
Numerical simulation model. Table 1. Geological parameters of the pond bank slope. Elastic Modulus
(MP )
Poisson’s Ratio
Unit Weight
(KN/
3)
Cohesion
(KP )
Internal Friction
A
l (°)
Figure 3. Actual photos and numerical model of the trailing pond slope. (a) Actual photos. (b) Nu-
merical simulation model. As Figure 3 shows, AB and FE are the fixed water level boundaries, which are 130 m
and 10 m high, respectively. The initial groundwater level was determined according to the
survey data. CD is the rainfall infiltration interface, AF is the impervious boundary, and (a) (a) (b) (b) photos and numerical model of the trailing pond slope. (a) Actual photos. (b)
lation model. Figure 3. Actual photos and numerical model of the trailing pond slope. (a) Actual photos. (b) Nu-
merical simulation model. 3 R
lt
3.1. Temporal Changes of the Seepage in the Saturated and Unsaturated Slopes
3.1. Temporal Changes of the Seepage in the Saturated and Unsaturated Slopes 3 R
lt
3.1. Temporal Changes of the Seepage in the Saturated and Unsaturated Slopes
3.1. Temporal Changes of the Seepage in the Saturated and Unsaturated Slopes 3. Results
3.1. Temporal Changes of the Seepage in the Saturated and Unsaturated Slopes
Figure 5a shows the seepage field of the bank slope on 1 April. Negative pore water
pressure appears in the area above the initial water level. The shear strength parameters
above and below the water level are different. As all strength parameters are stored in one
unit, only the distribution of gravity is shown in Figure 5a. Figure 5b is the seepage field
of the slope on 27 July. We adjust the strength parameters of saturated and unsaturated
Figure 5a shows the seepage field of the bank slope on 1 April. Negative pore water
pressure appears in the area above the initial water level. The shear strength parameters
above and below the water level are different. As all strength parameters are stored in one
unit, only the distribution of gravity is shown in Figure 5a. Figure 5b is the seepage field of
the slope on 27 July. We adjust the strength parameters of saturated and unsaturated slopes
to obtain the gravity distribution in Figure 5b. After the rainy season the seepage change
continues, resulting in a variation of shear strength parameters [15]. Figure 5a shows the seepage field of the bank slope on 1 April. Negative pore water
pressure appears in the area above the initial water level. The shear strength parameters
above and below the water level are different. As all strength parameters are stored in one
unit, only the distribution of gravity is shown in Figure 5a. Figure 5b is the seepage field
of the slope on 27 July. We adjust the strength parameters of saturated and unsaturated
slopes to obtain the gravity distribution in Figure 5b. After the rainy season the seepage
change continues, resulting in a variation of shear strength parameters [15]. slopes to obtain the gravity distribution in Figure 5b. After the rainy season the seepage
change continues, resulting in a variation of shear strength parameters [15]. (a)
(b)
(c)
(d)
(a)
(b)
(c)
(d)
Figure 5. Variation of pore water pressure and gravity distribution on (a,b) 1 April and (c,d) 27 July
2019. Figure 5. Variation of pore water pressure and gravity distribution on (a,b) 1 April and (c,d) 27 July 2019. 3.2. 3 R
lt
3.1. Temporal Changes of the Seepage in the Saturated and Unsaturated Slopes
3.1. Temporal Changes of the Seepage in the Saturated and Unsaturated Slopes Impacts of Real Rainfall and Pond Water Level Changes on Slope Stability slopes to obtain the gravity distribution in
change continues, resulting in a variation
(a) hear strength parameters [15]. (b) rameters [15]. (a) (b) (a)
(c) (b)
(d) (d) (c) (c)
(d)
Fi u e 5 Va iatio
of
o e
ate
e
u e a d
a ity di t ibutio
o
(a b) 1 A
il a d ( d) 27 July
Figure 5. Variation of pore water pressure and gravity distribution on (a,b) 1 April and (c,d) 27 July
2019. Figure 5. Variation of pore water pressure and gravity distribution on (a,b) 1 April and (c,d) 27 July 2019. 3.2. Impacts of Real Rainfall and Pond Water Level Changes on Slope Stability (c)
(d)
Fi
5 V
i ti
f
t
d
it
di t ib ti
( b) 1 A
il
d ( d) 27 J l
2019. 3.2. Impacts of Real Rainfall and Pond Water Level Changes on Slope Stability (c)
(d)
Fi
5 V
i ti
f
t
d
it
di t ib ti
( b) 1 A
i
2019. 3.2. Impacts of Real Rainfall and Pond Water Level Changes on Slope Stability Figure 5. Variation of pore water pressure and gravity distribution on (a,b) 1 April and (c,d) 27 July
2019. We analyzed the impacts of real rainfall, pond water level changes, and the combined
impacts on the slope stability [16]. The results are shown in Figures 6–8. In these figures,
“by FDSRM” means the results obtained by the improved FISH function and the FDSRM,
and “by LEM” means the results obtained by the Morgenston-Price based LEM method. If
the two results are close, the stability analysis procedure is reliable. Figure 5. Variation of pore water pressure and gravity distribution on (a,b) 1 April and (c,d) 27 July
2019. We analyzed the impacts of real rainfall, pond water level changes, and the combined
impacts on the slope stability [16]. The results are shown in Figures 6–8. In these figures,
“by FDSRM” means the results obtained by the improved FISH function and the FDSRM,
and “by LEM” means the results obtained by the Morgenston-Price based LEM method. If
the two results are close, the stability analysis procedure is reliable. Figure 6 shows the slope safety factor and the pore water pressure at the monitoring
points under the real rainfall data. e original FISH function cannot resto
ed the FISH fu
tio
by Pytho
o
2.6. Establish a Finite Difference Model l photos and numerical model of the trailing pond slope. (a) Actual photos. (b)
ulation model. Figure 3. Actual photos and numerical model of the trailing pond slope. (a) Actual photos. (b) Nu-
merical simulation model. ical parameters of the pond bank slope. s
Poisson’s Ratio
Unit Weight
(KN/m3)
Cohesion
(KPa)
Internal Friction
Angle (°)
0.4
20.5
15
32
As Figure 3 shows, AB and FE are the fixed water level boundaries, which are 130 m
and 10 m high, respectively. The initial groundwater level was determined according to the
survey data. CD is the rainfall infiltration interface, AF is the impervious boundary, and
BC and DE are the pervious boundaries. To study the seepage characteristics, we selected
6 measuring points (points 1–6) 10 m below the surface. 0.41
21
10
30
0.38
22.4
93.6
33.3
0.40
23.5
77.4
28.9
0.36
27
100
30
0.35
26.5
120.0
35.0
We focused on the seepage changes under the impacts of rainfall infiltration and
pond water level fluctuation [14]. The boundary conditions of the slope surface should
be adjusted timely to simulate the seepage field changes and slope seepage. In Figure 4,
we simulate the separate and coupled influences of hourly rainfall and groundwater level
changes on the slope stability from 0:00 on 1 April to 0:00 on 27 July 2019. 7 of 13
er level Appl. Sci. 2023, 13, 5268 4/1
4/8
4/15
4/22
4/29
5/6
5/13
5/20
5/27
6/3
6/10
6/17
6/24
7/1
7/8
7/15
7/22
7/29
8
10
12
14
16
18
20
22
Reservoir water level(m)
Rainfall intensity(mm)
Time(h)
Reservoir water level(m)
0
2
4
6
8
Rainfall (mm)
Figure 4. Rainfall and reservoir water level of the trailing pond during 1 April 2019 and 29 July
2019
Figure 4. Rainfall and reservoir water level of the trailing pond during 1 April 2019 and 29 July 2019. 3 Results
4/1
4/8
4/15
4/22
4/29
5/6
5/13
5/20
5/27
6/3
6/10
6/17
6/24
7/1
7/8
7/15
7/22
7/29
8
10
12
14
16
18
20
Reservoir water level(m)
Rainfall intensity(mm)
Time(h)
Reservoir water level(m)
0
2
4
6
8
Rainfall (mm)
Figure 4. Rainfall and reservoir water level of the trailing pond during 1 April 2019 and 29 July
2019. 3. e original FISH function cannot resto
ed the FISH fu
tio
by Pytho
o
2.6. Establish a Finite Difference Model Results 4/1
4/8
4/15
4/22
4/29
5/6
5/13
5/20
5/27
6/3
6/10
6/17
6/24
7/1
7/8
7/15
7/22
7/29
8
10
12
14
16
18
20
22
Reservoir water level(m)
Rainfall intensity(mm)
Time(h)
Reservoir water level(m)
0
2
4
6
8
Rainfall (mm)
4/1
4/8
4/15
4/22
4/29
5/6
5/13
5/20
5/27
6/3
6/10
6/17
6/24
7/1
7/8
7/15
7/22
7/29
8
10
12
14
16
18
20
Reservoir water level(m)
Rainfall intensity(mm)
Time(h)
Reservoir water level(m)
0
2
4
6
Rainfall (mm)
Figure 4. Rainfall and reservoir water level of the trailing pond during 1 April 2019 and 29 July Figure 4. Rainfall and reservoir water level of the trailing pond during 1 April 2019 and 29 July
Figure 4. Rainfall and reservoir water level of the trailing pond during 1 April 2019 and 29 July 2019. 2019. 3 Results Figure 4. Rainfall and reservoir water level of the trailing pond during 1 April 2019 and 29 July
2019. Figure 4. Rainfall and reservoir water level of the trailing pond during 1 April 2019 and 29 July 2019. 3. Results
2019. 3. Results 3 R
lt
3.1. Temporal Changes of the Seepage in the Saturated and Unsaturated Slopes
3.1. Temporal Changes of the Seepage in the Saturated and Unsaturated Slopes The slope safety factor is negatively related with the
rainfall, that is, it decreases with the increasing rainfall and goes up when the rain stops, Appl. Sci. 2023, 13, 5268 8 of 13 8 of 13 but the pore water pressure of the six monitoring points is related to the rainfall infiltration. Point 5 and Point 6 at the slope toe near the gravel pile are sensitive to the pore water
pressure. After penetrating the slope body, the water will flow to the toe, which will
increase the pore water pressure. VIEW
9 of 14 but the pore water pressure of the six monitoring points is related to the rainfall infiltration. Point 5 and Point 6 at the slope toe near the gravel pile are sensitive to the pore water
pressure. After penetrating the slope body, the water will flow to the toe, which will
increase the pore water pressure. VIEW
9 of 14 but the pore water pressure of the six monitoring points is related to the rainfall infiltration. Point 5 and Point 6 at the slope toe near the gravel pile are sensitive to the pore water
pressure. After penetrating the slope body, the water will flow to the toe, which will
increase the pore water pressure. VIEW
9 of 14 pressure. After penetrating the slope body, the water will flow to the toe, which will
increase the pore water pressure. Figure 6. The impact of rainfall on slope safety factor and pore water pressure at the monitoring
points. FOS stands for factor of safety. Figure 6. The impact of rainfall on slope safety factor and pore water pressure at the monitoring
points. FOS stands for factor of safety. Figure 7 reflects impacts of water level change on the slope stability. The pore water Figure 6. The impact of rainfall on slope safety factor and pore water pressure at the monitoring
points. FOS stands for factor of safety. Figure 6. The impact of rainfall on slope safety factor and pore water pressure at the monitoring
points. FOS stands for factor of safety. Figure 7 reflects impacts of water level change on the slope stability. The pore water
pressure at Points 4–6 at the slope toe correlates to the pond water level. The closer the
point to the water, the stronger the correlation. 3 R
lt
3.1. Temporal Changes of the Seepage in the Saturated and Unsaturated Slopes
3.1. Temporal Changes of the Seepage in the Saturated and Unsaturated Slopes Points 1–3 are far away from the water, so
their pore water pressure changes slightly (≤0.01%) with the water level. Figure 7 reflects impacts of water level change on the slope stability. The pore water
pressure at Points 4–6 at the slope toe correlates to the pond water level. The closer the
point to the water, the stronger the correlation. Points 1–3 are far away from the water, so
their pore water pressure changes slightly (≤0.01%) with the water level. Appl. Sci. 2023, 13, 5268 9 of 13 Figure 7. The impact of the pond water level on the safety factor and the pore water pressure
monitoring points. Figure 7. The impact of the pond water level on the safety factor and the pore water pressure at the
monitoring points. ure 7. The impact of the pond water level on the safety factor and the pore water pressure
nitoring points. Figure 7. The impact of the pond water level on the safety factor and the pore water pressure at the
monitoring points. Figure 8 shows the combined impacts of rainfall and water level on the slope stability. The increase of water level enhances rainfall infiltration and further reduces the bank slope
safety factor. In addition, the rainfall infiltration raises the water level. The changes of water
level have greater impacts on the safety factor than that of rainfall, as the pore pressure
changes at Points 1–3 under the coupled condition are similar to that under only the water
level change. Appl. Sci. 2023, 13, 5268
Sci. 2023, 13, x FOR PEE 10 of 1311 Figure 8. The coupled impacts of rainfall and pond water level on the safety factor and the p
water pressure at the monitoring points. Figure 8. The coupled impacts of rainfall and pond water level on the safety factor and the pore
water pressure at the monitoring points. gure 8. The coupled impacts of rainfall and pond water level on the safety factor and the
ter pressure at the monitoring points. Figure 8. The coupled impacts of rainfall and pond water level on the safety factor and the pore
water pressure at the monitoring points. Discussion
4. Discussion The bank slope safety factor calculated by the TSFDM and the displacement vector (a) under
the initial conditions, (b) under the rainfall impact alone at the end of the observation, (c) under the
pond water level change impact alone at the end of the observation, and (d) under the impacts of
both conditions at the end of the observation. 5. Conclusions
This paper proposed a time-scale varying finite difference method (TSFDM) to
analyze the influence of rainfall infiltration and water level fluctuation on the stability of
the bank slope in the Yunnan Pulang tailings pond. We used Python programming to
improve the FISH function in FLAC3D. The improved function can adjust the shear
strength parameters of rock-soil mass according to pore pressure, and also solve the
restoration failure in the old FISH function. Using the improved FISH function for the
On 27 July 2019, affected by rainfall alone, the safety factor was 0.9089 (Figure 9b). A
new shear strain surface was formed on the weathering interface of carbonaceous slate. The rear part of the slope was in tension [17]. The strongly weathered carbonaceous slate
and gravel soil front above the weathering zone have an obvious downward trend. This
is because the rock and soil above the weathering zone were deformed under the effects
of rainwater seepage, softening, and suspension. The safety factor is smaller than that in
April, which requires attention [18]. p
q
On 27 July 2019, impacted by the pond water level change alone, the safety factor of
the bank slope was 0.9068 (Figure 9c). The water level changes softened the rock-soil mass
on the weathering zone. The shear strain rate of the weathering zone increased, and some
parts of the front edge of the bank slope became unstable [19,20]. As Figure 9d shows, at the end of the observation period, under the combined impacts
of rainfall and water level change, the slope safety factor was 0.9062. The weathering
zone had softened seriously, and the shear strain rate of the weathering zone increased
significantly. Some parts of the front edge of the bank slope became unstable, requiring
treatments [21]. The safety factor calculated by TSFDM was generally smaller than that calculated
by LEM. Discussion
4. Discussion Discussion
Using the proposed time-scale varying finite difference method (TSFDM),
lculated the safety factor of the bank slope of a tailing pond from 1 April to 27 July 2
onsidering the rainfall and pond water level change (Figure 9). On 1 April 2019, un
e impacts of both rainfall and water level changes, the safety factor was 0.914, and
otential sliding surface was in the middle and lower part of the slope, across the stro
Using the proposed time-scale varying finite difference method (TSFDM), we calcu-
lated the safety factor of the bank slope of a tailing pond from 1 April to 27 July 2019,
considering the rainfall and pond water level change (Figure 9). On 1 April 2019, under
the impacts of both rainfall and water level changes, the safety factor was 0.914, and the
potential sliding surface was in the middle and lower part of the slope, across the strongly
weathered carbonaceous slate and gravel soil (Figure 9a). Appl. Sci. 2023, 13, 5268 11 of 13
d water
athering Figure 9. The bank slope safety factor calculated by the TSFDM and the displacement vector (a)
under the initial conditions, (b) under the rainfall impact alone at the end of the observation, (c)
under the pond water level change impact alone at the end of the observation, and (d) under the
impacts of both conditions at the end of the observation. Figure 9. The bank slope safety factor calculated by the TSFDM and the displacement vector (a) under
the initial conditions, (b) under the rainfall impact alone at the end of the observation, (c) under the
pond water level change impact alone at the end of the observation, and (d) under the impacts of
both conditions at the end of the observation. Figure 9. The bank slope safety factor calculated by the TSFDM and the displacement vector (a)
under the initial conditions, (b) under the rainfall impact alone at the end of the observation, (c)
under the pond water level change impact alone at the end of the observation, and (d) under the
impacts of both conditions at the end of the observation. Figure 9. Discussion
4. Discussion This is because the TSFDM underestimates the shear strength parameters of
the saturated slope, which is in line with reality, but the LEM uses the parameters of
unsaturated slope, generating larger shear strength parameters, and a higher safety factor. The program automatically reads the slope seepage field file by batch, performs finite
difference calculation, and realizes the intelligent calculation of the geological model. Using
TSFDM to study the slope stability of a tailings pond under heavy rainfall and water level
fluctuations obtains the evolution features of the stress and strain of the weathering zone
and gravel belt [22]. Appl. Sci. 2023, 13, 5268 12 of 13 12 of 13 Institutional Review Board Statement: Not applicable. Institutional Review Board Statement: Not applicable. Institutional Review Board Statement: Not applicable. Informed Consent Statement: Not applicable. Informed Consent Statement: Not applicable. Informed Consent Statement: Not applicable. Data Availability Statement: The source data can be obtained in the article. Conflicts of Interest: The authors declare no conflict of interest. Conflicts of Interest: The authors declare no conflict of interest. 5. Conclusions This paper proposed a time-scale varying finite difference method (TSFDM) to analyze
the influence of rainfall infiltration and water level fluctuation on the stability of the bank
slope in the Yunnan Pulang tailings pond. We used Python programming to improve the
FISH function in FLAC3D. The improved function can adjust the shear strength parameters
of rock-soil mass according to pore pressure, and also solve the restoration failure in the
old FISH function. Using the improved FISH function for the numerical simulation of the
tailing pond slope can greatly reduce the code operation time and achieve long-time and
high frequency stability monitoring. g
q
y
y
g
The TSFDM was compared with the LEM in calculating the safety factor of the tailing
pond slope. The results are similar, so the TSFDM is reliable. The TSFDM gets a smaller
result, which is in line with reality and can better reflect the stress and strain evolution of
the slope. p
Compared with rainfall, water level fluctuation has a greater impact on the safety
factor. The safety factor decreases with the increase in the shear strain rate of the weathering
zone. The potential sliding surface goes through the bottom of the gravel soil and the
weathering zone of carbonaceous slate. The safety factors under different conditions are
all smaller than 1, so corresponding measures should be taken. For example, spraying
waterproof concrete on the slope surface to reduce the infiltration of water on the one hand
and strengthen the stability of the slope on the other. Author Contributions: Conceptualization, G.Z. and G.L.; formal analysis, C.X. and T.L.; writing—
original draft, D.B. and G.L.; writing—review and editing, G.Z. and T.L. All authors have read and
agreed to the published version of the manuscript. Funding: This work was financially supported by the National Natural Science Foundation of
China under Grant No. 41974148 and Grant No. 52004327; Hunan Provincial Key Research and
Development Program, Grant No. 2020SK2135; Natural Resources Research Project of Hunan
Province, Grant No. 2021-15; and Science and Technology Project of Hunan Provincial Department of
Transportation Grant No. 202012. 10.
Song, Y.S.; Chae, B.G.; Lee, J. A method for evaluating the stability of an unsaturated slope in natural terrain during rainfall. Eng.
Geol. 2016, 210, 84–92. [CrossRef] References 1. Chen, G.Q.; Huang, R.Q.; Shi, Y.C.; Xu, Q. Stability Analysis of Slope Based on Dynamic and Whole Strength Reduction Methods. Chin. J. Rock Mech. Eng. 2014, 33, 243–256. 2. Cao, Z.Y. Construction mechanics and time-varying mechanics in civil engineering. China Civ. Eng. J. 2001, 34, 41–46. 3. Zhang, D.; Jian, W.B.; Ye, Q.; Lin, W. A time-varying analytic model of tailings slope and its application. Rock Soil Mech. 2014, 35,
835–840. y
g
g
g
g
3. Zhang, D.; Jian, W.B.; Ye, Q.; Lin, W. A time-varying analytic model of tailings slope and its application. Rock Soil Mech. 2014, 35,
835–840. 4. Jiang, Z.M.; Xiong, X.H.; Zeng, L. Unsaturated seepage analysis of slope under rainfall condition based on FLAC3D. Rock Soil
Mech. 2014, 35, 855–861. 5. Chu, W.J.; Xu, W.Y.; Yang, S.Q.; Zhou, W.Y. Secondary development of a viscoelasto-plastic rheological constitutive model of rock
based on FLAC3D. Rock Soil Mech. 2006, 27, 2005–2010. 6. He, Z.; Dai, B. Secondary Development of a Nonlinear Creep Model Based on Fractional Derivative in FLAC3D. In Proceedings
of the 2018 11th International Conference on Intelligent Computation Technology and Automation (ICICTA), Changsha, China,
22–23 September 2018. p
7. Wang, C.B.; Ding, W.Q.; Qiao, Y.F. Development and application of hardening soil constitutive model in FLAC(3D). Chin. J. Rock
Mech. Eng. 2014, 33, 199–208. j
f
J
y
J
9. Ou, X.P.; Bai, K.; Zhu, Y.S.; Yuan, C.; Wang, J.; Liu, D.T.; Liu, H.J. The Strength Reduction Method for Stability Analysis of Slope
Based on FLAC-3D. J. Wuhan Univ. Technol. 2009, 31, 59–61. 10. Song, Y.S.; Chae, B.G.; Lee, J. A method for evaluating the stability of an unsaturated slope in natural terrain during rainfall. Eng. Geol. 2016, 210, 84–92. [CrossRef] 13 of 13 13 of 13 Appl. Sci. 2023, 13, 5268 11. Genuchten, M.V. A Closed-form Equation for Predicting Hydraulic Conductivity for Unsaturated Soils. Soil Sci. Soc. Am. J. 1980,
44, 892–898. [CrossRef] [
]
12. Chen, X.; Fu, J.J.; Zhao, H.B.; Huang, T.P.; Shi, K. Analysis of Slope Reliability by Finite-Difference Strength-Reduction Method
Considering the Thought of Monte Carlo Method. J. Yangtze River Sci. Res. Inst. 2011, 28, 36–40. g
g
g
13. Jiang, X.L.; Yang, H.; Cao, P. FLAC(3D) analysis of interaction between goaf and opencut mining slope with 3D geological model
using SURPAC. Rock Soil Mech. 2011, 32, 1234–1240. 14. Liu, C.H.; Chen, C.X.; Feng, X.T.; Xiao, G.F. References Effect of groundwater on stability of slopes at reservoir bank. Rock Soil Mech. 2005, 26,
419–422. 15. Liu, J.X.; Liu, Y.T.; Hu, Q.J. Stability of embankment slope subjected to rainfall infiltration considering both runoff-underground
seepage and fluid-solid coupling. Rock Soil Mech. 2010, 31, 903–910. p
g
rty, D. Numerical analysis of rainfall effects in external overburden dump. Int. J. Min. Sci. Technol. 2016, 26 16. Koner, R.; Chakravarty, D. Numerical analysis of rainfall effects in external overburden dump. Int. J. M
825–831. [CrossRef] 17. Mousavi, S.M. Landslide Susceptibility in Cemented Volcanic Soils, Ask Region, Iran. Indian Geotech. J. 2016, 47, 115–130. [CrossRef] 18. Nguyen, T.S.; Borgesson, L.; Chijimatsu, M.; Rutqvist, J.; Fujita, T.; Hernelind, J.; Kobayashi, A.; Ohnishi, Y.; Tanaka, M.; Jing, L. Hydro-mechanical response of a fractured granitic rock mass to excavation of a test pit—The Kamaishi Mine experiment in Japan. Int. J. Rock Mech. Min. Sci. 2001, 38, 79–94. [CrossRef] 19. Wang, J.X.; Li, Z.; Chen, W. Lower bound analysis of soil slope stability using finite elements subjected to pore water pressure. Rock Soil Mech. 2005, 16, 1258–1262+1268. ,
,
20. Wang, L.F.; Li, L.G.; Yang, X. Instability initiation mechanism of gravel soil slope in Three Gorges Reservoir: Case study of
Hongyanzi landslide in Wushan county. Chin. J. Geot. Eng. 2018, 40, 209–214. gy
y
21. Wang, M.H.; Yan, E.C. Study on influence of reservoir water impounding on reservoir landslide. Rock Soil Mech. 2007, 28,
2722–2725. 22. Zhang, W.; Goh, A.T. Reliability assessment on ultimate and serviceability limit states and determination of critical factor of safety
for underground rock caverns. Tunn. Undergr. Space Technol. 2012, 32, 221–230. [CrossRef] 22. Zhang, W.; Goh, A.T. Reliability assessment on ultimate and serviceability limit states and determination of critical factor of safety
for underground rock caverns. Tunn. Undergr. Space Technol. 2012, 32, 221–230. [CrossRef] Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content. Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). References MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.
|
https://openalex.org/W4365563287
|
https://uu.diva-portal.org/smash/get/diva2:1689594/FULLTEXT01
|
English
| null |
Us and them: Privileged emotions of Cape Town’s urban water crisis
|
Geoforum
| 2,023
|
cc-by
| 14,311
|
Elisa Savelli a a Department of Earth Sciences, Air, Water and Landscape Science, Uppsala University, Uppsala, Sweden
b Centre of Natural Hazards and Disaster Science, CNDS, Uppsala, Sweden A B S T R A C T Keywords:
Elite
Emotions
Subjectivities
Water crisis
Cape Town Political ecology has already engaged with emotions in order to reveal the intimate, unconscious and unexplored
power dynamics which characterise patterns of water use and control. Similar explorations have mostly focused
on the emotional struggles of structurally disadvantaged people rather than on the emotions of those with
privilege: the elite. This oversight becomes problematic when it conceals disproportionate shares of power and
the implications that such power has on the sustainable use and just distribution of water resources. The 2018
water crisis which affected Cape Town’s metropolitan area constitutes the empirical context of this paper, which
sets out to address the aforementioned research gap. Focusing on the elite’s emotional responses to Cape Town’s
drought and subsequent water crisis, this paper seeks to advance political ecology’s understanding of urban
water crises by retracing the emotional geography of Cape Town’s most privileged urban dwellers. In particular,
this work leverages the concept of subjectivity to explain the way emotions are constructed and come to
materially and discursively reproduce historical power dynamics. These findings reveal that fear, anger, and a
sense of pride felt by wealthier Capetonians results from and perpetuates the privileged conditions of those elite. Rooted in colonial and apartheid past, Capetonians’ privileged emotions end up perpetuating the main causes of
the water crisis and eventually excluding the most disadvantaged inhabitants from future use and control of
water resources. Ultimately, by connecting with privileged emotions, it is possible to challenge certain sub
jectivities and create space for more just and sustainable urban-water imaginaries. 1. Day Zero: The water crisis of affluent Capetonians? People now really get it how water is important. This is the only shift that
can be permanent, is that people got it emotionally not just intellectually.” People now really get it how water is important. This is the only shift that
can be permanent, is that people got it emotionally not just intellectually.” People now really get it how water is important. This is the only shift that
can be permanent, is that people got it emotionally not just intellectually.” “We go through a process of change in the same way that we go through a
process of grief so that (…) people (are) first shocked, then, they do not
believe, gradually, they get angry, and eventually, they accept the reality. We went through those stages when responding to the water crisis. People
were initially shocked and horrified, then, they got angry, they wanted to
know who to blame […] Then, they got to the point of “ok so the rain is
not falling, we need to actually do something about that”. So, people
started adopting totally new behaviours and installing new infrastructure
at their own cost” Direct quote by: KKC These introductory quotes describe the manner in which Capeto
nians have faced up to the water crisis that affected Cape Town urban
area after 3 years of severe meteorological drought. KSWC insists that
Cape Town’s inhabitants have first lived the water crisis emotionally
before reacting to it and eventually changing their everyday practices of
water use. In particular, KSWC describes the emotions of fear and anger
that prevailed amongst people refusing to accept the idea of future water
shortages. Similarly, KKC also believes that if there has been a change in
people’s behaviour, that change was mostly a result of the emotions that
people felt during the latest water crisis. The most intense emotional
reactions were triggered on the 18th January 2018, when the executive
major of Cape Town, Patricia De Lille, declared that unless residents
drastically cut water consumption, the City would face a ‘Day Zero’ and
run out of water on the 22nd of April 2018. Thankfully, after six months
of severe water restrictions, on June 2018, the Municipality decided to Geoforum 141 (2023) 103746 Geoforum 141 (2023) 103746 Available online 12 April 2023
0016-7185/© 2023 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
https://doi.org/10.1016/j.geoforum.2023.103746
Received 29 March 2022; Received in revised form 20 March 2023; Accepted 23 March 2023 * Address: Uppsala University, Department of Earth Sciences, Villavagen. 16, 75236 Uppsala, Sweden.
E-mail addresses: elisa.savelli@geo.uu.se, savelli.elisa@gmail.com. Us and them: Privileged emotions of Cape Town’s urban water cri Elisa Savelli a,b,* Available online 12 April 2023
0016-7185/© 2023 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
* Address: Uppsala University, Department of Earth Sciences, Villavagen. 16, 75236 Uppsala, Sweden.
E-mail addresses: elisa.savelli@geo.uu.se, savelli.elisa@gmail.com.
https://doi.org/10.1016/j.geoforum.2023.103746
Received 29 March 2022; Received in revised form 20 March 2023; Accepted 23 March 2023 Direct quote by: KSWC Thus, to provide a thorough inter
pretation of water crises, their genesis and future implications, this
article enters the households of more affluent Capetonians, the elite, and
studies their emotional experience of the drought and the water crisis. A
focus on emotion is needed to make visible the most intimate, uncon
scious and unexplored dynamics which differentiate subjectivities and
eventually reshape the everyday and extraordinary experience of water
(in)security (Doshi, 2017; Truelove, 2019; 2011; Sultana, 2020). To
make sense of emotions, this paper leverages the notion of subjectivity
as it represents what neurologists would refer to as emotion concepts
(Barrett, 2017). Emotion concepts designate reference models that the
brain uses to categorize external sensations and in turn, construct
emotional responses to the present from experiences of the past. In a
similar way, subjectivities relate to the way people understand them
selves and their relationship to their context (Mansfield et al., 2020;
Morales and Harris, 2014; Blackman et al., 2008; Gill, 2008). Subjectivity is largely conceived as the condition or the way subjects
feel, perceive, believe, desire or experience themselves and their relation
with the rest of the world (Mansfield et al., 2020; Morales and Harris,
2014; Nightingale, 2011; Blackman et al., 2008; Gill, 2008). Morales and
Harris’ (2014, p.706) refer to subjectivity as “what it means and feels like
to exist within a specific place, time or set of relationships”. Subjectivities
reflect what people have experienced, learned about, and interpreted
throughout their lives and prior generations (Bonilla-Silva, 2019). Urban political ecology explains that subjectivities internalize the
complex workings of power, thereby becoming active agents that shape
and at the same time, are shaped by prevailing social, cultural and po
litical spaces (Blackman et al., 2008). Through this power laden
assemblage of socioenvironmental relations, subjectivity can either
reinforce or resist domination (Blackman et al., 2008; Morales and
Harris, 2014; Nightingale, 2011). Thus, by focusing on subjectivity,
urban political ecology is able to highlight how different urban subjects
can visualize (or not) different modes of interaction between the human
and the non-human (e.g. water), thereby enabling the creation of one
kind of urban environment instead of another (Gabriel, 2014). To study the privileged emotional experiences of a water crisis, this
paper first unravels the theoretical background to frame the emotional
experience of urban water crisis, and subsequently identifies the
research gap this work aims to address. Direct quote by: KSWC Across the first two research phases (Heynen’s (2014; 2016) progress
reports), urban political ecology was not paying considerable attention
to the subjective dimensions of urban water struggles whilst mostly
privileging urban scales of analysis (Truelove, 2019; Doshi, 2017; Grove,
2009). Similar studies have often used Marx, Lefebvre and Harvey’s
theorizations of urban socionatures to describe cities as sites of
convergence for capitalist economies of production (Gabriel, 2014) and
circulations of water as embedded in the “political ecology of power
through which the urbanization process unfolds” (Swyngedouw, 1997,
p.311). Thus, whilst the urban became the most significant scale to
explore the capitalist production of nature, at the same time, capitalism
became the main metabolism and power structure to characterize the
urban sphere and urban water flows (Gabriel, 2014; Grove, 2009). This
predominant urban focus has also entailed the inevitable disavowal of
“struggles over meanings and practices of nature and the city that shape
identities that make some forms of urban metabolisms possible while fore
closing others” (Escobar, 1999, p.209). Thus, to reveal the many di
mensions of the urban water struggle, urban political ecologists have
started exploring other scales of analysis—e.g. the households and the
body— whilst also developing a closer engagement with post-structural
theories (Gabriel, 2014). This post-structural turn has brought the
attention of urban political ecologies toward discourse, environmental
imaginary and subjectivity in the constitution of urban spaces (Doshi,
2017; Gabriel, 2014; Grove, 2009). In this way, urban political ecology
has moved away from conceptualization of power that emphasize power
over to capture instead, the way multiple or nonlinear power dynamics
become embedded in complex assemblages or web of socionatural re
lationships (Ahlborg and Nightingale, 2018). This multidimensional
understanding of power conceives subjects as produced and existing
within a set of socioenvironmental relations infused with power
(Nightingale, 2011). Political ecology and other critical studies have started to account for
emotional responses, and the manner in which emotions reshape (the
experience of) water use, access and control (Truelove, 2019; Sultana,
2015, 2011). Yet, those studies tend to focus on the everyday struggle of
poor or low-income dwellers rather than on the emotions of the elite. This dominant focus becomes problematic when it limits understanding
of the way elites can disproportionately control and use natural re
sources at the expense of the environment and less privileged social
groups (Beaverstock et al., 2004). Direct quote by: KSWC “All these little things (change practices of water use) take more time and
if you are not motivated emotionally, it becomes a burden. The fantastic
thing of the drought in South Africa, was that we have genuinely felt the
fact that we were going to run out of water. The drought made us see it as
very serious, and understand it. So there has been a permanent shift. * Address: Uppsala University, Department of Earth Sciences, Villavagen. 16, 75236 Uppsala, Sweden. E-mail addresses: elisa.savelli@geo.uu.se, savelli.elisa@gmail.com. E. Savelli Geoforum 141 (2023) 103746 suspend the Day Zero alert. According to local water institutions and
international media, apart from an unexpected spell of rainfall, the city
avoided a water shortage thanks to the sacrifice of local residents who
complied with stringent water restrictions, altered their consumption
practices, and paid heed to their daily water consumption (Baker, 2018). power relations through which water resources are produced, distrib
uted and transformed within a city. Across similar studies, urban water
crises can be explained as mechanisms of exclusion from and access to
water (Swyngedouw, 1997) or as material and discursive production of
water scarcity (Bakker, 2000, 2003; Budds, 2004; Loftus, 2007; Mehta,
2010, 2003; Kaika, 2003; Otero et al., 2011; Millington, 2018). Overall,
political ecological explanations of water crises, which encompass ex
periences of drought, water shortage, or uneven water access, are always
produced within certain political spaces and thus result from prevailing
power structures (Swyngedouw et al., 2002). i One of the main aspects highlighted by national and international
media about Day Zero was the fact that the water crisis had seemingly
affected a large urban area of affluent inhabitants. According to pre
dominant interpretations of Cape Town water crisis, well-off inhabitants
were the main social group affected by the drought and sacrificed the
most relative to other social groups (Warner and Meissner, 2021; Baker,
2018). Other in-depth analyses of Cape Town’s water crisis reveal a
more irregular and uneven picture of the crisis: where poor and black
suburbs experienced heavy shortages while white and more privileged
households pay for additional consumption and rely on alternative
water sources (Enqvist et al., 2022; Savelli et al., 2021; Robins, 2019). Nonetheless, the water crisis provoked a significant emotional response
among affluent Capetonians, who, as the initial quotes point out, had to
adjust their everyday water use practices (K¨ohlin et al., 2018). Direct quote by: KSWC Then, this study describes the
privileged emotional geographies of Cape Town’s water crisis alongside
discussing the socioenvironmental implications of the elite’ emotional
responses. Next, the paper elaborates on the materiality and temporality
of privileged emotions to explain their genesis as well as their discursive
and material implications. This work concludes by highlighting the
scholarly potential of a focus on emotions and, in particular, the emo
tions felt by the elite. By revealing the unconscious, privileged emotions
can help retrace the root cause of the unjust and unsustainable water
practices of the elite. 2. Theoretical framework 2.1. Urban political ecology of water crises: From the urban to the body
scale 2.2. The place of emotions in urban water struggles Inspired by post-structuralists ideas on power and subjectivities,
urban political ecology is able to explore how different power relations
and inequalities are continuously (re)produced throughout subjective
imaginaries, perceptions, beliefs, and everyday interactions with socio-
natures (Gonz´alez-Hidalgo and Zografos, 2020; Nightingale, 2011,
2013; Sultana, 2009; Blackman et al., 2008). Theories of affects in
human geography conceptualize this subjective process through the
notion of affect, or the transpersonal capacity that a body has to be
affected and to affect (Anderson, 2006). Affect is distributed between,
and can happen outside, bodies which are not exclusively human, and
might include technologies, things, non-human matter, or discourses
(Curti et al., 2011; Lorimer, 2008; Massumi, 2002). According to this
theory, subjectivities are then not contained within the body or the
psyche, but emerge relationally. One way in which subjectivities emerge
(and affects occur) is emotionally (Massumi, 2002). Precisely, emotions
are the most intimate and intense way in which subjects experience and
respond to affective modification to themselves (Lorimer, 2008;
Anderson, 2006; Massumi, 2002). The human geographers Anderson
and Smith (2001, p.7), define emotions as those “key set of relations
through which lives are lived and societies made”. This paper sets to unravel
how emotions work through and, both materially and discursively, (re)
shape socioenvironmental relations (Askins and Swanson, 2019; Doshi,
2017). Urban political ecologists have also started to consider emotions as
their object of inquiry (Gonz´alez-Hidalgo and Zografos, 2020; Truelove,
2019; Doshi, 2017; Gonz´alez-Hidalgo, 2017; Ryan, 2016; Sultana, 2009,
2015, 2011, 2020,2021; Kearney and Bradley, 2009). In her plea for an
embodiment of urban political ecologies, Doshi (2017) describes
emotional geographies as fundamental to understand how socionatures
are materially and discursively produced at a body scale. Similarly,
Sultana (2015) and Truelove (2019) reflect on the way emotions come to
matter in the production of resources access, use, and control. They
show that emotions are able to influence and therefore reshape the
everyday water practices (Truelove, 2019; Sultana, 2020, 2015, 2011). For instance, emotional distress can affect women choices between
drinking contaminated water or paying additional costs for slightly
better-quality water (Sultana, 2020). Thus, these scholars emphasize the
role that emotions play in re-negotiating and re-producing water-society
relations. Gonz´alez-Hidalgo and Zografos (2020) argue that emotions provide
a conceptual ground for investigating environmental conflicts. Gonz´alez-Hidalgo (2017) builds on this idea, describing the relationship
between emotions, power, and environmental conflicts in Chile and
Mexico. 2.1. Urban political ecology of water crises: From the urban to the body
scale Recently, the body, i.e. the place of subjective experience, has
become a recurring scale of inquiry in the urban political ecology of
water (Truelove, 2019). Doshi (2017, p.126) calls embodied political
ecologies those studies that have enriched capitalists socionatures with
the “fleshy, messy and indeterminate stuff of everyday”. A focus on the body
makes visible the intimate, unconscious, and power laden processes Work on the (urban) political ecology of water has largely explained
the production of urban water crises by politicizing understandings of
water flows across intra-urban spaces (Loftus, 2009). Mostly concerned
with water injustice, (urban) political ecology of water unearths the 2 E. Savelli Geoforum 141 (2023) 103746 silenced, belittled and undermined by both academia and in the public
sphere (Askins, 2019). Emotions were judged as too personal or irra
tional and therefore irrelevant for the study of human geography
(Davidson and Milligan, 2004). Thinking emotionally has been “implic
itly cast as a source of subjectivity which clouds vision and impairs judge
ment” (Anderson and Smith, 2001, p.7). Nevertheless, Anderson and
Smith (2001) argue that neglecting emotions as an object of study,
generates an incomplete understanding of the world: one that overlooks
a multitude of social relations that reshape every subjectivity and
eventually produce the society we live in. Since Anderson and Smith’s
(2001) “emotional turn” in geography, many scholars have started
connecting with emotions and their intersection with subjectivities,
power and ultimately, socionatures (Askins, 2019; Askins and Swanson,
2019; Pile, 2010; Sharp, 2009; Bondi, 2014, 2006; Nightingale, 2011;
Thien, 2005; Bennett, 2004, 2009; Davidson and Milligan, 2004). These
scholars stress the fact that all human experiences and interactions with
other people, objects or places are constituted through emotional ge
ographies. As Bennet (2004) explains, every rational or irrational action
necessitates a facilitating emotion to be initiated and therefore achieved. That is: without emotion there is no action. Thus, if on the one hand,
emotions are essential because they create the lived experience (Ben
nett, 2004, 2009), on the other, emotions matter because they affect the
way humans interact with places (Bondi, 2006). As Bondi (2006) argues,
accounting for emotions allows human geographies to develop more
nuanced and integrated understanding of the complexity of human re
lations with space. which differentiate subjectivities and their everyday experiences (Doshi,
2017). 2.1. Urban political ecology of water crises: From the urban to the body
scale For instance, it is within embodied experiences and everyday
practices that the intersectional axes of gender, class, race, religion and
other social differences become visible and produce subjective experi
ences. Amongst other feminist political ecologists, Truelove (2011) and
Sultana (2020) have focused on the body to map the everyday struggle
over access to water in India and Bangladesh respectively. Sultana
(2020) relates daily water practices with gender, class and other vul
nerabilities, to demonstrate that water access is essentially an embodied
process. Similarly, Truelove (2011) focuses on the embodied experience
of water insecurities to explain how and why inequalities perpetuate or
exacerbate themselves, even when both water quantities and qualities
are improved at larger scales — e.g. the urban. These examples of an
embodied political ecology of water disclose undertheorized dimensions
and dynamics in the production of water insecurities and inequalities
(Truelove, 2019). Nightingale (2011) explains that attention to
everyday, seemingly mundane, spatial practices gives insight into how
people produce a particular relationship with ‘others’ including their
environments, that are rarely ecologically neutral. Ultimately, a focus on
embodied subjectivities serves to unravel the power-laden web of soci
oenvironmental relations that produces urban spaces and in turn,
facilitate more transparent and just political transformations (Doshi,
2017; Gabriel, 2014). 2.3. The need to connect with privileged emotions To delineate and characterize the privileged emotional geographies
of an extreme water crisis, this paper refers to Cape Town’s elite as the
empirical foci for such analysis. The selection of Cape Town as a case in
point is motivated first of all by the fact that from 2015 until 2017 the
City has experienced a prolonged drought that in 2018, culminated in a
water crisis whose implications are still visible at the time of this writing
(Twidle, 2021; Maxmen, 2018). Secondly, the case of Cape Town is also
representative of unequal urban development which manifested through
accumulation of privilege and capital in white areas, and segregation of
black or coloured inhabitants to the outskirts of the city (Lemanski,
2007). More recently, the rise of neoliberal ideology and policy-making,
further concentrated development and investments in the richest and
more profitable areas of the city, whilst neglecting the rest of the city’s
space and people (McDonald, 2012; Miraftab, 2007; 2012) This unequal
urbanization has led to the establishment of elitist neighbourhoods
which are able to benefit from privileged locations sheltered from major
urban risks, and enjoy relatively higher level of public services (Mir
aftab, 2007, 2012). A focus on Cape Town’s elites provides this study
with valuable insights into how the emotional and social dimensions of a
water crisis intersect with broader patterns of power and inequality in
the city alongside their historical legacy. i The prevailing focus on the emotional struggle of the oppressed re
flects the emphasis that political ecology and human geography have
put on the study of those who live at the margins, as well as their re
lationships to socionature. Most of these works examine the experience
of people who are structurally disadvantaged rather than exposing the
lives and practices of those with privilege, the elite (Norgaard, 2012). An
explanation of such tendency can be found in the work of the anthro
pologist Nyamnjoh (2012) who observed that white privileged South
African have often been beyond ethnographic contemplation because
they have the same genealogy as the majority of anthropologists. The
glaring invisibility of the elite in environmental studies is often justified
by the fact that (super)rich appear somewhat inaccessible to researchers
or because privilege is also something that academic scholars often
enjoy (Argüelles, 2021; Savage and Williams, 2008; Beaverstock et al.,
2004). These aspects make critical analysis of privilege more difficult if
not uncomfortable (Argüelles, 2021). 2.2. The place of emotions in urban water struggles According to her thesis, thinking emotionally reveals how and
why certain subjectivities come to reproduce or subvert hegemonic
powers and their actions. Askins and Swanson, (2019) explain that
emotions reproduce the multiple and non-linear power relationships
that have shaped subjectivities in the first place. They assert that
oppression and injustice are, and have to be understood, as deeply
emotional. In this sense, thinking emotionally becomes essential to fully
grasp “the embodied and unconscious dimensions of oppression”. Bonilla-Silva (2019) takes this argument further, focusing specifically on
racial oppression. He states that to eradicate racism it is necessary to
uproot the visible component of oppression as well as to demolish its
emotional skeleton. Neuroscientists’ theory of constructed emotion offers valuable in
sights into the fundamental link between subjectivities and emotions
(Barrett, 2017). Instead of subjectivities, these scientists use the notion
of emotion concepts to define the past references that the brain uses to
make sense of incoming sensory inputs and in turn, construct emotions. By this account, emotions are not uncontrollable or hardwired brain
reactions but rather responses to external stimuli constructed upon past
experiences or stereotypical conceptions of reality (Yowei and Natisse,
2017; Barrett, 2017; 2018). This understanding highlights that emotions
would not exists without emotion concepts or subjectivities. Furthermore,
this conceptualization is particularly useful to explain how power
emerges through emotions and in turn, how socioenvironmental trans
formations of the urban waterscape occurs. Hence, emotions and affects
are crucial to explain issues of social and spatial justice (Askins and
Swanson, 2019). Throughout these case studies political ecology explains how emo
tions are relationally produced through interactions between people and
places (Sultana, 2015). Emotions represent the connective tissues that
link human experiences with geographies of places (Gonz´alez-Hidalgo
and Zografos, 2020; Kearney and Bradley, 2009; Davidson and Milligan,
2004). More precisely, emotions mediate in a significant way the re
lationships between the psychic and the social, the individual and the
collective, the human and the non-human. The significance of studying Despite being an integral part of the human world, the study of
emotion in human geography and political ecology has largely been 3 Geoforum 141 (2023) 103746 E. Savelli emotions lies within this relationality. As such, rather than abstract
concepts, emotions should be considered as connective medium that
reveal the human relationships within others and with the context of
these interactions (Bennet, 2009). 2.2. The place of emotions in urban water struggles The relationality of emotions also
harmonizes with neuroscientists’ theory of emotion concepts and con
structed emotions, for it discloses the socio, cultural, historical and po
litical milieu that produces them (Bonilla-Silva, 2019; Ryan, 2016;
Bennet, 2009; Pile, 2010). It is through such relationality that emotions
reveal power dynamics and their temporality. Colonialism (Kearney and
Bradley, 2009), racism (Bonilla-Silva, 2019), class discrimination (Bes
wick, 2020) and patriarchy (Sultana, 2011) are some of the legacies that
emotional geographies have been able to retrace. reduce risks for the privileged at the expense of the environment and
disadvantaged communities (Finewood, 2012; Otero et al., 2011). Whilst crucial to expose elites’ privileged conditions, these works do not
yet reveal who elites are, how they think and feel, nor how they un
derstand, justify, and maintain their power (Savage and Williams,
2008). In turn, urban studies have yet to reveal every dimension of
power that coalesces into the elites’ privileged patterns of control and
use of resources. This paper attempts to address this research gap by
closely investigating the emotional experience of privileged Capeto
nians, in order to examine “how” and “why” elites are able to perpetuate
their privileges both materially and discursively, and ultimately exac
erbate socioenvironmental injustices (Norgaard, 2012). The importance of understanding privilege lies in the fact that the
elites bear the utmost responsibility for reshaping the world we live in
due to their disproportionate control and use of resources (Beaverstock
et al., 2004). To study elites, then, is to study their material and
discursive control over, value of, and distribution of resources (Bea
verstock et al., 2004). Without accounting for this perspective, political
ecologists and other critical scholars risk hampering their strive for
justice by rendering invisible the disproportionate power of the elite and
the socioenvironmental implications of this power (Argüelles, 2021;
Freudenburg, 2005). As Norgaard (2012) claims, the denial or natural
ization of privilege inevitably silences the needs and voices of the
oppressed. Thus, by going from marginalised areas to places where
privilege is enjoyed, scholars can advance their understanding of
(environmental) injustice with insights about subtler forms of in
equalities and the conditions that enable, justify and sustain those in
justices in the long term (Argüelles, 2021; Park and Pellow, 2011). By highlighting the relationality and temporality of emotions,
emotional political ecologies have shown that emotions can channel and
reproduce deeply rooted power dynamics. 2.2. The place of emotions in urban water struggles In particular, this literature
focuses on the emotions of those that struggle whilst largely neglecting the
emotions of those who don’t. For instance, Kearney and Bradley (2009) engage with the emotional
geographies of indigenous people in North-eastern Australia, whose
lives have been profoundly reshaped by colonial invasions. Similarly,
Gonz´alez-Hidalgo (2017) retraces negative emotions of indigenous
people in the Mapuche territories of Chile to understand the way anger
and sorrow enable local resistance to extractivism. Truelove (2011) and
Sultana (2011) discuss the emotional violence, hardship, and distress
that characterise water insecurity of underprivileged women. Overall,
by exclusively exposing the struggle of the disadvantaged, these studies
have left uncharted and in turn, undisputed the emotions of those who
hold privilege and power. As Nyamnjoh (2012) argues, these scholars
have provided an incomplete interpretation of the relationships and
structures of dominance. 2.3. The need to connect with privileged emotions Source: elaboration of the author using data from COCT, 20 levels of water consumption which most privileged households usually
sustain with private boreholes or rainwater harvesting systems in
addition to the municipal water supply (WWF, 2020; McDonald, 2012;
Savelli et al., 2021). helped overcome certain power imbalances (Rusca, 2018; Castel´an
Cargile, 2016; Roberts and Lunch, 2015). The semi-structured in
terviews performed in low-income areas were mostly facilitated and
supervised by non-governmental organizations. Instead, when working
with elites, the researcher acted more independently. To gain access to
elite households, she reached out to a number of residents and tax
payers’ associations of different wealthier suburbs. With the help of
these associations she contacted several potential respondents and
continued her research using a snowball sampling technique. In addi
tion, to ensure the trust and respect of these interviewees, the researcher
had to establish her credibility and authenticity as a knowledgeable
academic researcher whilst also conveying the importance of this
research and the crucial role that the interviewees had in this process
(Lillie and Ayling, 2021). Ultimately, whilst performing the interviews,
the researcher deliberately chose not to discuss the opinions of the in
terviewees, in order to avoid making them feel uncomfortable by chal
lenging their assumptions or preconceived ideas (Mason-Bish, 2019). To
maintain the anonymity of the interviewees, the interviews are refer
enced in the main text and footnotes by using a unique code with name
and neighbourhood initials of the interviewee. This study combines sets of qualitative and quantitative data. Pri
mary qualitative data was collected through semi-structured interviews
and field observations carried out in Cape Town from May 2019 until
March 2020. Overall, the researcher interviewed 32 affluent Capeto
nians. Despite post-apartheid policies of socio-economic integration, the
social status of Capetonians continues to be deeply intertwined with
race, and more affluent neighbourhoods remain predominantly white
(Turok et al., 2021). In addition, about 7 additional interviews with
water authorities provided a comprehensive understanding of the way
the water crisis was handled and water restrictions implemented. The
semi-structured interviews aimed to explore the personal and intimate
experience of the participants during the drought. The questions focused
on their daily water usage, along with experiences, coping strategies and
emotions they felt throughout the water crisis. The participants were
given ample time to reflect, delve deeper into their experiences, and
share their perspectives. 2.3. The need to connect with privileged emotions Nevertheless, some geographers, environmental sociologists and
environmental justice scholars have begun to confront and analyse
privilege along with its socioenvironmental entanglements (Argüelles,
2021; Norgaard, 2012; Savage and Williams, 2008; Freudenburg, 2005;
Beaverstock et al., 2004). These studies have come to view privilege as
both the exclusive accumulation of environmental amenities and the
protection from those environmental harms that other less privileged
groups are constantly exposed to (Park and Pellow, 2011; Freudenburg,
2005). Whilst examining the socially constructed and disproportionate
patterns of access to and control of the biophysical environment, this
research does not directly engage with the embodied or subjective ex
periences of privilege (Savage and Williams, 2008; Freudenburg, 2005). Similarly, urban political ecology has mostly considered privilege within
processes of urbanization, suburbanization, or gentrification (Atkinson
et al., 2017; Finewood, 2012; Otero et al., 2011; Atkinson and Flint,
2004). For instance, these scholars describe urbanization or develop
ment trajectories as elitist processes which accumulate advantages and The selection of the elites as study sample is based on the official
Socio-Economic Index employed by the Municipality of Cape Town
(COCT, 2014). The Socio-Economic Index classifies Cape Town’s social
groups on the basis of their income levels, education, type of housing,
and access to basic services (COCT, 2014). Based on these variables,
each neighbourhood attains different scores which range from very
good, good, average, needy and very needy socioeconomic conditions
(Fig. 1). This paper considers elites only the urban areas classified as
“very good”. The spatial representation of the Socio-Economic Index
indicates that these elitist neighbourhoods are mostly located in the
western suburbs of the city and their inhabitants enjoy very high levels
of income. Usually, these areas enjoy a number of privileges which
include larger houses, more than one car, gardens and/or swimming
pools (McDonald, 2012). These amenities can lead to unsustainable 4 Geoforum 141 (2023) 1037
Fig. 1. Spatial representation of Socio-Economic Index of Cape Town neighbourhoods. Source: elaboration of the author using data from COCT, 2014. Savelli E. Savelli E. Savelli Geoforum 141 (2023) 103746 Fig. 1. Spatial representation of Socio-Economic Index of Cape Town neighbourhoods. Source: elaboration of the author using data from COCT, 2014. Fig. 1. Spatial representation of Socio-Economic Index of Cape Town neighbourhoods. Source: elaboration of the tation of Socio-Economic Index of Cape Town neighbourhoods. 2.3. The need to connect with privileged emotions In the next section, the analysis of these diverse sets of data will
delineate the privileged emotional geography of Cape Town’s water
crisis. Whilst wealthy Capetonians did experience a multitude of emo
tions with different intensities and at times contradictory connotations,
this paper analyses and focuses on the prevailing emotions that have
more extensively reshaped their responses to the drought and water
crisis. The first and more prominent emotional response reveals strong
apprehension and fear. Secondly, privileged emotions further escalate
into frustration and anger. Ultimately, a sense of pride and satisfaction
seems to expand across most privileged households. Qualitative interviews were combined with recorded images or
videos to provide a visual account of the context characterizing the
water practices of elites in their households (see Fig. 2). Furthermore,
this material account of drought experiences was enriched with quan
titative assessments of daily water consumption (averaged annually) at
the urban and household levels. Finally, primary data was cross-
referenced with information obtained from recent literature and the
City of Cape Town’s data portal. Throughout the entire research process, the author acknowledged
her positionality and its effects on both data collection and data analysis. In general, the researcher remained reflexive and thus conscious of her
white race, European origin and relatively privileged socio-economic
status. To avoid that her subjectivity would overly affect the research
process, she undertook most of the work based on the advice and support
of local contacts: such as civil society, non-governmental organizations
and South African academics. Filming the interviewees in their house
holds also established a more collaborative research approach and 4. The privileged emotional geography of Cape Town water
crisis I can’t do that!”.2 In addition,
having to become conscious of the need to ration water became a serious
fear and a reality perceived as impossible by some elites.3 “We were
worried of using buckets of water to flush the toilets” declared a father of
two kids living in an oceanfront mansion.4 Whilst many well-to-do
Capetonians expressed apprehension about the use of buckets to flush
their toilets, or not having the option to take a bath or a shower,5 others
were particularly concerned about their gardens and swimming pools.6
“The simple idea of not having water for a full bath, watering my garden or
filling up my swimming pool scared me” said LKGP. emotional response of the Capetonian elite was characterized by fear
and apprehension. While the intensity of each such emotion was
expressed differently, these emotions reflect the elite’s anxiety linked to
the prospect of having to temporarily give up on their relative level of
accustomed comfort—in this case, access to a limitless supply of water
within the premises of their household.1 The interviews consistently
show that the mere idea of standing in a queue with others, waiting for a
tank to be filled, or carrying the water buckets home, elicited a sense of
fear and apprehension among the Capetonian elite. As a white and
wealthy woman openly declared: “I was thinking, how will I be standing in
a queue and carry 25 L of water myself? I can’t do that!”.2 In addition,
having to become conscious of the need to ration water became a serious
fear and a reality perceived as impossible by some elites.3 “We were
worried of using buckets of water to flush the toilets” declared a father of
two kids living in an oceanfront mansion.4 Whilst many well-to-do
Capetonians expressed apprehension about the use of buckets to flush
their toilets, or not having the option to take a bath or a shower,5 others
were particularly concerned about their gardens and swimming pools.6
“The simple idea of not having water for a full bath, watering my garden or
filling up my swimming pool scared me” said LKGP. 7 Direct quote by: BMC.
8 Direct quote by: HMTL.
9 Direct quote by: APC. 4. The privileged emotional geography of Cape Town water
crisis 4.1. Fear and apprehension Triggered by the news of an impending Day Zero, the primary 5 E. Savelli Geoforum 141 (2023) 103746 Geoforum 141 (2
Fig. 2. Videography material used to characterize Cape Town’s most privileged households and neighbourhoods. Image: The Author. Fig. 2. Videography material used to characterize Cape Town’s most privileged households and neighbourhoods. Image: T The sudden possibility of losing access to water and the urgent need
to rapidly change their water use habits, had a profound emotional
impact on wealthy Capetonian. However, in most instances, these
intense emotions were not instigated by an actual shortage of water. In
fact, this study found that while low-income households experienced
recurring water shortages, wealthy households never faced any shortage
of water, nor did they ever have their public water supply cut off. When
asked about the material experience of the water crisis, a woman from a
wealthy neighbourhood admitted “We (the well-to-do classes), only had
restrictions, that’s all! It didn’t affect us much”.7 That said, even though
each privileged household interviewed professed to never having
experienced a time without access to water; fears and apprehensions
continued to pervade their daily lives. emotional response of the Capetonian elite was characterized by fear
and apprehension. While the intensity of each such emotion was
expressed differently, these emotions reflect the elite’s anxiety linked to
the prospect of having to temporarily give up on their relative level of
accustomed comfort—in this case, access to a limitless supply of water
within the premises of their household.1 The interviews consistently
show that the mere idea of standing in a queue with others, waiting for a
tank to be filled, or carrying the water buckets home, elicited a sense of
fear and apprehension among the Capetonian elite. As a white and
wealthy woman openly declared: “I was thinking, how will I be standing in
a queue and carry 25 L of water myself? 1 Reference from interviews with: LCC; GCC; CNC; HMTL; BPC; GJGP; LKGP;
SMGP; EHHB.
2 Direct quote by: GJGP.
3 Reference from interviews with: GJGP; BPC.
4 Direct quote by: SMGP.
5 Reference from interviews with: CNC; LKGP; SMGP; EHHB.
6 Reference from interviews with: LKGP. 4. The privileged emotional geography of Cape Town water
crisis “We (the wealthy Capetonians) were frightened at the idea of queuing for
water or waiting for (water) tanks like those poor people”.8 This quote ex
emplifies the fact that rather than the material scarcity of water, it was
the prospect of becoming like them —the low-income dwellers that al
ways struggle to access water—to amplify the fear of drought amongst
the elite. It was such fear that compelled the elites to take urgent actions
and avoid those unthinkable imaginaries. In fact, each well-to-do
Capetonian interviewed declared to have found an alternative to the
municipal water. A very wealthy man in Constantia admits that “When
the municipality started with these restrictions, we decided to install our own
systems”.9 Apart from stocking hectolitres of drinking water bottles in 6 E. Savelli Geoforum 141 (2023) 103746 well-to-do Capetonians and eventually, reconfigured Cape Town’s urban
space. their garages, wealthier Capetonians were able to install new infra
structure to increment their water availability. Depending on their
financial means and the location of their households, some drilled new
boreholes, whilst others added water tanks on their rooftops and pur
chased new water pumps or filters to ensure additional water would be
suitable for consumption.10 Through the material transformations of elites’ physical space and
water access, the emotions of fear and apprehension end up reproducing
and exacerbating existing urban inequalities in multiple ways. Firstly, by
building private water systems, the elite can exclusively benefit from
additional public water at the expense of the rest of the population and
local water resources. Secondly, their exclusive water security can in
fluence the urban water demand and eventually stop public policies that
would instead benefit the entire population and the environment. Thirdly, the construction of security systems in the elites’ households
can create physical and social barriers, exacerbating a sense of exclusion
from the wider community, while also leading to reduced investments in
public safety measures. Thus, the fears perceived by Capetonian elites carry with them the
material power of reshaping the urban space. This study reveals how a
palpable sense of fear and apprehension among elites triggered them to
install new personal water systems, drill boreholes on private premises
and, ultimately, transform Cape Town’s hydrosocial configurations. Precisely, the emotions of fear prompted the unsustainable and unjust
processes of enclosure and dispossession of common water resources
(mostly groundwater) for the sole disposal and benefit of privileged
users. 4.2. Frustration and anger Emotions of fear and apprehension amongst the elite gradually
developed into and intertwined with emotions of frustration which often
escalate to become anger. In general, these emotional responses were
directed to the Western Cape government in part, but mostly the town
ships, informal settlements or black neighbourhoods as they were seemingly
unable to save water the way they should. From the interviews it be
comes clear how the elite was angry toward the government because it
did not prevent the water crisis but in particular, as the government did
not stop what the elite considered the disrespectful behaviour of the
black low-income population.18 Some privileged Capetonians did share
their concerns for low-income dwellers and their limited capacity to face
such a crisis.19 However, most of the interviewees, directed emotions of
anger and frustration towards “these people”.20 Irrespective of underly
ing dynamics of power and differences between identities, race, or social
status, the elite wantonly accuse them of not complying to the city’s
restrictions, wasting water, being uneducated, not caring for environ
ment or community, having too many children, and invading the city.21
By fuelling similar accusations, emotions of anger and frustration also
contributed to reinforce racial and class dominance thereby exacer
bating the pre-existing tensions between the elite and the black
population. The other most prevalent fear and apprehension shared by the elite
was related to the safety and the security of their property or physical
person. Many interviewees declared to be worried about the possibility
of having to share their water with other households.12 A woman clearly
stated that she was not worried to run out of water but only “afraid of
those people who don’t have water”.13 Indeed, a young white businessman
recalls the fear among his parents and neighbours “because they were
concerned that in time of drought, people might break into their properties just
to try to take hold of their water”. The young man concludes that “to
prevent any aggression, these families decided to install screens to protect
their boreholes and water tanks14”. 4. The privileged emotional geography of Cape Town water
crisis A hydro-census performed by the World Wildlife Fund in New
lands, one of Cape Town’s wealthier suburbs, reported a steep increase
in the number of boreholes drilled by residents over the course of the
drought and the subsequent water crisis, with around 50 % of boreholes
surveyed in the neighbourhood drilled between 2017 and 2019 (WWF,
2020). Similarly, the owner of a water-well drilling company stated that,
during the drought and the water crisis, all his clients had completely
ceased to use municipal water and relied exclusively on borehole water. In fact, many of the new installations that his company completed
during this period, consisted of water pumps or filters to ensure that all
alternative water was safe to drink.11 These hydrosocial reconfigura
tions of the city epitomize the link that emotions constitute between
subjective experience and geographies of places (Gonz´alez-Hidalgo and
Zografos, 2020; Kearney and Bradley, 2009). 10 Reference from interviews with: GCC; APC; SWC; BPHB.
11 Reference from interview with: SWC.
12 Reference from interviews with: KKC; APC; CNC.
13 Direct quote by: KKC.
14 Direct quote by: SWC.
15 Reference from interviews with: KCC; ADC; SWC.
16 Reference from interviews with: SWC; KKC; APC.
17 Reference from interview with: KKC. 18 Reference from interviews with: GJGP; HMTL.
19 Reference from interviews with: HMTL; BLBFB.
20 Reference from interviews with: ADC; GCC; CNC; KKC; EHHB.
21 Reference from interviews with: GCC; ADC; CNC;
22 Direct quote by: KKC.
23 Reference from interviews with: GCC; BPC; LKGP; KKC; ADC; CNC; EHHB.
24 Direct quote by: CNC.
25 Direct quote by: GCC. 4.2. Frustration and anger A white woman, resident in the elite neighbourhood Clovelly,
seemed frustrated at the idea that: “In really poor areas […] they are not
going to turn the taps off even on Day Zero, (thus) other people get tense when
they look at these people and think it is unfair that they sacrifice while these
people get water without restriction”.22 Frustration amongst the wealthier
Capetonians was caused by what they perceived as uncivilized water
practices of poorest residents, whom, they argued, did not comply with
municipal water restrictions. This perceived unfairness unsettled the
elite, who felt they were trying their best to adapt to such an extreme
situation, whilst these people (in townships or informal areas), would just
leave their (communal) tap open, keep the water running, establish
illegal connection and “do what they like” because “they just do not
care”.23 A coloured and wealthy woman in Constantia bitterly declared
that “we are not all (raised) by the same system. It is unfair that we are saving
water so that other people can waste it. We should all have the same mindset. (We should all be) saving for our city”.24 The prevalent idea amongst the
most privileged Capetonians was that every-one “need(ed) to change their
habits, not just the rich people.”25 Once more, these emotions were not rooted in concrete experience of
thefts, intrusion or aggression. Instead, the interviews revealed that the
main source of fear and apprehension was the idea that scarce water
resources may trigger envy and hatred for the elite by less privileged
individuals.15 The elite imagined that their privileged access to addi
tional water resources would raise envy and discontent amongst low-
income groups. These fears and apprehensions seemed to persist and
increase even after having secured new boreholes or additional rain
water tanks. Most elites would still fear intrusion, robbery or physical
aggression during the water crisis. In turn, several elites decided to
install alarms, video security systems, or other technologies to protect
both homes and water infrastructure from the possibility of intrusion or
attack.16 One privileged Capetonian even went so far as to suggest the
need to create water walls which she imagined as physical barriers which
separate each household’s water source from other households.17 Ulti
mately, the fear that lower-income dwellers could attack or intrude upon
the elite physical space, materially reshaped the properties of many 18 Reference from interviews with: GJGP; HMTL. 4.2. Frustration and anger A study found
that whilst informal settlements use about 4 % of the total water
available, well-off inhabitants account for almost 70 % of domestic
water consumption (Robins, 2019). In an earlier publication, Mc Donald
(2012) reports that, of the water consumed by affluent suburbs, 35 %
was used to water gardens, and over 10 million litres per day to fill up
about 68.000 swimming pools. Elite households very often exceed 2000
Litres per day, whilst households in informal settlements typically
consume about 40 Litres per day (Savelli et al., 2023). Furthermore,
during drought restrictions, a large number of township dwellers or low-
income households were temporarily cut off from the water supply
because they were unable to pay for increased water tariffs. Unlike the
elite, these households could not rely on alternative water sources and
very often ran out of water (Enqvist et al., 2022; Savelli et al., 2021;
Robins, 2019, Millington and Scheba, 2021). These quantitative ac
counts, explain the manner in which privileged emotions discounted the
elite of the responsibility they have in consuming the largest amount of
water across the city (McDonald, 2012). In this sense, the anger and
frustration felt by the elite seem to disavow existing inequalities, and
eventually justify their unsustainable water consumptions. By doing so,
these privileged emotions can reinforce the racial and class-based power
dynamics that allow the Cape Town’s elite to maintain their privileged
status across the city. The frustration and anger of the elite was also exacerbated by the fact
that low-income neighbourhoods do not always pay for the water they
use because they either access water by communal water points or
receive a certain amount of water for free each month. Thus, they
seemingly constitute a financial burden for wealthier households that
have to subsidize water to low-income dwellers of Cape Town.26 Some
elites felt frustrated or even angry about the fact that the South African
constitution accords the right to drinking water to every citizen. As this
human right entails that the one that can afford “have to pay the water
for”27 the most disadvantaged population. A white and wealthy resident
of Hout Bay is convinced that “The problem is amongst the black people
who don’t care because they don’t have to pay for water. 4.2. Frustration and anger She also
added that “We don’t have that many dams available for the influx of people
coming into Cape Town […] and using our water source, there is just not
enough to go around for everybody”.37 Thus, some of the elite wish these
people to leave and go back to their home.38 Capetonians point at the lack of education of informal settlers, town
ships or black people. What upsets and frustrates some privileged
Capetonians is the fact these people are not well informed, uneducated,
and as a result, they do not save water nor care for the environment.34 As
one young woman living in a wealthy neighbourhood stated “people
living in townships and in other impoverished areas are not listening to the
radio, and if they are listening, then they will not listen to the news but rather
listen to the music”.35 From similar emotional statements, lack of edu
cation emerges as a human condition rather than a product of historical
and unchallenged injustices. This lacking in education is also used to
point to the high fertility rate and the ‘higher’ water demand amongst
black townships or informal settlements. “The population is exploding
amongst the black people”36 says a white male living in Constantia who
then adds with visible frustration “They produce children without having a
care or concern about education, clothing, food, or water. They just produce
children. Africa is extremely badly behaved and South Africa is not
different”. Similar to ‘uncontrolled procreation’, also migration repre
sents a source of frustration amongst wealthier Capetonians. According
to them, people coming from other regions or Southern African coun
tries constitute an additional burden on local water sources. Another
woman from Constantia concluded that “there are too many people living
in the city […] feeding into the poor water resources we have”. She also
added that “We don’t have that many dams available for the influx of people
coming into Cape Town […] and using our water source, there is just not
enough to go around for everybody”.37 Thus, some of the elite wish these
people to leave and go back to their home.38 From such declarations one can sense how the elite frustration
gradually escalates into anger toward other people’s presumptive
habits. Yet these emotions seem almost paradoxical when compared
with real data on water consumption across Cape Town. 4.2. Frustration and anger […] black people
have a culture of refusing to pay and (the government) cannot let them
pay”.28 The frustration with black ‘entitlement’ to free water services
often extends to other public services provided in informal settlements
or black areas: “They think the government has to do everything for them. They get their grants. The young girls get grants when they are 18 […] Then
the whole family who is not working gets grants. 17 million people in this
country get grants”.29 Beside racial bias, similar frustrations also reveal
the elite neoliberal ideas about their city: a city where rights and power
should belong to those who have capital. In the neoliberal city imagined
by Capetonian elites, low-income dwellers should not receive any water
if they are unable to pay for it.30 It is only through the payment of water
fees that one can understand the value of water and eventually saving
it.31 According to their neoliberal ideology, water rights and other in
terventions aimed at fostering equality are considered either counter
productive or morally unacceptable.32 For the elite in Cape Town, it is
the market that restores justice by making sure that every-one is
rewarded for what they deserve.33 Whilst frustration and anger seemed to prevail amongst the elite,
elites gave space to more sympathetic emotional reactions toward black
low-income dwellers. When explaining her interpretation of the Day
Zero crisis, a white woman living in The Lake, a residential and affluent
area in Noordhoek, declared: “the people living in townships were laughing
at us (the elite). Domestic workers and shank dwellers where saying to me:
now you know what does it mean living like us… You think (that) we throw
away the water we use during the shower? We shower in a bucket and then we
use that water to wash our floors or clothes. We use it three times before it
goes away down the system”.39 To some extent such example reflects an
awareness of the harsher conditions faced in black neighbourhoods. Similarly, another wealthier Capetonian living in the well-off suburb
Fish Hoek, also mentioned the fact that people living in townships didn’t
know about the drought. She literally said that “the drought didn’t really
change anything (for low-income dwellers)” because “for them every day is
a day zero”.40 4.2. Frustration and anger 19 Reference from interviews with: HMTL; BLBFB. 20 Reference from interviews with: ADC; GCC; CNC; KKC; EHHB. 21 Reference from interviews with: GCC; ADC; CNC;
22 Direct quote by: KKC. 23 Reference from interviews with: GCC; BPC; LKGP; KKC; ADC; CNC; EHHB. 24 Direct quote by: CNC. 25 Direct quote by: GCC. 10 Reference from interviews with: GCC; APC; SWC; BPHB. 11 Reference from interview with: SWC. 12 Reference from interviews with: KKC; APC; CNC. 13 Direct quote by: KKC. 14 Direct quote by: SWC. 15 Reference from interviews with: KCC; ADC; SWC. 16 Reference from interviews with: SWC; KKC; APC. 17 Reference from interview with: KKC. 7 E. Savelli Geoforum 141 (2023) 103746 Capetonians point at the lack of education of informal settlers, town
ships or black people. What upsets and frustrates some privileged
Capetonians is the fact these people are not well informed, uneducated,
and as a result, they do not save water nor care for the environment.34 As
one young woman living in a wealthy neighbourhood stated “people
living in townships and in other impoverished areas are not listening to the
radio, and if they are listening, then they will not listen to the news but rather
listen to the music”.35 From similar emotional statements, lack of edu
cation emerges as a human condition rather than a product of historical
and unchallenged injustices. This lacking in education is also used to
point to the high fertility rate and the ‘higher’ water demand amongst
black townships or informal settlements. “The population is exploding
amongst the black people”36 says a white male living in Constantia who
then adds with visible frustration “They produce children without having a
care or concern about education, clothing, food, or water. They just produce
children. Africa is extremely badly behaved and South Africa is not
different”. Similar to ‘uncontrolled procreation’, also migration repre
sents a source of frustration amongst wealthier Capetonians. According
to them, people coming from other regions or Southern African coun
tries constitute an additional burden on local water sources. Another
woman from Constantia concluded that “there are too many people living
in the city […] feeding into the poor water resources we have”. 34 Reference from interviews with: ADC; APC; CNC.
35 Reference from interview with: CNC.
36 Direct quote by: APC.
37 Direct quote by: CNC.
38 Reference from interview with: APC.
39 Direct quote by: HMTL.
40 Direct quote by: JWFH.
41 Reference from interviews with: LCC; BPC; KKC; EHHB.
42 Reference from interviews with: LCC; GJGP; KKC; LKGP. 4.3. Satisfaction and pride When remembering the drought and the subsequent water crisis,
some women described it sad or heart-breaking. This feeling was mostly
generated by the fact they had to endure seeing the plants or lawns of
their gardens dying. They describe pulling dead grass from their garden
as a truly painful experience.41 Yet, besides these instances, wealthier
inhabitants tend to remember the water crisis as an event that led to
some positive outcomes. The reason being is that whilst taking the elite
out of their comfort zone, the municipal restrictions also made them
aware of the value of water and, ultimately, triggered them to prepare
for future water crises.42 “The fantastic thing of the drought” says KKC Often, when explaining their anger and frustration, well-to-do 34 Reference from interviews with: ADC; APC; CNC. 35 Reference from interview with: CNC. 36 Direct quote by: APC. 37 Direct quote by: CNC. 38 Reference from interview with: APC. 39 Direct quote by: HMTL. 40 Direct quote by: JWFH. 41 Reference from interviews with: LCC; BPC; KKC; EHHB. 42 Reference from interviews with: LCC; GJGP; KKC; LKGP. 26 Reference from interviews with: KKC; EEHB; ADC. 27 Direct quote by: KKC; EHHB. 28 Direct quote by: EHHB. 29 Direct quote by: EHHB. 30 Reference from interviews with: EHHB; CNC; APC; GCC. 31 Reference from interview with: EHHB. 32 Reference from interviews with: EHHB; CNC; APC; GCC. 33 Reference from interviews with: EHHB, APC; ADC; GCC; KKC. 27 Direct quote by: KKC; EHHB. 28 Direct quote by: EHHB. 38 Reference from interview with: APC. 30 Reference from interviews with: EHHB; CNC; APC; GCC 31 Reference from interview with: EHHB. 8 E. Savelli Geoforum 141 (2023) 103746 reinforce elites’ neoliberal merits thereby producing hydrosocial
reconfigurations that are both unjust and unsustainable. Specifically, the
increasing use of private boreholes risks triggering what ecological
economists and sociohydrologists define as the supply–demand cycle i.e. an unforeseen increase in water demand as a result of the expansion or
construction of additional water infrastructure (Di Baldassarre et al.,
2018; Kallis, 2010). In this case, the development of private boreholes
by elites could produce a supply–demand cycle which in the longer-
term, would deplete the local aquifers and in turn, reduce the future
water available for the entire population. 5. Discussion and conclusions: the temporality and materiality
of privileged emotions It was
indeed this sense of pride for preventing Day Zero that reinforced the
entitlement of well-to-do Capetonians to consume alternative water
sources even when consumption was unsustainable, illegal, and lacked
government authorization.53 For instance, when asked by the municipal
authorities to register his borehole, a wealthy man in Constantia refused
to do so and told them “it is nothing to do with you, it is not your water, it is
my water and I am independent of the council”.54 Race is central to understand the way affluent Capetonians perceive
themselves and their surroundings. I argue that if on the one hand these
racialized subjectivities originate from Cape Town colonial and apart
heid history, on the other they also draw upon a disavowal of the same
colonial and apartheid oppressions (Hook, 2005; Riggs and Augoustinos,
2005). What Hook (2005) defines racist disavowal is crucial here to un
derstand the ways the elite refuse to acknowledge their unsettling past
and in turn, persevere the acts of dispossession and segregation that
started during colonial and apartheid regimes. For instance, the fearful
ways well-to-do residents predicated their ownership of Cape Town
water sources, their proud assertions of their rightness of doing, together
with their angry rejection of black South African pre-eminence within
the city, are but refusals of the past. The problem with such disavowal is
not only the production of ill-founded emotions but also the fact that
such emotions risk confirming and justifying the elite racial and class
privilege. In this way, existing social injustices are reinforced and
perpetuated into the future. Along with racial bias, the emotions of well-to-do Capetonians also
reflect a neoliberal image of society made up of self-interested in
dividuals. Privileged emotions do not give space to collective or shared
feelings but rather reflect individualized, competitive, and market-based
ways of thinking. This neoliberal connotation is evident both in the
anger felt toward low-income dwellers unable to pay for water as well as
in the satisfaction that wealthier residents felt when isolating themselves
from future water crises. Significant in this respect are also the actions
triggered by privileged emotions like for instance, the enclosure of
otherwise common water resources along with the development of off-
the-grid households and gated communities which enabled well-to-do
residents to avoid rather than address a shared problem like water
shortage. 5. Discussion and conclusions: the temporality and materiality
of privileged emotions The privileged emotional geography of Cape Town water crisis re
veals the manner in which well-off Capetonians have hinged upon their
subjectivities or emotion concepts to make sense of the water crisis
(Barrett, 2017, 2018). What characterized privileged experiences of Day
Zero with emotions of fear, anger, and pride was, I argue, the distinctive
way the elite perceive themselves and their relation with the rest of the
city. Thus, the emotions depicted in the previous analysis are neither
hardwired reactions of the brain, nor indications of something objective
about the crisis. Rather, they mostly result from the privileged condi
tions of the elite which bring about preconceived images or moral ste
reotypes concerning the city, its water resources and inhabitants
(Bonilla-Silva, 2019). More specifically, this analysis has shown how the
subjectivities of Capetonian elites are deeply ensnared by racial bias and
neoliberal ideologies. l In most cases, Capetonians felt very proud of themselves for having
helped the municipality to halve water consumption and avoid Day
Zero.49 Amongst the wealthier households, people felt particularly
proud for having developed their own alternative water systems thereby
achieving independence from the municipal supply.50 In some cases,
they justified such investments as conscious efforts or ethical choices in
favour of the environment or less wealthier inhabitants.51 As the CEO of
a water-well drilling company stated, “I think another factor influencing
people’s decision to look for alternative water sources was that they are
ethically doing their best to take their bit to take pressure off the grid, so if I
can treat my borehole water […] that every 50 L taken from my borehole is
50 L of municipal water that I can give to somebody else in the townships or in
the more impoverished areas. So, I think there is that ethical play for it”.52
However, ‘ethical’ this choice might be, studies have shown that in the
long term, the usage of private boreholes or off-the-grid water systems
risks becoming unsustainable and threatening the future availability of
water resources (Srinivasan et al., 2013; Otero et al., 2011). 4.3. Satisfaction and pride Ultimately, privileged emo
tions risk perpetuating a sense of entitlement among the wealthy
Capetonians and contribute to a culture of individualism and exclusion
that undermines collective efforts to address water crises and other
environmental challenges. “was that (Capetonians) genuinely felt the fact that (they were) going to run
out of water. In this way, the drought made people see and understand how
water is important”. Most Capetonian elites felt a sense of pride and satisfaction when
remembering about the water crisis. According to the interviewees, one
of the triggers of such emotions, was the fact that the water crisis united
the City against Day Zero, their common enemy.43 However, this sense
of unification did not involve every Capetonian. More than anything,
what these wealthier Capetonians were feeling was a reunification of the
elite which eventually deepened the already existing racial and class
divisions. Many of the wealthy interviewees recall meeting with friends
and neighbours either to better understand what should be done or to
find common solutions to face Day Zero and future water crises. Some
even recall their excitement in organizing events, interviewing pro
fessors or other responsible authorities as a way to gather useful infor
mation, thereby supporting their (privileged) communities.44 In general,
more than privileged, elites consider themselves satisfied with how easily
they have escaped the water crisis without too much stress or sacrifice.45
Privileged Capetonians acknowledge their luck of living close to alter
native water sources,46 the financial abilities they have to buy additional
water sources47 or in some cases, the knowledge they have acquired and
that eased the installation of water technologies in their households.48
Here privileged emotions reflect again the neoliberal ideology of well-
to-do Capetonian, who persuade themselves that they have acquired
their water security through merit, thereby ignoring the structural
advantage that has helped the elite achieve their security (Harvey,
2007). 43 Reference from interviews with: KKC; LKGP; BLBFB.
44 Reference from interviews with: KKC; LKGP; BLBFB; APC.
45 Reference from interviews with: APC; SWC; SMGP; BMC; ADC; KKC.
46 Reference from interviews with: BMC; APC.
47 Reference from interviews with: SMGP; BMC; SWC.
48 Reference from interviews with: KKC; GCC, SWC.
49 Reference from interviews with: LCC; GCC; APC; SWC; KKC; LKGP; SMGP;
ADC; EHHB; BLBFB.
50 Reference from interviews with: KKC; GCC; APC; HMTL.
51 Reference from interviews with: SWC; KKC; HMTL; LKGP.
52 Direct quote by: SWC.
53 Reference from interviews with: HMTL; APC; GCC.
54 Direct quote by: APC. 51 Reference from interviews with: SWC; KKC; HMTL; LKGP References Sultana (2015) had already shown the manner in which emotions
come to shape the access, use and control of water resources in a city. Yet
whilst she argues that emotions matter in the everyday struggle for re
sources (Sultana, 2015), here I argue that emotions matter and signifi
cantly, also in the extra-ordinary privileges enjoyed by the elite. What
distinguishes privileged and extra-ordinary emotions from everyday
struggles is both their power and intensity. Whilst the crisis makes
socioenvironmental changes occur more quickly and intensely, the
disproportionate privilege of the elite marks the power of such changes. In fact, privilege emotions of water crisis have increased the control and
use of water resources of the elite, whilst maintaining other social
groups in their everyday racial and water struggles. Ahlborg, H., Nightingale, A.J., 2018. Theorizing power in political ecology: the’where’of
power in resource governance projects. J. Political Ecol. 25 (1), 381–401. Anderson, B., 2006. Becoming and being hopeful: towards a theory of affect. Environ. Planning d: Soc. Space 24 (5), 733–752. Anderson, K., Smith, S.J., 2001. Emotional geographies. Trans. Inst. Br. Geogr. 26 (1),
7–10. Argüelles, L., 2021. Privileged socionatures and naturalization of privilege: untangling
environmental privilege dimensions. Prof. Geogr. 73 (4), 650–661. https://doi.org/
10.1080/00330124.2021.1924804. Askins, K., Swanson, K., 2019. Holding onto emotions: a call to action in academia. Emot. Space Soc. 33, 100617. Askins, K., 2019. Emotions. In: Keywords in Radical Geography: Antipode at 50, pp. 107–112. Atkinson, R., Flint, J., 2004. Fortress UK? Gated communities, the spatial revolt of the
elites and time–space trajectories of segregation. Hous. Stud. 19 (6), 875–892. elites and time–space trajectories of segregation. Hous. Stud. 19 (6), 875–892. Atkinson, R., Parker, S., Burrows, R., 2017. Elite formation, power and space in
contemporary London. Theory Cult. Soc. 34 (5–6), 179–200. Atkinson, R., Parker, S., Burrows, R., 2017. Elite formation, power and space in
contemporary London. Theory Cult. Soc. 34 (5–6), 179–200. In conclusion, the focus on extra-ordinary and privileged emotions
enables a thorough explanation of a water crisis, its genesis, and future
implications. In turn, privileged emotional geographies can potentially
advance political ecology’s explanations of water crises in three mean
ingful ways. First, an engagement with emotions can help retrace the
most intimate and unconscious manifestations of power dynamics which
characterize patterns of resources use and control. Second, what I named
here extra-ordinary emotions, mark a point of continuity (or disconti
nuity) of existing path dependencies. Funding This research has received funding by the European Research
Council (ERC) within the project ‘HydroSocialExtremes: Uncovering the
Mutual Shaping of Hydrological Extremes and Society’, ERC Consoli
dator Grant No. 771678. E. Savelli E. Savelli Geoforum 141 (2023) 103746 An in-depth analysis of emotions and the subjectivities that construct
them, reveals the temporality of emotions, which signifies the way the
past emerges, surfaces in the present and it is projected into the future. By stretching emotion over longer time spans it is possible to consider
them as conjunctures that emerge from certain power dynamics which
have endured history and will eventually persist in the future (Beswick,
2020; Bonilla-Silva, 2019; Holloway et al., 2019). Thus, I argue that the
temporality of emotions helps on the one hand explaining their genesis,
but on the other it also forwards emotions into the future by revealing
the way they can repeat or create ruptures from the past (Holloway
et al., 2019). In the case of Cape Town, for instance, stretching emotions
over time reveals the ways they reassert the elites’ privilege, and por
tends to future imaginaries of a deeply racial and neoliberal city. One
that ends up excluding low-income dwellers from citizens’ right on the
basis of their race and financial inadequacies. Such future is also unjust
because enables the elites to overconsume water at the expenses the
most vulnerable population and the environment. By reproducing
distinctive power dynamics and triggering unsustainable practices,
privilege emotions risk shifting water crises towards the future eventu
ally restoring the uneven and unsustainable hydrosocial configurations
that have contributed to the water crisis in the first place. with privileged emotions it is possible to challenge their subjectivities,
dismantle the resulting imaginaries and give space to more just and
sustainable futures. Data availability Data will be made available on request. i
Privileged emotions, have thus the potential to trap the future into
past socionatural configurations and eventually ruin it (Millington and
Scheba, 2021). What is more, is that such a form of ruination does not
only have imaginary potentials but also material implications. Indeed,
the privileged emotional geography of Cape Town water crisis have
shown the manner in which power laden emotions become imprinted in
space when they materially affect surrounding people and places
(Bonilla-Silva, 2019). Fear, anger, and pride have reconfigured the
elite’s relationship with space and water resources for they have trig
gered amongst others, the enclosure and dispossession of common re
sources, the unsustainable exploitation of (ground)water sources,
alongside the development of security systems and gated communities. Together, these material reconfigurations of private space end up
endangering social justice and the future availability of water resources
within the city. Acknowledgements This research has received funding by the European Research
Council (ERC) within the project ‘HydroSocialExtremes: Uncovering the
Mutual Shaping of Hydrological Extremes and Society’, ERC Consoli
dator Grant No. 771678. I would like to express my sincere appreciation
to the interviewees and participants for sharing their personal experi
ences with me throughout the duration of this research. I also want to
thank Maria Rusca for her engagement and insightful comments during
the conception of the article, and Sami Halabi for his helpful feedback
and encouragement throughout the writing process. Declaration of Competing Interest The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
the work reported in this paper. 5. Discussion and conclusions: the temporality and materiality
of privileged emotions In turn, such neoliberal emotions and the resulting actions
help individuals to strategize for themselves, sustaining a society that
will, in the long-term, only benefit those who are financially suitable
whilst excluding the possibility of any collective transformation (Read,
2009). By legitimising a doing what I want attitude, privileged emotions 43 Reference from interviews with: KKC; LKGP; BLBFB. 44 Reference from interviews with: KKC; LKGP; BLBFB; APC. 45 Reference from interviews with: APC; SWC; SMGP; BMC; ADC; KKC. 46 Reference from interviews with: BMC; APC. 47 Reference from interviews with: SMGP; BMC; SWC. 48 Reference from interviews with: KKC; GCC, SWC. 49 Reference from interviews with: LCC; GCC; APC; SWC; KKC; LKGP; SMGP;
ADC; EHHB; BLBFB. 50 Reference from interviews with: KKC; GCC; APC; HMTL. 51 Reference from interviews with: SWC; KKC; HMTL; LKGP. 52 Direct quote by: SWC. 53 Reference from interviews with: HMTL; APC; GCC. 54 Direct quote by: APC 9 References Singap. J. Trop. Geogr. 25 (3), 322–342. Millington, N., 2018. Producing water scarcity in S˜ao Paulo, Brazil: the 2014–2015 water
crisis and the binding politics of infrastructure. Polit. Geogr. 65, 26–34. Castel´an Cargile, A., 2016. Can empathy improve concern for secondary group members? Testing an emotionally engaging video intervention. Communication Research
Reports 33 (3), 265–268. Millington, N., Scheba, S., 2021. Day zero and the infrastructures of climate change:
water governance, inequality, and infrastructural politics in Cape Town’s water
crisis. Int. J. Urban Reg. Res. 45 (1), 116–132. Miraftab, F., 2007. Governing post apartheid spatiality: implementing city improvement
districts in Cape Town. Antipode 39 (4), 602–626. COCT, 2014. Cape Town 2011 Census Socio-Economic Index, City of Cape Town
Development Information and GIS Department, 61 Pages. Available from: <http://
resource.capetown.gov.za/documentcentre/Documents/City%20research%
20reports%20and%20review/16429%20COCT%20State%20of%20Cape%20Town
%20Report%202016%20FINAL.pdf> (accessed the: 16 December 2019). Miraftab, F., 2012. Colonial present: legacies of the past in contemporary urban practices
in Cape Town, South Africa. J. Plan. Hist. 11 (4), 283–307. rt%202016%20FINAL.pdf> (accessed the: 16 December 2019). Morales, M.C., Harris, L.M., 2014. Using subjectivity and emotion to reconsider
participatory natural resource management. World Dev. 64, 703–712. Ni h i
l
A J
2011 B
d d
i
i
i l
bj
i i
i
d h
(i ) Curti, G.H., Aitken, S.C., Bosco, F.J., Goerisch, D.D., 2011. For not limiting emotional
and affectual geographies: a collective critique of Steve Pile’s’ Emotions and affect in
recent human geography’. Trans. Inst. Br. Geogr. 36 (4), 590–594. Nightingale, A.J., 2011. Beyond design principles: subjectivity, emotion, and the (ir)
rational commons. Soc. Nat. Resour. 24 (2), 119–132. Davidson, J., Milligan, C., 2004. Embodying emotion sensing space: introducing
emotional geographies. Soc. Cult. Geogr. 5 (4), 523–532. Nightingale, A.J., 2011. Bounding difference: Intersectionality and the material
production of gender, caste, class and environment in Nepal. Geoforum 42 (2),
153–162. Di Baldassarre, G., Wanders, N., AghaKouchak, A., Kuil, L., Rangecroft, S., Veldkamp, T. I., Garcia, M., van Oel, P.R., Breinl, K., Van Loon, A.F., 2018. Water shortages
worsened by reservoir effects. Nat. Sustainability 1 (11), 617–622. Nightingale, A., 2013. Fishing for nature: the politics of subjectivity and emotion in
Scottish inshore fisheries management. Environment and Planning A 45 (10),
2362–2378. Doshi, S., 2017. Embodied urban political ecology: Five propositions. Area 49 (1),
125–128. Norgaard, K.M., 2012. Climate denial and the construction of innocence: reproducing
transnational environmental privilege in the face of climate change. Race Gender
Class 80–103. References ‘Day Zero’, hydraulic citizenship and the defence of the commons in
Cape Town: a case study of the politics of water and its infrastructures (2017–2018). J. South. Afr. Stud. https://doi.org/10.1080/03057070.2019.1552424. Gonz´alez-Hidalgo, M., Zografos, C., 2020. Emotions, power, and environmental conflict:
expanding the ‘emotional turn’ in political ecology. Prog. Hum. Geogr. 44 (2),
235–255. Rusca, M., 2018. Visualizing urban inequalities: The ethics of videography and
documentary filmmaking in water research. Wiley Interdisciplinary Reviews: Water
5 (4), e1292. Grove, K., 2009. Rethinking the nature of urban environmental politics: security,
subjectivity, and the non-human. Geoforum 40 (2), 207–216. lib
li
d h
i
di
i
i l
i Grove, K., 2009. Rethinking the nature of urban environmental politics: security,
subjectivity, and the non-human. Geoforum 40 (2), 207–216. Harvey, D., 2007. Neoliberalism and the City. Studies in Social Justice 1, 2–13. Heynen, N., 2014. Urban political ecology I: The urban century. Prog. Hum. Geogr. 38
(4), 598–604. Ryan, K., 2016. Incorporating emotional geography into climate change research: a case
study in Londonderry, Vermont, USA. Emot. Space Soc. 19, 5–12. Harvey, D., 2007. Neoliberalism and the City. Studies in Social Justice 1, 2 13. Heynen, N., 2014. Urban political ecology I: The urban century. Prog. Hum. Geogr. 38
(4), 598–604. Savage, M., Williams, K., 2008. Elites: remembered in capitalism and forgotten by social
sciences. Sociol. Rev. 56, 1–24. Heynen, N., 2016. Urban political ecology II: The abolitionist century. Prog. Hum. Geogr. 40 (6), 839–845. Savelli E., et al; 2023 ’Urban water crises driven by elites’ unsustainable consumption’,
Nature Sustainability. DOI: 10.1038/s41893-023-01100-0 , https://www.nature.co
m/articles/s41893-023-01100-0. Holloway, S.L., Holt, L., Mills, S., 2019. Questions of agency: capacity, subjectivity,
spatiality and temporality. Prog. Hum. Geogr. 43 (3), 458–477. Savelli, E., Rusca, M., Cloke, H., Di Baldassarre, G., 2021. Don’t blame the rain: social
power and the 2015–2017 drought in Cape Town. J. Hydrol. 594, 125953. y
y
g
g
Hook, D., 2005. Paradoxes of the other: (Post) colonial racism, racial difference,
stereotype-as-fetish. Psychology in society 31, 9–30. Kaika, M., 2003. Constructing scarcity and sensationalising water politics: 170 days that
shook Athens. Antipode 35 (5), 919–954. Sharp, J., 2009. Geography and gender: what belongs to feminist geography? Emotion,
power and change. Prog. Hum. Geogr. 33 (1), 74–80. Kallis, G., 2010. Coevolution in water resource development: the vicious cycle of water
supply and demand in Athens, Greece. Ecol. Econ. 69 (4), 796–809. References Enqvist, J., Ziervogel, G., Metelerkamp, L., van Breda, J., Dondi, N., Lusithi, T.,
Mdunyelwa, A., Mgwigwi, Z., Mhlalisi, M., Myeza, S., Nomela, G., 2022 Jan 2. Informality and water justice: community perspectives on water issues in Cape
Town’s low-income neighbourhoods. International Journal of Water Resources
Development. 38 (1), 108–129. Nyamnjoh, F.B., 2012. Blinded by sight: Divining the future of anthropology in Africa. Africa Spectrum 47 (2–3), 63–92. Otero, I., Kallis, G., Aguilar, R., Ruiz, V., 2011. Water scarcity, social power and the
production of an elite suburb: the political ecology of water in Matadepera,
Catalonia. Ecol. Econ. 70 (7), 1297–1308. Escobar, A., 1999. After nature: steps to an antiessentialist political ecology. Curr. Anthropol. 40 (1), 1–30. Park, L.S.H., Pellow, D.N., 2011. The Slums of Aspen: Immigrants vs. The Environment in
America’s Eden. New York University Press, New York. Finewood, M.H., 2012. Feeling the squeeze: a political ecology of race and amenity-
based development in coastal Bluffton. South Carolina. Local Environment 17 (9),
991–1011. Pile, S., 2010. Emotions and affect in recent human geography. Trans. Inst. Br. Geogr. 35
(1), 5–20. Freudenburg, W.R., 2005. Privileged access, privileged accounts: toward a socially
structured theory of resources and discourses. Soc. Forces 84 (1), 89–114. Gabriel, N., 2014. Urban political ecology: Environmental imaginary, governance, and
the non-human. Geogr. Compass 8 (1), 38–48. Freudenburg, W.R., 2005. Privileged access, privileged accounts: toward a socially
structured theory of resources and discourses. Soc. Forces 84 (1), 89–114. Read, J., 2009. A genealogy of homo-economicus: neoliberalism and the production of
subjectivity. Foucault Studies 6, 25–36. https://doi.org/10.22439/fs.v0i0.2465. structured theory of resources and discourses. Soc. Forces 84 (1), 89–114. Gabriel, N., 2014. Urban political ecology: Environmental imaginary, governance, and
the non-human. Geogr. Compass 8 (1), 38–48. Gabriel, N., 2014. Urban political ecology: Environmental imaginary, governance, and
the non-human. Geogr. Compass 8 (1), 38–48. Riggs, D.W., Augoustinos, M., 2005. The psychic life of colonial power: racialised
subjectivities, bodies and methods. J. Community Appl. Soc. Psychol. 15 (6),
461–477. Gill, R., 2008. Culture and subjectivity in neoliberal and postfeminist times. Subjectivity
25 (1), 432–445. l Roberts, T., Lunch, C., 2015. Participatory video. The international encyclopedia of
digital communication and society 1–6. Gonz´alez-Hidalgo, M., 2017. The politics of reflexivity: Subjectivities, activism,
environmental conflict and Gestalt Therapy in southern Chiapas. Emot. Space Soc. 25, 54–62. l Robins, S., 2019. References Srinivasan, V., Seto, K.C., Emerson, R., Gorelick, S.M., 2013. The impact of urbanization
on water vulnerability: a coupled human–environment system approach for Chennai,
India. Global Environ. Change 23 (1), 229–239. Kearney, A., Bradley, J.J., 2009. ‘Too strong to ever not be there’: place names and
emotional geographies. Soc. Cult. Geogr. 10 (1), 77–94. Sultana, F., 2009. Fluid lives: subjectivities, gender and water in rural Bangladesh. Gend. Place Cult. 16 (4), 427–444. K¨ohlin, G., Whittington, D., and Visser, M., 2018. Beyond day zero in cape town
economic instruments for water-scarce cities. Efd initiative blog post, 06 June 2018. Available from: <https://www.efdinitiative.org/blog/beyond-day-zero-cape-town-
economic-instruments-water-scarce-cities> (accessed the 07 March 2023). Sultana, F., 2011. Suffering for water, suffering from water: emotional geographies of
resource access, control and conflict. Geoforum 42 (2), 163–172. l
Sultana, F., 2020. Embodied intersectionalities of urban citizenship: water,
infrastructure, and gender in the global south. Ann. Am. Assoc. Geogr. 110 (5)
1407–1424. https://doi.org/10.1080/24694452.2020.1715193. Lemanski, C., 2007. Global Cities in the South: Deepening Social and Spatial Polarisation
in Cape Town. Cities 24 (6), 448–461. https://doi.org/10.1016/j.cities.2007.01.011. Lillie, K., Ayling, P., 2021. Revisiting the un/ethical: the complex ethics of elite studies
research. Qual. Res. 21 (6), 890–905. Lemanski, C., 2007. Global Cities in the South: Deepening Social and Spatial Polarisation
in Cape Town. Cities 24 (6), 448–461. https://doi.org/10.1016/j.cities.2007.01.011. Sultana, F., 2021. Political ecology 1: From margins to center. Prog. Hum. Geogr. 45 (1),
156–165. Lillie, K., Ayling, P., 2021. Revisiting the un/ethical: the complex ethics of elite studies
research. Qual. Res. 21 (6), 890–905. Sultana, F., 2015. Emotional political ecology. In: The International Handbook of
Political Ecology. Edward Elgar Publishing. Loftus, A., 2007. Working the socio-natural relations of the urban waterscape in South
Africa. Int. J. Urban Reg. Res. 31 (1), 41–59. Loftus, A., 2009. Rethinking political ecologies of water. Third World Q. 30 (5), 953–968. Lorimer, H., 2008. Cultural geography: non-representational conditions and concerns. Prog. Hum. Geogr. 32 (4), 551–559. i Swyngedouw, E., 1997. Power, nature, and the city. The conquest of water and the
political ecology of urbanization in Guayaquil, Ecuador: 1880–1990. Environ Plan A
29 (2), 311–332. Swyngedouw, E., Kaika, M., Castro, E., 2002. Urban water: a political-ecology
perspective. Built Environ. (1978-), 124–137. Mansfield, N., Fensham, R., Threadgold, T., Tulloch, J., 2020. Subjectivity: Theories of
the Self from Freud to Haraway. Routledge. l Thien, D., 2005. After or beyond feeling? A consideration of affect and emotion in
geography. References This entails that extra-ordinary
emotions can reveal the manner in which distinctive power dynamics
can be reinforced or resisted during a crisis. Lastly and most importantly,
the exploration of privileged emotions is key to understand how the
disproportional control and use of water resources is rationalised,
justified and reinforced by the elite. Privileged emotional geographies
ultimately reveal the power through which elites (materially and
discursively) maintain their privileges whilst compromising the avail
ability of natural resources for the less advantaged population and
eventually intensifying environmental degradation. Only by connecting Baker, A. (2018). "What It’s Like to Live through Cape Town’s Massive Water Crisis."
TIME Magazine. TIME’s Africa. Available at: https://time.com/cape-town-south-
africa-water-crisis/ (Accessed 16/11/2020). Bakker, K.J., 2000. Privatizing water, producing scarcity: the Yorkshire drought of 1995. Econ. Geogr 76 (1), 4–27. Bakker, K.J., 2003. A political ecology of water privatization. Stud. Political Econ. 70 (1),
35–58. Barrett, L.F., 2017. The theory of constructed emotion: an active inference account of
interoception and categorization. Soc. Cogn. Affect. Neurosci. 12 (1), 1–23. https://
doi.org/10.1093/scan/nsw154. Barrett, L.F., 2018. You aren’t at the mercy of your emotions – your brain creates them
[Video]. TED Conferences. Available from: <https://www.youtube.com/watch? v=0gks6ceq4eQ>. Beaverstock, J.V., Hubbard, P., Short, J.R., 2004. Getting away with it? Exposing the
geographies of the super-rich. Geoforum 35 (4), 401–407. Bennett, K., 2004. Emotionally intelligent research. Area 36 (4), 414–422. Bennett, K., 2009. Challenging emotions. Area 41 (3), 244–251. i Beswick, K., 2020. Feeling working class: affective class identification and its
implications for overcoming inequality. Stud. Theatr. Perform. 40 (3), 265–274. https://doi.org/10.1080/14682761.2020.1807194. Blackman, L., Cromby, J., Hook, D., Papadopoulos, D., Walkerdine, V., 2008. Creating
subjectivities. Subjectivity 22, 1–27. https://doi.org/10.1057/sub.2008.8. 10 E. Savelli Geoforum 141 (2023) 103746 Bondi, L., 2014. Understanding feelings: engaging with unconscious communication and
embodied knowledge. Emot. Space Soc. 10, 44–54. i McDonald, D.A., 2012. World City Syndrome: Neoliberalism and Inequality in Cape
Town. Routledge. Bondi, L., 2006. Emotional Geographies (J. Davidson, Ed.). first ed. Routledge. doi:
10.4324/9781315579245. Mehta, L., 2003. Contexts and constructions of water scarcity. Econ. Pol. Wkly
5066–5072. Mehta, L., 2010. The social construction of scarcity: the case Mehta, L., 2010. The social construction of scarcity: the case of water in western India. In: Global Political Ecology. Routledge, pp. 385–400. Bonilla-Silva, E., 2019. Feeling race: theorizing the racial economy of emotions. Am. Sociol. Rev. 84 (1), 1–25. Budds, J., 2004. Power, nature and neoliberalism: the political ecology of water in Chile. WWF, 2020. “Cape Town’s groundwater under the spotlight” Online Webinar. Available
from: <https://africa.panda.org/food_footer/?uNewsID=32522> (Accessed the: 19
November 2020). Yowei, S., Natisse, K.M., 2017. Emotions, part one [Audio podcast episode]. In:
Invisibilia, NPR. Available from: <https://www.npr.org/2017/06/01/530928414/
emotions-part-one?t=1645031693206> (accessed the: 03 February 2022). Turok, I., Visagie, J., Scheba, A., 2021. Social inequality and spatial segregation in Cape
Town. In: Urban Socio-Economic Segregation and Income Inequality. Springer,
Cham, pp. 71–90.
Twidle, H., 2021. Shadow of a drought: notes from Cape Town’s water crisis.
Interventions 1–5. https://doi.org/10.1080/1369801X.2021.2015708.
Warner, J.F., Meissner, R., 2021. Cape Town’s “Day Zero” water crisis: a manufactured
media event? Int. J. Disaster Risk Reduct. 64, 102481. References Area 37 (4), 450–454. Mason-Bish, H., 2019. The elite delusion: reflexivity, identity and positionality in
qualitative research. Qual. Res. 19 (3), 263–276. Massumi. B., 2002a. Parables for theVirtual: Movement, Affect and Sensation. Duke
University Press, Durham, NC. Truelove, Y., 2011. (Re-) Conceptualizing water inequality in Delhi, India through a
feminist political ecology framework. Geoforum 42 (2), 143–152. T
l
Y
2019 R thi ki
t
i
it
i
lit
d i f
t
t
th
h Truelove, Y., 2011. (Re-) Conceptualizing water inequality in Delhi, India through a
feminist political ecology framework. Geoforum 42 (2), 143–152. Truelove, Y., 2019. Rethinking water insecurity, inequality and infrastructure through an
embodied urban political ecology. Wiley Interdiscip. Rev. Water 6 (3), e1342. Maxmen, A., 2018. As Cape Town water crisis deepens, scientists prepare for ’Day Zero’. Nature 554 (7690), 13–15. Truelove, Y., 2019. Rethinking water insecurity, inequality and infrastructure through an
embodied urban political ecology. Wiley Interdiscip. Rev. Water 6 (3), e1342. 11 E. Savelli Geoforum 141 (2023) 103746 Turok, I., Visagie, J., Scheba, A., 2021. Social inequality and spatial segregation in Cape
Town. In: Urban Socio-Economic Segregation and Income Inequality. Springer,
Cham, pp. 71–90. Yowei, S., Natisse, K.M., 2017. Emotions, part one [Audio podcast episode]. In:
Invisibilia, NPR. Available from: <https://www.npr.org/2017/06/01/530928414/
emotions-part-one?t=1645031693206> (accessed the: 03 February 2022). Twidle, H., 2021. Shadow of a drought: notes from Cape Town’s water crisis. Interventions 1–5. https://doi.org/10.1080/1369801X.2021.2015708. Warner, J.F., Meissner, R., 2021. Cape Town’s “Day Zero” water crisis: a manufactured
media event? Int. J. Disaster Risk Reduct. 64, 102481. 12
|
https://openalex.org/W4205806965
|
https://www.e3s-conferences.org/10.1051/e3sconf/202233600070/pdf
|
English
| null |
A novel real-time monitoring system for a PVWPS
|
E3S web of conferences
| 2,022
|
cc-by
| 2,727
|
1 Introduction Current
energy
consumption
in
agriculture
is
unsustainable. Modern agriculture production are highly
dependent on fossil fuels, consuming about 30% of the
world's available energy and producing more than 20%
of world’s greenhouse gas emissions [1]. In the next
decades there will be significant and simultaneous
increases in water, energy and food needs. These will
lead to a degraded and depleted natural resource base,
and increasing climate change challenges. Modernizing
agriculture production by increasing the use of fossil
fuels will be neither an affordable nor a sustainable
option because of climate change and the impact of high
and volatile fossil fuel prices on production costs and
food prices. when farms are not connected to the electricity grid. Photovoltaic water pumping (PVWP) technology for
irrigation is an innovative and sustainable solution to
curb the grassland degradation. At the same time it can
promote the conservation of farmland, especially in
remote areas of Tunisia. In remote areas, the need of
monitoring solar pumping control system is very
necessary to ensure stable operation. This paper
describes the hardware design for the solar pumping
control system. The monitoring system is equipped with
different sensors especially to measure the DC electrical
output parameters of the PV panel and AC parameters
of the inverter. Previous work shows that PVWPS can be monitored
using wireless sensor networks [4]. There are many
drawbacks when it used to transmit data. In deed, it
consumes more energy to operate. To produce more food using less energy but also with
cleaner sources of energy, food and agricultural systems
will progressively need to decouple from fossil fuel
dependence and adopt renewable energy sources [2]. The term "Renewable energies" includes all forms of
energy: solar also known as photovoltaic energy, wind
power, biomass, hydraulic [3]. These sources of energy
are projected to assist in the improvement of the security
of the worldwide energy supply system and in granting
the developing nations with an opportunity to use
affordable energy. For these reasons, the utilization of
solar energy, known by photovoltaic systems, is rapidly
turning into one of the most debated issues of the Other transmission techniques can be used for
monitoring the remote operated systems as Ethernet
network (uses network cable to transmit data), RF
module, Zigbee wireless network... * Corresponding author:sallem.amin@yahoo.com A novel real-time monitoring system for a PVWPS Amine Sallem1*,Jean Marie Soubeiga2, Soufiane Abi3, Bachir Benhala3and Nouri Masmoudi
1ISGIS, University of Sfax, Tunisia
2LETI-ENIS, University of Sfax, Tunisia
3IESII, Faculty of Sciences, University of Moulay Ismail, Morocco Amine Sallem1*,Jean Marie Soubeiga2, Soufiane Abi3, Bachir Benhala
1ISGIS, University of Sfax, Tunisia
2LETI-ENIS, University of Sfax, Tunisia
3IESII, Faculty of Sciences, University of Moulay Ismail, Morocco Amine Sallem1*,Jean Marie Soubeiga2, Soufiane Abi3, Bachir Benhala3and Nouri Masmoudi2 1ISGIS, University of Sfax, Tunisia
2LETI-ENIS, University of Sfax, Tunisia
3IESII, Faculty of Sciences, University of Moulay Ismail, Morocco Abstract.The Tunisian Solar Plan, a renewable energy development plan elaborated by the government, foresees a
30% share of renewable in the electricity mix by 2030. The solar photovoltaic capacity that can be installed for pumping
water for irrigation is estimated at 24 MW by 2030. The water pumping system in the agricultural sector is considered
among the main processes requiring the use of photovoltaic solar energy particularly in areas without an electric grid. The Photovoltaic Water Pumping System (PVWPS) contributes, in addition to the optimization of water pumping, to
improve their energy performance in order to reduce their water consumption for irrigation. In Tunisia, PVWPS was
adopted by the Ministry of Agriculture as a trustworthy alternative solution to cover the potable water needs of people
in faraway areas, particularly in southern Tunisia. A novel real-time monitoring system for a PVWPS is proposed in
this paper. The purpose of the monitoring system is to provide a useful tool for the operation and management of the
PVWPS using the MODBUS communication protocol and the SIM900 module connected to ARDUINO to extend the
control of the PV inverter to a distance greater than 1800m. modern day. The production of electricity using
photovoltaic systems is also an interesting possibility ICEGC'2021
E3S Web of Conferences 336, 00070 (2022) ICEGC'2021
E3S Web of Conferences 336, 00070 (2022) https://doi.org/10.1051/e3sconf/202233600070 © The Authors, published by EDP Sciences. This is an open access article distributed under the terms of the Creative Commons Attribution License 4.0
(http://creativecommons.org/licenses/by/4.0/). © The Authors, published by EDP Sciences. This is an open access article distributed under the terms of the Creative Commons Attribution License 4.0
(http://creativecommons.org/licenses/by/4.0/). DP Sciences. This is an open access article distributed under the terms of the Creative Commons Attribution License 4.0
ses/by/4 0/) 1 Introduction But Ethernet
network is affected by geographical environment [5],
transmission bit error rate of RF module is high and less
reliable [6] and Zigbee wireless network is more costly
compared to other module [7]. p
GSM (Global System for Mobile Communication) is a
standard set used to describe protocols for digital E3S Web of Conferences 336, 00070 (2022) ICEGC'2021
E3S Web of Conferences 336, 00070 (2022) https://doi.org/10.1051/e3sconf/202233600070 cellular networks. GSM network has low error rate, low
costs and wide signal coverage. Users can communicate
easily using GSM to perform monitoring anytime and
anywhere. Micro-
controller
SERVER
Solenoid
Valves
Inverter
phone
PC
(FARMER)
TABLET ... WPF C#
MySQL
PHP Symphony
JavaScript:
-React Native
-React
-Node
RS 485
Relay card
SMS
USB
GSM/GPS/GPRS
C/C++
Arduino
USB
Fig.2. Monitoring solar pumping control system flow chart. MySQL
PHP Symphony A technical solution for the monitoring and tracking of
PVWPS is shown in this work, and a novel quasi-real-
time monitoring system for a PVWPS. Relay card In this work, a stable and reliable system is built using
ARDUINO microcontroller and GSM (SIM900)
module to monitor and control performance of the solar
pumping control system. The figure below summarizes our objective for the
proposed work: ENERGY-SMART
DECISION SYSTEM
Renewable Energies
(PV system)
Farmer's
needs
On / Off and
monitoring of
pumps
Fig. 1. Energy-smart decision system: The objective idea of
the proposed work. Fig.2. Monitoring solar pumping control system flow chart. INVT inverter [8] was used in this work and it allows
fairly direct control via the MODBUS communication
protocol. Fig. 3, 4 and 5 show an example of
communication and the type of control possible with this
protocol. Fig. 1. Energy-smart decision system: The objective idea of
the proposed work. The rest of the paper is organized as follows: section 2
presents an overview of the Modbus communication
protocol. Section 3 describes the experimental
equipment used in this study. The development of the
mobile application is presented in section 4. Finally,
Section 5 presents the conclusions of this work and
perspectives for future work. Fig. 3. Transmission technique using the MODBUS Protocol. 3.3 Relay module The drive data on the ARDUINO board viewed through
the computer are shown as follows: The drive data on the ARDUINO board viewed through
the computer are shown as follows: Relay module was connected to the ARDUINO MEGA
and allows turning on and off the pump or any other
devices, both AC and DC, based on the received SMS
from the authorized user. -The operating frequency in Hz is: 20.50 -The set frequency in Hz is: 20.50 -The voltage in Volt is: 324.10 The orders received by the relay module that has been
sent by ARDUINO are used to turn on or off the system
controlling switch. Figure 8 shows photo for the
prototype including the project and the user end. The
smart mobile acts as user end. -Output voltage in Volt is: 129 -The output current in Ampere is: 0.00 -The speed of rotation in RPM is: 65535 -The output power is %: 0.00 Fig.8. Relay Module -The output torque is %: 6553.20 Fig. 7. SIMCOM SIM900 module. Attention (AT) commands are used to control
functionality of GSM modem. The different AT
commands for module configuration are: Attention (AT) commands are used to control
functionality of GSM modem. The different AT
commands for module configuration are: In the Fig. 6, we present the different module of the
proposed monitoring solar pump control system. In the Fig. 6, we present the different module of the
proposed monitoring solar pump control system. Fig. 6. Photo shows the proposed system, project end. •
ATD <number>: Checks number availability
•
AT+CMGS: Send SMS
•
AT+ CSQ: Checking signal quality
•
AT+CLTS: Get Local Timestamp
•
AT+CREG: Network Registration Status
•
AT+CCLK: Clock signal
•
set the SIM900 to text mode: AT+CMGF=1\r
•
send SMS to a number: AT+CMGS =
PHONE_NUMBER(in international format)
•
read
the
first
SMS
from
the
inbox:
AT+CMGR=1\r
•
read the second SMS from the inbox:
AT+CMGR 2\r •
AT+CMGS: Send SMS •
read
the
first
SMS
from
the
inbox:
AT+CMGR=1\r •
read the second SMS from the inbox:
AT+CMGR=2\r •
read
all
SMS
from
the
inbox:
AT+CMGR=ALL\r Fig. 6. Photo shows the proposed system, project end. Fig. 6. Photo shows the proposed system, project end. In addition, we can make the command by the computer
and make a reading of the data. In addition, we can make the command by the computer
and make a reading of the data. 2 Modbus communication protocol The monitoring solar pumping control system prototype
consists of smart mobile, ARDUINO MEGA, SIM900,
GPS/GSM Neo7M module, MAX-485 TTL to RS-485
converter module, Solenoid valve connected to a relay
card. Fig. 3. Transmission technique using the MODBUS Protocol. Fig. 4. A detailed example of MODBUS communication. Fig. 5. Example of control command sent to start the inverter. The user communicates with the microcontroller
through SMS using the GSM network. A SIM card is
installed in the GSM module which is on the field and
connected to the microcontroller. The GSM module
sends the data received to ARDUINO MEGA which
also continuously receives the data from the relay card
(Solenoid Valve are connected to a relay card to be
controlled by ARDUINO) to check the condition of the
motor pump. The flow chart of the system is shown in fig.2 Fig. 4. A detailed example of MODBUS communication. g
p
Fig. 5. Example of control command sent to start the inverter. Fig. 5. Example of control command sent to start the inverter. 2 ICEGC'2021
E3S Web of Conferences 336, 00070 (2022) https://doi.org/10.1051/e3sconf/202233600070 E3S Web of Conferences 336, 00070 (2022) Fig. 7. SIMCOM SIM900 module. In the next part, we will replace the computer (i.e., the
master) by an ARDUINO MEGA where we will
translate the command shown in Fig. 5. In the next part, we will replace the computer (i.e., the
master) by an ARDUINO MEGA where we will
translate the command shown in Fig. 5. 3.1 ARDUINO board programming ARDUINO is an open-source platform and easy-to-use
hardware and software. The selected ARDUINO
MEGA is a microcontroller board based on the
ATmega2560. To facilitate sending commands to the INVT inverter
through the ARDUINO MEGA, we will use the
MODBUS MASTER library. Fig. 7. SIMCOM SIM900 module. 3.4 Ublox NEO-7M GPS Module Ublox NEO-7M is a pretty advanced GPS module
recommend for user who are interested in navigation,
tracking, the vast world of GPS. This module can
receive data and then calculate the geographical position
with a very high accuracy and speed. In addition the
module has internal memory to save settings. This
module is compatible with ARDUINO and can be used
in any project. For this project, you will need parts
which include: •
1 Ublox Neo-7N GPS Module •
1 Arduino MEGA •
Jumper Wires (Male to Female) •
1 USB Data Cable •
1 Female SMA antenna (optional; if needed)
Utilising one main library, the TinyGPS++ library by
Mikal Hart, it help massively in using this Ublox GPS
module with ARDUINO. The pinout of this module is presented in the figure
below. Fig.9. NEO-7M GPS Module Fig.9. NEO-7M GPS Module 3.2 SIMCOM SIM900 Module SIMCOM SIM900 module is a GSM/GPS-compatible SIMCOM SIM900 module is a GSM/GPS-compatible
Quad-band cell phone, which works on a frequency of
850/900/1800/1900MHz [9]. We use this module,
shown in Fig. 7, to receive and send text messages,
from/to the user mobile phone. Fig.8. Relay Module Fig.8. Relay Module 3 ICEGC'2021
E3S Web of Conferences 336, 00070 (2022) https://doi.org/10.1051/e3sconf/202233600070 E3S Web of Conferences 336, 00070 (2022) 4 Mobile application development In this section, a mobile application with REACT
NATIVE [10] has been developed and it is compatible
with IOS and ANDROID for controlling the variable
speed drive. We will use the direct sending of SMS through the
phone without going through the writing phase, the user
will just have to click on the different keys for the
command. In our example we will take the case of a user
with two watering sites. The first pages (images in figure 10) are dedicated to
configuring the application. The user (farmer) gives: • The number of sites that the farmer owns • The name of its sites • The name of its sites • The number of pumps he owns per site • The name assigned to the pump as well as its module
number. Let us illustrate this configuration through a user who
owns 02 sites: Sfax (in the south of Tunisia) and Tunis
(in the north). Let us illustrate this configuration through a user who
owns 02 sites: Sfax (in the south of Tunisia) and Tunis
(in the north). # Sfax: 02 pumps 5 Conclusion PVWPS offers an alternate means to meet the electricity
demand for irrigation and livestock watering. Under the
circumstances of inadequate supply of conventional
sources of energy, the solar water pump can play a
significant role. In this present work, we have realized a novel real-time
monitoring system based on Simcom SIM900 module
and ARDUINO broad, and by a mobile application
development for Photovoltaic Water Pumping System
(PVWPS). We have monitored by sending and receiving
SMS through the variable-speed drives of the Sfax and
the Tunis watering sites. g
In our future work, we will try to improve the general
vision of the study by integrating the geolocation of
reading, sending of data to the server and the
development of a web site for the administrator and the
clients. # Sfax: 02 pumps Name of Pump1: Watering vegetables Name of Pump1: Watering vegetables
Pump 1 number: 95788294 Pump 1 number: 95788294 Name of Pump 2: Watering the flowers
Pump 2 number: 95788294 Name of Pump 2: Watering the flowers Pump 2 number: 95788294 # Tunis: 01 pump Name of Pump 1: Fruit watering Pump 1 number: 95788294. 4 ICEGC'2021
E3S Web of Conferences 336, 00070 (2022) https://doi.org/10.1051/e3sconf/202233600070 Zigbee wireless sensor network. Journal of Beijing
Forestry University. 122-124. 8. http://www.invt.su/assets/files/GD350/ruk_na_exp
lGD350.pdf
9. https://simcom.ee/documents/SIM900/SIM900_A
T%20Command%20Manual_V1.11.pdf
10. https://reactnative.dev/docs/getting-started Zigbee wireless sensor network. Journal of Beijing
Forestry University. 122-124. 8. http://www.invt.su/assets/files/GD350/ruk_na_exp
lGD350.pdf
9. https://simcom.ee/documents/SIM900/SIM900_A
T%20Command%20Manual_V1.11.pdf
10. https://reactnative.dev/docs/getting-started Zigbee wireless sensor network. Journal of Beijing
Forestry University. 122-124. 8. http://www.invt.su/assets/files/GD350/ruk_na_exp
lGD350.pdf
9. https://simcom.ee/documents/SIM900/SIM900_A
T%20Command%20Manual_V1.11.pdf
10. https://reactnative.dev/docs/getting-started Fig. 10. Extract from the application developed in React
Native. Fig. 10. Extract from the application developed in React
Native. The user can thus on the home page: •
Set the frequency of each pump •
Activate / Deactivate the solenoid valves •
Start / Stop the pump •
View the parameters of each pump (voltage,
current, speed, frequency) References 1. W. Yamaka, R. Phadkantha, and P. Rakpho,
“Economic and energy impacts on greenhouse gas
emissions: A case study of China and the USA,”
Report Energy, 6th International Conference on
Advances on Clean Energy Research, ICACER
2021 April 15–17, 2021, Barcelona, Spain 2. John Vourdoubas, Olivier Dubois, Energy and agri-
food systems: production and consumption,
Mediterra 2016, Chapter 7, pp. 159-179, Presses de
Sciences Po, Paris, 2016. 3. R Sims, A Flammini, M Puri, S Bracco,
“Opportunities for agri-food chains to become
energy-smart”,
UN
Food
and
Agricultural
Organisation, 2015, Rome and GIZ. ISBN: 978-92. 4. Ranhotigamage C., and Mukhopadhyay, S. C.,
2011. Field trials and performance monitoring of
distributed solar panels using a low-cost wireless
sensor network for domestic application. IEEE
Sensors Journal 11, 2583-2590. 5. Dargie,
W.,
and
Poellabauer,
C.,
2010
Fundamentals of Wireless Networks: Theory and
Practice, Wiley, NJ, USA. 6. Liu Hui, Wang Mao-hua, and et al., 2008. Development of farmland soil moisture and
temperature monitoring system based on wireless
sensor network. Journal of Jilin University
(Engineering and Technology Edition). 604-608. 7. Junguo ZHAN, Wenbin LI, Ning HAN, Jiangming
KAN, 2008. Forest fire detection system based on 5
|
https://openalex.org/W2945633203
|
https://zenodo.org/record/2647815/files/ETASR_V9_N2_pp_3965-3970.pdf
|
English
| null |
A Satellite-based Remote Sensing Technique for Surface Water Quality Estimation
|
Engineering, Technology and Applied science research/Engineering, Technology and Applied Science Research
| 2,019
|
cc-by
| 5,270
|
Vol. 9, No. 2, 2019, 3965-3970 Vol. 9, No. 2, 2019, 3965-3970 Engineering, Technology & Applied Science Research 3965 A Satellite-based Remote Sensing Technique for
Surface Water Quality Estimation Michelle V. Japitana
School of Engineering, University of San Carlos, Cebu City
and College of Engineering and GeoSciences, Caraga State
University, Butuan City, Philippines
mvjapitana@carsu.edu.ph Marlowe Edgar C. Burce
School of Engineering,
University of San Carlos
Cebu City, Philippines
mcburce@yahoo.com Abstract—Remote sensing provides a synoptic view of the earth
surface that can provide spatial and temporal trends necessary
for comprehensive water quality (WQ) monitoring and
assessment. This study explores the applicability of Landsat 8
and regression analysis in developing models for estimating WQ
parameters such as pH, dissolved oxygen (DO), total dissolved
solids (TDS), total suspended solids (TSS), biological oxygen
demand (BOD), turbidity, and conductivity. The input image was
radiometrically-calibrated using fast line-of-sight atmospheric
analysis (FLAASH) and then atmospherically corrected to obtain
surface reflectance (SR) bands using FLAASH and dark object
subtraction (DOS) for comparison. SR bands derived using
FLAASH and DOS, water indices, band ratio, and principal
component analysis (PCA) images were utilized as input data. Feature vectors were then collected from the input bands and
subsequently regressed together with the WQ data. Forward
regression results yielded significant high R2 values for all WQ
parameters except TSS and conductivity which had only 60.1%
and 67.7% respectively. Results also showed that the regression
models of pH, BOD, TSS, DO, and conductivity are highly
significant to SR bands derived using DOS. Furthermore, the
results of this study showed the promising potential of using RS-
based WQ models in performing periodic WQ monitoring and
assessment. management strategies. Also, though this in-situ measurement
offers high accuracy, it is not feasible to provide a
simultaneous WQ database on a regional scale [3]. Fortunately,
the emerging technologies of geoinformatics, particularly
remote sensing (RS) and geographic information system (GIS),
provide useful tools for a comprehensive WQ assessment and
management. Remote sensing datasets offer a synoptic view of
ongoing earth surface processes which allows an evaluation of
the pressures placed on aquatic ecosystems [4]. Scientific
findings towards understanding WQ by employing RS
techniques are highly effective. Such techniques are valuable in
monitoring upstream land use and land cover changes and
spread [5], deriving RS empirical models to aid WQ
monitoring and assessment[2, 6-8], and mapping the spatial
distribution of WQ parameters using GIS [2, 9]. Japitana & Burce: A Satellite-based Remote Sensing Technique for Surface Water Quality Estimation A Satellite-based Remote Sensing Technique for
Surface Water Quality Estimation Moreover,
satellite images have been used successfully in water
management protocols such as to conduct inventory and water
balance assessment [10-12], to assess flood areas [13], and for
WQ change detection and monitoring [14, 15]. There is no restriction on obtaining and using Landsat
satellite imagery [16]. Landsat satellite series are widely used
to establish relationships between WQ parameters of surface
waters and spectral reflectance [17-20]. Authors in [21] showed
that satellite-based regression models can be derived using
images acquired with a maximum of ±15 days offset from the
date of in-situ water sampling. However, there are
recommendations of a narrower time difference of ±5 days
[22], ±4 days [23], ±3 days [21], and ±1 day [22, 24] between
the sampling date and the acquisition date of the image. For
this study, an offset of ±5 or lesser days is set in the selection of
satisfactory (less cloud cover) satellite images. Regression
analysis is one of the widely used methods to determine
spectral reflectance and WQ parameter relationship by
selecting derived regression models with high R2 value [6, 16,
19, 25]. A wide range of studies proved the potential of using
satellite datasets in deriving WQ models which include
physicochemical water parameters [18, 24, 26], nutrient
concentrations [20], and heavy metals [27]. For instance,
authors in [17] applied regression modelling to develop
algorithms for estimating TSS, pH, BOD, and chlorophyll-a. Hence, in this study, WQ algorithms are developed using a
combined regression and remote sensing approach. The
objective of this study is to determine the potential of using Keywords-spectral;
reflectance;
radiance;
water
quality
modelling; geoinformatics Keywords-spectral;
reflectance;
radiance;
water
quality
modelling; geoinformatics Corresponding author: Michelle V. Japitana II.
STUDY AREA Tubay River is one of the principal rivers in Agusan del
Norte, Philippines which is classified as a Class A water body. Its headstream is the main outlet of Lake Mainit, and the river
traverses within the municipalities of Jabonga, Santiago, and
Tubay. There are two major tributaries of the river, one at its
upstream portion and the other at the mid-downstream part. The river plays a vital role in the economic development of the
said municipalities, as manifested by its diverse uses that
include irrigation and other agricultural uses, fisheries,
livestock production, and various domestic uses and tourism. The river is also a receptor of domestic solid and liquid wastes
and other non-point sources of polluted waters [28]. Figure 1
shows the location of Tubay River and the terrain within the
study area. Fig. 3. The procedure flow for pre-processing the Landsat 8 OLI image. In the pre-processing part, we employed radiometric
correction to remove errors in the raw pixel values (digital
numbers, DN) of the Landsat image and convert the pixels into
radiance or reflectance using the FLAASH tool. A detailed
procedure of transforming DN into at-sensor radiance and
surface reflectance is covered entirely in [29]. The general form
of the equation for computing TOA Reflectance (RTOA) is
shown in (1) and (2): Fig. 1. The study area showing the location of Tubay, Agusan del Norte
and the location of EMB water quality monitoring points. 𝐿′𝜆= 𝑀𝜌∗𝑄𝐶𝐴𝐿+ 𝐴𝜌
(1)
𝐿𝜆= 𝐿′𝜆cos 𝜃
⁄
𝑆𝐸
(2) (1) (2) where L΄λ is TOA spectral radiance, QCAL is the quantized
calibrated pixel values in DN, Mρ is the band-specific
multiplicative rescaling factor, band-specific Aρ is the additive
scaling factor, θSE is the local sun elevation angle, and Lλ is the
TOA reflectance. Then, atmospheric correction was employed to transform
TOA radiance or reflectance bands into surface reflectance
using two types of pre-processing methods to investigate which
method is the most appropriate for extracting RS-based WQ
models. The first method is to apply the FLAASH atmospheric
correction [30] automatically while the second method employs
DOS [31]. Band math computations were finally applied to
surface reflectance bands using the two approaches to complete
the post-FLAASH and post-DOS procedure. Then the raw,
radiance and reflectance images were used to perform PCA and
band ratios for deriving two water indices. II.
STUDY AREA To enhance the
spectral appearance of water bodies, modified normalized
difference water index (MNDWI) and normalized difference
water index (NDWI) variants were derived using (3)-(5). Band
ratio (BR) was also obtained using Band 1 and Band 5 by
applying (6): Fig. 1. The study area showing the location of Tubay, Agusan del Norte
and the location of EMB water quality monitoring points. INTRODUCTION Water is among the most important resources. However,
water bodies are under acute seasonal scarcity due to the
increased rate of human intervention and human-induced
modification of natural processes [1]. Periodic monitoring and
assessment of water quality (WQ) help to develop management
strategies to control surface water pollution [2]. Government-
sponsored WQ monitoring programs usually employ field
measurements and collection of water samples for subsequent
laboratory analysis in a traditional way. While such
conventional approach to WQ monitoring data is accurate at a
specific location and time, in most cases, it cannot provide
enough information on overall WQ. Water monitoring data and
reports in the Philippines usually only have point-specific
datasets which lack spatiotemporal trends that are rather vital in
the monitoring, assessment, and identification of water Corresponding author: Michelle V. Japitana www.etasr.com Vol. 9, No. 2, 2019, 3965-3970 Engineering, Technology & Applied Science Research 3966 Fig. 3. The procedure flow for pre-processing the Landsat 8 OLI image. Landsat 8 images combined with field measurements to predict
selected WQ indicators. To derive WQ regression models, this
study utilizes various raw, pre-processed, and post-processed
Landsat 8 bands and in-situ WQ data and employs regression
analysis. where 𝜌1 is the ultra-blue (coastal/aerosol) band, 𝜌3 is the
green band, 𝜌5 is the NIR band, and 𝜌6 is the SWIR 1 band of
the Landsat 8 OLI image. Japitana & Burce: A Satellite-based Remote Sensing Technique for Surface Water Quality Estimation www.etasr.com IV.
EXPERIMENTAL RESULTS Quarterly WQ monitoring data for the years 2014 and 2015
were available during this study. From these datasets, only 7
physical WQ parameters were considered during this study:
pH, DO, BOD, TSS, TDS, conductivity and turbidity. Among
the available satellite images in the study area, only the Landsat
8 OLI images acquired during April 19, 2015 were utilized
since they qualified the ±5 day difference and the minimal
cloud cover criteria. Hence, the EMB April 14, 2015 WQ data
(shown in Table I) were utilized for statistical RS modelling. A
total of 7 raw bands, 7 pre-processed bands using FLAASH, 7
pre-processed bands using DOS, 18 principal component (PC)
bands, 3 water indexes, and 1 BR were used in this study. This
results in a total of 43 variables to be considered in the
regression analysis. Statistical results showed that 6 variables
or image bands were found to be highly correlated to all
physical under consideration WQ parameters. These variables
include mostly image bands atmospherically corrected using
DOS. These variables were fitted in the respective WQ
regression models using the enter method in SPSS which all
resulted to an R2 value of 100%. Table II shows all the
regression models (enter method) derived using various RS
datasets. SR1=surface reflectance bands calibrated using post-FLAASH, SR2=surface reflectance bands
calibrated using post-DOS, B#=band number, PC=principal component Then, we evaluated back the WQ datasets for DO and
turbidity since the first run of the forward regression analysis
did not yield models for these WQ parameters. The WQ data in
Table I showed that it has the lowest level (1.21) of turbidity at
Station 1 which is located at the mouth of the river. This
suggests low flows during the measurement and that the water
was relatively clear. This situation is also evident even 5 days
after the WQ measurement as shown in the zoomed-in images
(true color image and histogram-stretched true color image)
from the Landsat (Figure 4). At the mouth of the river, we can
observe in Figure 4(a) that the pixels depict greenish color
while it is light blue in Figure 4(b). In the upper stream of the
river, turbid water appears dark (near black) in the normal RGB
image. On the other hand, it is very dark blue in the enhanced
RGB image applied with histogram equalization. TABLE I. WATER QUALITY DATA OF TUBAY RIVER, APRIL 14, 2015
Stn. III.
MATERIALS AND METHODS The 2014-2015 WQ monitoring data of the Environmental
Management Bureau (EMB) were utilized in this study. To
explore the relationships between WQ indicators and spectral
data, we employed the latest addition in the Landsat series
which is the Landsat 8 OLI. Landsat 8 images with acquisition
dates with a ±5-day difference to the water sampling dates of
the EMB were downloaded through USGS Earth Explorer. The
general methodology applied in this study is shown in Figure 2. 𝑀𝑁𝐷𝑊𝐼= (𝜌3 − 𝜌6) (𝜌3 + 𝜌6)
⁄
(3)
𝑁𝐷𝑊𝐼1 = (𝜌3 − 𝜌5) (𝜌3 + 𝜌5)
⁄
(4)
𝑁𝐷𝑊𝐼2 = (𝜌5 − 𝜌6) (𝜌5 + 𝜌6)
⁄
(5)
𝐵𝑅= 𝜌1 𝜌5
⁄
(6) Fig. 2. High level abstraction of processes and data flow. (6) where 𝜌1 is the ultra-blue (coastal/aerosol) band, 𝜌3 is the
green band, 𝜌5 is the NIR band, and 𝜌6 is the SWIR 1 band of
the Landsat 8 OLI image. g. 2. High level abstraction of processes and data flow. www.etasr.com Vol. 9, No. 2, 2019, 3965-3970 Engineering, Technology & Applied Science Research 3967 All images were then resized to contain only the region of
Tubay River as study area. We then extracted feature vectors
from each raw image and transformed satellite images using
the geographic locations of the monitoring stations of EMB. Then, all pixel values were organized and tabulated to be used
for regression analysis. During the regression, two fitting
methods were applied using the SPSS 16.0 software, the enter
method and the forward (step-up) method. In the first method,
all variables with high correlation to the WQ parameter are
fitted in the regression model while the forward method only
includes variables that explain a significant amount of
additional variance in the model and output all possible
regression models with significant R2 values. conductivity. Also, the PC reflectance band 4 is highly
significant to both BOD and TSS regression models, although
the R2 value of the TSS regression model is only 60.1%. Moreover, forward regression analysis for TDS resulted in two
models. The first model with the highest R2 value of 96.8% has
BR and PC1_SR1 as predictors while the second model shows
that TDS can also be estimated using only BR as a predictor
with an R2 value of 77.6%. TABLE II. REMOTE SENSING-BASED REGRESSION MODELS (ENTER
METHOD) FOR SELECTED WQ PARAMETERS. Model
No. III.
MATERIALS AND METHODS Dependent
Predictors and coefficients
R2 %
1
DO
7.961+65.873×PC5_SR1-
38.644×SR2B1+58.405×PC4_SR2+1.56
0×NDWI2-1.378×BR+9.507×RTOAB7
100
2
pH
8.128+42.442×PC5_SR1-
13.18×SR2B1+4.891×PC4_SR2+0.943×
NDWI2-0.754×BR+7.343×RTOAB7
100
3
BOD
0.285+10.063×PC5_SR1-
7.258×SR2B1+32.07×PC4_SR2+0.573×
NDWI2+0.017×BR+3.089×RTOAB7
100
4
TSS
15.184+7311.307×PC5_SR1-
2681.941×SR2B1-
3467.17×PC4_SR2+168.64×NDWI2+2
1.174×BR+1525.531×RTOAB7
100
5
TDS
75475.527×PC5_SR1-
23045.975×SR2B1-
55739.039×PC4_SR2+2118.343×NDWI
2+4404.119×BR+24777.549×RTOAB7
100
6
Turbidity
183.783+3031.118×PC5_SR1-
480.841×SR2B1+4903.518×PC4_SR2-
64.447×NDWI2-258.509×BR-
681.372×RTOAB7
100
7
Conductivity
2.726+18.552×PC5_SR1+33.248×SR2B
1-110.85×PC4_SR2–
12.795×NDWI2+2.775×BR-
19.44×RTOAB7
100
SR1=surface reflectance bands calibrated using post-FLAASH, SR2=surface reflectance bands
calibrated using post-DOS, B#=band number, PC=principal component TABLE II. REMOTE SENSING-BASED REGRESSION MODELS (ENTER
METHOD) FOR SELECTED WQ PARAMETERS. Japitana & Burce: A Satellite-based Remote Sensing Technique for Surface Water Quality Estimation IV.
EXPERIMENTAL RESULTS In particular, these are the surface
reflectance Band 1, PC Band 4, NDWI2, and BR and then, 2
more bands, TOA Reflectance Band 7 and PC 5 of surface
reflectance, using Post-FLAASH (PC5_SR1). This implies that
the surface reflectance bands processed using post-DOS
correction, the TOA Reflectance Band 7, and PC5_SR1 are
practically useful in developing WQ algorithms using enter
regression analysis. However, in Table III, when forward
regression analysis was employed, only 1 to 2 image bands
were found to be significant for each RS-based WQ model. Two of the 6 significant bands (i.e. TOA Reflectance Band 7
and the NDWI2) using enter regression were already not
included as predictors in the models determined by forward
regression analysis. It can be observed that the resulting
predictors of each regression model can be easily associated
with the type of calibration method used. For example, the RS-
based regression models for pH, BOD, TSS, DO, and
conductivity involve predictors representing image bands
derived using post-DOS calibration. The second regression
model for TDS, on the other hand, involved surface reflectance
bands derived from both post-DOS and post-FLAASH. While
TDS (1st model) and turbidity regression models include image
bands derived using post-FLAASH method. amount of oxygen dissolved in the water to support aquatic life. The very extreme turbidity and TDS values in Station 1 may
indicate a different scenario since Station 1 is located at the
mouth of the river. To avoid any uncertainties during the
modelling process, Station 1 data for turbidity and TDS were
excluded. Considering
the
relationship
between
WQ
parameters, we also included them in each run of the regression
analysis. Indeed, a new set of regression models were derived
for DO and Turbidity. Table III shows all the derived RS-based
models using regression analysis (forward method) including
DO and turbidity. (a)
(b)
Fig. 4. The subset or zoomed in images of Landsat 8 acquired April 19,
2015, showing the downstream river: (a) true color RGB image (b) enhanced
true color RGB composite image by employing histogram equalization. (b) (a) (a) (a) (b) Fig. 4. The subset or zoomed in images of Landsat 8 acquired April 19,
2015, showing the downstream river: (a) true color RGB image (b) enhanced
true color RGB composite image by employing histogram equalization. IV.
EXPERIMENTAL RESULTS DO
(mg/L)
BOD
(mg/L)
TSS
(mg/L)
TDS
(mg/L)
pH
Conductivity
(mS/cm)
Turbidity
(NTU)
1
7.3
0.7
5
1,510
7.65
2.25
1.21
2
7.8
0.7
4
112
7.93
1.72
81.6
3
8.1
0.8
8
116
7.95
0.165
85.3
4
8.0
0.8
24
86
8.20
0.125
92.5
5
7.4
0.6
12
92
8.03
0.140
66.3
6
7.4
0.7
0.0
97
8.03
0.149
51.0
7
7.6
0.5
48
97
8.14
0.149
55.3
Source: EMB (2015) TABLE I. WATER QUALITY DATA OF TUBAY RIVER, APRIL 14, 2015
O
O
SS
S
C
i i
i i Source: EMB (2015) The first run of forward regression analysis did not output a
statistical model for DO and turbidity. In this method, out of
the 6 significant variables identified in previously, only 3 were
found to be substantial in the derived WQ regression models. These are the BR, PC4 of surface reflectance bands calibrated
using DOS (PC4_SR2), and the PC5 of surface reflectance
bands calibrated using FLAASH (PC5_SR1). Forward
regression analysis results showed that the models for TDS,
BOD, and pH yielded the highest R2 values of 96.8%, 89.3%,
and 89.0%, respectively. The resulting regression models also
showed the strong association of BR to water pH, TDS, and TDS level at Station 1 is extremely high and exceeds the
1000mg/L minimum standard. This TDS concentration may
indicate pollution and may pose danger to aquatic life. The
TDS is a measure of dissolved solid materials in the river water
that includes salts, some organic materials, and a wide range of
other things from nutrients to toxic substances [32]. Despite the
high concentration of TDS at Station 1, its DO level passed the
minimum criteria of DENR, indicating that there is enough www.etasr.com www.etasr.com Vol. 9, No. 2, 2019, 3965-3970 Engineering, Technology & Applied Science Research 3968 turbidity with highest R2=100% has 4 predictors which are PC
3 and PC 5 of surface reflectance bands corrected using post-
FLAASH method (PC3_SR1 and PC5_SR1), DO, and PC 2 of
surface reflectance bands corrected using DOS method
(PC2_SR2). With these regression modelling results for
turbidity and DO, the significant predictors that were included
in the models confirmed the high correlations found between
turbidity and DO. As observed in Table II, using enter
regression, 4 of the most significant variables are band images
applied with post-DOS correction and its derivatives (e.g.,
PCA, NDWI2, and BR). V.
CONCLUSION AND RECOMMENDATIONS [7] D. G. Hadjimitsis, M. G. Hadjimitsis, L. Toulios, C. Clayton, “Use of
space technology for assisting water quality assessment and monitoring
of inland water bodies”, Physics and Chemistry of the Earth, Parts
A/B/C, Vol. 35, No. 1-2, pp. 115-120, 2010 This study aimed to develop RS-based models of Landsat 8
images by employing regression analysis. The results of this
study showed a strong correlation between in-situ WQ readings
and RS-based datasets in Tubay River. Forward regression
analysis results showed that the models for turbidity, TDS, DO,
BOD, and pH yielded the highest R2 values of 99.3%, 96.8%,
94.1%, 89.3%, and 89.0%, respectively, while TSS and
conductivity have R2 values of only 60.1% and 67.7%
respectively. These models will be beneficial in estimating WQ
parameters and predicting seasonal changes. This study
demonstrated promising results in estimating WQ parameters
in which if further validated with multi-temporal datasets, can
be considered for the development of an operational RS-based
monitoring and assessment system that can be escalated at a
regional and national scale. From the model comparison
results, this study can conclude that it is not necessary to
acquire high resolution and commercial satellite images since a
free satellite image like Landsat series can be a reliable input
image if the right pre-processing method is employed. We
further suggest that the government’s WQ monitoring
campaigns will be planned to coincide with the dates of
Landsat image acquisitions in the respective areas. In this way,
more water characterization data with matching satellite images
can be utilized in deriving more reliable empirical WQ models,
as well as in the model validation. [8] V. Markogianni, D. Kalvas, G. P. Petropoulos, E. Dimitriou, “An
appraisal of the potential of Landsat 8 in estimating chlorophyll-a,
ammonium concentrations and other water quality indicators”, Remote
Sensors, Vol. 10, No. 7, pp. 1-22, 2018 [9] Y. A. El-Amier, A. A. Elnaggar, M. A. El-Alfy, “Evaluation and
mapping spatial distribution of bottom sediment heavy metal
contamination in Burullus Lake, Egypt”, Egyptian Journal of Basic and
Applied Sciences, Vol. 4, No. 1, pp. 55-66, 2017 [10] A. S. Jasrotia, A. Majhi, S. Singh, “Water Balance Approach for
Rainwater Harvesting using Remote Sensing and GIS Techniques,
Jammu Himalaya, India”, Water Resources Management, Vol. 23, No. 14, pp. 3035-3055, 2009 [11] P. H. Gowda, J. L. Chavez, P. D. Colaizzi, S. R. Evett, T. A. Howell, J. A. REFERENCES [1] F. Mushtaq, A. C. Pandey, “Assessment of land use/land cover dynamics
vis-a-vis hydrometeorological variability in Wular Lake environs
Kashmir Valley, India using multitemporal satellite data”, Arabian
Journal of Geosciences, Vol. 7, No. 11, pp. 4707-4715, 2014 [2] F. Mushtaq, M. G. Nee Lala, A. C. Pandey, “Assessment of pollution
level in a Himalayan Lake, Kashmir, using geomatics approach”,
International Journal of Environmental Analytical Chemistry, Vol. 95,
No. 11, pp. 1001-1013, 2015 [3] M. Gholizadeh, A. Melesse, L. Reddi, “A Comprehensive Review on
Water
Quality
Parameters
Estimation
Using
Remote
Sensing
Techniques”, Sensors, Vol. 16, No. 8, p. 1298, 2016 [4] S. J. Goetz, N. Gardiner, J. H. Viers, “Monitoring freshwater, estuarine
and near-shore benthic ecosystems with multi-sensor remote sensing: An
introduction to the special issue”, Remote Sensing of Environment, Vol. 112, No. 11, pp. 3993-3995, 2008 [5] J. Kibena, I. Nhapi, W. Gumindoga, “Assessing the relationship between
water quality parameters and changes in landuse patterns in the Upper
Manyame River , Zimbabwe”, Physics and Chemistry of the Earth, Parts
A/B/C, Vol. 67-69, pp. 153-163, 2014 [6] E. Alparslan, C. Aydoner, V. Tufekci, H. Tufekci, “Water quality
assessment at Omerli Dam using remote sensing techniques”,
Environmental Monitoring and Assessment, Vol. 134, No. 1-3, p. 391,
2007 TABLE III.
REMOTE SENSING BASED REGRESSION MODELS
(FORWARD METHOD) FOR SELECTED WQ PARAMETERS. Model
No. Dependent
Predictors and
coefficients
R2 %
Standard
estimate error
1
DO
7.446+0.008×
Turbidity-
11.59×SR2B1
88.5
0.828
2
pH
8.339-0.827×BR
89.0
0.060
3
BOD
0.382+28.746×PC4
_SR2
90.0
0.041
4
TSS
53.257-
3633.719×PC4_SR2
60.1
11.55
5
TDS
90.124-
1911.396×PC6_SR1
84.6
5.113
6
TDS
84.734-
3465.448×PC6_SR1+4
07.034×PC3_SR2
99.5
1.025
7
Turbidity
37.985×DO+4742.646
×PC5_SR1-
295.309×PC3_SR1-
42.973×PC2_SR2-
220.624
100.0
0.026
8
Turbidity
39.985×DO+4548.48
6×PC5_SR1-
213.951×PC3_SR1-
228.288
99.8
1.051
9
Turbidity
36.176×DO+3334.53
6×PC5_SR1-198.601
96.5
4.076
10
Conductivity
3.682×BR-0.882
67.7
0.570
SR1=surface reflectance bands calibrated using post-FLAASH, SR2=surface reflectance bands
calibrated using post-DOS, B#=band number, PC=principal component Regarding the goodness of fit, WQ models obtained
remarkable R2 values except TSS and conductivity with 60.1%
and 67.7%, respectively. This proves the potential of
employing image processing and regression modelling in
deriving RS-based WQ models. We compared the performance
of the developed models in this study to other regression
models derived from remote sensing data with moderate
resolution like Landsat series [14, 16, 19, 23], and from other
satellite sensors with high resolutions like Worldview-2 [18]
and ASTER [26]. We also took note of the atmospheric
correction method and other post-processing techniques. As
observed in Table IV, the R2 values of the algorithms
developed in this study for pH, turbidity, TSS, and BOD are
better than the ones from the other models presented in
previous studies using Landsat images. This holds true even
when our developed models for DO, pH, and TSS are
compared to regression models based on a 2-meter resolution
Worldview-2 image. It is interesting to note that the turbidity
model performance of this study is comparable to the model
derived using ASTER with a 15-meter resolution. Overall, we
can conclude that the developed regression models using DOS-
corrected Landsat 8 bands can reliably estimate water
parameters such as DO, pH, TSS, and BOD, while TDS and
turbidity can be estimated using FLAASH-corrected Landsat 8
images. This comparison result suggests that a free satellite =surface reflectance bands calibrated using post-FLAASH, SR2=surface reflectance bands
calibrated using post-DOS, B#=band number, PC=principal component The DO regression model yielded an R2 of 88.5% with
turbidity and surface reflectance Band 1 derived using post-
DOS method (SR2 B1) as predictors, while the regression for
turbidity resulted in 3 models with very high R2 of 100%,
99.8%, and 96.5%, respectively. The regression model for Japitana & Burce: A Satellite-based Remote Sensing Technique for Surface Water Quality Estimation www.etasr.com Vol. 9, No. TABLE III.
REMOTE SENSING BASED REGRESSION MODELS
(FORWARD METHOD) FOR SELECTED WQ PARAMETERS. 2, 2019, 3965-3970 Engineering, Technology & Applied Science Research 3969 image like Landsat series can be a reliable input image if the
right pre-processing method is employed. ACKNOWLEDGMENT This paper was supported by an Engineering Research and
Development for Technology Grant of the Philippine’s
Department of Science and Technology. Authors would like to
thank James Earl B. Cubillas for his technical assistance and
the Center for Resource Assessment, Analytics, and Emerging
Technologies of Caraga State University for providing a
workspace to download and pre-process the Landsat images. right pre processing method is employed. TABLE IV. COMPARISON OF REGRESSION MODEL PERFORMANCES
BETWEEN WQ MODELS
Dependent
Variable
Processing
method(s)
Model Performance
Reference
Predictors
R2 %
DO
DOS
Turbidity and B1
88.5
This study
DOS
B1, B3, B4, B5, B7
93.0
[16]
N/A
B3, B6, B7
44.0
[18]
pH
DOS, Band Ratio
B1, B5
89.0
This study
DOS
B3, B4, B5, B6
81.53
[16]
FLAASH
B8, B9
95.06
[26]
N/A
B3, B4, B5, B6, B7
60.7
[18]
Turbidity
FLAASH, PCA
DO, PC3, PC5
99.8
This study
DOS
B4, B5
64.2
[16]
FLAASH
B2, B3, B4, B5
98.6
[23]
FLAASH
B1, B2, B3
99.8
[26]
DOS
B1, B2, B3
46.1
[19]
Conductivity
DOS, Band
Ratio
B1, B5
67.7
This study
DOS
B2, B3, B4, B6
69.9
[16]
FLAASH
B3, B4, B8
99.6
[26]
TSS
DOS, PCA
PC4
60.1
This study
DOS
B1, B2, B3
24.4
[19]
FLAASH
B2, B3, B4, B5
98.8
[23]
N/A
B1, B3, B5 B6, B7
38.8
[18]
TDS
DOS/FLAASH,
Band Ratio, PCA
B1, B5, PC2_SR1
96.8
This study
FLAASH
B3
99.7
[26]
BOD
DOS, PCA
PC4
89.3
This study
DOS
B1, B2, B3
70.7
[14]
N/A
B2, B6, B7
47.6
[18] TABLE IV. COMPARISON OF REGRESSION MODEL PERFORMANCES
BETWEEN WQ MODELS Japitana & Burce: A Satellite-based Remote Sensing Technique for Surface Water Quality Estimation www.etasr.com V.
CONCLUSION AND RECOMMENDATIONS Tolk, “ET mapping for agricultural water management: present status
and challenges”, Irrigation Science, Vol. 26, No. 3, pp. 223-237, 2008 [12] I. Klein, A. J. Dietz, U. Gessner, A. Galayeva, A. Myrzakhmetov, C. Kuenzer, “Evaluation of seasonal water body extents in Central Asia
over thepast 27 years derived from medium-resolution remote sensing
data”, International Journal of Applied Earth Observation and
Geoinformation, Vol. 26, No. 1, pp. 335-349, 2014 [13] C. Ye, “Extraction of water body in before and after images of flood
using Mahalanobis distance-based spectral analysis”, Remote Sensing,
Vol. 31, No. 4, pp. 293-302, 2015 [14] Y. Wang, H. Xia, J. Fu, G. Sheng, “Water quality change in reservoirs of
Shenzhen, China: detection using LANDSAT/TM data”, The Science of
the Total Environment, Vol. 328, No. 1-3, pp. 195-206, 2004 [15] R. Swain, B. Sahoo, “Improving river water quality monitoring using
satellite data products and a genetic algorithm processing approach”, www.etasr.com Engineering, Technology & Applied Science Research Vol. 9, No. 2, 2019, 3965-3970 3970 Sustainability of Water Quality and Ecology, Vol. 9-10, pp. 88-114,
2017 of Philippines. Her research interests include resource mapping,
environmental monitoring, and RS-GIS applications. of Philippines. Her research interests include resource mapping,
environmental monitoring, and RS-GIS applications. of Philippines. Her research interests include resource mapping,
environmental monitoring, and RS-GIS applications. [16] L. C. Gonzalez-Marquez, F. M. Torres-Bejarano, A. C. Torregroza-
Espinosa, I. R. Hansen-Rodriguez, H. B. Rodriguez-Gallegos, “Use of
LANDSAT 8 images for depth and water quality assessment of El
Guajaro reservoir, Colombia”, Journal of South American Earth
Sciences, Vol. 82, pp. 231-238, 2018 Marlowe Edgar C. Burce is an Associate Professor of the University of
San Carlos. He finished his PhD (2012) and MSc (2009) in Biosystems
Sustainability at Hokkaido University, Japan. He also received his MSc
in Information Technology in 2005 and BSc in Computer Engineering at
the University of San Carlos. His research interests include precision
agriculture, field robotics, remote sensing, management information
systems, application and systems programming, programmable logic
controllers, microcontrollers and embedded systems. [17] A. El-Zeiny, S. El-Kafrawy, “Assessment of water pollution induced by
human activities in Burullus Lake using Landsat 8 operational land
imager and GIS”, The Egyptian Journal of Remote Sensing and Space
Science, Vol. 20, Suppl. 1, pp. S49-S56, 2017 [18] A. M. El Saadi, M. M. Yousry, H. S. Jahin, “Statistical estimation of
Rosetta branch water quality using multi-spectral data”, Water Science,
Vol. 28, No. 1, pp. V.
CONCLUSION AND RECOMMENDATIONS 18-30, 2014 [19] A. Kulkarni, “Water Quality Retrieval from Landsat TM Imagery”,
Procedia Computer Science, Vol. 6, pp. 475-480, 2011 [20] V. Markogianni, D. Kalivas, G. Petropoulos, E. Dimitriou, “Analysis on
the feseability of L8 for WQPs assessment”, International Journal of
Environmental, Chemical, Ecological, Geological and Geophysical
Engineering, Vol. 11, No. 9, pp. 906-914, 2017 [21] E. Kontopoulou, P. Kolokoussis, K. Karantzalos, “Water quality
estimation in Greek lakes from Landsat 8 multispectral satellite data”,
European Water, Vol. 58, pp. 191-196, 2017 [22] M. Waxter, Analysis of Landsat Satellite Data to Monitor Water Quality
Parameters in Tenmile Lake, Oregon, MSc Thesis, Portland State
University, 2014 [23] T. S. Kapalanga, Assessment and Development of Remote Sensing
Based Algorithms For Water Quality Monitoring in Olushandja Dam,
North-Central Namibia, MSc Thesis, University of Zimbabwe, 2015 [24] D. Barrett, A. Frazier, “Automated Method for Monitoring Water
Quality Using Landsat Imagery”, Water, Vol. 8, No. 6, p. 257, 2016 [25] V. Kumar, A. Sharma, A. Chawla, R. Bhardwaj, A. K. Thukral, “Water
quality assessment of river Beas, India, using multivariate and remote
sensing techniques”, Environmental Monitoring and Assessment, Vol. 188, No. 3, p. 137, 2016 [26] K. W. Abdelmalik, “Role of statistical remote sensing for Inland water
quality parameters prediction”, The Egyptian Journal of Remote Sensing
and Space Science, Vol. 21, No. 2, pp. 193-200, 2018 [27] R. Swain, B. Sahoo, “Mapping of heavy metal pollution in river water at
daily time-scale using spatio-temporal fusion of MODIS-aqua and
Landsat satellite imageries”, Journal of Environmental Management,
Vol. 192, pp. 1-14, 2017 [28] Environmental Management Bureau-Caraga Region, Tubay River Water
Quality Assessment, 2015 [29] G. Chander, B. Markham, D. Helder, “Summary of Current Radiometric
Calibration Coefficients for”, Remote Sensing of Environment, Vol. 1,
No. 2009, pp. 1-24, 2009 [30] S. M. Adler-Golden, A. Berk, L. S. Bernstein, S. Richtsmeier, P. K. Acharya, M. W. Matthew, G. P. Anderson, C. L. Allred, L. S. Jeong, J. H. Chetwynd, “FLAASH, a MODTRAN4 atmospheric correction
package for hyperspectral data retrievals and simulations”, in:
Proceedings of the 7th Annual JPL Airborne Earth Science Workshop,
Vol. 97, JPL Publication, 1998 [31] P. S. Chavez, “An improved dark-object subtraction technique for
atmospheric scattering correction of multispectral data”, Remote Sensing
of Environment, Vol. 24, No. 3, pp. 459-479, 1988 [32] G. Srivastava, P. Kumar, “Water quality index with missing
parameters”, International Journal of Research in Engineering and
Technology, Vol. 2, No. 4, pp. 609-614, 2013 Japitana & Burce: A Satellite-based Remote Sensing Technique for Surface Water Quality Estimation www.etasr.com AUTHORS PROFILE Michelle V. Japitana is an Associate Professor the College Engineering
and GeoSciences (CEGS), Caraga State University and is currently a
Doctor of the Engineering Program of the University of San Carlos. She
received her BSc degree in Geodetic Engineering from Caraga State
University in 2003 and her MSc in Remote Sensing from the University www.etasr.com
|
https://openalex.org/W4293193655
|
https://ieeexplore.ieee.org/ielx7/8566058/9669085/09763833.pdf
|
English
| null |
Mutual Coupling and Channel Imbalance Calibration of Colocated MIMO Radars
|
IEEE Open Journal of Antennas and Propagation
| 2,022
|
cc-by
| 11,171
|
Mutual Coupling and Channel Imbalance
Calibration of Colocated MIMO Radars ABSTRACT
The Multiple-Input Multiple-Output (MIMO) radar combines the signals from multiple
transmit and receive elements separately to form a virtual array. Consequently, system imperfections
in the MIMO radar caused by channel imbalance and mutual coupling are repeatedly present in the
virtual array, affecting the beamforming performance. In this paper, the beam pattern deviations caused
by these imperfections are described mathematically in relation to a single point target. Furthermore,
a nonparametric calibration technique to estimate and compensate for these effects at different angles
is presented by exploiting that the same imperfections are present multiple times in the virtual array. The proposed calibration method is applied to measurements with an 8 × 8 colocated X-band MIMO
radar, and the peak sidelobe level (PSL) and the integrated sidelobe level (ISL) are used as performance
parameters. The calibration technique is found to improve both the PSL and the ISL by approximately
15 dB within the whole field-of-view (FOV) for antennas that are not minimum scatterers. For angles
above ±30◦from the antenna boresight, the performance of the proposed calibration method is found to
surpass other calibration methods. This is advantageous for MIMO radars as they typically cover a wide
FOV. INDEX TERMS Multiple-input multiple-output (MIMO), radar system, calibration, channel imbalance,
mutual coupling, colocated, beamforming. INDEX TERMS Multiple-input multiple-output (MIMO), radar system, calibration, channel imbalance,
mutual coupling, colocated, beamforming. some cases, antenna mutual coupling can be adequately
described by a scattering matrix containing the scattering
parameters (S-parameters) of the array, or equivalently by
embedded element radiation patterns. Early studies describe
mutual coupling of phased-array radars, demonstrating a
scan angle dependency of the S-parameters. If the element
patterns within an array are identical, the total radiated
field can be estimated by multiplying with the array pat-
tern without considerable errors [10]. This is typically the
case for large arrays with identical elements, even with the
presence of mutual coupling where the embedded radia-
tion pattern differs from that of an isolated element. If the
element radiation patterns are not identical due to the geo-
metrical array environment, the array pattern multiplication
cannot be applied without serious degradation of the system
performance. Particularly, the end-elements of linear arrays
are highly affected by their asymmetric surroundings, and This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/ INDEX TERMS Multiple-input multiple-output (MIMO), radar system, calibration, channel imbalance,
mutual coupling, colocated, beamforming. VOLUME 3, 2022 Received 4 March 2022; revised 30 March 2022; accepted 10 April 2022. Date of publication 27 April 2022; date of current version 5 May 2022. Digital Object Identifier 10.1109/OJAP.2022.3170795 Digital Object Identifier 10.1109/OJAP.2022.3170795 Digital Object Identifier 10.1109/OJAP.2022.3170795 I. INTRODUCTION I I
N RELATION to the number of physical elements,
the multiple-input and multiple-output (MIMO) radar
offers the enhanced angular resolution within a large
field-of-view
(FOV)
and
hence
new
capabilities
for
multichannel radar systems, including identification and
surveillance [1]–[5]. When applying beamforming to the vir-
tual array, system imperfections in a single channel, whether
originating from channel imbalance or influence from nearby
channels, occur repeatedly since each transmit channel is
combined with every receive channel. Typically, a small element spacing of the virtual array is
desired to reduce the angular ambiguities, and inevitably
some of the physical antenna elements are located in
close proximity. Consequently, these elements are subject
to mutual coupling, which has led to a considerable number
of studies within the topic on antenna arrays [6]–[9]. In 511 VOLUME 3, 2022 GROVE et al.: MUTUAL COUPLING AND CHANNEL IMBALANCE CALIBRATION OF COLOCATED MIMO RADARS RF System
Source
RF System
Collector
Transmit
Receive
w0
wR
FIGURE 1. Block diagram representation of a MIMO radar signal model including
channel imbalance and mutual coupling from generation of ideal waveforms to
collection of the received imperfect signals. RF System
Collector
Receive
wR RF System
Source
Transmit
w0 instead the measured or computed active element pattern of
each element is to be considered, when estimating the total
radiated field [11], [12]. In contrast to phased-array radar, a MIMO radar ideally
emits orthogonal waveforms from the transmit elements to
form the virtual array, such that the total field is not steered
towards a specified scan angle. Similarly, each receive ele-
ment is accessed separately without scanning in a certain
angle, which preserves all information for a price of an
increased amount of data and system complexity. Hereby,
digital beamforming can be applied, and the scattering matrix
becomes constant. Nevertheless, depending on the selec-
tion of waveforms, the scattering matrix can potentially
vary as a function of time as documented for phase-coded
waveforms in [13]–[15]. The effects on the beam pattern
from system imperfections in a MIMO radar system have
been investigated for worst-case boundaries through statisti-
cal analysis [16]. Furthermore, mutual coupling effects are
analyzed by implementing a MIMO communication model
that also considers impedance matching in [17]. Receive FIGURE 1. Block diagram representation of a MIMO radar signal model including
channel imbalance and mutual coupling from generation of ideal waveforms to
collection of the received imperfect signals. II. SIGNAL MODEL II. SIGNAL MODEL A block diagram of a MIMO radar signal model including
system imperfections is illustrated in Fig. 1, where the dif-
ferent contributions to imperfections are separated. It forms
the basis of the signal model described in the following. The derivation of the model reflects the signal path, and
it is hence divided into a transmission, propagation, and
reception section. I. INTRODUCTION Evidently, calibration is required, and several studies have
proposed measurement-based calibration methods. By using
the measured channel deviations from a single calibration tar-
get, decent calibration coefficients are conveniently obtained
such as in [4]. The robustness can be increased by mea-
suring the reflected signal from several known targets to
estimate more suitable calibration coefficients. It is done
either from the relative average amplitude and phase devia-
tion throughout each virtual channel [18] or from the phase
slope and offset through a set of linear equations and apply-
ing matrix inversion [19]. In general, these methods perform
well for calibration of channel imbalance, however, they
assume that the effects of system imperfections are angle
independent. This is not the case for mutual coupling even
if the S-parameters are constant. Without taking this matter
into account, the calibration performance often decreases for
increasing angles, as the embedded element patterns between
the end- and center-elements diverge. were minimum scatterers, they would not necessarily behave
as such, e.g., if the surrounding environment also affects the
elements. Finally, mutual coupling can also occur within the
RF system. For MIMO radar systems using an RF mixing
stage, the use of a shared local oscillator may lead to RF
mutual coupling through the power splitters. In this paper, a calibration method to compensate for
channel imbalance and mutual coupling is proposed using
several observations of the relative radiation pattern obtained
within a single MIMO measurement. As such a measure-
ment includes all system errors, this method performs well
even if the antenna elements are not MSAs, corrects for
impedance mismatch, and takes the potential environment
effects on the elements into account. In the following sec-
tion, a system level model including channel imbalance and
mutual coupling is presented with appropriate assumptions. Section III describes some calibration considerations,
emphasizing the importance to calibrate for both the mutual
coupling and channel imbalance. These considerations moti-
vate the proposed method, which is described in Section IV. Here an X-band MIMO system with colocated elements is
used to document the increased calibration accuracy of the
proposed calibration technique. In this paper, a calibration method to compensate for
channel imbalance and mutual coupling is proposed using
several observations of the relative radiation pattern obtained
within a single MIMO measurement. I. INTRODUCTION As such a measure-
ment includes all system errors, this method performs well
even if the antenna elements are not MSAs, corrects for
impedance mismatch, and takes the potential environment
effects on the elements into account. In the following sec-
tion, a system level model including channel imbalance and
mutual coupling is presented with appropriate assumptions. Section III describes some calibration considerations,
emphasizing the importance to calibrate for both the mutual
coupling and channel imbalance. These considerations moti-
vate the proposed method, which is described in Section IV. Here an X-band MIMO system with colocated elements is
used to document the increased calibration accuracy of the
proposed calibration technique. Estimating the direction of targets in presence of mutual
coupling is described in [20]–[22] using super-resolution
techniques such as MUSIC and ESPRIT. These techniques
have high computational complexity and require multiple
samples realized by a series of pulses which affects the
Doppler domain. A calibration method which applies the
coupling matrix from the measured S-parameters is intro-
duced in [14], [23], but the method is not applied to a full
MIMO system. This is done in [24], [25] with a least-
squares approach, but the method does not utilize that
the effect of a physical element impacts several virtual
elements. Additionally, the relation between the coupling
matrix and the embedded element radiation pattern only
applies to minimum scattering antennas (MSA) [26]–[28]. An MSA assumes that the scattering of an antenna under
matched conditions, referred to as residual or structural scat-
tering, has a pattern identical to that of the element, which
is typically not the case. Furthermore, even if the antennas A. TRANSMISSION Initially, a source generates the number of ideally orthogonal
waveforms corresponding to the number of transmitters, NT. 512 VOLUME 3, 2022 VOLUME 3, 2022 In the absence of noise, the waveforms can be written as In the absence of noise, the waveforms can be written as
w0(t) =
w1(t), w2(t), . . . , wNT(t)
T,
(1) bsence of noise, the waveforms can be written as
w0(t) =
w1(t), w2(t), . . . , wNT(t)
T,
(1) k is the wave number, while ER and ET are square diagonal
matrices of order NR and NT, respectively. They correspond
to the isolated element patterns of the receive and transmit
array elements for the given direction angle in azimuth, θ,
and elevation, φ, measured from the normal to the plane of
the MIMO system antenna constellation. A is the NR × NT
attenuation matrix describing the propagation attenuation as
the signal propagates from each transmitter to every receiver,
including the radar cross section of the target at range r. Similarly, T is the NR × NT matrix describing the phase
change due to the signal propagation as a function of the
receive and transmit element position and the target posi-
tion, assuming a propagation speed equal to the speed of
light. Thus, the channel matrix P represents the ideal system
without imperfections. (1) where (T) denotes the transpose and wi(t) is the time-varying
signal of channel i. As the signals traverse through the dif-
ferent components in the RF system, undesired amplitude
and phase variations across the channels occur, which is
also referred to as channel imbalance. These effects can be
described by a diagonal square matrix of order NT: CT,RF =
⎡
⎢⎢⎢⎢⎣
c11
0
· · ·
0
0
c22
... 0
... ... ... ... 0
0
· · ·
cNTNT
⎤
⎥⎥⎥⎥⎦
,
(2) (2) where the cii is the complex channel imbalance coefficient
of element i. For the purpose of this paper, it is assumed that
there is no RF mutual coupling, which otherwise would be
included in the off-diagonal elements in (2). As the generated
signal reaches the antenna elements, the presence of adjacent
elements will additionally lead to scattering, which can be
described by the square coupling matrix of order NT Depending on the application for the model, several sim-
plifications can be made, which will also be the case here. A. TRANSMISSION First, it is assumed that the transmit and receive array are
parallel such that the elevation angle, φ, can be omitted. Next, the target is assumed a single point target in the far-
field of the array with identical radar cross section for all NT
transmitted signals such that A can be ignored. Finally, it is
assumed that the system is colocated such that the direction-
of-arrival (DOA) angle is identical for all elements. Hereby,
T can be split in two phase-terms. First a constant phase-
term by the two-way propagation path from a reference point
on the MIMO array, which can be ignored. Second, a rela-
tive phase change due to the array element positions, which
is of interest for direction estimation [29]. Disregarding the
element patterns, the amplitude-normalized channel matrix
becomes CT,A =
1
√ZT
⎛
⎜⎜⎜⎜⎝
INT −
⎡
⎢⎢⎢⎢⎣
S11
S12
· · ·
S1NT
S21
S22
... S2NT
... ... ... ... SNT1
SNT2
· · ·
SNTNT
⎤
⎥⎥⎥⎥⎦
⎞
⎟⎟⎟⎟⎠
, (3) where
ZT
is
the
termination
impedance
or
reference
impedance, Sij describes the S-parameter from the excited
channel j to channel i, and INT is the identity matrix of order
NT. The diagonal elements of the scattering matrix in (3) are
known as the reflection coefficients, whereas the off-diagonal
elements are the antenna mutual coupling coefficients. The
characteristics of the antenna mutual coupling depend on
the antenna type, the array geometry and antenna orienta-
tion. By combining (2) and (3), an imperfection matrix for
the transmit array can be described as P(θ) ∼= exp
−jksin(θ)y
exp
−jksin(θ)xT
,
(6) (6) with x and y being the position vectors describing the relative
distance of each element in the azimuth direction of the
transmit and receive array, respectively x =
x1, x2, . . . , xNT
T,
(7)
y =
y1, y2, . . . , yNR
T,
(8) (7)
(8) CT = CT,ACT,RF,
(4) (4) (8) such that P(θ) becomes an NR × NT matrix. which is a square matrix of order NT. For an ideal system the
imperfection matrix equals the identity matrix: CT = INT. In
(4) it is assumed that the system is frequency independent
throughout the entire signal bandwidth, which is typically
the case for narrowband signals. C. RECEPTION Similar to the transmit array, mutual coupling in the receive
array also occurs together with channel imbalance, which
again is implemented by a square matrix, CR, of order NR. It
should be noted that on transmit side the channel imbalance
of the RF system occurs prior to the antenna mutual coupling,
and in the opposite order on the receive side. Combining
(1), (4) and (6), the signal collected on the receive side as
seen in Fig. 1, can then be described as It can be expressed as (12) wout(θ) = vec(Wout(θ)),
(12) with vec(·) denoting the vectorization, which converts a
matrix to a column vector through a linear transform by
stacking the columns. Furthermore, it should be noted that if
the array of transmitters represents the full array, a transpose
of Wout(θ) is required prior to vectorization. Otherwise, the
vectorization will stack the virtual elements from each sparse
element consecutively instead of the full array, resulting in
nonconsecutive spatial sampling of the virtual array. with vec(·) denoting the vectorization, which converts a
matrix to a column vector through a linear transform by
stacking the columns. Furthermore, it should be noted that if
the array of transmitters represents the full array, a transpose
of Wout(θ) is required prior to vectorization. Otherwise, the
vectorization will stack the virtual elements from each sparse
element consecutively instead of the full array, resulting in
nonconsecutive spatial sampling of the virtual array. If the imperfection matrices have been estimated, the ideal
signal output can easily be computed by multiplication of
the inverse of the transmit and receive imperfection matrices In the following, three calibration considerations are
described. The simulations that exemplify the considerations
will build on an X-band SMA system as illustrated in Fig. 2,
which is also used to demonstrate the proposed calibration
technique experimentally. The system consists of commercial
off-the-shelf components, features transmission of arbitrary
waveforms with adjustable data collection up to 4 GB, and
has its bandwidth limited to 200 MHz due to the RF subsys-
tem. The antenna elements of the MIMO system are standard
WR90 open-ended waveguides with NF = NS = 8 with a
sufficient length to attenuate higher modes in the waveguide. Open-ended waveguides are suitable for a MIMO radar due
to their broad element pattern, which allows a large FOV. Additionally, absorbers surrounding the elements are added
to reduce backlobe reflections, which will also affect the
embedded element radiation pattern. Wcal(θ) = C−1
R Wout(θ)C−1
T ,
(11) (11) which leads to a calibrated system. In principle, CR and CT
can be ill-conditioned, but diagonally dominant matrices are
often well-conditioned such that the inverse can be estimated. The challenge is of course to obtain the calibration matrix
either by simulations or measurements, such that the effects
of the imperfections are minimized. B. PROPAGATION After transmission, the signals propagate from the transmit-
ting elements, and reflections from targets and clutter will be
collected by the NR receiving elements. For a point target,
the resulting signal is described by the channel matrix P(r, θ, φ) = ER(θ, φ)A(r, θ, φ)
⊙exp
−jkT(r, θ, φ)
ET(θ, φ),
(5) wR(t, θ) = CRP(θ)CTw0(t) + e(t),
(9) (9) (5) such that wR becomes a column vector with NR elements,
and e(t) is the noise vector with independent Gaussian white
noise consisting of NR elements as well. where ⊙is the Hadamard product, i.e., the elementwise
multiplication, exp[ · ] denotes the elementwise exponential, 513 GROVE et al.: MUTUAL COUPLING AND CHANNEL IMBALANCE CALIBRATION OF COLOCATED MIMO RADARS FIGURE 2. The experimental X-band MIMO array, designed and constructed at DTU
Space, used for calibration verification, and consisting of a sparse and full array of
open-ended waveguide antennas. The virtual antenna array is formed by cross-correlating
the collected signals at each receiver in wR(t, θ), with a
copy of all transmitted waveforms in w0(t), also referred to
as matched filtering. Assuming that the transmitted signals
are orthogonal, the normalized matched filter output will be
one. It is also assumed that range migration is negligible (or
corrected for), i.e., the line-of-sight dimension of the array is
much less than the range resolution. From (9), the maximum
normalized matched filter output, including imperfections,
becomes Wout(θ) = CRP(θ)CT + ε,
(10) (10) with Wout(θ) being another NR × NT matrix, and ε being
the normalized additive noise matrix. As the noise in (10)
is added posterior to matched filtering, the SNR is typically
increased equal to the pulse-compression ratio compared to
the additive noise in (9). Depending on the MIMO array con-
figuration, direct coupling between the transmit and receive
elements can occur. It is a rather simple operation to mask
out this coupling after matched filtering, because the signal
will be at near-zero range. Nevertheless, the propagation loss
is near-zero for the direct coupling between the transmit and
receive elements, such that even low coupling coefficients
can saturate the system or increase the sidelobe level from
imperfect waveform orthogonality. This can become critical
for especially coded waveforms, where the sidelobe level is
restricted by the cross-correlation between the waveforms. FIGURE 2. B. PROPAGATION The experimental X-band MIMO array, designed and constructed at DTU
Space, used for calibration verification, and consisting of a sparse and full array of
open-ended waveguide antennas. III. CALIBRATION CONSIDERATIONS Power [dB]
1 2 3 4 5 6 7 8
Channel j
1
2
3
4
5
6
7
8
Channel i
-2
0
2
Phase [rad]
(a)
(b)
FIGURE 4. Normalized amplitude (a) and phase (b) of the modelled S-parameters
estimated from the model in Fig. 3. -1
-0.5
0
0.5
1
sin( )
-2
0
2
4
6
8
Rad. Pattern [dB]
Isolated Antenna
S-parameters + MSA
Embed. Rad. Pat. FIGURE 5. Comparison of element radiation patterns of an isolated open-ended
waveguide from a CEM model, an end-element of the array model estimated from the
obtained scattering parameters, and directly from the CEM model. For consistency, all
models are without surrounding absorbers. where S is the scattering matrix. For most systems, S can be
measured with a Vector Network Analyzer, and the reference
impedance, ZT, is generally 50 . For MSAs, the array radiation pattern is a function of the
mutual impedance [27]: -1
-0.5
0
0.5
1
sin( )
-2
0
2
4
6
8
Rad. Pattern [dB]
Isolated Antenna
S-parameters + MSA
Embed. Rad. Pat. FIGURE 5. Comparison of element radiation patterns of an isolated open-ended
waveguide from a CEM model, an end-element of the array model estimated from the
obtained scattering parameters, and directly from the CEM model. For consistency, all
models are without surrounding absorbers. FIGURE 3. Full array model of the experimental MIMO system applied in a full-wave
simulator. -1
-0.5
0
0.5
1
sin( )
-2
0
2
4
6
8
Rad. Pattern [dB]
Isolated Antenna
S-parameters + MSA
Embed. Rad. Pat. FIGURE 3. Full array model of the experimental MIMO system applied in a full-wave
simulator. FIGURE 3. Full array model of the experimental MIMO system applied in a full-wave
simulator. 1 2 3 4 5 6 7 8
Channel j
1
2
3
4
5
6
7
8
Channel i
-2
0
2
Ph
[
d]
(b) 1 2 3 4 5 6 7 8
Channel j
1
2
3
4
5
6
7
8
Channel i
-40
-20
0
Norm. Power [dB]
(a) Phase [rad] FIGURE 5. Comparison of element radiation patterns of an isolated open-ended
waveguide from a CEM model, an end-element of the array model estimated from the
obtained scattering parameters, and directly from the CEM model. For consistency, all
models are without surrounding absorbers. where S is the scattering matrix. III. CALIBRATION CONSIDERATIONS For most systems, S can be
measured with a Vector Network Analyzer, and the reference
impedance, ZT, is generally 50 . (a) (b) FIGURE 4. Normalized amplitude (a) and phase (b) of the modelled S-parameters
estimated from the model in Fig. 3. For MSAs, the array radiation pattern is a function of the
mutual impedance [27]: E(θ) = Eiso(θ)as(θ)(Z + ZTIN)−1v,
(17) the sparse array even without the surrounding absorbers. Hence, the sparse array has a relatively small influence on
the mutual coupling in the system. (17) where Eiso(θ) is the element pattern of an isolated element
and as is the so-called steering vector using an end-element
as reference: III. CALIBRATION CONSIDERATIONS The presented system model can be applied to all array
configurations. In this paper, a common MIMO system is
however assumed; one array consists of a uniform linear
array (ULA) with λ/2 spacing, dF, where λ is the free
space wavelength, and the subscript “F” refers to the array
being full. This array has also been referred to as a standard
linear array (SLA), because the maximum element distance
is achieved without forming any angular grating lobes when
performing beamforming. The second array in the MIMO
system is a ULA with a spacing of dS = NFdF, where
the subscript “S” refers to array being sparse. In the fol-
lowing, this MIMO system will be addressed as a standard
MIMO array (SMA), since an NF ×NS SMA forms a virtual
SLA array with NFNS elements. For such a system it is nec-
essary to vectorize the output matrix prior to beamforming. Fig. 3 illustrates a model of the full array of the experimen-
tal MIMO radar for computational electromagnetics (CEM)
using CST Microwave Studio. A transient solver is used to
simulate the scattering matrix depicted in Fig. 4, and the
element radiation patterns to investigate the antenna mutual
coupling effects. Since the mutual coupling is highly related
to the element spacing relative to λ, its influence on the
sparse array will depend on NF for an SMA, such that
antenna mutual coupling can be neglected for large arrays. In
this specific case with open-ended waveguides and NF = 8,
the simulated coupling was found to be below −30 dB for 514 VOLUME 3, 2022 VOLUME 3, 2022 FIGURE 3. Full array model of the experimental MIMO system applied in a full-wave
simulator. 1 2 3 4 5 6 7 8
Channel j
1
2
3
4
5
6
7
8
Channel i
-40
-20
0
Norm. Power [dB]
1 2 3 4 5 6 7 8
Channel j
1
2
3
4
5
6
7
8
Channel i
-2
0
2
Phase [rad]
(a)
(b)
FIGURE 4. Normalized amplitude (a) and phase (b) of the modelled S-parameters
estimated from the model in Fig. 3. FIGURE 3. Full array model of the experimental MIMO system applied in a full-wave
simulator. 1 2 3 4 5 6 7 8
Channel j
1
2
3
4
5
6
7
8
Channel i
-40
-20
0
Norm. B. EMBEDDED RADIATION PATTERN Large arrays are often assumed to be infinite, omitting the
effects of the end-elements, where the embedded element
patterns differ from the center-elements due to the asymmet-
ric array environment. However, for an SMA MIMO system,
the impact of mutual coupling on the end-elements of the
full array will repeat itself periodically NS times through
the virtual array. The resulting sidelobes will have their
power concentrated in NS equidistant peaks in the angu-
lar frequency domain as described in [31]. Naturally, the
effect also decreases with an increasing number of elements,
since the percentage of virtual elements affected decreases
assuming NF = NS. A. MINIMUM SCATTERING ANTENNA as(θ) =
1e−jkd sin(θ) · · · e−jkd(N−1)sin(θ)
,
(18) When calibrating for antenna mutual coupling using S-
parameters, it is assumed that the antenna array consists
of MSAs. For an antenna element, the absorption efficiency
is defined as [28] (18) where d is the distance between the elements. In (17), v is
a column vector of the generator voltages or open circuit
voltages, if referring to either a transmitting or receiving
array. Considering only the embedded element pattern of a
single element such that all other elements are terminated,
the unit voltage array becomes η = 1
2
1 −
1 −Da
Ds
,
(13) (13) where Da and Ds are the element directivity and scattering
directivity, respectively. For an MSA, an equal amount of
received energy is absorbed and scattered such that η = 0.5
with a thin dipole being the most common example of an
MSA. The coupling matrix for an MSA with respect to
impedance can be described as [27] V(p)
0q =
0,
V0 = 1
if p ̸= q
if p = q,
(19) (19) when
evaluating
element
q and
exciting
element
p. Combining (19) and (17), it is thus possible to compute the
embedded radiation pattern from the isolated element pattern
and the scattering matrix assuming minimum scatterers. CA = (ZT + ZA)(Z + ZTIN)−1,
(14) (14) Fig. 5 illustrates the radiation patterns of three different
scenarios similar to the modelled array in Fig. 3, but without
the surrounding absorbers. First, the isolated radiation pattern
which has a symmetric amplitude around boresight. Second,
the embedded radiation pattern of an end-element estimated
from the CEM scattering matrix using the MSA approxi-
mation (17) with all other elements terminated. Third, the
embedded radiation pattern estimated with the same CEM
again with sequential excitation. It is obvious, that the
embedded radiation pattern from the minimum scattering
approximation is not identical to the simulated pattern, and with ZT being the reference impedance, and ZA being the
antenna impedance computed from [30]: ZA,n = ZT,n
1 + Snn
1 −Snn
,
(15) (15) where Snn is the diagonal element n of the scattering matrix,
such that ZA = ZT in the case of complex conjugated
impedance matching. A. MINIMUM SCATTERING ANTENNA Of course, the change of the sidelobe levels depends
on the window function since the ratio of the system error
magnitude and sidelobe level plays a significant role. Hence,
for a Dolph-Chebyshev window with a theoretical sidelobe
level of −30 dB, the maximum sidelobe level of 8 × 8 and
16×16 SMAs would be −20 dB and −23 dB, respectively. Therefore, the impact of the channel imbalance caused by
the embedded radiation patterns of the end-elements, cannot
be ignored even for large systems. Additionally, if the array
is not an MSA as shown previously, the correction of the
element pattern will not yield optimal beamforming results. level has risen to about −23 dB, whereas with sixteen ele-
ments the sidelobe level is about −29 dB, which is still
significantly higher than the Hamming weighted sidelobe
level. Of course, the change of the sidelobe levels depends
on the window function since the ratio of the system error
magnitude and sidelobe level plays a significant role. Hence,
for a Dolph-Chebyshev window with a theoretical sidelobe
level of −30 dB, the maximum sidelobe level of 8 × 8 and
16×16 SMAs would be −20 dB and −23 dB, respectively. Therefore, the impact of the channel imbalance caused by
the embedded radiation patterns of the end-elements, cannot
be ignored even for large systems. Additionally, if the array
is not an MSA as shown previously, the correction of the
element pattern will not yield optimal beamforming results. A. MINIMUM SCATTERING ANTENNA Finally, Z is the mutual impedance
matrix described by [30] Z = ZT(IN + S)(IN −S)−1,
(16) (16) 515 VOLUME 3, 2022 GROVE et al.: MUTUAL COUPLING AND CHANNEL IMBALANCE CALIBRATION OF COLOCATED MIMO RADARS -1
-0.5
0
0.5
1
sin( b)
-50
-40
-30
-20
-10
0
Norm. Power [dB]
N = 8
N = 16
FIGURE 6. Simulated SMA system beam pattern for an 8 × 8 and
16 × 16 SMA system with a target at θ0 = 70◦. An amplitude deviation of 1.6 of an
end-element in the full array is assumed with θb indicating the beamforming angle. consequently, the calibration performance of mutual coupling
is degraded when assuming MSAs. Since the antenna array
of open-ended waveguides has a small airgap between each
element, it is not considered an MSA, as the absorption effi-
ciency was estimated to be about 0.25 [28]. Furthermore, in
the specific case with surrounding absorbers as illustrated
in Fig. 2 and Fig. 3, the radiation pattern is addition-
ally altered, such that the behavior of the antennas further
deviates from minimum scatterer condition. This is not nec-
essarily undesired, as the absorbers are found to smooth the
radiation pattern for especially larger angles. Without the
gaps between the elements, the array would become a slot-
type antenna, which is an example of an MSA with respect
to admittance [27]. Fig. 5 only addresses antenna mutual coupling, but as
mentioned mutual coupling can also occur in the RF sub-
system. Assuming only a single element being excited, the
passive elements will additionally radiate an attenuated and
phase shifted copy of the generated signal, depending on the
path length deviation in the RF subsystem. Independently of
the antenna type, the resulting embedded radiation patterns
will deviate from the radiation patterns that are estimated
from the S-parameters. FIGURE 6. Simulated SMA system beam pattern for an 8 × 8 and
16 × 16 SMA system with a target at θ0 = 70◦. An amplitude deviation of 1.6 of an
end-element in the full array is assumed with θb indicating the beamforming angle. level has risen to about −23 dB, whereas with sixteen ele-
ments the sidelobe level is about −29 dB, which is still
significantly higher than the Hamming weighted sidelobe
level. A. PROPOSED CALIBRATION TECHNIQUE 4, and finally the combined chan-
nel imbalance and mutual coupling is applied as described
above. The channel imbalance and mutual coupling sepa-
rately increases the sidelobe level, but when the imbalance is
−0.15 as in Fig. 7, the phase of the imperfections will add up
constructively in accordance with the matrix multiplication
in (10), since the S-parameter of the neighboring element has
a negative phase. This element has the highest contribution
to the beam pattern deviations due to the high amplitude. Hereby the resulting sidelobe level further increases, and a
calibration of either channel imbalance or mutual coupling
improves the beamforming performance. This is opposite
in Fig. 8, where the channel imbalance phase is positive,
such that the imperfections add destructively. Consequently, to its maximum value, instead of the ideal array, since the
relative sidelobes are the ones of interest. First, no mutual
coupling is present, but a channel imbalance of 0.15 radians
is applied to the first channel of the full array similar to (20)
but with CF(1, 1) = 0.9888 ± 0.1494i. Second, mutual cou-
pling is applied assuming no channel imbalance using the
S-parameters from Fig. 4, and finally the combined chan-
nel imbalance and mutual coupling is applied as described
above. The channel imbalance and mutual coupling sepa-
rately increases the sidelobe level, but when the imbalance is
−0.15 as in Fig. 7, the phase of the imperfections will add up
constructively in accordance with the matrix multiplication
in (10), since the S-parameter of the neighboring element has
a negative phase. This element has the highest contribution
to the beam pattern deviations due to the high amplitude. Hereby the resulting sidelobe level further increases, and a
calibration of either channel imbalance or mutual coupling
improves the beamforming performance. This is opposite
in Fig. 8, where the channel imbalance phase is positive,
such that the imperfections add destructively. Consequently, The steps for estimating the calibration coefficients are
illustrated in Fig. 9. Starting with a single measurement, the
matched filtering in range is applied to the collected signals. Considering only the range bin of the calibration target, the
matrix Wout is obtained from (10), where each row repre-
sents a receive element and each column a transmit element. C. COMBINED CHANNEL IMBALANCE AND MUTUAL
COUPLING From (4), it is seen that the imperfection matrix is a combina-
tion of channel imbalances and mutual coupling. In principle,
it is possible to correct for the two imperfection effects inde-
pendently, but depending on the RF system, it is either not
possible or cumbersome to calibrate for channel imbalance
without calibration of mutual coupling. If the antenna array
consists of MSAs, the mutual coupling can be compensated
for directly from obtained S-parameters, followed by a cali-
bration of the channel imbalance. Nonetheless, the minimum
scattering assumption does not apply to many antenna arrays. Another problem arises when only one of the two imper-
fections are considered, because they can either affect each
other constructively or destructively, such that a calibration
of only one of the imperfections potentially yields a worse
outcome. Fig. 6 illustrates the simulated normalized power of the
beam pattern using a Fast Fourier Transform (FFT) of (12). A relative amplitude deviation of 4 dB of an end-element in
the full array is assumed, omitting the small phase deviation
due to a change of the element phase center. Hence, mutual
coupling is also disregarded in this example, resulting in the
following diagonal imperfection matrix of the full array CF,nn =
1.6
1
if n = 1
otherwise,
(20) (20) for element n. The applied amplitude deviation corresponds
to the embedded radiation pattern measured with the MIMO
system seen in Fig. 2, which is illustrated in Section IV. The
virtual array has been Hamming weighted resulting in an
ideal maximum sidelobe level of −42.5 dB. Weighting has
been applied, because an unweighted system has very few
practical uses due to the high sidelobe levels. With a system
of eight elements in the full and sparse array, the sidelobe Fig. 7 and Fig. 8 illustrate the Hamming-weighted beam
pattern for three different imperfection scenarios for an
8 × 8 SMA. For all scenarios, a point target is located
at boresight, i.e., θ0 = 0, and it is assumed that the sparse
array is ideal. Each curve in both figures has been normalized VOLUME 3, 2022 516 -1
-0.5
0
0.5
1
sin( b)
-50
-40
-30
-20
-10
0
Norm. Power [dB]
Chan. Imb. Mut. Coup. Combined
FIGURE 7. Normalized simulated 8 × 8 MIMO radar beam pattern with an
imbalanced channel of −0.15 radians, causing constructive interference. IV. MIMO RADAR SYSTEM CALIBRATION It is evident, that several aspects of system imperfections in
a MIMO radar system are to be considered to increase the
performance in the angular dimension of the radar system. C. COMBINED CHANNEL IMBALANCE AND MUTUAL
COUPLING -1
-0.5
0
0.5
1
sin( b)
-50
-40
-30
-20
-10
0
Norm. Power [dB]
Chan. Imb. Mut. Coup. Combined
FIGURE 8. Normalized simulated 8 × 8 MIMO radar beam pattern with an
imbalanced channel of 0.15 radians, causing destructive interference. -1
-0.5
0
0.5
1
sin( b)
-50
-40
-30
-20
-10
0
Norm. Power [dB]
Chan. Imb. Mut. Coup. Combined
FIGURE 7. Normalized simulated 8 × 8 MIMO radar beam pattern with an
imbalanced channel of −0.15 radians, causing constructive interference. the sidelobe levels for some of the angular frequencies are
greatly reduced. Conclusively, when a system is affected from both chan-
nel imbalance and mutual coupling, but only one of these
imperfections is corrected for, the system performance can
be degraded. IV. MIMO RADAR SYSTEM CALIBRATION A. PROPOSED CALIBRATION TECHNIQUE With only a single measurement, several estimates of devi-
ations from an ideal system can be obtained for each
channel. This is possible, since NRNT virtual elements are
formed from only NR + NT physical elements, such that
an over-determined solution can be formed. However, as
described in Section III, the effects of mutual coupling
changes with respect to the DOA. Hence, a set of mea-
surements of a calibration target with a MIMO system at
different DOA angles is required to achieve high calibration
performance for large angles. This can be attained either
by rotation of the radar system or by changing the location
of the target. For each angle, a set of transmit and receive
calibration coefficients are estimated and used for calibra-
tion. In this way, they correct for all imperfection effects
at a given angle including mutual coupling and presence
of the surrounding environment with the price of having
angular-dependent calibration coefficients, which increases
the computational complexity. Thereby, the measurements
should ideally cover the whole FOV, where the angular
step-size determines the quantity of estimated calibration
coefficient sets. In principle, this is a question of computa-
tional complexity on one hand and system performance on
the other, as additional sets of calibration coefficients offer
better calibration performance. Furthermore, a relatively high
SNR is needed, as no averaging is performed between mea-
surements. However, calibration scenarios are in general set
up to ensure a strong return signal. FIGURE 7. Normalized simulated 8 × 8 MIMO radar beam pattern with an
imbalanced channel of −0.15 radians, causing constructive interference. -1
-0.5
0
0.5
1
sin( b)
-50
-40
-30
-20
-10
0
Norm. Power [dB]
Chan. Imb. Mut. Coup. Combined
FIGURE 8. Normalized simulated 8 × 8 MIMO radar beam pattern with an
imbalanced channel of 0.15 radians, causing destructive interference. FIGURE 8. Normalized simulated 8 × 8 MIMO radar beam pattern with an
imbalanced channel of 0.15 radians, causing destructive interference. to its maximum value, instead of the ideal array, since the
relative sidelobes are the ones of interest. First, no mutual
coupling is present, but a channel imbalance of 0.15 radians
is applied to the first channel of the full array similar to (20)
but with CF(1, 1) = 0.9888 ± 0.1494i. Second, mutual cou-
pling is applied assuming no channel imbalance using the
S-parameters from Fig. VOLUME 3, 2022 A. PROPOSED CALIBRATION TECHNIQUE Next, the angle-dependent linear phase-term originating from
the channel matrix in (6) is removed assuming a target in
the far-field or applying a corresponding near-field phase
compensation: Wres(θ) = Wout(θ) ⊘P(θ),
(21) (21) with ⊘denoting the Hadamard division. The remaining
signal, Wres(θ), now corresponds to the NR × NT matrix
containing the residual signals of all virtual elements for a
given angle. Hence, it is not possible to distinguish between VOLUME 3, 2022 517 GROVE et al.: MUTUAL COUPLING AND CHANNEL IMBALANCE CALIBRATION OF COLOCATED MIMO RADARS FIGURE 9. Diagram for estimating the calibration coefficients from the proposed
calibration method. the different origins of the signal deviations of the virtual
elements Under the assumption that the imperfections affect
1
2
3
4
5
6
7
8
0.8
0.9
1
1.1
1.2
Relative Magnitude
TX1
TX2
TX3
TX4
TX5
TX6
TX7
TX8
Ref
1
2
3
4
5
6
7
8
Receive Element #
-0.1
-0.05
0
0.05
0.1
0.15
Relative Phase [rad]
FIGURE 10. Simulated deviations in amplitude (top) and phase (bottom) caused by
the receive array mutual coupling for a target at θ0 = 50◦and SNR of 30 dB. FIGURE 9. Diagram for estimating the calibration coefficients from the proposed
calibration method. 1
2
3
4
5
6
7
8
0.8
0.9
1
1.1
1.2
Relative Magnitude
TX1
TX2
TX3
TX4
TX5
TX6
TX7
TX8
Ref
1
2
3
4
5
6
7
8
Receive Element #
-0.1
-0.05
0
0.05
0.1
0.15
Relative Phase [rad] 1
2
3
4
5
6
7
8
0.8
0.9
1
1.1
1.2
Relative Magnitude
TX1
TX2
TX3
TX4
TX5
TX6
TX7
TX8
Ref
1
2
3
4
5
6
7
8
Receive Element #
-0.1
-0.05
0
0.05
0.1
0.15
Relative Phase [rad]
FIGURE 10. Simulated deviations in amplitude (top) and phase (bottom) caused by
the receive array mutual coupling for a target at θ0 = 50◦and SNR of 30 dB. FIGURE 9. Diagram for estimating the calibration coefficients from the proposed
calibration method. FIGURE 10. Simulated deviations in amplitude (top) and phase (bottom) caused by
the receive array mutual coupling for a target at θ0 = 50◦and SNR of 30 dB. the different origins of the signal deviations of the virtual
elements. Under the assumption that the imperfections affect-
ing each of the NR + NT physical elements are constant, the
virtual elements can be combined. A. PROPOSED CALIBRATION TECHNIQUE Starting with the receive
array, each column of Wres(θ) is normalized by its mean
value mutual coupling) of the receive channels at a given angle. An imperfection matrix of the receive array can then be
formed from WR,norm(θ) = Wres(θ)diag
Wres(θ)
−1
,
(22) (22) CR(θ) = diag(cR(θ)),
(24) (24) with ( ) denoting the column average and diag(x) the diag-
onal matrix with the vector x as diagonal. Hereby, for each
receive channel, NT estimates of the calibration coefficients
are obtained such that WR,norm(θ) is another NR × NT
matrix. Since each column has been normalized, the devi-
ations due to transmit imperfections are removed, resulting
in NT estimates of the receive imperfections. An example
of WR,norm(θ) for an 8 × 8 MIMO system is illustrated in
Fig. 10 using the simulated S-parameters from Fig. 4 for the
receive array assuming no other imperfections. An interesting
observation can be seen for the end-elements, whose ampli-
tude and phase deviate from the mean values. Across the
receive elements, opposite tendencies are observed in par-
ticular for the amplitude, which is expected due to the array
geometry and θ0 = 50◦. In this way, the mutual coupling
effects can be considered directly as a channel amplitude
and phase deviation for a given angle. These values are
finally averaged to estimate the calibration coefficients of
the receive array: with ( ) denoting the column average and diag(x) the diag-
onal matrix with the vector x as diagonal. Hereby, for each
receive channel, NT estimates of the calibration coefficients
are obtained such that WR,norm(θ) is another NR × NT
matrix. Since each column has been normalized, the devi-
ations due to transmit imperfections are removed, resulting
in NT estimates of the receive imperfections. An example
of WR,norm(θ) for an 8 × 8 MIMO system is illustrated in
Fig. 10 using the simulated S-parameters from Fig. 4 for the
receive array assuming no other imperfections. An interesting
observation can be seen for the end-elements, whose ampli-
tude and phase deviate from the mean values. Across the
receive elements, opposite tendencies are observed in par-
ticular for the amplitude, which is expected due to the array
geometry and θ0 = 50◦. In this way, the mutual coupling
effects can be considered directly as a channel amplitude
and phase deviation for a given angle. A. PROPOSED CALIBRATION TECHNIQUE These values are
finally averaged to estimate the calibration coefficients of
the receive array: which is an angle-dependent counterpart of CR found in (9)
with all contributions of system imperfections gathered in the
diagonal. The counterpart of the transmit imperfection matrix
CT from (4) can be estimated in a similar manner from
(22)-(24) by applying averaging in the opposite dimension. Hence, in (22) and (23) the row and column averaging,
respectively, is applied to achieve CT(θ) as illustrated in
Fig. 9. With CR(θ) and CT(θ) being diagonal matrices, they
are not ill-conditioned, and the inverse can be determined. Finally, the system is then calibrated as from (11), followed
by beamforming on the resulting virtual array. In the following, the proposed method will be compared to
the calibration method described in [18], where the calibra-
tion coefficient of each virtual element is estimated from the
average of several measurements at different DOA angles,
again with the linear, DOA-related phase-term removed. This method will in the following be referred to as the
average-method. B. EXPERIMENTAL RESULTS cR(θ) = WR,norm(θ),
(23) (23) (23) As mentioned in Section III, an SMA radar system is used
to demonstrate the calibration technique. The antenna array
configuration is illustrated in Fig. 2, and naturally the full
array will to a larger extent be influenced by mutual coupling with (
) denoting the row average resulting in a col-
umn vector with NR elements. Hence, these calibration
coefficients represent the total imperfection (imbalance and VOLUME 3, 2022 518 FIGURE 11. Measurement setup in the anechoic chamber from the point of view of
the radar system with the calibration target located at boresight. compared to the sparse array. Nevertheless, the proposed
method is nonparametric, i.e., it can be applied to any array
configuration. The system was put on top of a turntable,
where measurements were performed in the angle interval
1
2
3
4
5
6
7
8
0.6
0.8
1
1.2
1.4
1.6
Relative Magnitude
Sparse1
Sparse2
Sparse3
Sparse4
Sparse5
Sparse6
Sparse7
Sparse8
Ref
1
2
3
4
5
6
7
8
Full Element #
-0.5
-0.25
0
0.25
0.5
Relative phase [rad]
FIGURE 12. Measured amplitude (top) and phase (bottom) for a target at
θ0 = 70 degrees for the transmit array constituting the full array of the SMA. 1
2
3
4
5
6
7
8
0.6
0.8
1
1.2
1.4
1.6
Relative Magnitude
Sparse1
Sparse2
Sparse3
Sparse4
Sparse5
Sparse6
Sparse7
Sparse8
Ref
1
2
3
4
5
6
7
8
Full Element #
-0.5
-0.25
0
0.25
0.5
Relative phase [rad]
FIGURE 12. Measured amplitude (top) and phase (bottom) for a target at
θ0 = 70 degrees for the transmit array constituting the full array of the SMA. 1
2
3
4
5
6
7
8
0.6
0.8
1
1.2
1.4
1.6
Relative Magnitude
Sparse1
Sparse2
Sparse3
Sparse4
Sparse5
Sparse6
Sparse7
Sparse8
Ref
1
2
3
4
5
6
7
8
Full Element #
-0.5
-0.25
0
0.25
0.5
Relative phase [rad] FIGURE 11. Measurement setup in the anechoic chamber from the point of view of
the radar system with the calibration target located at boresight. -
Relative phase [rad] FIGURE 11. Measurement setup in the anechoic chamber from the point of view of
the radar system with the calibration target located at boresight. FIGURE 12. Measured amplitude (top) and phase (bottom) for a target at
θ0 = 70 degrees for the transmit array constituting the full array of the SMA. (23) compared to the sparse array. Nevertheless, the proposed
method is nonparametric, i.e., it can be applied to any array
configuration. The system was put on top of a turntable,
where measurements were performed in the angle interval
of ±70◦from boresight with 1◦interval. From simulations
the open-ended waveguides were found to have a half-power
beamwidth of 110◦, such that the resulting FOV of 140◦
covers the scene of interest. If clutter is present during the
calibration of the system, this will naturally affect the cali-
bration coefficients, resulting in a degraded performance if
the scene changes. Hence, the measurements were conducted
in an anechoic chamber to reduce the amount of clutter. the final result is found by averaging over all sparse elements. It is seen that the amplitude of the end-elements has a rel-
ative high amplitude deviation, which is expected due the
embedded radiation patterns. Since the propagation phase is
removed, the relative phase differences mainly arise from
channel imbalances and deviations in the phase centers due
to the antenna mutual coupling, in particular for the end-
elements. Smaller, seemingly random, deviations are seen in
both amplitude and phase caused mainly by system noise. Compared to Fig. 10, it is evident that the mutual coupling
in the presented MIMO system is stronger than the mod-
elled coupling, as the same trend is observed just with higher
variation from the average value. The calibration target was a corner reflector with an
RCS of 2.2 dBsm, located in the far end of the chamber
at a distance of approximately 7 meters from the system
as illustrated in Fig. 11. Since the distance to the target
is within the near-field of the system, a near-field phase
correction was applied to the data before estimating the
calibration coefficients. Additionally, background subtrac-
tion has been applied to further reduce the clutter level. This was implemented by simply subtracting the raw data
without the calibration target, from those with the calibra-
tion target at an angle-to-angle basis. To accommodate the
assumption of orthogonal waveforms and avoid time-variant
scattering parameters time-division multiplexing, TDM, is
selected, where only a single transmitter is radiating at a
time. Linear Frequency-Modulated (LFM) waveforms are
used to achieve pulse-compression where the full array is
used for transmission and the sparse array for reception. g
Fig. (23) ISL for the uncalibrated system, when applying the proposed and
average-method, and for a Hamming weighted ideal point target response. -0.8
-0.4
0
0.4
0.8
sin( 0)
-45
-40
-35
-30
-25
-20
-15
-10
PSL [dB]
Uncal. Prop. Avg. Ref. -1
-0.5
0
0.5
1
-40
-30
-20
-10
0
Norm. Power [dB]
Uncalibrated
Proposed Method
Average-Method
-1
-0.5
0
0.5
1
sin( b)
-40
-30
-20
-10
0
Norm. Power [dB]
FIGURE 13. Normalized beam pattern with the 8 × 8 MIMO radar when uncalibrated,
calibrated with the proposed method, and with the average-method at boresight (top)
and at θ0 = 70◦(bottom). FIGURE 14. PSL for the uncalibrated system, when applying the proposed and
average-method, and for a Hamming weighted ideal point target response. FIGURE 14. PSL for the uncalibrated system, when applying the proposed and
average-method, and for a Hamming weighted ideal point target response. -0.8
-0.4
0
0.4
0.8
sin( 0)
-35
-30
-25
-20
-15
-10
-5
ISL [dB]
Uncal. Prop. Avg. Ref. FIGURE 15. ISL for the uncalibrated system, when applying the proposed and
average-method, and for a Hamming weighted ideal point target response. FIGURE 13. Normalized beam pattern with the 8 × 8 MIMO radar when uncalibrated,
calibrated with the proposed method, and with the average-method at boresight (top)
and at θ0 = 70◦(bottom). FIGURE 15. ISL for the uncalibrated system, when applying the proposed and
average-method, and for a Hamming weighted ideal point target response. small imperfections will have a relatively high impact on the
sidelobes. At θ0 = 70◦, large sidelobes occur for the uncalibrated
system mainly because of the embedded radiation patterns. The average-method does not correct for this effect effi-
ciently, even though the sidelobe level between the peaks
is significantly decreased. On the other hand, the proposed
method corrects for these effects, resulting in a more constant
sidelobe level throughout the whole beam pattern. where the border between main lobe and sidelobe is at the
first null from the main lobe, i.e., where the first local mini-
mum is found. The deviation of the DOA estimation before
and after system calibration is below 0.1◦for all angles,
which is also below the angular resolution of the radar system
and will not be assessed further in this paper. However, it is
important to notice that the DOA error is to be considered
for MIMO radar systems with few elements. (23) 13 illustrates the normalized beam patterns for three
different calibration scenarios at boresight and at θ0 = 70◦. The first scenario is the single point target response for
the uncalibrated system, the second scenario when apply-
ing the proposed calibration method, and the third scenario
when applying the average-method in [18]. At boresight,
the sidelobe level is clearly reduced for both calibration
methods. This is also expected, since the mutual coupling
effects at larger angles to some degree cancel out, assuming
that the embedded radiation patterns between the two end-
elements in the full array are typically mirrored. Furthermore,
the sidelobes that do appear are seemingly located periodi-
cally, indicating that small channel deviations still occur. It
should be noted, that with a very low sidelobe level even An example of the amplitude and phase values of
WT,norm(θ) with a calibration target at θ0 = 70◦is illustrated
in Fig. 12. The results are relative to the average values, and 519 VOLUME 3, 2022 VOLUME 3, 2022 GROVE et al.: MUTUAL COUPLING AND CHANNEL IMBALANCE CALIBRATION OF COLOCATED MIMO RADARS -1
-0.5
0
0.5
1
-40
-30
-20
-10
0
Norm. Power [dB]
Uncalibrated
Proposed Method
Average-Method
-1
-0.5
0
0.5
1
sin( b)
-40
-30
-20
-10
0
Norm. Power [dB]
FIGURE 13. Normalized beam pattern with the 8 × 8 MIMO radar when uncalibrated,
calibrated with the proposed method, and with the average-method at boresight (top)
and at θ0 = 70◦(bottom). small imperfections will have a relatively high impact on the
-0.8
-0.4
0
0.4
0.8
sin( 0)
-45
-40
-35
-30
-25
-20
-15
-10
PSL [dB]
Uncal. Prop. Avg. Ref. FIGURE 14. PSL for the uncalibrated system, when applying the proposed and
average-method, and for a Hamming weighted ideal point target response. -0.8
-0.4
0
0.4
0.8
sin( 0)
-35
-30
-25
-20
-15
-10
-5
ISL [dB]
Uncal. Prop. Avg. Ref. FIGURE 15. ISL for the uncalibrated system, when applying the proposed and -0.8
-0.4
0
0.4
0.8
sin( 0)
-45
-40
-35
-30
-25
-20
-15
-10
PSL [dB]
Uncal. Prop. Avg. Ref. FIGURE 14. PSL for the uncalibrated system, when applying the proposed and
average-method, and for a Hamming weighted ideal point target response. -0.8
-0.4
0
0.4
0.8
sin( 0)
-35
-30
-25
-20
-15
-10
-5
ISL [dB]
Uncal. Prop. Avg. Ref. FIGURE 15. (23) In [17] it was
found that a 2x4 MIMO radar system had an angle deviation
up to 1◦for large angles. To assess the performance of the calibration technique
in combination with FFT beamforming, two performance
parameters are used, namely the peak sidelobe level (PSL)
and the integrated sidelobe level (ISL). The PSL converted
to dB is simply computed as Fig. 14 and Fig. 15 illustrate the performance parame-
ters of the uncalibrated system together with the calibrated
system with the proposed technique and with the average-
method for all the angles in the measurement series. Fig. 14 depicts the PSL, which is almost constant throughout
the whole FOV for the uncalibrated scenario. A significant
reduction of the PSL is observed for both calibration meth-
ods, where the average-method tends to perform slightly
better around boresight, but the proposed method greatly
outperforms the average-method for angles above ±30◦from
boresight, i.e., |sin(θ0)| > 0.5. Even if the PSL is reduced PSL = 20log10
max(|SLL|)
max(|MLL|)
,
(25) (25) with SLL being the sidelobe level, and MLL the mainlobe
level. ISL is calculated as the ratio between the energy within
the mainlobe and the sidelobes ISL = 10log10
|SLL|2
|MLL|2
,
(26) (26) 520 VOLUME 3, 2022 with about 15 dB, the PSL is still approximately 10 dB
above the theoretical PSL of the Hamming weighted ideal
beam pattern. REFERENCES [1]
D. W. Bliss
and
K. W. Forsythe,
“Multiple-input
multiple-
output (MIMO) radar and imaging: Degrees of freedom and reso-
lution,” in Proc. 37th Asilomar Conf. Signals Syst. Comput., vol. 1,
2003, pp. 54–59. p
Fig. 15 depicts the ISL, where the ISL of the uncali-
brated data lies around −9 dB for most angles but increases
to about −6 dB towards the larger angles. Similar to the
PSL, both calibration methods perform well for all angles,
reducing the ISL with about 11 dB for the average-method
and 16 dB for the proposed method. Again, for large angles
relative to boresight the proposed method performs signif-
icantly better with ISL values up to 10 dB lower than for
the average-method. Like the PSL, the ISL is still not close
to the theoretical Hamming weighted ISL after calibration. V. CONCLUSION This paper presents a detailed model of the system imper-
fections in a MIMO radar system including both channel
imbalances and mutual coupling. It is found that calibra-
tion of both types of system imperfections is necessary to
achieve high performance of the MIMO radar within its
whole FOV. Thus, a calibration technique has been presented,
utilizing that a single MIMO measurement offers several
observations of the relative radiation pattern of each chan-
nel to efficiently calibrate for the system imperfections. Calibration coefficients are estimated experimentally with
an 8 × 8 X-band MIMO radar using open-ended waveguide
antennas. An improvement of the peak sidelobe level and
integrated sidelobe level of approximately 15 dB is found
across the whole FOV. This corresponds to the achieved
performance of another calibration method used for com-
parison at boresight. However, at large angles relative to the
antenna boresight, the proposed calibration method performs
particularly well even if the antennas are not minimum scat-
terers. This is an advantage for MIMO radar systems as they
typically cover a wide FOV. [17] B. T. Arnold and M. A. Jensen, “The effect of antenna mutual coupling
on MIMO radar system performance,” IEEE Trans. Antennas Propag.,
vol. 67, no. 3, pp. 1410–1416, Mar. 2019. [18] F. Belfiori, W. van Rossum, and P. Hoogeboom, “Array calibration
technique for a coherent MIMO radar,” in Proc. 13th Int. Radar Symp.,
2012, pp. 122–125. [19] A. Durr et al., “On the calibration of mm-Wave MIMO radars using
sparse antenna arrays for DoA estimation,” in Proc. 16th Eur. Radar
Conf., 2019, pp. 349–352. [20] B. Friedlander and A. J. Weiss, “Direction finding in the presence
of mutual coupling,” IEEE Trans. Antennas Propag., vol. 39, no. 3,
pp. 273–284, Mar. 1991. [21] I. J. Gupta and A. A. Ksienski, “Effect of mutual coupling on
the performance of adaptive arrays,” IEEE Trans. Antennas Propag.,
vol. AP-31, no. 5, pp. 785–791, Sep. 1983. [22] C. Zhang, H. Huang, and B. Liao, “Direction finding in MIMO radar
with unknown mutual coupling,” IEEE Access, vol. 5, pp. 4439–4447,
2017. [23] H. Steyskal and J. S. Herd, “Mutual coupling compensation in small
array antennas,” IEEE Trans. Antennas Propag., vol. 38, no. 12,
pp. 1971–1975, Dec. 1990. [24] C. M. Schmid, R. Feger, C. Wagner, and A. Stelzer, “Mutual cou-
pling and compensation in FMCW MIMO radar systems,” Frequenz,
vol. 65, nos. 9–10, pp. 297–302, 2011. [Online]. (23) Again, this is a consequence of the low sidelobe levels of the
window function, such that even very small inaccuracies of
the calibration coefficients such as noise and discretization
cause the relative sidelobe level to increase significantly. [2]
P. Stoica and J. Li, MIMO Radar Signal Processing. Hoboken, NJ,
USA: Wiley, 2008. [3]
P. Stoica and J. Li, “MIMO radar with colocated antennas,” IEEE
Signal Process. Mag., vol. 24, no. 5, pp. 106–114, Sep. 2007. [4]
R. Feger, C. Wagner, S. Schuster, S. Scheiblhofer, H. Jäger, and
A. Stelzer, “A 77-GHz FMCW MIMO radar based on an SiGe sin-
glechip transceiver,” IEEE Trans. Microw. Theory Techn., vol. 57,
no. 5, pp. 1020–1035, May 2009. [5]
J. Klare, O. Biallawons, and D. Cerutti-Maori, “UAV detection with
MIMO radar,” in Proc. Int. Radar Symp., 2017, pp. 1–8. [6]
S. Edelberg and A. A. Oliner, “Mutual coupling effects in large antenna
arrays: Part 1—Slot arrays,” IRE Trans. Antennas Propag., vol. 8,
no. 3, pp. 286–297, May 1960. [7]
J. H. Richmond, “A reaction theorem and its application to antenna
impedance calculations,” IRE Trans. Antennas Propag., vol. 9, no. 6,
pp. 515–520, Nov. 1961. [8]
J. L. Allen and B. Diamond, Mutual Coupling in Array Antennas,
MIT Lincoln Lab., Lexington, MA, USA, 1966. C. CALIBRATION VARIATIONS [9]
J. B. Andersen and R. G. Vaughan, “Transmitting, receiving, and scat-
tering properties of antennas,” IEEE Antennas Propag. Mag., vol. 45,
no. 4, pp. 93–98, Aug. 2003. The proposed calibration technique increases the computa-
tional complexity because different calibration coefficients
are applied to different DOA angles. In practice, this
means that beamforming must be applied on an angle-by-
angle basis, for instance with a backprojection algorithm. However, it is possible to reduce the computational com-
plexity by applying the same coefficients for a broader
angular frequency interval, either by averaging over a selec-
tion of angles or simply measure fewer sets of calibration
coefficients. This will of cause decrease the beamforming
performance, so in the end it becomes a matter of prioritiza-
tion. For instance, the average calibration coefficients above
30◦, below −30◦and in the interval ±30◦could be used,
since the benefit of the proposed method is observed for
large angles. [10] C. A. Balanis, Antenna Theory, Analysis and Design, 4th ed. Hoboken,
NJ, USA: Wiley, 2016. [11] D. F. Kelley and W. L. Stutzman, “Array antenna pattern modeling
methods that include mutual coupling effects,” IEEE Trans. Antennas
Propag., vol. 41, no. 12, pp. 1625–1632, Dec. 1993. [12] D. M. Pozar, “The active element pattern,” IEEE Trans. Antennas
Propag., vol. 42, no. 8, pp. 1176–1178, Aug. 1994. [13] N. Colon-Diaz, J. G. Metcalf, D. Janning, and B. Himed, “Mutual
coupling analysis for colocated MIMO radar applications using
CEM modeling,” in Proc. IEEE Radar Conf. (RadarConf), 2017,
pp. 441–446. [14] G. Babur, P. J. Aubry, and F. Le Chavalie, “Antenna coupling effects
for space-time radar waveforms: Analysis and calibration,” IEEE
Trans. Antennas Propag., vol. 62, no. 5, pp. 2572–2586, May 2014. [15] G. Babur, P. Aubry, and F. Le Chevalier, “Improved calibration tech-
nique for the transmit beamforming by a coherent MIMO radar with
collocated antennas,” in Proc. 31st URSI Gen. Assembly Sci. Symp. (URSI GASS), 2014, pp. 1–4. [16] C. M. Schmid, A. Stelzer, R. Feger, S. Schuster, and C. M. Schmid,
“On the effects of calibration errors and mutual coupling on the beam
pattern of an antenna array,” IEEE Trans. Antennas Propag., vol. 61,
no. 8, pp. 4063–4072, Aug. 2013. V. CONCLUSION Available: https://doi-
org.proxy.findit.cvt.dk/10.1515/freq.2011.044 VOLUME 3, 2022 521 GROVE et al.: MUTUAL COUPLING AND CHANNEL IMBALANCE CALIBRATION OF COLOCATED MIMO RADARS [25] C. M. Schmid, C. Pfeffer, R. Feger, and A. Stelzer, “An FMCW
MIMO radar calibration and mutual coupling compensation approach,”
in Proc. Eur. Microw. Week 10th Eur. Radar Conf., 2013, pp. 13–16. JØRGEN DALL (Member, IEEE) received the
M.Sc. degree in electrical engineering and the
Ph.D. degree from the Technical University of
Denmark in 1984 and 1989, respectively, where
he is a Professor of Microwave Remote Sensing
Systems. He has contributed to the development
of the Danish airborne SAR, EMISAR, notably
he led the development of a dedicated real-time
SAR processor, was responsible for the offline
data processing, and organized the EMISAR data
acquisition campaigns in a five-year period. Later,
he led the development of ESA’s airborne Ice Sounding radar and SAR,
POLARIS. Most recently, he led the development of an experimental X-band
MIMO radar. His research interests include various aspects of ice sheet
mapping using InSAR, PolInSAR, and TomoSAR techniques as well as ice
sounding. [26] W. K. Kahn and H. Kurss, “Minimum-scattering antennas,” IEEE
Trans. Antennas Propag., vol. 3, no. 5, pp. 671–675, Sep. 1965. [27] B. Clerckx, C. Craeye, D. Vanhoenacker-Janvier, and C. Oestges,
“Impact of antenna coupling on 2 × 2 MIMO communications,”
IEEE Trans. Veh. Technol., vol. 56, no. 3, pp. 1009–1018, May 2007. [28] J. B. Andersen and A. Frandsen, “Absorption efficiency of receiv-
ing antennas,” IEEE Trans. Antennas Propag., vol. 53, no. 9,
pp. 2843–2849, Sep. 2005. [29] P. Stoica and R. L. Moses, “Spatial methods,” in Spectral Analysis of
Signals. Upper Saddle River, NJ, USA: Pearson/Prentice Hall, 2005,
pp. 263–319. pp
[30] D. M. Pozar, Microwave Engineering. New Delhi, India: Wiley, 201 [31] R. L. Grove and J. Dall, “On the impact of channel imbalance on
MIMO radar performance,” in Proc. 17th Eur. Radar Conf. (EuRAD),
2021, pp. 330–333. POUL
LETH-ESPENSEN
received
the
M.Sc. degree in electrical engineering and telecommu-
nications from Aalborg University, Denmark, in
1994, where he was with Aalborg University,
Center for Person Kommunikation as a Research
Assistant and as an Assistant Research Professor
from 1994 to 1998 and worked with signal pro-
cessing and direction of arrival algorithms for
speech signals and direction of arrival estimation
algorithms in mobile phone antenna array systems. V. CONCLUSION He joined TERMA A/S in 1998, where he has
worked as an Electrical Engineer and as a Senior Engineer of Microwave
Electronics, where he is currently a Specialist of Microwave Electronics. His current interests are microwave circuits and antennas. POUL
LETH-ESPENSEN
received
the
M.Sc. degree in electrical engineering and telecommu-
nications from Aalborg University, Denmark, in
1994, where he was with Aalborg University,
Center for Person Kommunikation as a Research
Assistant and as an Assistant Research Professor
from 1994 to 1998 and worked with signal pro-
cessing and direction of arrival algorithms for
speech signals and direction of arrival estimation
algorithms in mobile phone antenna array systems. He joined TERMA A/S in 1998, where he has
worked as an Electrical Engineer and as a Senior Engineer of Microwave
Electronics, where he is currently a Specialist of Microwave Electronics. His current interests are microwave circuits and antennas. RICARD L. GROVE received the B.S. and M.S. degrees in electrical engineering within the Earth
and Space Physics and Engineering Program
from
the
Technical
University
of
Denmark,
Copenhagen, Denmark, in 2015 and 2017, respec-
tively, where he is currently pursuing the Ph.D. degree in signal processing. His current research
interests include system imperfection such as
channel
imbalance
and
mutual
coupling
on
multiple-input multiple-output radars, the effects
of the imperfections and calibration techniques to
compensate for the imperfections. RICARD L. GROVE received the B.S. and M.S. degrees in electrical engineering within the Earth
and Space Physics and Engineering Program
from
the
Technical
University
of
Denmark,
Copenhagen, Denmark, in 2015 and 2017, respec-
tively, where he is currently pursuing the Ph.D. degree in signal processing. His current research
interests include system imperfection such as
channel
imbalance
and
mutual
coupling
on
multiple-input multiple-output radars, the effects
of the imperfections and calibration techniques to
compensate for the imperfections. 522 VOLUME 3, 2022
|
https://openalex.org/W4242042130
|
https://hal.science/hal-03043603/document
|
English
| null |
Incorporating moisture content in surface energy balance modeling of a debris-covered glacier
| null | 2,019
|
cc-by
| 19,339
|
Alexandra Giese, Aaron Anthony Boone, Patrick Wagnon, Robert Hawley To cite this version: Alexandra Giese, Aaron Anthony Boone, Patrick Wagnon, Robert Hawley. Incorporating moisture
content in surface energy balance modeling of a debris-covered glacier. The Cryosphere, 2020, 14,
pp.1555 - 1577. 10.5194/tc-14-1555-2020. hal-03043603 Incorporating moisture content in surface energy balance modeling
of a debris-covered glacier Alexandra Giese1, Aaron Boone2, Patrick Wagnon3, and Robert Hawley1
1Department of Earth Sciences, Dartmouth College, Hanover, NH, USA
2CNRM-GAME – Groupe d’Étude de l’Atmosphère Météorologique, Toulouse, France
3Univ. Grenoble Alpes, CNRS, IRD, Grenoble-INP, IGE, 38000 Grenoble, France Alexandra Giese1, Aaron Boone2, Patrick Wagnon3, and Robert Hawley1
1Department of Earth Sciences, Dartmouth College, Hanover, NH, USA
2CNRM-GAME – Groupe d’Étude de l’Atmosphère Météorologique, Toulouse, France
3Univ. Grenoble Alpes, CNRS, IRD, Grenoble-INP, IGE, 38000 Grenoble, France Alexandra Giese1, Aaron Boone2, Patrick Wagnon3, and Robert Hawley1
1Department of Earth Sciences, Dartmouth College, Hanover, NH, USA
2CNRM-GAME – Groupe d’Étude de l’Atmosphère Météorologique, Toulouse, France
3Univ. Grenoble Alpes, CNRS, IRD, Grenoble-INP, IGE, 38000 Grenoble, France Correspondence: Alexandra Giese (algiese@gmail.com) Received: 12 July 2019 – Discussion started: 10 September 2019
Revised: 2 March 2020 – Accepted: 26 March 2020 – Published: 13 May 2020 Received: 12 July 2019 – Discussion started: 10 September 2019
Revised: 2 March 2020 – Accepted: 26 March 2020 – Published: 13 May 2020 Abstract. Few surface energy balance models for debris-
covered glaciers account for the presence of moisture in the
debris, which invariably affects the debris layer’s thermal
properties and, in turn, the surface energy balance and sub-
debris melt of a debris-covered glacier. We adapted the inter-
actions between soil, biosphere, and atmosphere (ISBA) land
surface model within the SURFace EXternalisée (SURFEX)
platform to represent glacier debris rather than soil (referred
to hereafter as ISBA-DEB). The new ISBA-DEB model in-
cludes the varying content, transport, and state of moisture in
debris with depth and through time. It robustly simulates not
only the thermal evolution of the glacier–debris–snow col-
umn but also moisture transport and phase changes within the
debris – and how these, in turn, affect conductive and latent
heat fluxes. We discuss the key developments in the adapted
ISBA-DEB and demonstrate the capabilities of the model,
including how the time- and depth-varying thermal conduc-
tivity and specific heat capacity depend on evolving tempera-
ture and moisture. Sensitivity tests emphasize the importance
of accurately constraining the roughness lengths and surface
slope. Emissivity, in comparison to other tested parameters,
has less of an effect on melt. ISBA-DEB builds on existing
work to represent the energy balance of a supraglacial debris
layer through time in its novel application of a land surface
model to debris-covered glaciers. Incorporating moisture content in surface energy balance modeling
of a debris-covered glacier Comparison of measured
and simulated debris temperatures suggests that ISBA-DEB
includes some – but not all – processes relevant to melt under
highly permeable debris. Future work, informed by further
observations, should explore the importance of advection and
vapor transfer in the energy balance. 1
Introduction Enhancing the melt of underlying ice when thin and in-
hibiting it when thick (Östrem, 1959), supraglacial debris is
known to affect the surface energy balance and retreat pat-
terns of mountain glaciers. Supraglacial debris covers 11 %
of glacier area in high-mountain Asia (HMA) (Kraaijenbrink
et al., 2017), a region that contains the highest volume of ice
on Earth outside the polar regions and where glacier melt
flows into rivers that deliver water to 800 million people
(Pritchard, 2019). Understanding sub-debris melt is crucial
for making informed projections of climate change impacts
and associated water security issues in HMA. Sub-debris ablation is fundamentally a function of the tem-
perature at the surface of the debris and the ability of the
debris to conduct heat to its base at the ice–debris inter-
face. Therefore, the amount of ice melt under debris is deter-
mined by local meteorological conditions and physical prop-
erties of the debris itself. The efficiency with which a debris
layer conducts heat is determined by its thermal diffusivity κ
(m2 s−1). κ is given by thermal conductivity K (W m−1 K−1)
normalized by the volumetric heat capacity (J m−3 K−1), it-
self a product of density ρ (kg m−3) and heat capacity c
(J kg−1 K−1). Debris properties beyond thickness are inher-
ently difficult to constrain; a debris layer is comprised of
rock clasts of different sizes, angularities, and lithologies that
are distributed and sorted heterogeneously over the ablation
zone. A debris layer’s interstitial spaces may be comprised
of air or percolating water, which itself undergoes phase
changes as a function of temperature. HAL Id: hal-03043603
https://hal.science/hal-03043603v1
Submitted on 7 Dec 2020 L’archive ouverte pluridisciplinaire HAL, est
destinée au dépôt et à la diffusion de documents
scientifiques de niveau recherche, publiés ou non,
émanant des établissements d’enseignement et de
recherche français ou étrangers, des laboratoires
publics ou privés. HAL is a multi-disciplinary open access
archive for the deposit and dissemination of sci-
entific research documents, whether they are pub-
lished or not. The documents may come from
teaching and research institutions in France or
abroad, or from public or private research centers. The Cryosphere, 14, 1555–1577, 2020
https://doi.org/10.5194/tc-14-1555-2020
© Author(s) 2020. This work is distributed under
the Creative Commons Attribution 4.0 License. A. Giese et al.: Modeling moisture content in debris cover energy balance Table 1. Thermally relevant properties of dry debris, in which interstitial pore spaces are filled with air, water-saturated debris, and ice-
saturated debris of porosity (φ) = 0.39. References and calculation details are given in a complementary table in the Supplement (Table S1). Air density is a function of elevation, air temperature, and air moisture. Thermal conductivity presented by Reid and Brock (2010) is an
“effective” value, from measurements, that is a function of debris’ unspecified porosity and any moisture content at the time of measurement
(Collier et al., 2016). Brock et al. (2010) used a published value of specific heat (948 J kg−1 K−1). We assume that these values of thermal
conductivity and heat capacity (here listed for “dry debris”) are valid for dry debris on West Changri Nup glacier and subsequently perform
sensitivity tests. More details about this assumption and its implications can be found in the Discussion. Thermal
Specific
Volumetric
Debris porosity
conductivity
heat capacity
Density
heat capacity
Diffusivity
(φ) = 0.39
(W m−1 K−1)
(J kg−1 K−1)
(kg m−3)
(J m−3 K−1)
(m2 s−1)
Dry debris
0.94
948
1690
1 602 129
5.867 ×10−7
Air
0.024
1006
∼0.67
–
–
Water-saturated debris
1.16
–
–
3 247 140
3.572 ×10−7
Water
0.57
4218
1000
–
–
Ice-saturated debris
1.81
–
–
2 355 289
7.680 ×10−7
Ice
2.2
2106
917
–
– Moisture has been largely unaddressed in glacier models,
despite the fact that water and ice affect the thermal prop-
erties of a debris layer. Table 1 contrasts bulk thermal con-
ductivity, heat capacity, and density of dry debris with de-
bris of the same porosity (φ = 0.39) that has water-filled and
ice-filled interstitial spaces. A number of studies (e.g., Con-
way and Rasmussen, 2000; Reznichenko et al., 2010; Nichol-
son and Benn, 2012; Collier et al., 2014) have emphasized
the importance of moisture to the thermal properties of de-
bris, particularly in transition seasons. Rounce and McKin-
ney (2014) found a dramatic increase in conductivity from
the top 10 cm of debris on Khumbu region glaciers to the
deeper depths; they attribute this difference to water content,
noting that Nicholson and Benn (2006) found the conduc-
tivity of fully saturated debris to be a factor of 2–3 greater
than that of dry debris. A. Giese et al.: Modeling moisture content in debris cover energy balance Importantly, water content shows an
association with grain size, too: coarser sediments are less
likely to have wet surfaces because fine-grained sediments
have small void spaces and, thus, greater capacity for water
retention (Juen et al., 2013; Blum et al., 2018). on relative humidity and rain (Rounce et al., 2015). Evatt
et al. (2015) advanced debris-covered glacier modeling by
accounting for the evaporative heat flux at the base of the de-
bris and, in doing so, the wind speed above and within a de-
bris layer. Including the thickness-dependent wind dynamics
in their energy balance model contributed to their reproduc-
tion of the thickness–ablation curve of Östrem (1959). How-
ever, their model does not account for moisture beyond that
which is evaporated; like the other models, it assumes that
melt runs off and does not affect the system’s energy (except
in the case of evaporation). Collier et al. (2014) introduced the first energy-balance
model that included an evolving, partially saturated debris
layer. The model treated moisture through a reservoir ap-
proach and calculated the water vapor partial pressure gra-
dient to inform calculations of latent heat fluxes within the
debris. This study laid the groundwork for modeling mois-
ture and identified the need for a physically based approach
to incorporating vertical transport processes (i.e., capillary
action, hydraulic gradient-driven flow, etc.) and to prognos-
ing the distribution and phase changes of moisture with depth
and through time. Further, evaporation and sublimation will lower the sur-
face temperature and remove mass from the system. Con-
densation and deposition have the opposite effect. Sakai et al. (2004) suggest that neglecting evaporation in energy balance
computations can cause an overestimation of sub-debris melt
rates by a factor of 2. Here, we introduce a model that, to our knowledge, is the
first to incorporate moisture with consideration of its verti-
cal transport processes and distribution in debris; ISBA-DEB
is capable of representing vertical moisture fluxes, phase
changes, and moisture retention. We adapt the interactions
between soil, atmosphere, and biosphere (ISBA) soil model
housed within the SURFace EXternalisée (SURFEX) plat-
form of Météo-France to include boundary conditions, ther-
mal properties, hydraulic properties, and runoff parameteri-
zations appropriate for supraglacial debris. The ISBA-DEB
model is capable of solving not only the heat equation but
also moisture transport and retention via the mixed-form
Richards’ equation. Published by Copernicus Publications on behalf of the European Geosciences Union. 1556 2.2
In situ measurements An automatic weather station (AWS) located at 5360 m a.s.l. on a 0.03 km2 debris-covered area of West Changri Nup
glacier (Sherpa et al., 2017; Vincent et al., 2016, Fig. 1)
supplied the meteorological driving data for the model. The
AWS was also the site of debris temperature and point mass
balance measurements used for model calibration and vali-
dation. Half-hourly meteorological measurements on 6 De-
cember 2012, 15:00–28 November 2014, 13:30 local Nepal
time provided all necessary data to force the model, and ad-
ditional half-hourly data included ultrasonic depth from an
SR50 adjacent to the AWS. During the December 2012–
November 2014 period used for this study, there were four
resistance temperature probes installed at distributed depths
(5, 7.5, 10, and 12.5 cm) in the 12.5 cm thick debris 40 cm
from the AWS. The bottom temperature sensor was placed at
the debris–ice interface. Figure 1. A map of West Changri Nup and North Changri Nup
glaciers, showing the location of the AWS (Sect. 2.2), which is also
the location of the measurements of debris temperature and point
mass balance. Source: Esri, DigitalGlobe, GeoEye, Earthstar Geo-
graphics, CNES/Airbus DS, USDA, USGS, AeroGRID, IGN, and
the GIS user community. The glacier outlines are from Sherpa et al. (2017). In this paper, we show capabilities of the model, evalu-
ate its performance, and conduct a series of sensitivity tests
on input parameters. We ran ISBA-DEB by driving it with
2 years of gap-filled in situ meteorological data from West
Changri Nup glacier and compared output to debris measure-
ments over the same period. In addition to the ultrasonic depth gauge installed at the
AWS directly above the resistance temperature probes, there
was an 8 m long bamboo mass balance stake installed 5 m
uphill (west) of the AWS, in ice covered by ∼10 cm de-
bris. Field measurements on 22 April 2015 supplied addi-
tional measurements of debris density and porosity local to
the AWS: the mass of debris samples in filled, known-volume
buckets provided the density, and the volume of water that
filled interstitial pore spaces gave the porosity. Quantities
used for modeling are averages of nine samples. The mea-
surements for each sample give means (standard deviations)
of 1691 (65) kg m−3 and 0.37 (0.04) for density and porosity,
respectively. 2
Field site and data The nearest direct measurement of precipitation is from
a Geonor T200B all-weather sensor at Pyramid Research
Station (5035 m a.s.l., 4.3 km southeast of the AWS). Sherpa
et al. (2017) provide details on the precipitation data acquisi-
tion and correction; the dataset begins on 6 December 2012
and extends beyond the end of our study. We assumed cor-
rected total precipitation at Pyramid Research Station equiv-
alent to total precipitation on West Changri Nup glacier. Be-
cause of differences in elevation and local microclimates, we
repartitioned phase based on AWS local temperature follow-
ing Wagnon et al. (2009), with subsequent, first-order adjust-
ments to the phase to match the timing of major snowfall
events detected by the SR50. Table 2 summarizes available
data from these stations and indicates which drive the model. 2.2
In situ measurements We highlight the important physical processes that need to
be accounted for in any debris-covered glacier melt model,
such as conduction and phase change of water and ice in the
debris. We also discuss the limitations of our model and pro-
pose some further considerations for making improvements. A. Giese et al.: Modeling moisture content in debris cover energy balance Most existing models have assumed a dry debris layer,
with rain, snowmelt, and glacier melt running off instan-
taneously (e.g., Lejeune et al., 2013; Rounce and McKin-
ney, 2014). The few studies that do address moisture focus
on end-member cases (Nakawo and Young, 1981; Nicholson
and Benn, 2006), explicitly account for moisture only when
relative humidity is 100 % (Reid et al., 2012; Fyffe et al.,
2014), incorporate a thickness-dependent “wetness factor”
(Fujita and Sakai, 2014), or parameterize latent heat based www.the-cryosphere.net/14/1555/2020/ The Cryosphere, 14, 1555–1577, 2020 1557 A. Giese et al.: Modeling moisture content in debris cover energy balance times that of the glacier tongue, do not compensate for the ab-
lation reduction impact of the debris on North Changri Nup
glacier (Brun et al., 2018). Field measurements and observa-
tions confirm the presence of water in debris: density mea-
surements at four sites show that deeper debris retains more
moisture, and water has been observed to both wet the debris
and pool within it. Figure 1. A map of West Changri Nup and North Changri Nup
glaciers, showing the location of the AWS (Sect. 2.2), which is also
the location of the measurements of debris temperature and point
mass balance. Source: Esri, DigitalGlobe, GeoEye, Earthstar Geo-
graphics, CNES/Airbus DS, USDA, USGS, AeroGRID, IGN, and
the GIS user community. The glacier outlines are from Sherpa et al. (2017). 2.1
Field site: West Changri Nup glacier West Changri Nup glacier (Fig. 1, 27.97◦N, 86.76◦E), also
known as White Changri Nup glacier, has an area of 0.92 km2
(measured 2013), ranges in elevation from 5330 to 5690 m,
and has a small debris-covered area despite being mostly
composed of clean ice. The debris is a granitic metamor-
phic mix, likely consisting of gneissic clasts eroded from the
surrounding cliffs (Searle et al., 2003); see Lejeune et al. (2013) for further details and a photograph of the debris. West Changri Nup lies 200 m southwest of North Changri
Nup glacier (Sherpa et al., 2017; Vincent et al., 2016) in
the Mt. Everest region of Nepal. The ablation zone of North
Changri Nup glacier is dominated by a debris cover that has
an insulating effect on mass balance (Vincent et al., 2016). Ice cliffs, despite imparting a localized ablation rate of ∼3 The Cryosphere, 14, 1555–1577, 2020 The Cryosphere, 14, 1555–1577, 2020 A. Giese et al.: Modeling moisture content in debris cover energy balance 1558 Table 2. Meteorological quantities and debris characteristics measured by the AWS on West Changri Nup glacier on 6 December 2012,
09:15–28 November 2014, 07:45 UTC. All sensors give values every 30 min; these values are 30 min averages of data with 30 s scanning
intervals for all values except the ultrasonic SR50 and wind direction, which are sampled every 30 min. The Kipp & Zonen CNR4 net
radiometer measures the spectral band 0.305 < wavelength (λ) < 2.8 µm for shortwave radiation and 5 < λ < 50 µm for longwave radiation. The SR50 senses surface height changes and, thus, can indicate the occurrence of ablation or accumulation but not measure it directly. Quantity
Data gaps
Instrument
Accuracy according
(%)
to the manufacturer
West Changri Nup AWS (5360 m a.s.l.)
Air temperaturea,b (◦C)
24
Temperature and relative humidity probe
±0.2 ◦C
(Vaisala HMP45C)
Relative humiditya,b (%)
24
Temperature and relative humidity probe
±2 %
(Vaisala HMP45C)
Wind direction (◦) and speedb (m s−1)
43
Anemometer (Young 05103-5)
±3◦and ±0.3 m s−1
Incident shortwave radiationb (W m−2)
18
Net radiometer (Kipp & Zonen CNR4)
±3 %
Reflected shortwave radiation (W m−2)
18
Net radiometer (Kipp & Zonen CNR4)
±3 %
Incoming longwave radiationb (W m−2)
24
Net radiometer (Kipp & Zonen CNR4)
±3 %
Outgoing longwave radiation (W m−2)
24
Net radiometer (Kipp & Zonen CNR4)
±3 %
Debris temperature at
24
Resistance temperature probes (TCA PT100)
±0.1 ◦C
5, 7.5, 10, and 12.5 cmc (◦C)
Ablation/accumulationc (m)
33
Ultrasonic depth gauge (Campbell SR50A),
±1 cm
Mass balance stake
Pyramid Research Station (5035 m a.s.l.)
Precipitation (kg m−2)b
0
Geonor T200B
±0.1 mm
An a indicates measurements that must be gathered with artificial aspiration in the daytime, a b denotes quantities used to drive ISBA-DEB, and a c marks variables
used for calibration or validation. Precipitation was measured hourly at Pyramid Research Station. An a indicates measurements that must be gathered with artificial aspiration in the daytime, a b denotes quantities used to drive ISBA-DEB, and a c marks variables
used for calibration or validation. Precipitation was measured hourly at Pyramid Research Station. 2.3
Model inputs the portion of data plotted with black in Fig. 2a and b and the
percentage of data gaps in Table 2 for the extent of missing
AWS data over the period used in this study. SURFEX must be forced with temporal and geographic
specifications, as well as continuous meteorological vari-
ables (complete list at http://www.umr-cnrm.fr/surfex/spip. php?article215, last access: 27 April 2020, and in the Sup-
plement): atmospheric temperature (K), atmospheric hu-
midity (Qair, kg kg−1), atmospheric pressure (Pa), rainfall
rate (kg m−2 s−1), snowfall rate (kg m−2 s−1), wind speed
(m s−1) and direction (degrees), incoming longwave radia-
tion (W m−2), direct shortwave radiation (W m−2), diffuse
shortwave radiation (W m−2), and near-surface CO2 concen-
tration (kg m−3). As is clear from the superscript “b” in Ta-
ble 2, which denotes the measurements used to drive the
model, not all of the parameters required to run SURFEX
listed above impact the fluxes relevant to integrated models
ISBA or ISBA-DEB. Missing meteorological values were approximated by the
monthly averages of values at the missing time step dur-
ing a longer period of data acquisition at the AWS than
used for this study: October 2010–November 2016. Every
missing value was filled with the corresponding time step’s
mean monthly value. Using values specific to timestamps
preserved both diurnal and seasonal variability in the gap-
filled dataset. This method proved inappropriate for wind
speed, whose amplitude and variability could not be con-
served with averages. For the whole series, the wind speed
data were gap-filled by the wind speed at the same times-
tamp in a different year of the AWS’s operation, randomly
selected. When the same timestamp in all years is missing a
wind speed, we chose the closest later timestamp with data
in any year. The 2-year period used in this study contains data gaps of
various lengths affecting different sensors (see Table 2). For
example, a battery problem prevented nighttime data read-
ings between April and December 2013 (Fig. 5), installation
problems made the wind readings questionable for several
months, and station tilt compromised the quality of some
measurements but not others. Because the forcing file for
SURFEX must be continuous, it was necessary to fill such
data gaps and periods when data were deemed suspect. See www.the-cryosphere.net/14/1555/2020/ A. Giese et al.: Modeling moisture content in debris cover energy balance 1559 Figure 2. (a, b) Continuous half-hourly forcing data in black, with in situ data overlaid in blue. The gaps are apparent where the black data
points are displayed; Sect. 2.3 describes the procedure used to assign these values. (c, d) The continuous precipitation dataset from Pyramid
(interpolated half-hourly), with phase partitioned by Tair at West Changri Nup glacier (top panel of a). Note the different y axis scales on (c)
and (d). Figure 2. (a, b) Continuous half-hourly forcing data in black, with in situ data overlaid in blue. The gaps are apparent where the black data
points are displayed; Sect. 2.3 describes the procedure used to assign these values. (c, d) The continuous precipitation dataset from Pyramid
(interpolated half-hourly), with phase partitioned by Tair at West Changri Nup glacier (top panel of a). Note the different y axis scales on (c)
and (d). 3.2
Model structure is a physically based scheme that solves both time- and
depth-dependent heat and moisture diffusion numerically
through mass- and heat-conserving implicit time schemes. It
provides a convenient basis for simulating the surface energy
balance of a supraglacial debris layer, after making modifi-
cations to account for the differences between soil and de-
bris. This work builds on that of Lejeune et al. (2013), who
used one of the SURFEX snow models, Crocus, to repre-
sent dry debris year-round, accounting for snowfall and sub-
freezing glacier temperatures during the accumulation sea-
son. However, we instead adapted the diffusive version of
SURFEX’s soil model (ISBA option DIF). As the full de-
tails of the ISBA-DIF option for heat and moisture transfer
and water phase changes within soil are presented in a series
of publications (Boone et al., 1999, 2000; Decharme et al.,
2011, 2013), they will only be summarized here to provide
context for the detailed modifications in the supraglacial de-
bris model, ISBA-DEB. By adapting ISBA, we have built a
model that not only simulates a supraglacial debris layer’s
temperature and moisture but also computes glacier melt. ISBA-DEB computes temperature and moisture in a snow–
debris–ice column. Temperature and moisture evolution are
calculated for 10–15 debris layers with user-specified thick-
nesses. Debris layers are assigned thermal, hydraulic, and
physical properties of glacial debris as informed by field
measurements on West Changri Nup glacier or, when un-
known, by the debris-covered glacier literature. The underly-
ing layers (up to 20 total layers are permitted by the model)
approximate a glacier. In ISBA, the glacier layers must be
soil, but in ISBA-DEB we assigned them a porosity of 99.9 %
and specified that they be ice-saturated. Since 99.9 % of the
volume of these layers is filled with solid ice, glacier layers
have an effective porosity of zero. ISBA-DEB computes temperature and moisture in a snow–
debris–ice column. Temperature and moisture evolution are
calculated for 10–15 debris layers with user-specified thick-
nesses. Debris layers are assigned thermal, hydraulic, and
physical properties of glacial debris as informed by field
measurements on West Changri Nup glacier or, when un-
known, by the debris-covered glacier literature. The underly-
ing layers (up to 20 total layers are permitted by the model)
approximate a glacier. 3.1
Model overview The ISBA land surface model (Noilhan and Planton, 1989)
within the SURFEX community-based open-source software
platform maintained by Météo-France (Masson et al., 2013) The Cryosphere, 14, 1555–1577, 2020 www.the-cryosphere.net/14/1555/2020/ A. Giese et al.: Modeling moisture content in debris cover energy balance et al.: Modeling moisture content in debris cover energy balance www.the-cryosphere.net/14/1555/2020/ 3.3
Physical processes ing variables can be found in Sect. 2.3. Figure 3 schemati-
cally shows the configuration of the domain and summarizes
fluxes and processes in the one-dimensional ISBA-DEB. 3.2
Model structure In ISBA, the glacier layers must be
soil, but in ISBA-DEB we assigned them a porosity of 99.9 %
and specified that they be ice-saturated. Since 99.9 % of the
volume of these layers is filled with solid ice, glacier layers
have an effective porosity of zero. Glacier meltwater enters the debris at the base, and rain
and snowmelt water enter the debris at the surface. Precipita-
tion, wind, air temperature and humidity, and incoming long-
wave and shortwave radiation measured on West Changri
Nup glacier drive the model. We neglect energy carried by
precipitation, an assumption supported by other work in the
Himalayas (Azam et al., 2014) and on the nearby Tibetan
Plateau (Huintjes et al., 2015). A discussion of the forc- www.the-cryosphere.net/14/1555/2020/ The Cryosphere, 14, 1555–1577, 2020 1560 A. Giese et al.: Modeling moisture content in debris cover energy balance Figure 3. General scheme of ISBA-DEB with fluxes and physical processes. Note that ISBA-DEB is a point-scale model but that the
schematic is shown in 2D for interpretability. Precipitation comprises part of the system’s mass flux and affects debris surface properties;
however, the heat it carries is not included in the surface energy budget. This schematic shows the user options for ISBA-DEB. In this study,
we use 13 debris layers of 1 cm, an ice surface layer of 1 cm, and six additional ice layers (total 20 layers) of varying thicknesses reaching to
a depth of 60 m. Figure 3. General scheme of ISBA-DEB with fluxes and physical processes. Note that ISBA-DEB is a point-scale model but that the
schematic is shown in 2D for interpretability. Precipitation comprises part of the system’s mass flux and affects debris surface properties;
however, the heat it carries is not included in the surface energy budget. This schematic shows the user options for ISBA-DEB. In this study,
we use 13 debris layers of 1 cm, an ice surface layer of 1 cm, and six additional ice layers (total 20 layers) of varying thicknesses reaching to
a depth of 60 m. 3.3.1
Heat diffusion ISBA-DEB, like ISBA, solves the temperature in all lay-
ers of the domain; the temperature profile is then passed to
a routine that computes energy fluxes, including evaporation
and glacier melt. The volume of glacier melt and the tem-
perature profile, which has been updated with any melt that
occurred during the time step, pass into the hydrology rou-
tines that calculate water volume and location in all allowed
layers – as well as its phase according to temperature. Given
that the measured debris thickness of 12.5 cm is accurate to
±1 cm, we use thirteen 1 cm layers of debris in ISBA-DEB. The prognostic state variables are assumed to be located at
the midpoint of each layer; accordingly, the uppermost simu-
lated temperature is at 0.5 cm depth in the debris, not the sur-
face. Under the 13 debris layers are seven layers of ice, with
increasing thicknesses. The layer boundaries in the glacier
are at 0.14, 0.16, 0.45, 2.25, 7.00, 20.0, and 30.0 m in depth. Forced with repeated years of 2013 meteorological data, the
model reaches steady state after 40 years of spin-up; this
gives an initial uniform temperature of 268.35 K and an ini-
tial uniform liquid soil water index of 0.1 m3 m−3. Other ini-
tial conditions require a longer spin-up. Above the debris is
a transient snowpack, represented by a dynamic 0–12 layers. The snow scheme used in ISBA-DEB is ISBA-ES (Boone
and Etchevers, 2001; Decharme et al., 2016). ISBA-DEB, like ISBA, solves the temperature in all lay-
ers of the domain; the temperature profile is then passed to
a routine that computes energy fluxes, including evaporation
and glacier melt. The volume of glacier melt and the tem-
perature profile, which has been updated with any melt that
occurred during the time step, pass into the hydrology rou-
tines that calculate water volume and location in all allowed
layers – as well as its phase according to temperature. Given
that the measured debris thickness of 12.5 cm is accurate to
±1 cm, we use thirteen 1 cm layers of debris in ISBA-DEB. The prognostic state variables are assumed to be located at
the midpoint of each layer; accordingly, the uppermost simu-
lated temperature is at 0.5 cm depth in the debris, not the sur-
face. Under the 13 debris layers are seven layers of ice, with
increasing thicknesses. 3.3.3
Moisture inputs and diffusion Water entering the debris from glacier melt and precipitation
moves with a vertical flow rate F (m s−1) and acts as a source
or sink of latent heat 8 (J m−3 s−1) from changes in phase as
a function of temperature. Mass leaves the system through
latent heat mass fluxes and runoff (R). The amounts of liquid
water (wl) and ice (wi), respectively, are given by ISBA-DEB calculates temperature for the snow–debris–
ice column continuously. However, since the glacier can-
not exceed 0 ◦C, we introduce a condition for the ice lay-
ers that follows an analogous scenario for snow in Boone
and Etchevers (2001). Only the top layer of ice contributes
to the glacier melt term. Underlying ice layers’ temperatures
are prevented from exceeding freezing by concentrating all
above-freezing energy into the melt of the top ice layer. The
top layer is 1 cm thick, far greater than the melt possible in a
single 15 min time step. ∂wl
∂t = −∂F
∂z −
8
Lmρw
−Sl
ρw
−R
ρw
(wmin ≤wl ≤wsat −wi),
(3)
∂wi
∂t =
8
Lmρw
−Si
ρw
(0 ≤wi ≤wsat −wmin),
(4) ∂wl
∂t = −∂F
∂z −
8
Lmρw
−Sl
ρw
−R
ρw
(wmin ≤wl ≤wsat −wi),
(3) (3) ∂wi
∂t =
8
Lmρw
−Si
ρw
(0 ≤wi ≤wsat −wmin),
(4) (4) where Lm is the latent heat of fusion (J kg−1), ρw is the den-
sity of water, and wsat is the water concentration at satura-
tion. Sl and Si are the external latent heat source–sink terms
(kg m−2 s−1) for water and ice, i.e., evaporation for water
and sublimation for ice. Values of important physical con-
stants and West Changri Nup glacier-specific parameters are
listed in Table A1. A minimum water content wmin = 0.0001
is retained for numerical stability; wmin in ISBA (0.001) was
decreased by an order of magnitude in ISBA-DEB given the
importance of the exact water content in heat and moisture A. Giese et al.: Modeling moisture content in debris cover energy balance 1561 Figure 4. Flow of the processes in the ISBA-DEB model. The red text indicates major changes introduced to the ISBA code in the creation
of ISBA-DEB. Table A1 contains physical constants and model parameters used for the runs on West Changri Nup glacier. Figure 4. Flow of the processes in the ISBA-DEB model. The red text indicates major changes introduced to the ISBA code in the creation
of ISBA-DEB. Table A1 contains physical constants and model parameters used for the runs on West Changri Nup glacier. end of the time step (Fig. 4). annual melt to compare with field measurements. The melt-
ing layer is implicitly refilled at the end of the time step such
that the 1 cm thick top layer begins every model iteration at
full ice saturation. cg
∂Tg
∂t = ∂G
∂z + 8
(2) (2) In Eq. (2), conduction flux G represents the term in brack-
ets on the right-hand side of Eq. (1). A zero flux at depth pro-
vides the Neumann lower boundary condition, and surface
flux from the energy balance provides the upper boundary
condition. Shortwave radiation, longwave radiation, and tur-
bulent fluxes together comprise the surface energy balance. Daily average energy balance components are shown for both
winter and summer in Fig. S2. 3.3.1
Heat diffusion The layer boundaries in the glacier
are at 0.14, 0.16, 0.45, 2.25, 7.00, 20.0, and 30.0 m in depth. F
d
i h
d
f 2013
l
i
l d
h The ISBA scheme assumes that heat flow along the ther-
mal gradient is the dominant first-order process and neglects
other heat transfer processes such as advection within the
soil. Such an assumption is common to land surface models
currently used in operational numerical weather prediction
and general circulation model applications. Heat capacity (c)
and thermal conductivity (K) are weighted averages of the
respective volumetric proportions of air, rock, water, and ice
(note that the latter is a difference from ISBA). ISBA-DEB updates the temperature profile for the entire
column each time step using the heat equation in one dimen-
sion: cg
∂Tg
∂t = ∂
∂z
K ∂Tg
∂z
,
(1) (1) Forced with repeated years of 2013 meteorological data, the
model reaches steady state after 40 years of spin-up; this
gives an initial uniform temperature of 268.35 K and an ini-
tial uniform liquid soil water index of 0.1 m3 m−3. Other ini-
tial conditions require a longer spin-up. Above the debris is
a transient snowpack, represented by a dynamic 0–12 layers. The snow scheme used in ISBA-DEB is ISBA-ES (Boone
and Etchevers, 2001; Decharme et al., 2016). where K is thermal conductivity (W m−1 K−1), cg volumet-
ric ground specific heat capacity (J m−3 K−1), z depth (m),
and Tg ground temperature (K). Temperature in debris lay-
ers evolves not only by conductive heat transfer but also by
latent heat from phase changes between water and ice in the
debris (8, J m−3 s−1) that are added to the right-hand side
of Eq. (1), giving Eq. (2). These phase changes are calcu-
lated subsequently to the heat transfer routine; the temper-
ature profile is updated accordingly as an adjustment at the The Cryosphere, 14, 1555–1577, 2020 www.the-cryosphere.net/14/1555/2020/ A. Giese et al.: Modeling moisture content in debris cover energy balance A. Giese et al.: Modeling moisture content in debris cover energy balance We assume this to be true at temperatures above 0 ◦C, also,
and follow the vapor parameterization of ISBA, where vapor
transport is not explicitly modeled on the basis of its being
small compared to heat transfer along the thermal gradient
and mass flow governed by Darcy’s law (i.e., on the basis of
scaling arguments). It should be noted that ISBA, in treating
vapor transport solely as diffusive, does include an additive
term for vapor conductivity. Vertical soil water flux is Here, k (m s−1) is hydraulic conductivity and ψ (m) is soil
matric potential, the potential energy attributed to the adhe-
sion of water to soil grains. There have been no observations
of ice growth at the surface in subfreezing temperatures on
West Changri Nup glacier (as on Mullins glacier, Antarctica,
by Kowalewski et al., 2011), suggesting that vapor is not a
dominant transport mechanism at subfreezing temperatures. We assume this to be true at temperatures above 0 ◦C, also,
and follow the vapor parameterization of ISBA, where vapor
transport is not explicitly modeled on the basis of its being
small compared to heat transfer along the thermal gradient
and mass flow governed by Darcy’s law (i.e., on the basis of
scaling arguments). It should be noted that ISBA, in treating
vapor transport solely as diffusive, does include an additive
term for vapor conductivity. Vertical soil water flux is The values of matric potential (m) at saturation, hydraulic
conductivity (m s−1) at saturation, and shape parameter (di-
mensionless) of the soil–water retention curve (ψsat, ksat, and
b, respectively) are typically calculated according to the con-
tinuous pedotransfer functions (PTFs) of Noilhan and Lacar-
rere (1995), which compute key hydraulic parameters based
upon soil composition. For PTF equations, see Appendix C1
of Decharme et al. (2011). Power curves of Brooks and Corey
(1966) relate matric potential, hydraulic conductivity, and
volumetric liquid water content to the variables computed by
PTFs. Values used in our simulations are listed in Table A1. Instead of using a PTF to calculate ksat, ISBA-DEB adopts
gravel’s ksat value (0.03 m s−1;
Domenico and Schwartz,
1998) throughout the debris except for at the bottommost
layer, where ksat = 0 m s−1. This supplies a flux of 0 for the
lower boundary condition, while rainfall and snowmelt pro-
vide the upper boundary condition. Equation (3) is solved
with a Crank–Nicolson implicit time scheme. A. Giese et al.: Modeling moisture content in debris cover energy balance F = −k ∂
∂z(ψ + z) −Dvψ
ρw
∂ψ
∂z
(6) (6) (Boone et al., 2000), where Dvψ is the isothermal vapor con-
ductivity (kg m−2 s−1) as in Braud et al. (1993). 3.3.2
Glacier melt If the top layer of ice exceeds freezing, melt is computed and
temperature reset to 0 ◦C. Subsurface ice temperatures (i.e.,
layers 14–20) are subsequently recalculated with this 0 ◦C
boundary condition, precluding melt from occurring in the
subsurface layers. Energy is conserved, and the amount of
water melted in the top layer of the glacier in each time step
is added to the overlying debris and tracked for a cumulative www.the-cryosphere.net/14/1555/2020/ The Cryosphere, 14, 1555–1577, 2020 1562 A. Giese et al.: Modeling moisture content in debris cover energy balance such that ice lowers debris porosity and enhances the matric
potential and vertical upward suction of water. diffusion calculations (Decharme et al., 2011; LeMoigne,
2018). The change in the liquid moisture in debris is, then,
the sum of vertical flow, phase change, inflow and/or evap-
oration, and runoff; that of ice is the phase change minus
sublimation. When there is ice in the debris, Eq. (6) is rewritten F = −κ ∂ψ
∂z −k,
(7) (7) Vertical soil water flux is given by the Richards’ equa-
tion and an additive term to account for water vapor. The
Richards’ equation is an expression derived from Darcy’s law
that represents water diffusion arising from pressure gradi-
ents in partially saturated media. where κ = ℘
k + Dvψ
ρw
and ℘= 10−α℘wi/w (Boone et al.,
2000)
i t
d th “i
i
d
ffii
t”
hi h i 2000). ℘is termed the “ice impedance coefficient”, which in-
hibits upward movement of water towards the freezing front,
and α℘is the “ice impedance factor”, equal to 6 in ISBA
(Johnsson and Lundin, 1991) and ISBA-DEB. The form of
Eq. (7) emphasizes that there is a drainage term k and dif-
fusion along a potential κ which includes isothermal vapor
pressure. 2000). ℘is termed the “ice impedance coefficient”, which in-
hibits upward movement of water towards the freezing front,
and α℘is the “ice impedance factor”, equal to 6 in ISBA
(Johnsson and Lundin, 1991) and ISBA-DEB. The form of
Eq. (7) emphasizes that there is a drainage term k and dif-
fusion along a potential κ which includes isothermal vapor
pressure. ∂wl
∂t = ∂
∂z
k(wl)
∂ψ
∂z + 1
(5) (5) Here, k (m s−1) is hydraulic conductivity and ψ (m) is soil
matric potential, the potential energy attributed to the adhe-
sion of water to soil grains. There have been no observations
of ice growth at the surface in subfreezing temperatures on
West Changri Nup glacier (as on Mullins glacier, Antarctica,
by Kowalewski et al., 2011), suggesting that vapor is not a
dominant transport mechanism at subfreezing temperatures. www.the-cryosphere.net/14/1555/2020/ A. Giese et al.: Modeling moisture content in debris cover energy balance 1563 Figure 5. Measured (a) and modeled (b) debris temperatures at depths of 5, 7.5, 10, and 12.5 cm in a 12.5 cm thick debris layer, displayed
to show that the period between the vertical dashed lines, 9 April 2014–23 October 2014, was most informative for comparison in tuning. There was a battery problem causing no nighttime temperature recordings in April–December 2013 (period indicated by the black box), and
the clear temperature disagreement in late 2014 results from a problem in the meteorological forcing file for ISBA-DEB (having insufficient
snowfall to produce the observed persistent snow cover). The model runs in (b) were carried out with τα = 30 and τmax = 48 h, and a closer
look at modeled and measured temperatures during the period of comparison is given in Fig. S3. Figure 5. Measured (a) and modeled (b) debris temperatures at depths of 5, 7.5, 10, and 12.5 cm in a 12.5 cm thick debris layer, displayed
to show that the period between the vertical dashed lines, 9 April 2014–23 October 2014, was most informative for comparison in tuning. There was a battery problem causing no nighttime temperature recordings in April–December 2013 (period indicated by the black box), and
the clear temperature disagreement in late 2014 results from a problem in the meteorological forcing file for ISBA-DEB (having insufficient
snowfall to produce the observed persistent snow cover). The model runs in (b) were carried out with τα = 30 and τmax = 48 h, and a closer
look at modeled and measured temperatures during the period of comparison is given in Fig. S3. A τα > 30 does not change the shape of the runoff timescale
(green curve in Fig. S1) markedly, and it also does not im-
prove the RMSE significantly. 2014) to its maximum value (also tuned) at the base of the
debris, depth H (m). τα controls the distribution of mois-
ture, with larger values leading to a concentration of water
at the debris–ice boundary and smaller values leading to a
more even distribution. Increasing values of τα show steeper
curves, with an increasing number of subsurface layers hav-
ing the same moisture content as the surface (Fig. S1). A. Giese et al.: Modeling moisture content in debris cover energy balance All
values considered give an increase in water with depth, which
is to be expected with the combination of gravity and the fact
that debris clasts become finer (with a greater ability to retain
water; see Sect. 1) with depth. Energy and water budgets in ISBA-DEB are the same as
those in ISBA, with the exception of an additional term for
glacier melt (Mice). Both budgets close, and details are pre-
sented in the Supplement. 4
Tuning Of the December 2012–November 2014 series used in this
study, we used 2014 debris temperatures to tune parameters
and both seasons of ablation to assess the impact of mois-
ture inclusion in ISBA-DEB. We compared simulated debris
temperatures with measured ones from 9 April to 23 Octo-
ber 2014 (reason displayed in Fig. 5), using an RMSE cal-
culation to capture the magnitude of temperature. We tested
five runoff timescale shape factors (τα, Fig. S1) and maxi-
mum runoff timescale (τmax) values of 3, 6, 18, 24, 48, 72,
and 96 h. The RMSE metric suggested a shallow minimum
for τα = 30 and negligible differences for different τmax val-
ues. We used τα = 30 and τmax = 48 for our modeling work,
despite the shallow minima, because they are highly plausi-
ble values. This parameterization is necessarily simple in the absence
of field measurements but corroborated by gravel’s high hy-
draulic conductivity (Domenico and Schwartz, 1998) and
the observed changes in debris’ grain size distribution with
depth. Debris grains tend to be smaller in size at the ice sur-
face than at the top of the debris layer, thereby imparting
more of a damming effect on entrained water lower in the
debris column. The timescale of sand draining is on the or-
der of a day or two (Blum et al., 2018), indicating an ap-
proximate magnitude to inform τmax tuning tests. Further,
debris permeability field tests show that after 10 s, ∼95 %
of a 100 mL volume of water poured into gravel and cobbles
drains. However, for fine particulates sampled at the ice inter-
face (< 5 mm in diameter), only ∼20 % of the water drains
in the same amount of time. Since it takes a saturated sandy soil 24–48 h to drain to its
field capacity, 48 h for τmax is consistent with measurements
of the kinds of particles at the base of a debris layer. A shape
factor of 30 is consistent with observations of wetted debris
right at the debris–ice interface (Nakawo and Young, 1981;
Conway and Rasmussen, 2000; Nicholson and Benn, 2012). 3.3.4
Water runoff The Richards’ equation (Eq. 5) includes both diffusion and
drainage terms. Observations suggest that moisture transport
in glacier debris is neither completely reservoir-like (as pa-
rameterized in Collier et al., 2014) nor fully governed by
Darcy’s law (as in the original ISBA for soil) but rather some
of both simultaneously. A number of studies (e.g., Collier
et al., 2014; Nicholson and Benn, 2012) mention a saturated
basal layer of debris, and Rounce and McKinney (2014) dis-
cuss deeper, wet debris overlain by dry debris; our own field
observations are consistent. The concentration of wetness at
the debris base is due both to the fact that debris coarsens
upward (Reid and Brock, 2010) and to the fact that the per-
meability of the overlying debris is greater than that of the
debris at the interface (precipitation quickly moves through
the debris until it reaches the impermeable ice surface). By
solving the Richards’ equation and using an appropriate hy-
draulic conductivity (Table A1), ISBA-DEB simulates both
diffusion and pooling. The pebble-sized to gravel-sized grains comprising the debris
cannot hold liquid water long-term, and water runs off with a
slope-dependent timescale (Zuo and Oerlemans, 1996; Rei-
jmer and Hock, 2008). The timescale is a linear function of
glacier surface slope, with values of 1 h−1 for 0◦and 0 h−1
for 90◦(Collier et al., 2014) at the surface and an increasing
value with depth. Runoff (kg m−2 s−1) can be expressed as R = wl,jρw1zj
τj
θ
90
,
(8) (8) where θ is glacier surface slope, measured from the hori-
zontal, and z is the thickness (m) of each layer j. Runoff
timescale τj must be ≤dt. τj = τmin + (τmax −τmin)
"
exp
τα
zj
H
exp(τα)
#
(9) (9) Moisture changes phase as a function of available mass
and energy (Boone et al., 2000; Giard and Bazile, 2000). As
soil freezes, ice is assumed to become part of the soil matrix τα is a tunable shape parameter defining the runoff timescale
from its minimum value at the surface (1 h; Collier et al., τα is a tunable shape parameter defining the runoff timescale
from its minimum value at the surface (1 h; Collier et al., The Cryosphere, 14, 1555–1577, 2020 A. Giese et al.: Modeling moisture content in debris cover energy balance 5
Results and discussion In this section, we present the results (and describe the be-
havior) of model simulations for nearly 2 years of meteoro-
logical forcing, describe key physical processes related to the www.the-cryosphere.net/14/1555/2020/ presence of debris, and show results from a series of sensi-
tivity tests related to parameter uncertainties. Its middle four panels show how the liquid and solid mois-
ture contents change with temperature in the top two and bot-
tom two layers of debris (i.e., layers 1, 2, 12, and 13). ISBA-
DEB simulates temperature evolution throughout the entire
debris–glacier model domain (Fig. 7a); the domain is 60 m
total, including the 13 debris layers, each 1 cm thick. Output
shows temperature amplitude attenuation and phase lag with
depth (clearly seen in Fig. 7b). The above-freezing tempera-
tures propagating into the ice cause melt (Fig. 6). Cumulative
melt at each time step (Fig. 8, blue dots) gives the total melt
(Fig. 8, red line). www.the-cryosphere.net/14/1555/2020/ www.the-cryosphere.net/14/1555/2020/ The Cryosphere, 14, 1555–1577, 2020 A. Giese et al.: Modeling moisture content in debris cover energy balance 1564 1564
A. Giese et al.: Modeling moisture content in debris cover energy balance
Figure 6. Time-evolving water content, ice content, and temperature of the top two and bottom two layers of debris, shown during a 2-week
period when water enters the debris both at its surface (from rain and snowmelt, top panel) and at its base (from glacier melt, bottom panel). The phase of the moisture changes as a function of temperature. Note that the moisture and temperature y scales vary between the layers. Figure 6. Time-evolving water content, ice content, and temperature of the top two and bottom two layers of debris, shown during a 2-week
period when water enters the debris both at its surface (from rain and snowmelt, top panel) and at its base (from glacier melt, bottom panel). The phase of the moisture changes as a function of temperature. Note that the moisture and temperature y scales vary between the layers. Figure 7. (a) Temperatures in the debris and underlying glacier throughout the period used for this study. The black line indicates the debris–
ice interface. The depth scales for debris and ice differ; debris thickness is 0.125 m whereas the ice extends to 60 m (note the nonlinear y scale
in the ice). Y labels correspond to the mid-layer depth of the 20 discrete layers used in ISBA-DEB. The temperatures simulated throughout
the whole 60 m column over the entire forcing period show phase lag and attenuation with depth, characteristics that are more clearly seen
in (b), which shows the temperature of various debris depths during an arbitrary day (20 February 2014). Figure 7. (a) Temperatures in the debris and underlying glacier throughout the period used for this study. The black line indicates the debris–
ice interface. The depth scales for debris and ice differ; debris thickness is 0.125 m whereas the ice extends to 60 m (note the nonlinear y scale
in the ice). Y labels correspond to the mid-layer depth of the 20 discrete layers used in ISBA-DEB. The temperatures simulated throughout
the whole 60 m column over the entire forcing period show phase lag and attenuation with depth, characteristics that are more clearly seen
in (b), which shows the temperature of various debris depths during an arbitrary day (20 February 2014). 5.1
Model simulation characteristics For layer 13 at the ice-debris interface (12.5 cm), con-
ductivity is greatest at the transition into winter, when the
water filling the pore spaces freezes (Fig. S5a). Conductivity
is greater for ice-saturated debris than for water-saturated de-
bris, which is still greater than for dry debris. Heat capacity is
greatest for water-saturated debris and less for ice-saturated
debris (still less for dry). As expected, heat capacity is greater
in the summer than winter in the bottommost debris layer
(Fig. S5b). The temporal and spatial evolution of these pa-
rameters throughout the debris column as a function of water
and ice contents is a strength of ISBA-DEB. Figure 9. Temporal evolution of (a) thermal conductivity and
(b) volumetric heat capacity according to debris moisture amount,
phase, and gradient. Figure S5 gives a different perspective on the
evolution: temporal changes in the debris surface, middle, and base
layers, as well as the 2013 monsoon season (JJAS) diurnal averages
for each quantity. 5.1
Model simulation characteristics During the model simulations, glacier melt, snowmelt, and
rain enter the debris base or surface. The moisture in each
layer evolves with time, and the phase of the moisture
changes as a function of temperature. Figure 6 illustrates de-
bris water input in the top (surface) and bottom (debris–ice
interface) panels. The Cryosphere, 14, 1555–1577, 2020 www.the-cryosphere.net/14/1555/2020/ A. Giese et al.: Modeling moisture content in debris cover energy balance A. Giese et al.: Modeling moisture content in debris cover energy balance 1565 Figure 8. Daily totals of modeled glacier melt (blue dots), overlaid by modeled cumulative melt (red line) for the entire period Decem-
ber 2012–November 2014 simulated by ISBA-DEB. For a visual depiction of how and when melt occurs in ISBA-DEB, see Fig. 6. Figure 8. Daily totals of modeled glacier melt (blue dots), overlaid by modeled cumulative melt (red line) for the entire period Decem-
ber 2012–November 2014 simulated by ISBA-DEB. For a visual depiction of how and when melt occurs in ISBA-DEB, see Fig. 6. Figure 9. Temporal evolution of (a) thermal conductivity and
(b) volumetric heat capacity according to debris moisture amount,
phase, and gradient. Figure S5 gives a different perspective on the
evolution: temporal changes in the debris surface, middle, and base
layers, as well as the 2013 monsoon season (JJAS) diurnal averages
for each quantity. layer may hold more water, with moisture concentrated in
more of the lower layers than solely layer 13.) During the
summer, the glacier is melting, and the bottommost layer
of debris is almost always saturated with liquid water, such
that it has a thermal conductivity (K) of ∼1.16 W m−1 K−1
(Table 1, Fig. S5a, c). The thermal conductivity of layer 13
changes little throughout the day; layer 1 shows slight vari-
ation in conductivity because its water content experiences
variation via condensation and evaporation. For conductivity
in the summer (Fig. S5c), a higher value means more water
content while a lower value indicates dryness. The summer
monsoon season (JJAS) mean diurnal patterns in conductiv-
ity (Fig. S5c) are similar to those in specific heat capacity
(Fig. S5d) because both are functions of water content, and
both thermal conductivity and heat capacity have higher val-
ues for the water-saturated debris in layer 13 than the drier
debris in the overlying layers – and vary little throughout the
day. 5.2
Wet vs. dry debris We ran an experiment to contrast the sub-debris melt un-
der totally dry, partially saturated, and fully saturated de-
bris layers forced with the same meteorological conditions
(Fig. 2) measured on West Changri Nup glacier between
December 2012 and November 2014. The “partially satu-
rated” scenario uses parameters listed in Table A1. In order As the debris’ moisture content and phase vary, its thermal
conductivity and heat capacity evolve accordingly (Fig. 9;
extreme values are listed in Table 1). Layers 1–12 look sim-
ilar because the moisture is concentrated in layer 13, which
is just above the ice–debris interface. (Note that the AWS on
West Changri Nup has a slope of 5◦and that a flatter debris A. Giese et al.: Modeling moisture content in debris cover energy balance Table 3. Mass balance components of three model runs (dry debris, partially saturated debris, and fully saturated debris) compared with the
available measured point mass balance from 5 December 2012 to 19 July 2014 (N/A indicates that data are not available). The observation-
driven model behavior and output presented in previous figures is from what is termed the “partially saturated scenario” here. Note that
runoff has not been taken into account for this comparison. Mass balance
Completely dry
Partially saturated
Fully saturated
Measurements
components
scenario
scenario
scenario
(mm w.e.)
(mm w.e.)
(mm w.e.)
(mm w.e.)
(mm w.e.)
5 December 2012–2 December 2013
Cumulated solid
289
289
289
precipitation
Melt
1241
1237
771
Sublimation
40
40
220
Evaporation
0
81
549
Ablation
1281
1359
1540
Point mass
−992
−1069
−1251
−1080 (at SR50)
balance
−753 (at stake)
2 December 2013–28 November 2014
Cumulated solid
368
368
368
precipitation
Melt
975
983
605
Sublimation
15
15
441
Evaporation
0
66
553
Ablation
990
1064
1599
Point mass
−622
−696
−1231
N/A
balance
9 April–19 July 2014
Cumulated solid
142
142
142
precipitation
Melt
500
495
267
Sublimation
8
8
338
Evaporation
0
47
362
Ablation
508
550
967
Point mass
−366
−408
−825
−760 (at SR50)
balance sublimation and debris water sublimation that occurs when a
snow cover is absent in subfreezing temperatures. to achieve fully dry debris, all rain and melted snow was
assumed to run off immediately, although solid precipita-
tion (snow) could persist and sublimate. Table 3 shows the
three computed point mass balance values (not accounting
for runoff) compared to available measurements. A bamboo
stake, which carries an uncertainty of ±200 mm w.e. (Vin-
cent et al., 2016), and SR50-detected surface height changes,
using a snow density of 200 km m−3, supply the measure-
ments. In 2014 the stake broke, and the SR50 was operational
from only 9 April to 19 July 2014. to achieve fully dry debris, all rain and melted snow was
assumed to run off immediately, although solid precipita-
tion (snow) could persist and sublimate. Table 3 shows the
three computed point mass balance values (not accounting
for runoff) compared to available measurements. A bamboo
stake, which carries an uncertainty of ±200 mm w.e. (Vin-
cent et al., 2016), and SR50-detected surface height changes,
using a snow density of 200 km m−3, supply the measure-
ments. www.the-cryosphere.net/14/1555/2020/ The Cryosphere, 14, 1555–1577, 2020 1566 The Cryosphere, 14, 1555–1577, 2020 A. Giese et al.: Modeling moisture content in debris cover energy balance In 2014 the stake broke, and the SR50 was operational
from only 9 April to 19 July 2014. The glacier under completely dry debris melts signifi-
cantly more than the glacier situated under fully saturated de-
bris in all three periods (Table 3). The glacier under partially
saturated debris gives a simulated melt close to that under dry
debris. In 2013, the SR50 data are not in agreement with the
single ablation stake only a few meters away because the spa-
tial variability of point mass balance can be very high over a
few meters of debris cover (Vincent et al., 2016), due to dif-
ferences in debris thicknesses, properties, and water content. Overall, there is a reasonable agreement between measured
and modeled point mass balances, lending confidence to the
simulations. Important characteristics of Table 3 include the dry debris’
zero evaporation but sublimation commensurate with that of
the partially saturated debris; this results from the assumption
that rain and melt run off instantaneously while ISBA-DEB
still models a snow cover that can sublimate. The sublimation
computed in the fully saturated scenario is a sum of snow Sub-debris melt is a function of the debris thickness, which
is the same for all three cases, and the thermal diffusivity
of the debris (K/cg in Eq. 1), which differs for all three The Cryosphere, 14, 1555–1577, 2020 www.the-cryosphere.net/14/1555/2020/ www.the-cryosphere.net/14/1555/2020/ A. Giese et al.: Modeling moisture content in debris cover energy balance A. Giese et al.: Modeling moisture content in debris cover energy balance 1567 A. Giese et al.: Modeling moisture content in debris cover e
Figure 10. Debris layer moisture by phase in (a) the partially satu-
rated ISBA-DEB and (b) the fully saturated ISBA-DEB scenarios. As shown in (b), the fully saturated debris is water-filled (having a
lower thermal diffusivity than dry debris) during the summers when
debris surface temperatures lead to glacier melt. The difference in
moisture in the debris surface layer accounts for the different sur-
face latent heat fluxes (Fig. 11). We refer the interested reader to the
2013 JJAS mean diurnal water and ice contents (Fig. S4). Figure 10. Debris layer moisture by phase in (a) the partially satu-
rated ISBA-DEB and (b) the fully saturated ISBA-DEB scenarios. As shown in (b), the fully saturated debris is water-filled (having a
lower thermal diffusivity than dry debris) during the summers when
debris surface temperatures lead to glacier melt. A. Giese et al.: Modeling moisture content in debris cover energy balance Debris layer moisture by phase in (a) the partially satu-
rated ISBA-DEB and (b) the fully saturated ISBA-DEB scenarios. As shown in (b), the fully saturated debris is water-filled (having a
lower thermal diffusivity than dry debris) during the summers when
debris surface temperatures lead to glacier melt. The difference in
moisture in the debris surface layer accounts for the different sur-
face latent heat fluxes (Fig. 11). We refer the interested reader to the
2013 JJAS mean diurnal water and ice contents (Fig. S4). as a function of the amount, phase, and location of mois-
ture. Completely water-saturated debris has a thermal diffu-
sivity that is less than half of the diffusivity for completely
ice-saturated debris. Dry debris’ diffusivity falls nearly mid-
way between the two (Table 1). The share of water and ice
in the interstitial spaces of the partially (Fig. 10a) and fully
(Fig. 10b) saturated debris differs significantly in amount and
distribution. Ice-saturated debris conducts heat much more
efficiently than water-saturated debris does; however, glacier
melt happens only when the glacier surface exceeds 0 ◦C, and
efficiently conductive, ice-laden debris overlying a melting
glacier is a physical impossibility. The fully saturated de-
bris conducts heat that leads to melt with an efficiency of
κ = 3.572×10−7 m2 s−1, while the dry debris has a diffusiv-
ity of 5.867 × 10−7 m2 s−1; hence, more melt occurs below
the dry debris in ISBA-DEB. ing radiative fluxes will vary mainly based on surface albedo
and surface temperature (they are both modulated by the ab-
sence or presence of snow cover). Latent heat varies most
between the scenarios and has the greatest impact on surface
energy balance and, thus, its residual, the conductive heat
flux into the debris and ultimately transferred to the underly-
ing glacier. (For completeness, note that sensible heat flux is
larger for a dry surface where less latent heat transfer is tak-
ing place than for a wet surface but is comparatively smaller
in magnitude.) Even in the fully saturated debris, the latent
heat from freeze–thaw of the debris is 2 orders of magni-
tude less than the latent heat from evaporation and sublima-
tion. A. Giese et al.: Modeling moisture content in debris cover energy balance The significant energy used for evaporation and subli-
mation leaves comparatively little energy for heat conduction
through the debris–ice interface when the debris layer is fully
saturated and water-filled (or, in winter, ice-filled); indeed,
when latent heat is significant, the surface does not heat up,
and the temperature gradient controlling conduction is weak. Therefore, not only does a wet debris layer transfer heat less
efficiently from its surface to its base than dry debris because
of a decreased thermal diffusivity, but it also has less energy As expected, the surface latent heat flux is much greater
over the saturated debris, and the latent heat flux due to phase
changes within the debris is also greatest for the saturated
debris (Fig. 11). In the scenarios compared here, incoming
shortwave and longwave components are the same. Outgo- A. Giese et al.: Modeling moisture content in debris cover energy balance The difference in
moisture in the debris surface layer accounts for the different sur-
face latent heat fluxes (Fig. 11). We refer the interested reader to the
2013 JJAS mean diurnal water and ice contents (Fig. S4). Figure 11. Cumulative heat fluxes in the three scenarios (dry debris,
partially saturated debris, and fully saturated debris). The surface
latent heat flux exchanged with the atmosphere (a) indicates how
much energy is removed from the system at the debris surface, while
the latent heat from phase changes within the debris (b) shows how
much energy is removed from the system in the subsurface of the
debris layer. Greater latent heat fluxes are balanced with a lower
conductive heat flux through the debris into the underlying glacier. Figure 11. Cumulative heat fluxes in the three scenarios (dry debris,
partially saturated debris, and fully saturated debris). The surface
latent heat flux exchanged with the atmosphere (a) indicates how
much energy is removed from the system at the debris surface, while
the latent heat from phase changes within the debris (b) shows how
much energy is removed from the system in the subsurface of the
debris layer. Greater latent heat fluxes are balanced with a lower
conductive heat flux through the debris into the underlying glacier. Figure 11. Cumulative heat fluxes in the three scenarios (dry debris,
partially saturated debris, and fully saturated debris). The surface
latent heat flux exchanged with the atmosphere (a) indicates how
much energy is removed from the system at the debris surface, while
the latent heat from phase changes within the debris (b) shows how
much energy is removed from the system in the subsurface of the
debris layer. Greater latent heat fluxes are balanced with a lower
conductive heat flux through the debris into the underlying glacier. Figure 10. Debris layer moisture by phase in (a) the partially satu-
rated ISBA-DEB and (b) the fully saturated ISBA-DEB scenarios. As shown in (b), the fully saturated debris is water-filled (having a
lower thermal diffusivity than dry debris) during the summers when
debris surface temperatures lead to glacier melt. The difference in
moisture in the debris surface layer accounts for the different sur-
face latent heat fluxes (Fig. 11). We refer the interested reader to the
2013 JJAS mean diurnal water and ice contents (Fig. S4). Figure 10. The Cryosphere, 14, 1555–1577, 2020 5.3
Sensitivity tests Aerodynamic roughness lengths are used to determine the
two exchange coefficients (CH, CD) in the stability correc-
tion for the bulk method of calculating turbulent heat fluxes
(i.e., fits to the Monin–Obukhov functions; see Noilhan and
Mahfouf, 1996). CD (for momentum) depends on z0,m, while
CH (for H and LE) depends on both z0,m and z0,h. The surface
roughness length due to momentum, z0,m, is the height above
a rough surface at which the horizontal wind speed is zero. It varies with time and snowfall, and it is notoriously poorly
constrained (Quincey et al., 2017) and difficult to compute
consistently with different approaches (Miles et al., 2017). The values of both roughness lengths are inherently diffi-
cult to measure and poorly known because they depend on
not only the local surface state but also meteorology and sur-
rounding surface features. We performed sensitivity tests on the six parameters listed in
Table 4 to explore and quantify uncertainty associated with
parameterizations in ISBA-DEB. In most cases, the tested
ranges were informed by literature. In the case of albedo,
which has been found to vary up to 0.6 on debris-covered
glaciers in the Everest region (Kayastha et al., 2000), we
tested values ranging from 0.1 to 0.5; Kayastha et al. (2000)
claimed that most albedo values fall in the 0.2–0.4 range,
while Nicholson and Benn (2012) showed that 62 % of their
measurements fell between 0.1 and 0.3. A midday mean of
the ratio of reflected to incoming shortwave radiation mea-
sured on West Changri Nup glacier gives an albedo of 0.2. Despite the fact that albedo has been measured on West
Changri Nup glacier, ISBA-DEB’s sensitivity to this parame-
ter is important to assess for future application of ISBA-DEB
to other debris covers. Studies that informed our range of tested values were In-
oue and Yoshida (1980) and Takeuchi et al. (2000) for 0.0035
and 0.0063 m on Khumbu glacier, respectively; Reid and
Brock (2010) for 0.016 m on Miage glacier; and Lejeune
et al. (2013) for 0.05 m determined through model tuning
on West Changri Nup glacier. We test 0.1 m, reasoning that
debris’ roughness can be approximated by that of rough ice
(Smeets and Van den Broeke, 2008). An upper end-member,
0.5 m, is taken from the value of Miles et al. (2017) for boul-
ders on Lirung glacier. A. Giese et al.: Modeling moisture content in debris cover energy balance 1568 to transfer in the first place because of the other energy fluxes
(mainly the surface latent heat) associated with the scenario. were “effective” values reported for Miage glacier. As Brock
et al. (2010) measured thermal conductivity in the abla-
tion season, the reported K of 0.94 W m−1 K−1 (Reid and
Brock, 2010) was likely higher than that of perfectly dry
Miage glacier debris since any moisture in the pore spaces
would have had K = 0.57 W m−1 K−1 (water) rather than
K = 0.024 W m−1 K−1 (air). Additionally, debris on Miage
glacier (Italy) may have a dramatically different lithology
than debris on Changri Nup glacier (Nepal). Reports of de-
bris conductivity on Khumbu glacier, adjacent to Changri
Nup, include 0.85 W m−1 K−1, 1.28 W m−1 K−1 (Conway
and Rasmussen, 2000), and 0.96 W m−1 K−1 (Rounce and
McKinney, 2014) and indicate that 0.94 W m−1 K−1 is not
inappropriate to apply to West Changri Nup. West Changri
Nup’s debris is most likely comprised of the sillimanite
gneiss that forms its surrounding mountains (Searle et al.,
2003). The US Geological Survey’s “Thermal Properties of
Rocks” (Robertson, 1988) gives a thermal conductivity of
2 W m−1 K−1 (see Figs. 3 and 16 therein). For debris with
39 % porosity and air-filled pore spaces, a weighted-average
K for dry debris is 1.2 W m−1 K−1, which is within our
tested range. Thermal conductivity is difficult to measure in
the field, and it is not known how transferable the limited
available measurements are to other debris covers and con-
ditions. It is also not known whether a weighted average of
bedrock and air thermal properties is a valid representation
of porous debris. Accordingly, we intended to encompass the
true value(s) in the range for which we tested ISBA-DEB’s
response. Overall, our results show that including moisture in
supraglacial debris with ISBA-DEB over 2012–2014 on West
Changri Nup glacier does not significantly decrease sub-
debris glacier melt compared to a dry debris layer – until
the debris is fully saturated and the top layer holds a signif-
icant amount of moisture. A. Giese et al.: Modeling moisture content in debris cover energy balance In general with ISBA-DEB, any
change in melt under partially saturated debris is determined
by the distribution and amount of water in the debris be-
cause water decreases the debris layer’s bulk thermal diffu-
sivity and causes it to lose energy, which otherwise might be
used for ice melt, to latent heat fluxes. The amount of melt is
highly sensitive to the runoff parameterization, the assump-
tions made in ISBA-DEB, and the meteorological forcing. The partially saturated debris is predominantly dry, with the
exception of the lowermost layer (Fig. 10a, matching obser-
vations described in Sect. 3.3.3). With different runoff pa-
rameters, a flatter slope, and/or more water entering the de-
bris from precipitation, the partially saturated debris scenario
could yield an annual ablation closer to the value for satu-
rated debris in Table 3. ISBA-DEB follows ISBA’s calcula-
tion of atmospheric latent heat exchange from the top layer
only. Introducing an atmospheric latent heat flux within the
debris similar to that at the “saturated horizon” of Collier
et al. (2014) or a wind flow parameterization as in Evatt et al. (2015) would give lower glacier melt underlying a partially
saturated debris layer. www.the-cryosphere.net/14/1555/2020/ www.the-cryosphere.net/14/1555/2020/ The Cryosphere, 14, 1555–1577, 2020 A. Giese et al.: Modeling moisture content in debris cover energy balance A. Giese et al.: Modeling moisture content in debris cover energy balance et al.: Modeling moisture content in debris cover energy balance 1569 Table 4. Summary of sensitivity tests performed on ISBA-DEB on albedo (α), thermal conductivity (K), roughness lengths for momentum
(z0,m) and heat (z0,h), emissivity (ϵ), and slope (θ). An asterisk indicates values that are used in Sect. 5.1 and for the partially saturated
scenario in Sect. 5.2. Each parameter was varied while the others were held at their values with an asterisk. These values provide the basis of
comparison in columns 3 and 4. Table 4. Summary of sensitivity tests performed on ISBA-DEB on albedo (α), thermal conductivity (K), roughness lengths for momentum
(z0,m) and heat (z0,h), emissivity (ϵ), and slope (θ). An asterisk indicates values that are used in Sect. 5.1 and for the partially saturated
scenario in Sect. 5.2. Each parameter was varied while the others were held at their values with an asterisk. These values provide the basis of
comparison in columns 3 and 4. A. Giese et al.: Modeling moisture content in debris cover energy balance Parameter
Cumulative melt (mm)
% change in melt
% change in parameter range
Dec 2012–Nov 2014
(relative to ∗)
(from ∗)
α = 0.1
2425.10
9.25
-20
α = 0.2∗
2219.80
–
–
α = 0.3
2001.10
−9.85
20
α = 0.4
1766.60
−20.42
40
α = 0.5
1510.40
−31.96
60
K = 0.6 W m−1 K−1
1533.30
−30.93
−49
K = 0.7 W m−1 K−1
1760.80
−20.68
−34
K = 0.8 W m−1 K−1
1966.20
−11.42
−20
K = 0.94 W m−1 K−1∗
2219.80
–
–
K = 1.0 W m−1 K−1
2317.20
4.39
8.6
K = 1.1 W m−1 K−1
2466.40
11.11
23
K = 1.2 W m−1 K−1
2603.10
17.27
37
K = 1.3 W m−1 K−1
2730.70
23.02
51
z0,m = 0.0035 m
3081.90
38.84
−9.4
z0,m = 0.0063 m
2879.50
29.72
−8.8
z0,m = 0.016 m
2575.20
16.01
−6.9
z0,m = 0.05 m∗
2219.80
–
–
z0,m = 0.1 m
2008.40
−9.52
10
z0,m = 0.5 m
1502.30
−32.32
91
z0,h = 0.05 m
1426.20
−35.75
91
z0,h = 0.0125 m
1945.70
−12.35
15
z0,h = 0.0071 m
2115.30
−4.71
4.2
z0,h = 0.005 m∗
2219.80
–
–
z0,h = 0.002 m
2479.60
11.70
−6.0
z0,h = 0.001 m
2663.30
19.98
−8.1
z0,h = 0.0005 m
2827.60
27.38
−9.1
z0,h = 0.0003 m
2949.70
32.88
−9.5
ϵ = 0.9
2246.90
1.22
−40
ϵ = 0.94∗
2219.80
–
–
ϵ = 1
2179.90
−1.80
60
θ = 0◦
1479.40
−33.35
−50
θ = 1◦
1539.90
−30.63
−40
θ = 2◦
1694.30
−23.67
−30
θ = 3◦
1908.20
−14.04
−20
θ = 4◦
2094.60
−5.64
−10
θ = 5◦∗
2219.80
–
–
θ = 6◦
2284.30
2.91
10
θ = 10◦
2320.40
4.53
50
u glacier (0.0184 and 0.0243 m) fall within the
ted values. h
l
th
f h
t t
f
(
) i
i
range for ISBA-DEB given the uncertainty sur
value of this parameter. We test ratio values of 1
50 100
d 200 on Khumbu glacier (0.0184 and 0.0243 m) fall within the
range of tested values. range for ISBA-DEB given the uncertainty surrounding the
value of this parameter. We test ratio values of 1, 4, 7, 10, 25,
50, 100, and 200. The roughness length of heat transfer (z0,h) is incorpo-
rated into ISBA through the variable z0,m/z0,h, which must
be ≥1. A. Giese et al.: Modeling moisture content in debris cover energy balance The smaller this ratio, the larger z0,h and the larger
CH (and turbulent flux). z0,m/z0,h is commonly taken to be
= 10 (ISBA default, Mascart et al., 1995), but we test a wide Emissivity affects net longwave radiation and other sur-
face fluxes through feedbacks; we test the model’s response
to a wide range of values for this parameter (i.e., 0.9–1). Fi-
nally, we test how sensitive model-simulated melt is to the www.the-cryosphere.net/14/1555/2020/ 5.3
Sensitivity tests Their value for gravels (0.005 m) and
the recent measurements of Quincey et al. (2017) at two sites A study on Miage glacier, Italy, provided 0.94 W m−1 K−1
as a starting point for thermal conductivity tests (Reid
and Brock, 2010), though we varied the conductivity val-
ues throughout the range reported in the literature, 0.60
to 1.29 W m−1 K−1 (Rounce et al., 2015). This was a par-
ticularly important sensitivity test to perform because, as
noted in the caption of Table 1, the thermal values we as-
sumed valid for dry debris on West Changri Nup glacier The Cryosphere, 14, 1555–1577, 2020 www.the-cryosphere.net/14/1555/2020/ A. Giese et al.: Modeling moisture content in debris cover energy balance As shown in the subplots of Fig. S6 and associated Table 4,
the ISBA-DEB model would give significantly different melt
results on glaciers with a much more reflective debris cover
(i.e., a lithology with a higher albedo), a much flatter sur-
face, or different z0,m and z0,h values. Given the responsive-
ness of ISBA-DEB’s calculated melt to thermal conductivity,
we elected to compare simulated and measured debris tem-
peratures to glean information about which value of thermal
conductivity yields simulated debris temperatures that most
closely match measured ones in timing (via R2 of envelope
functions) and magnitude (via RMSE). Our tests do suggest
an optimal value of 1 W m−1 K−1, which agrees closely with
that of Reid and Brock (2010), though further tests over more
time periods with available debris temperatures are necessary
because neither of these tests yielded deep minima. Since H = ρair ×Cp ×CH ×V ×(Tsurf −Tair), where Cp is
heat capacity and V wind speed, an anomalously large H
implies that the surface debris in ISBA-DEB is overheat-
ing and evacuating too much heat. If any of the simulated
incoming components to the energy balance are too large,
the model could potentially compensate by overestimating
the debris surface temperature. The overestimated tempera-
ture reflects the fact that multiple parameter sets can provide
equally good model outputs (equifinality principle) and could
be due to a number of underlying factors. First, both the sen-
sitivity tests and the disagreement of ISBA-DEB’s sensible
heat magnitude with that of Steiner et al. (2018) suggest that
z0,m = 0.05 m and z0,h = 0.005 m are not spatially represen-
tative values. Additionally or alternatively, processes miss-
ing from ISBA-DEB could influence H. ISBA-DEB’s lack
of advection (rather than explicit inclusion of wind dynamics
like Evatt et al., 2015) and its highly simplified vapor trans-
port (rather than empirical fit to the curves of Monin and
Obukhov (1954)) may account for the anomalously large H. Finally, we assume our observed energy fluxes comprise a
closed budget, a condition that ISBA follows, but we cannot
rule out that energy budget errors in observations contribute
to the large H magnitude without a detailed future evalua-
tion. Robust assessment of H over debris-covered glaciers
requires more measurements using an eddy correlation ap-
proach. Figure 13 accompanies Fig. A. Giese et al.: Modeling moisture content in debris cover energy balance A. Giese et al.: Modeling moisture content in debris cover energy balance 1570 Figure 12. Cumulative glacier melt over the December 2012–
November 2014 period under the extreme values for each param-
eter listed in Table 4. An equivalent figure for each parameter may
be found in Fig. S6. by definition, has the greatest impact on the conductive heat
flux, the cumulative value of which varies from 1.343 ×
109 J m−2 for K = 0.6 W m−1 K−1 to 3.025 × 109 J m−2 for
K = 1.3 W m−1 K−1. Because roughness lengths are difficult to measure, they
are unlikely to be well constrained; nevertheless, they are
parameters to which ISBA-DEB is extremely sensitive. Ta-
ble 4 gives the percent change in melt resulting from the pa-
rameter change and the percent that the parameter value was
perturbed relative to the plausible range, which was tested. Because roughness lengths are so poorly known, we tested
a large range that was nonlinear in its distribution. The high
sensitivity to roughness length perturbations (a 9 % decrease
in the value of roughness length for momentum resulting in a
39 % increase in melt, for example) is partly reflective of the
size and distribution of the range tested but also underscores
the model’s sensitivity to these parameters and the impor-
tance of focusing future work on them. See Sect. 5.5 for a
discussion of distributed modeling. Figure 12. Cumulative glacier melt over the December 2012–
November 2014 period under the extreme values for each param-
eter listed in Table 4. An equivalent figure for each parameter may
be found in Fig. S6. user-specified slope that determines runoff. We test a range
from flat to a slope of 10◦. Figure 12 summarizes cumulative
melt over the entire 2-year period for the extreme parame-
ter values tested, and Fig. S6 shows cumulative melt for all
parameter values in Table 4. Our simulations showed that sensible heat flux (H) is 1
order of magnitude larger during the monsoon and a factor
of 7 larger during the pre- and post-monsoon than that mea-
sured using an eddy correlation approach over Lirung glacier
(Langtang area, Nepal, 4250 m a.s.l.; Steiner et al., 2018). Al-
though the sites differ in topography, meteorology, and ele-
vation and are, thus, not fully comparable, we suspect that
sensible heat flux is overestimated by ISBA-DEB on West
Changri Nup glacier. www.the-cryosphere.net/14/1555/2020/ The Cryosphere, 14, 1555–1577, 2020 A. Giese et al.: Modeling moisture content in debris cover energy balance 1571 gure 13. Cumulative energy fluxes most impacted by the six parameters perturbed through sensitivity tests, shown with the maximum and
inimum tested values. Subfigure locations of Fig. S6 correspond. Figure 13. Cumulative energy fluxes most impacted by the six parameters perturbed through sensitivity tests, shown with the maximum and
minimum tested values. Subfigure locations of Fig. S6 correspond. Some decrease in sensible heat magnitude was achieved
through adjusting the roughness lengths and albedo, although
further work is necessary to improve the sensible heat flux
calculated by ISBA-DEB when the simulated surface tem-
perature is greater than the prescribed air temperature for an
extended period (i.e., during unstable conditions). Because
an excessively large sensible heat flux removes heat from the
debris that could otherwise be put towards glacier melt, re-
solving the sensible heat overproduction would likely lead to
an increase in ISBA-DEB’s glacier melt calculation. indicates more runoff and less potential for water buildup and
turbulent heat exchange. A flatter slope gives a more water-
saturated debris layer, and it is useful to make a comparison
between the model runs with various slope values and model
runs with dry vs. saturated debris (Sect. 5.2). For slopes
greater than 5◦, the debris is drained sufficiently well that
it is no longer dominated by the thermal properties of water. Sub-debris melt on a slope of 0–4◦somewhat resembles that
under fully saturated debris, with the flatter debris having a
lower interface flux (Fig. 13f) and higher surface latent heat
flux than the more sloped and comparatively drier debris. The
flat debris does not show nearly the same magnitude of sur-
face latent heat flux as the saturated debris does; while its top
layer has more moisture than the debris on steeper slopes, it
is far from fully saturated. The configuration that holds more
water in the debris has a greater in-debris latent heat flux
(Fig. 11b). A. Giese et al.: Modeling moisture content in debris cover energy balance S6 in that it shows which
energy fluxes have the strongest impact on the variation
in the cumulative melt curves (shown in Fig. S6). Rough-
ness lengths determine the surface turbulent heat fluxes, and
albedo and emissivity affect net radiation. Thus, these four
parameters determine how much energy is available to enter
the debris. Both slope (which effectively alters the thermal
properties of the debris by controlling the residence time of
water) and thermal conductivity determine how much energy
reaches the ice surface to melt it. (Note that, by modulating
water content in the debris, slope has a large effect on the
flux of latent heat due to phase changes within the debris.)
When water is available at the surface, as with a flatter slope
or the fictional “fully saturated” wet scenario in Sect. 5.2,
ISBA-DEB’s simulation of sublimation and evaporation re-
moves energy and reduces ice melt. Thermal conductivity, www.the-cryosphere.net/14/1555/2020/ The Cryosphere, 14, 1555–1577, 2020 A. Giese et al.: Modeling moisture content in debris cover energy balance 5.4
Uncertainty Although slope is known for West Changri Nup glacier’s
AWS, the slope at other sites where ISBA-DEB could be ap-
plied will inevitably vary. A slight change in surface slope,
particularly if the slope is less than 5◦, has a dramatic impact
on the sub-debris melt calculated by ISBA-DEB. Runoff is
directly proportional to slope angle such that a greater slope The Cryosphere, 14, 1555–1577, 2020 The Cryosphere, 14, 1555–1577, 2020 A. Giese et al.: Modeling moisture content in debris cover energy balance In order for melt computed by ISBA-DEB to be within
10 % of true melt, albedo must not vary by more than
±0.1 and the thermal conductivity should stay within
±0.15 W m−1 K−1. Table 4 shows the model sensitivity to
roughness lengths and emphasizes the need to verify a site-
specific value before applying ISBA-DEB at that differ-
ent site. The varied measurements of momentum roughness
length (two-thirds on neighboring Khumbu glacier) increase
model melt by 16 %, 30 %, and 39 % over 2 years, while the
theoretically reasoned greater roughness length decreases it
by nearly 10 % and the value of Miles et al. (2017) for boul-
ders by over 30 %. The thermal roughness length is even
more poorly constrained than the roughness length for mo-
mentum, and our tests simply explored model response to a
range of ratios. They demonstrate how crucial this parameter,
which determines calculated latent heat fluxes, is to an en-
ergy balance model. Losing more energy to latent heat leaves
less for glacier melt. Increasing the ratio of momentum to
thermal roughness lengths from 10 to 25 increases melt by
more than 10 %. sites, that, in contrast to soil, supraglacial debris has a higher
permeability and lower evaporation rate. A lower evaporation
rate is consistent with the fact that debris stores moisture at
depth. Moisture deep in debris is less prone to evaporation
than moisture on the surface since permeability (and, thus,
ease of air flow) generally decreases into the debris (Evatt
et al., 2015), although some does evaporate. Evatt et al. (2015) designed a model that was novel in its
incorporation of the evaporative heat flux throughout the de-
bris layer. They justified their parameterization by noting that
the melt occurs at the debris base and pointed out that calcu-
lating evaporation lower in the debris requires accounting for
the air flow within the debris. Significant air flow is absent
in soils for which the original ISBA was designed. By ad-
justing the surface vapor pressure term based upon the water
content in the debris, Collier et al. (2014) also considered the
evaporative heat flux within the debris layer. Evaporative heat fluxes in ISBA are computed based on
moisture content of only the top layer. A. Giese et al.: Modeling moisture content in debris cover energy balance While such a parame-
terization may be reasonable in soils, it has no physical basis
in far more permeable debris, within which the atmosphere
can exchange heat and mass at different depths. Collier et al. (2014) found that the latent heat flux calculated at the sat-
urated horizon of their reservoir model was too low, offer-
ing that their computed saturated horizon itself was flawed
without accounting for capillary effects. Therefore, ISBA-
DEB provides an advancement in prognosing the location of
moisture, which both pools (as in the model of Collier et al. (2014)) and undergoes diffusion. Introducing the possibility
for latent heat fluxes to arise from within the debris is the
next step in advancing ISBA-DEB. Another source of uncertainty stems from the fact that the
debris temperatures come from sensors which can migrate
within the debris: the four sensors were installed at depths
of 5, 7.5, 10, and 12.5 cm in 12.5 cm of debris in Decem-
ber 2012. Before the end of their deployment in Novem-
ber 2014, their depths were checked only twice, Decem-
ber 2013 and April 2014. Therefore, while a portion of the
modeled–measured temperature mismatch (Fig. S3) is due
to shortcomings of ISBA-DEB, another portion is due to the
migration of the thermistors in the debris, which renders their
depths unknown. It is not possible to attribute the disagree-
ment of ISBA-DEB temperatures with measured ones en-
tirely to the model. Simulating neither evaporation nor advective heat transfer
within the debris accounts for the relative melt modeled un-
der 12.5 cm of wet, dry, and partially wet debris (Sect. 5.2,
Table 3). While ISBA-DEB can accommodate any debris
thickness, so long as the entire debris–glacier column is dis-
cretized into 20 or fewer layers, we expect sub-debris melt
values calculated by the model in its current version to match
measurements of melt under thinner debris more closely than
measurements under thicker debris. The effect of neglecting
evaporation deeper than the surface is decreased for thinner
debris, as is the effect of neglecting advection within the de-
bris. With thicker debris – or, more generally, with debris
covers in which subsurface evaporation and/or advection are
more favorable – we expect the departures from measured
sub-debris melt values due to these unrealistic assumptions
in ISBA-DEB to be more pronounced. www.the-cryosphere.net/14/1555/2020/ 1572 6
Conclusions tribute to the model’s overestimation of sensible heat flux
(Sect. 5.3). In this study, we used precipitation data from
Pyramid Research Station and partitioned phase based upon
AWS air temperature. The SR50 depth sensor provides ad-
ditional information: it may indicate snowfall when there
was no recorded precipitation at Pyramid, and it may con-
versely indicate no solid precipitation when subfreezing tem-
peratures at the Changri Nup AWS coincided with recorded
precipitation at Pyramid. Any remaining mismatch after the
basic site-specific adjustments performed in this study would
propagate error to the calculated ablation. Verifying snow
cover duration in the forcing is, therefore, an important un-
dertaking in future research using ISBA-DEB on Changri
Nup glacier and should be a priority when collecting data
with which to force a debris-covered glacier surface energy
and mass balance model – particularly if it includes moisture. tribute to the model’s overestimation of sensible heat flux
(Sect. 5.3). In this study, we used precipitation data from
Pyramid Research Station and partitioned phase based upon
AWS air temperature. The SR50 depth sensor provides ad-
ditional information: it may indicate snowfall when there
was no recorded precipitation at Pyramid, and it may con-
versely indicate no solid precipitation when subfreezing tem-
peratures at the Changri Nup AWS coincided with recorded
precipitation at Pyramid. Any remaining mismatch after the
basic site-specific adjustments performed in this study would
propagate error to the calculated ablation. Verifying snow
cover duration in the forcing is, therefore, an important un-
dertaking in future research using ISBA-DEB on Changri
Nup glacier and should be a priority when collecting data
with which to force a debris-covered glacier surface energy
and mass balance model – particularly if it includes moisture. While the introduction of advective heat transfer and atmo-
spheric exchanges deeper than the surface of the debris could
make the model more physically realistic, ISBA-DEB never-
theless provides an advancement in modeling the processes
in a debris layer. It reasonably simulates the temperature evo-
lution of a snow–debris–glacier column according to meteo-
rological forcing and evolving thermal properties year-round,
even when the ice temperature is subfreezing and a snowpack
is present on the debris. It successfully produces variations
in non-saturated water content, phase, and location, demon-
strating both diffusion and water pooling at the glacier sur-
face. 6
Conclusions It also computes glacier melt based on the processes of
heat and water transfer, their determination of thermal and
hydraulic properties, and their interplay with one another. ISBA-DEB is the first debris surface energy balance model
to integrate heat conduction with moisture diffusion. In its
simulations of West Changri Nup glacier, enhanced melt oc-
curs below dry debris due to a combination of greater thermal
diffusivity and little loss of energy to evaporation or subli-
mation. ISBA-DEB explicitly accounts for the atmosphere–
debris latent heat exchanges in the top (surface) layer of the
debris only. The large difference in glacier melt below dry
and saturated debris shows that latent heat is enormously
important in removing energy from the system. Account-
ing for moisture in the conductive heat flux alone is insuf-
ficient when modeling melt under a debris-covered glacier. It
is, therefore, an essential next step to examine and incorpo-
rate the latent heat exchanges of moisture at all depths in the
debris. p
y
Finally, with spatially distributed, glacier-wide modeling
in mind, our sensitivity tests show that it is important to
investigate not only values of roughness lengths that gov-
ern the turbulent heat fluxes but also the spatial distribution
of those values. Albedo and emissivity are of lesser impor-
tance to constrain beyond their plausible values. The model’s
sensitivity to thermal conductivity and slope perturbations,
particularly on flatter terrain, reflects its sensitivity to wa-
ter content. It is, therefore, important to compute thermal
conductivity correctly, taking thermal conductivity measure-
ments for dry debris as well as investigating how varying
amounts of liquid water change the bulk value. Roughness
lengths, dry debris thermal conductivity, and slope are enor-
mously important variables to constrain when performing
distributed surface energy balance modeling over a debris-
covered glacier (the slope controls the water content enough
such that, with an accurate measure of dry thermal conductiv-
ity and a known relationship between moisture and bulk con-
ductivity, spatially distributed values of thermal conductivity
can be calculated from slope). Debris surfaces are typically
very rough, with variable slopes over short distances. As a
result of this topography, it is common to see saturated de-
bris and pooled water in topographic lows and dry debris on
topographic highs. A. Giese et al.: Modeling moisture content in debris cover energy balance 1573 5.5
Future directions A central part of the ISBA structure is the neglect of advec-
tion based on the observation that advective heating makes
relatively small changes to the soil temperature compared
to conduction. In addition to the thermal properties, hy-
draulic properties, and hydrological processes accounted for
in ISBA-DEB, soil and debris also differ in the size of their
interstitial void spaces. In highly permeable debris, there is
ample space for air flow through the debris layer. Advective
heat transfer is not accounted for in ISBA or ISBA-DEB. Reznichenko et al. (2010) showed in a laboratory that
rain advects heat from warm, highly permeable debris to the
glacier surface. Sakai et al. (2004) showed that heat flux from
percolated water assigned the temperature of debris was only
9 % of the ice melt flux despite the fact that 75 % of rain-
fall percolated, whereas the evaporative flux equaled nearly
half of the net radiative flux, the main driver of glacier melt. They concluded that not accounting for the evaporative heat
flux would lead to a 2-fold overestimation of sub-debris melt. They also pointed out, in comparing their two data collection Correct representation of snow is extremely important
to ISBA-DEB’s performance. Snow is a highly reflective,
strong insulator, and any error in simulated occurrence of
snowfall will cause error in not only the surface energy bal-
ance and underlying debris temperature profile simulated by
ISBA-DEB (e.g., Figs. 5, S3) but also the water mass bud-
get of the debris. An error in snow cover timing or duration
affects the net radiation budget and could potentially con- www.the-cryosphere.net/14/1555/2020/ The Cryosphere, 14, 1555–1577, 2020 Appendix A Table A1. Physical constants as well as parameter values used in the baseline ISBA-DEB; sensitivity tests were performed on parameters
with a superscript “a”. Superscript “b” appears with values predicted by pedotransfer functions (PTFs) of Noilhan and Lacarrere (1995)
(using Clapp and Hornberger, 1978) based upon an input of 98 % sand and 2 % clay. The calculated porosity given by the PTFs is 0.39, close
enough to the measured porosity of 0.37 that we did not overwrite the PTF calculation. The designation of zero hydraulic conductivity of
the bottom debris layer simulates an impenetrable glacier surface and ensures no nonphysical drainage out of the debris into the glacier. The
third column of the table indicates the file in which these parameters are set, for the future user. Air density is a function, as described in the
caption of Table S1. Values for which no references are listed are the standard values used by SURFEX (LeMoigne, 2018). Model parameter or physical constant
Value
Source/note
Where specified
Ice density (kg m−3)
917
Air density (kg m−3)
0.644–0.720
Air thermal cond. (W m−1 K−1)
0.024
introduced in ISBA-DEB,
Haynes (2017)
Air heat capacity (J kg−1 K−1)
1006
introduced in ISBA-DEB,
Haynes (2017)
Water density (kg m−3)
1000
Water thermal conductivity (W m−1 K−1)
0.57
Water heat capacity (J kg−1 K−1)
4218
Ice thermal conductivity (W m−1 K−1)
2.22
Ice heat capacity (J kg−1 K−1)
2106
Gravitational acceleration (m s−2)
9.80665
Latent heat of vaporization of water (Lv) (J kg−1)
2.5008 × 106
Latent heat of sublimation of water (Ls) (J kg−1)
2.8345 × 106
Latent heat of fusion of water (Lm) (J kg−1)
3.337 × 105
Debris emissivity (ϵ)
0.94a
Reid and Brock (2010)
OPTIONS &
modd_isba_par.F90
Debris albedo (α)
0.2a
calc. from SW measurements
OPTIONS
Dry debris density (kg m−3)
1690
measured
modd_isba_par.F90
Dry debris thermal cond. (W m−1 K−1)
0.94a
Reid and Brock (2010)
thrmcondz.F90
Dry debris heat capacity (J kg−1 K−1)
948
Reid and Brock (2010)
modd_isba_par.F90
Debris vol. heat cap. (J m−3 K−1)
1 602 120
–
–
Debris saturated hydraulic
Domenico and Schwartz (1998)
init_veg_pgdn.F90
conductivity (ksat, m s−1), Layers 1–12
0.03
Debris saturated hydraulic
init_veg_pgdn.F90
conductivity (ksat, m s−1), Layer 13
0
Matric potential at saturation (9sat, m)
0.097b
init_veg_pgdn.F90
Shape parameter of soil–water retention curve (b)
3.8b
init_veg_pgdn.F90
Debris porosity (8)
0.388465b, 0.999
∼0.37 (measured)
init_veg_pgdn.F90
Debris surface z0,m (m)
0.05a
Lejeune et al. 6
Conclusions Slope is also crucial for identifying low-
lying troughs, where pooled water or saturated debris could
dominate the surface albedo and emissivity (e.g., lakes over
supraglacial debris have lower albedo; Miles et al., 2016). Distributing ISBA-DEB would, then, produce large spatial
variability of melt and sublimation at the glacier scale be-
cause of the large variability of not only debris thicknesses
and properties (Nicholson et al., 2018; Rowan et al., 2017;
Rounce et al., 2018) but also of the water content in debris. Accounting for advection may change the model’s simula-
tion of glacier ice melting most under completely dry debris. ISBA-DEB provides a basis for developing a model that
can be applied at the glacier scale by identifying not only the
importance of atmospheric exchanges throughout the debris
column but also the most sensitive parameters controlling the
melt at point scale. In addition to using accurate roughness
lengths, it is crucial to represent moisture sources and sinks
correctly. An important part of the latter is constraining the
lateral runoff timescale (through, for example, laboratory or
field-based experiments). ISBA-DEB may be used to explore past or future changes
in sub-debris melt. Reanalysis data, such as those of ERA
Interim, provide all variables necessary to drive the model. Running ISBA-DEB under various Representative Concen-
tration Pathway (RCP) emissions scenarios (Van Vuuren
et al., 2011; IPCC, 2014) would provide insight into the fate
of ice under debris, an increasingly important topic as de-
bris cover is increasing in a warming climate (Thakuri et al.,
2014; Kirkbride and Deline, 2013). The Cryosphere, 14, 1555–1577, 2020 www.the-cryosphere.net/14/1555/2020/ 1574 www.the-cryosphere.net/14/1555/2020/ Competing interests. The authors declare that they have no conflict
of interest. Braud, I., Noilhan, J., Bessemoulin, P., Mascart, P., Haverkamp,
R., and Vauclin, M.: Bare-ground surface heat and water ex-
changes under dry conditions: Observations and parameteriza-
tion, Bound.-Lay. Meteorol., 66, 173–200, 1993. Acknowledgements. Samuel Morin provided an introduction and
orientation to using SURFEX, and Matthieu Lafaysse (Météo-
France – Centre d’Etudes de la Neige, CEN) and Susan Schwarz
(Dartmouth College) answered many technical questions. Yves
Lejeune corrected precipitation data and together with Marie Du-
mont facilitated Alexandra Giese’s work during a 10-month stay at
CEN. Gabriel Lewis provided valuable feedback and, along with
Jonathan Chipman, created the map. Christian Vincent offered ad-
vice throughout the project, and Yves Arnaud helped define a direc-
tion and also gave feedback. We thank Sonam Futi Sherpa, Dibas
Shrestha, Christian Vincent, Yves Arnaud, Fanny Brun, Joseph
Shea, and many other people who cannot all be listed for their as-
sistance in the field. We thank the Pyramid staff and Glacier Safari
Treks for their logistical support. Finally, the two anonymous re-
viewers greatly contributed to improving the quality and scope of
this paper, particularly with regard to the implications of our results
for debris-covered glacier physics. Brock, B. W., Mihalcea, C., Kirkbride, M. P., Diolaiuti, G., Cutler,
M. E., and Smiraglia, C.: Meteorology and surface energy fluxes
in the 2005–2007 ablation seasons at the Miage debris-covered
glacier, Mont Blanc Massif, Italian Alps, J. Geophys. Res.-
Atmos., 115, D09106, https://doi.org/10.1029/2009JD013224,
2010. Brooks, R. H. and Corey, A. T.: Properties of porous media affecting
fluid flow, J. Irr. Drain. Div.-ASCE, 92, 61–90, 1966. Brun, F., Wagnon, P., Berthier, E., Shea, J. M., Immerzeel, W. W.,
Kraaijenbrink, P. D. A., Vincent, C., Reverchon, C., Shrestha,
D., and Arnaud, Y.: Ice cliff contribution to the tongue-wide
ablation of Changri Nup Glacier, Nepal, central Himalaya, The
Cryosphere, 12, 3439–3457, https://doi.org/10.5194/tc-12-3439-
2018, 2018. Clapp, R. B. and Hornberger, G. M.: Empirical equations for some
soil hydraulic properties, Water Resour. Res., 14, 601–604, 1978. Collier, E., Nicholson, L. I., Brock, B. W., Maussion, F., Essery,
R., and Bush, A. B. G.: Representing moisture fluxes and phase
changes in glacier debris cover using a reservoir approach, The
Cryosphere, 8, 1429–1444, https://doi.org/10.5194/tc-8-1429-
2014, 2014. Financial support. This work has been supported by the French
Observatory CRYOBSCLIM now part of the IR OZCAR, the
French National Research Agency (ANR) through ANR-13-SENV-
0005-04-PRESHINE, and by a grant from Labex OSUG@2020
(Investissements d’avenir – ANR10 LABX56). A. Giese et al.: Modeling moisture content in debris cover energy balance Review statement. This paper was edited by Tobias Bolch and re-
viewed by two anonymous referees. Review statement. This paper was edited by Tobias Bolch and re-
viewed by two anonymous referees. Code and data availability. <surfex_git2@V8_1_giese>
code
is available after registration at http://opensource.umr-cnrm.fr/
projects/surfex_git2/repository?utf8=?&rev=V8_1_giese
(last
access: April 2020, Giese and Boone, 2020). FORCING.nc and OPTIONS.nam files are available at https:
//glacioclim.osug.fr/ (last access: April 2020, Giese et al., 2020;
Giese and Wagnon, 2020). Appendix A (2013)
OPTIONS
Debris surface z0,m/z0,h ratio
10a
ISBA default,
OPTIONS &
Mascart et al. (1995)
ini_data_param.F90
Shape factor
30
tuned
hydro_soildif.F90
τmax (s)
86 400
tuned
hydro_soildif.F90
τmin (s)
3600
Collier et al. (2014)
hydro_soildif.F90
Slope (◦)
5a
measured
hydro_soildif.F90
α℘
6
Johnsson and Lundin (1991)
hydro_soildif.F90
wmin (m3 m−3)
0.0001
0.001 in ISBA
modd_isba_par.F90
Model time step (s)
900
with splitting in hydro_soildif.F90
OPTIONS
Number of calculation layers
13 debris, seven ice
measured
OPTIONS
Debris layer thickness (cm)
12.5
measured
Altitude of measurement site (m)
5360
measured
FORCING Model parameter or physical constant The Cryosphere, 14, 1555–1577, 2020
www.the-cryosphere.net/14/1555/2020/ www.the-cryosphere.net/14/1555/2020/ The Cryosphere, 14, 1555–1577, 2020 The Cryosphere, 14, 1555–1577, 2020 1575 Competing interests. The authors declare that they have no conflict
of interest. This study was
carried out within the framework of the Ev-K2-CNR project in
collaboration with the Nepal Academy of Science and Technol-
ogy as foreseen by the Memorandum of Understanding between
Nepal and Italy. Alexandra Giese was supported by NASA Space
Grant NNX15AH79H, NSF GRF grant DGE-1313911, and the
Chateaubriand Fellowship of the Office for Science & Technology
of the Embassy of France in the United States. Collier, E., Nicholson, L. I., Brock, B. W., Maussion, F., Essery,
R., and Bush, A. B. G.: Corrigendum to “Representing moisture
fluxes and phase changes in glacier debris cover using a reservoir
approach” published in The Cryosphere, 8, 1429–1444, 2014,
The Cryosphere, 2016. Conway, H. and Rasmussen, L.: Summer temperature profiles
within supraglacial debris on Khumbu Glacier, Nepal, IAHS
Publication no. 264, 89–97, 2000. References Azam, M. F., Wagnon, P., Vincent, C., Ramanathan, AL., Favier, V.,
Mandal, A., and Pottakkal, J. G.: Processes governing the mass
balance of Chhota Shigri Glacier (western Himalaya, India) as-
sessed by point-scale surface energy balance measurements, The
Cryosphere, 8, 2195–2217, https://doi.org/10.5194/tc-8-2195-
2014, 2014. Supplement. The supplement related to this article is available on-
line at: https://doi.org/10.5194/tc-14-1555-2020-supplement. Supplement. The supplement related to this article is available on-
line at: https://doi.org/10.5194/tc-14-1555-2020-supplement. Blum, W. E., Nortcliff, S., and Schad, P.: Essentials of soil science:
soil formation, functions, use and classification (world reference
base), Borntraeger, 2018. Author contributions. AG conceived the study, performed the mod-
eling work, and wrote the manuscript. AB assisted in modeling
work; he was instrumental in the coding of ISBA-DEB adaptations
and in integrating them into ISBA. PW made the meteorological
forcing data and debris temperature data available for this study,
and he provided invaluable feedback on the direction of the model-
ing work and on the specific model output. RH helped troubleshoot
many modeling hurdles and advised on the methodology for the
tuning and sensitivity tests. Boone, A. and Etchevers, P.: An intercomparison of three snow
schemes of varying complexity coupled to the same land surface
model: Local-scale evaluation at an Alpine site, J. Hydrometeo-
rol., 2, 374–394, 2001. Boone, A., Calvet, J.-C., and Noilhan, J.: Inclusion of a third soil
layer in a land surface scheme using the force–restore method, J. Appl. Meteorol., 38, 1611–1630, 1999. Boone, A., Masson, V., Meyers, T., and Noilhan, J.: The influ-
ence of the inclusion of soil freezing on simulations by a soil–
vegetation–atmosphere transfer scheme, J. Appl. Meteorol., 39,
1544–1569, 2000. Competing interests. The authors declare that they have no conflict
of interest. A. Giese et al.: Modeling moisture content in debris cover energy balance 1576 Decharme, B., Boone, A., Delire, C., and Noilhan, J.: Lo-
cal evaluation of the Interaction between Soil Biosphere At-
mosphere soil multilayer diffusion scheme using four pedo-
transfer functions, J. Geophys. Res.-Atmos., 116, D20126,
https://doi.org/10.1029/2011JD016002, 2011. Juen, M., Mayer, C., Lambrecht, A., Wirbel, A., and Kueppers, U.:
Thermal properties of a supraglacial debris layer with respect to
lithology and grain size, Geogr. Ann. A, 95, 197–209, 2013. Kayastha, R. B., Takeuchi, Y., Nakawo, M., and Ageta, Y.: Practi-
cal prediction of ice melting beneath various thickness of debris
cover on Khumbu Glacier, Nepal, using a positive degree-day
factor, IAHS Publication no. 264, 71-82, 2000. Decharme, B., Martin, E., and Faroux, S.: Reconciling soil ther-
mal and hydrological lower boundary conditions in land surface
models, J. Geophys. Res.-Atmos., 118, 7819–7834, 2013. Kirkbride, M. P. and Deline, P.: The formation of supraglacial de-
bris covers by primary dispersal from transverse englacial debris
bands, Earth Surf. Proc. Land., 38, 1779–1792, 2013. Decharme, B., Brun, E., Boone, A., Delire, C., Le Moigne, P.,
and Morin, S.: Impacts of snow and organic soils parameteri-
zation on northern Eurasian soil temperature profiles simulated
by the ISBA land surface model, The Cryosphere, 10, 853–877,
https://doi.org/10.5194/tc-10-853-2016, 2016. Kowalewski, D. E., Marchant, D. R., Swanger, K. M., and Head III,
J. W.: Modeling vapor diffusion within cold and dry supraglacial
tills of Antarctica: Implications for the preservation of ancient
ice, Geomorphology, 126, 159–173, 2011. Domenico, P. A. and Schwartz, F. W.: Physical and chemical hydro-
geology, vol. 506, Wiley New York, 1998. Kraaijenbrink, P., Bierkens, M., Lutz, A., and Immerzeel, W.: Im-
pact of a global temperature rise of 1.5 degrees Celsius on Asia’s
glaciers, Nature, 549, 257–260, 2017. Evatt, G. W., Abrahams, I. D., Heil, M., Mayer, C., Kingslake, J.,
Mitchell, S. L., Fowler, A. C., and Clark, C. D.: Glacial melt
under a porous debris layer, J. Glaciol., 61, 825–836, 2015. Lejeune, Y., Bertrand, J.-M., Wagnon, P., and Morin, S.: A physi-
cally based model of the year-round surface energy and mass bal-
ance of debris-covered glaciers, J. Glaciol., 59, 327–344, 2013. Fujita, K. and Sakai, A.: Modelling runoff from a Himalayan
debris-covered glacier, Hydrol. Earth Syst. Sci., 18, 2679–2694,
https://doi.org/10.5194/hess-18-2679-2014, 2014. LeMoigne, P. (Ed.): SURFEX Scientific Documentation, CNRM,
Toulouse, France, 304 pp., available at: https://www.umr-cnrm. fr/surfex/spip.php?rubrique11 (last access: April 2020), 2018. A. Giese et al.: Modeling moisture content in debris cover energy balance Fyffe, C., Reid, T., Brock, B., Kirkbride, M., Diolaiuti, G., Smi-
raglia, C., and Diotri, F.: A distributed energy-balance melt
model of an alpine debris-covered glacier, J. Glaciol., 60, 587–
602, 2014. Mascart, P., Noilhan, J., and Giordani, H.: A modified parameteri-
zation of flux-profile relationships in the surface layer using dif-
ferent roughness length values for heat and momentum, Bound.-
Lay. Meteorol., 72, 331–344, 1995. Giard, D. and Bazile, E.: Implementation of a new assimilation
scheme for soil and surface variables in a global NWP model,
Mon. Weather Rev., 128, 997–1015, 2000. Masson, V., Le Moigne, P., Martin, E., Faroux, S., Alias, A.,
Alkama, R., Belamari, S., Barbu, A., Boone, A., Bouyssel, F.,
Brousseau, P., Brun, E., Calvet, J.-C., Carrer, D., Decharme, B.,
Delire, C., Donier, S., Essaouini, K., Gibelin, A.-L., Giordani, H.,
Habets, F., Jidane, M., Kerdraon, G., Kourzeneva, E., Lafaysse,
M., Lafont, S., Lebeaupin Brossier, C., Lemonsu, A., Mahfouf,
J.-F., Marguinaud, P., Mokhtari, M., Morin, S., Pigeon, G., Sal-
gado, R., Seity, Y., Taillefer, F., Tanguy, G., Tulet, P., Vincendon,
B., Vionnet, V., and Voldoire, A.: The SURFEXv7.2 land and
ocean surface platform for coupled or offline simulation of earth
surface variables and fluxes, Geosci. Model Dev., 6, 929–960,
https://doi.org/10.5194/gmd-6-929-2013, 2013. Giese, A. and Boone, A.: V8_1_giese SURFEX Branch, CNRM,
Toulouse, France, Météo-France Surfex_Git2 repository, avail-
able
at:
https://opensource.umr-cnrm.fr/projects/surfex_git2/
repository?utf8=?&rev=V8_1_giese, last access: April 2020. Jidane, M., Kerdraon, G., Kourzeneva, E., Lafaysse, Giese, A. and Wagnon, P.: FORCING.nc, input file, IGE, Greno-
ble, France, GLACIOCLIM, available at: https://nextcloud.osug. fr/index.php/s/QzzLLTCjY4fZS3Q, last access: April 2020. Giese, A., Boone, A. and Wagnon, P.: OPTIONS.nam, input
file, IGE, Grenoble, France, GLACIOCLIM, available at: https:
//nextcloud.osug.fr/index.php/s/AYEkqEyyAg8XKSq, last ac-
cess: April 2020. Haynes, W. M. (Ed.): CRC Handbook of Chemistry and Physics,
97th edn., CRC Press, Taylor & Francis Group, 2017. Miles, E. S., Pellicciotti, F., Willis, I. C., Steiner, J. F., Buri,
P., and Arnold, N. S.: Refined energy-balance modelling of a
supraglacial pond, Langtang Khola, Nepal, Ann. Glaciol., 57,
29–40, 2016. Huintjes, E., Sauter, T., Schröter, B., Maussion, F., Yang, W.,
Kropáˇcek, J., Buchroithner, M., Scherer, D., Kang, S., and
Schneider, C.: Evaluation of a coupled snow and energy balance
model for Zhadang glacier, Tibetan Plateau, using glaciological
measurements and time-lapse photography, Arct. Antarct. Alp. Res., 47, 573–590, 2015. Miles, E. S., Steiner, J. F., and Brun, F.: Highly variable aerody-
namic roughness length (z0) for a hummocky debris-covered
glacier, J. Geophys. www.the-cryosphere.net/14/1555/2020/ The Cryosphere, 14, 1555–1577, 2020 A. Giese et al.: Modeling moisture content in debris cover energy balance 1577 Sakai, A., Fujita, K., and Kubota, J.: Evaporation and percolation
effect on melting at debris-covered Lirung Glacier, Nepal Hi-
malayas, 1996, Bull. Glaciol. Res., 21, 9–16, 2004. Nicholson, L. I., McCarthy, M., Pritchard, H. D., and Willis, I.:
Supraglacial debris thickness variability: impact on ablation and
relation to terrain properties, The Cryosphere, 12, 3719–3734,
https://doi.org/10.5194/tc-12-3719-2018, 2018. Searle, M., Simpson, R., Law, R., Parrish, R., and Waters, D.: The
structural geometry, metamorphic and magmatic evolution of the
Everest massif, High Himalaya of Nepal–South Tibet, J. Geol. Soc., 160, 345–366, 2003. Noilhan, J. and Mahfouf, J.-F.: The ISBA land surface parameteri-
sation scheme, Global Planet. Change, 13, 145–159, 1996. Noilhan, J. and Planton, S.: A simple parameterization of land sur-
face processes for meteorological models, Mon. Weather Rev.,
117, 536–549, 1989. Sherpa, S. F., Wagnon, P., Brun, F., Berthier, E., Vincent, C., Leje-
une, Y., Arnaud, Y., Kayastha, R. B., and Sinisalo, A.: Contrasted
surface mass balances of debris-free glaciers observed between
the southern and the inner parts of the Everest region (2007–15),
J. Glaciol., 63, 637–651, 2017. Noilhan, J. and Lacarrere, P.: GCM grid-scale evaporation from
mesoscale modeling, J. Climate, 8, 206–223, 1995. Östrem, G.: Ice melting under a thin layer of moraine, and the exis-
tence of ice cores in moraine ridges, Geogr. Ann., 41, 228–230,
1959. Smeets, C. and Van den Broeke, M.: The parameterisation of
scalar transfer over rough ice, Bound.-Lay. Meteorol., 128, 339,
https://doi.org/10.1007/s10546-008-9292-z, 2008. Pritchard, H. D.: Asia’s shrinking glaciers protect large populations
from drought stress, Nature, 569, 649–654, 2019. Steiner, J. F., Litt, M. H., Stigter, E. E., Shea, J. M., Bierkens, M. F.,
and Immerzeel, W. W.: The Importance of Turbulent Fluxes in
the Surface Energy Balance of a Debris-Covered Glacier in the
Himalayas, Frontiers in Earth Science, Front. Earth Sci., 6, 144,
https://doi.org/10.3389/feart.2018.00144, 2018. Quincey, D., Smith, M., Rounce, D., Ross, A., King, O., and Wat-
son, C.: Evaluating morphological estimates of the aerodynamic
roughness of debris covered glacier ice, Earth Surf. Proc. Land.,
42, 2541–2553, 2017. Takeuchi, Y., Kayastha, R. B., and Nakawo, M.: Characteristics of
ablation and heat balance in debris-free and debris-covered ar-
eas on Khumbu Glacier, Nepal Himalayas, in the pre-monsoon
season, IAHS Publication no. 264, 53–62, 2000. Reid, T. and Brock, B.: An energy-balance model for debris-covered
glaciers including heat conduction through the debris layer, J. Glaciol., 56, 903–916, 2010. A. Giese et al.: Modeling moisture content in debris cover energy balance Reid, T., Carenzo, M., Pellicciotti, F., and Brock, B.: In-
cluding
debris
cover
effects
in
a
distributed
model
of
glacier ablation, J. Geophys. Res.-Atmos., 117, D18105,
https://doi.org/10.1029/2012JD017795, 2012. Thakuri, S., Salerno, F., Smiraglia, C., Bolch, T., D’Agata, C., Vi-
viano, G., and Tartari, G.: Tracing glacier changes since the
1960s on the south slope of Mt. Everest (central Southern Hi-
malaya) using optical satellite imagery, Global Planet. Change,
8, 1297–1315, 2014. Reijmer, C. H. and Hock, R.: Internal accumulation on Stor-
glaciären, Sweden, in a multi-layer snow model coupled to a dis-
tributed energy-and mass-balance model, J. Glaciol., 54, 61–72,
2008. Van Vuuren, D. P., Edmonds, J., Kainuma, M., Riahi, K., Thomson,
A., Hibbard, K., Hurtt, G. C., Kram, T., Krey, V., Lamarque, J. F., Masui, T., Meinshausen, M., Nakicenovic, N., Smith, S. J.,
and Rose, S. K. : The representative concentration pathways: an
overview, Clim. Change, 109, 5, https://doi.org/10.1007/s10584-
011-0148-z, 2011. Reznichenko, N., Davies, T., Shulmeister, J., and McSaveney, M.:
Effects of debris on ice-surface melting rates: an experimental
study, J. Glaciol., 56, 384–394, 2010. Robertson, E. C.: Thermal properties of rocks, Tech. rep., US Geo-
logical Survey, 1988. Vincent, C., Wagnon, P., Shea, J. M., Immerzeel, W. W., Kraai-
jenbrink, P., Shrestha, D., Soruco, A., Arnaud, Y., Brun, F.,
Berthier, E., and Sherpa, S. F.: Reduced melt on debris-covered
glaciers: investigations from Changri Nup Glacier, Nepal, The
Cryosphere, 10, 1845–1858, https://doi.org/10.5194/tc-10-1845-
2016, 2016. Rounce, D. R. and McKinney, D. C.: Debris thickness of glaciers
in the Everest area (Nepal Himalaya) derived from satellite im-
agery using a nonlinear energy balance model, The Cryosphere,
8, 1317–1329, https://doi.org/10.5194/tc-8-1317-2014, 2014. Rounce, D. R., Quincey, D. J., and McKinney, D. C.: Debris-
covered glacier energy balance model for Imja–Lhotse Shar
Glacier in the Everest region of Nepal, The Cryosphere, 9, 2295–
2310, https://doi.org/10.5194/tc-9-2295-2015, 2015. Wagnon, P., Lafaysse, M., Lejeune, Y., Maisincho, L., Rojas, M.,
and Chazarin, J.-P.: Understanding and modeling the physical
processes that govern the melting of snow cover in a tropi-
cal mountain environment in Ecuador, J. Geophys. Res.-Atmos.,
114, D19113, https://doi.org/10.1029/2009JD012292, 2009. Rounce, D. R., King, O., McCarthy, M., Shean, D. E., and Salerno,
F.: Quantifying Debris Thickness of Debris-Covered Glaciers in
the Everest Region of Nepal Through Inversion of a Subdebris
Melt Model, J. Geophys. Res.-Earth, 123, 1094–1115, 2018. Zuo, Z. A. Giese et al.: Modeling moisture content in debris cover energy balance Res.-Atmos., 122, 8447–8466, 2017. Monin, A. and Obukhov, A.: Osnovnie haraktristiki turbulentnogo
peremeshivaniya v prizemnom sloe atmosferi [Main characteris-
tics of turbulent mixing in atmospheric boundary layer], Trudy
Inst. Geofiz. (Akad. Nauk SSSR), 24, 163–187, 1954. Inoue, J. and Yoshida, M.: Ablation and heat exchange over the
Khumbu Glacier, J. Jpn. Soc. Snow Ice, 41, 26–33, 1980. IPCC, 2014: Climate Change 2014: Synthesis Report, Contribution
of Working Groups I, II and III to the Fifth Assessment Report of
the Intergovernmental Panel on Climate Change, edited by: Core
Writing Team, Pachauri, R. K., and Meyer, L. A., IPCC, Geneva,
Switzerland, 151 pp., 2014. Nakawo, M. and Young, G. J.: Field experiments to determine the
effect of a debris layer on ablation of glacier ice, Ann. Glaciol.,
2, 85–91, 1981. Nicholson, L. and Benn, D. I.: Calculating ice melt beneath a debris
layer using meteorological data, J. Glaciol., 52, 463–470, 2006. Johnsson, H. and Lundin, L.-C.: Surface runoff and soil water per-
colation as affected by snow and soil frost, J. Hydrol., 122, 141–
159, 1991. Nicholson, L. and Benn, D. I.: Properties of natural supraglacial de-
bris in relation to modelling sub-debris ice ablation, Earth Surf. Proc. Land., 38, 490–501, 2012. The Cryosphere, 14, 1555–1577, 2020 www.the-cryosphere.net/14/1555/2020/ A. Giese et al.: Modeling moisture content in debris cover energy balance and Oerlemans, J.: Modelling albedo and specific balance of
the Greenland ice sheet: calculations for the Søndre Strømfjord
transect, J. Glaciol., 42, 305–317, 1996. Rowan, A. V., Quincey, D. J., Gibson, M. J., Glasser, N. F., Westoby,
M. J., Irvine-Fynn, T. D., Porter, P. R., and Hambrey, M. J.: The
sustainability of water resources in High Mountain Asia in the
context of recent and future glacier change, Geological Society,
London, Special Publications, 462, 189–204, 2017. The Cryosphere, 14, 1555–1577, 2020 www.the-cryosphere.net/14/1555/2020/
|
https://openalex.org/W4243303390
|
https://peerj.com/articles/1397v0.2/submission
|
English
| null |
Peer Review #1 of "Homeotic transformations and number changes in the vertebral column of Triturus newts (v0.1)"
| null | 2,015
|
cc-by
| 16,164
|
Manuscript to be reviewed Manuscript to be reviewed Manuscript to be reviewed 1
Homeotic transformations and number changes in the vertebral column of
2
Triturus newts
3
4
Maja D. Slijepčević1, Frietson Galis2, Jan W. Arntzen2 and Ana Ivanović2,3
5
6
1Institute for Biological Research “Siniša Stanković”, University of Belgrade, Serbia
7
2Naturalis Biodiversity Center, Leiden, The Netherlands
8
3Institute of Zoology, Faculty of Biology, University of Belgrade, Serbia
9
10
Corresponding author
11
Ana Ivanović,
12
ana@bio.bg.ac.rs
13
14
Abstract
15
We explored intraspecific variation in vertebral formulae, more specifically the variation in the
16
number of thoracic vertebrae and frequencies of transitional sacral vertebrae in Triturus newts
17
(Caudata: Salamandridae). Within salamandrid salamanders this monophyletic group shows the
18
highest disparity in the number of thoracic vertebrae and considerable intraspecific variation in
19
the number of thoracic vertebrae. Triturus species also differ in their ecological preferences,
20
from predominantly terrestrial to largely aquatic. Following Geoffroy St. Hilaire’s and Darwin’s
21
rule which states that structures with a large number of serially homologous repetitive elements
22
are more variable than structures with smaller numbers, we hypothesized that the variation in
23
vertebral formulae increases in more elongated species with a larger number of thoracic
24
vertebrae. We furthermore hypothesized that the frequency of transitional vertebrae will be
25
correlated with the variation in the number of thoracic vertebrae within the species. We also
26
investigated potential effects of species hybridization on the vertebral formula. The proportion of
27
individuals with a number of thoracic vertebrae different from the modal number and the range
28
of variation in number of vertebrae significantly increased in species with a larger number of
29
thoracic vertebrae. Contrary to our expectation, the frequencies of transitional vertebrae were not
30
correlated with frequencies of change in the complete vertebrae number. The frequency of
31
transitional sacral vertebra in hybrids did not significantly differ from that of the parental
32
species. Such a pattern could be a result of selection pressure against transitional vertebrae
33
and/or a bias towards the development of full vertebrae numbers. Although our data indicate Homeotic transformations and number changes in the
vertebral column of Triturus newts
Maja Slijepčević, Frietson F Galis, Jan W Arntzen, Ana Ivanović Maja Slijepčević, Frietson F Galis, Jan W Arntzen, Ana Ivanović We explored intraspecific variation in vertebral formulae, more specifically the variation in
the number of thoracic vertebrae and frequencies of transitional sacral vertebrae in
Triturus newts (Caudata: Salamandridae). Within salamandrid salamanders this
monophyletic group shows the highest disparity in the number of thoracic vertebrae and
considerable intraspecific variation in the number of thoracic vertebrae. Triturus species
also differ in their ecological preferences, from predominantly terrestrial to largely aquatic. Following Geoffroy St. Hilaire’s and Darwin’s rule which states that structures with a large
number of serially homologous repetitive elements are more variable than structures with
smaller numbers, we hypothesized that the variation in vertebral formulae increases in
more elongated species with a larger number of thoracic vertebrae. We furthermore
hypothesized that the frequency of transitional vertebrae will be correlated with the
variation in the number of thoracic vertebrae within the species. We also investigated
potential effects of species hybridization on the vertebral formula. The proportion of
individuals with a number of thoracic vertebrae different from the modal number and the
range of variation in number of vertebrae significantly increased in species with a larger
number of thoracic vertebrae. Contrary to our expectation, the frequencies of transitional
vertebrae were not correlated with frequencies of change in the complete vertebrae
number. The frequency of transitional sacral vertebra in hybrids did not significantly differ
from that of the parental species. Such a pattern could be a result of selection pressure
against transitional vertebrae and/or a bias towards the development of full vertebrae
numbers. Although our data indicate relaxed selection for vertebral count changes in more
elongated, aquatic species, more data on different selective pressures in species with
different numbers of vertebrae in the two contrasting, terrestrial and aquatic environments
are needed to test for causality. PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) Manuscript to be reviewed 34
relaxed selection for vertebral count changes in more elongated, aquatic species, more data on
35
different selective pressures in species with different numbers of vertebrae in the two contrasting,
36
terrestrial and aquatic environments are needed to test for causality. 37
38
Introduction
39
The vertebral column consists of repetitive, serially homologous skeletal elements – vertebrae. 40
Along the anterior-posterior axis, vertebrae are classified into regions, with a conserved order
41
and specific sizes and shapes (e.g. Starck, 1979). The strong regionalization of the vertebral
42
column is already present early in the evolution of tetrapods (Ahlberg, Clack & Blom, 2005). In
43
the early tetrapod Ichthyostega, five regions (cervical, thoracic, lumbar, sacral and caudal) can be
44
recognized as in many extant amniotes. 45
Vertebrae develop from embryonic segments (somites) that are generated from the presomitic
46
mesoderm in a sequential head-to tail order. This process involves a molecular oscillator, the
47
segmentation clock, that regulates the periodicity of segment formation (Cooke & Zeeman, 1976;
48
Palmeirim et al., 1997; Dequéant & Pourquié, 2008; Gomez & Pourquié, 2009). The duration of
49
segment formation (somitogenesis) and the speed of the segmentation clock determine the total
50
number of segments formed, and hence, the total number of vertebrae (Gomez et al., 2008,
51
2009). The determination of the identity of the vertebrae (e.g. cervical or thoracic) occurs as part
52
of the early head-to-tail patterning of the presomitic mesoderm and early somites. This head-to-
53
tail patterning involves complex genetic mechanisms that include various signaling molecules,
54
with an essential mediating role for the well-known Hox genes (e.g. Dubrulle, McGrew &
55
Pourquié, 2001; Diezdel-Corral et al., 2003; Aulehla & Pourquié, 2010; Mallo, Wellik &
56
Deschamps, 2010; Woltering, 2012; Wong et al., 2015). It is thought that the segmentation
57
process and the head-to-tail patterning of the segments by the Hox genes can be dissociated and
58
that this dissociation has allowed for the spectacular evolutionary diversification of vertebral
59
formulae (Carapuço et al., 2005; Gomez & Pourquié, 2009; Schroeter et al., 2010; Harima et
60
al.,2013; Wong et al., 2015). When there are shifts of vertebral boundaries, e.g. the cervico-
61
thoracic boundary, these shifts involve changes in the Hox patterning of the somites along the
62
head-to-tail axis. Manuscript to be reviewed If indeed the segmentation process and the head-to-tail patterning of the
63
segments are dissociated, the shifts of vertebral boundaries necessarily involve homeotic 34
relaxed selection for vertebral count changes in more elongated, aquatic species, more data on
35
different selective pressures in species with different numbers of vertebrae in the two contrasting,
36
terrestrial and aquatic environments are needed to test for causality. 34
relaxed selection for vertebral count changes in more elongated, aquatic species, more data on
35
different selective pressures in species with different numbers of vertebrae in the two contrasting,
36
terrestrial and aquatic environments are needed to test for causality. 37 1
Homeotic transformations and number changes in the vertebral column of
2
Triturus newts Maja D. Slijepčević1, Frietson Galis2, Jan W. Arntzen2 and Ana Ivanović2,3 PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) 38
Introduction 39
The vertebral column consists of repetitive, serially homologous skeletal elements – vertebrae. 40
Along the anterior-posterior axis, vertebrae are classified into regions, with a conserved order
41
and specific sizes and shapes (e.g. Starck, 1979). The strong regionalization of the vertebral
42
column is already present early in the evolution of tetrapods (Ahlberg, Clack & Blom, 2005). In
43
the early tetrapod Ichthyostega, five regions (cervical, thoracic, lumbar, sacral and caudal) can be
44
recognized as in many extant amniotes. 39
The vertebral column consists of repetitive, serially homologous skeletal elements – vertebrae. 40
Along the anterior-posterior axis, vertebrae are classified into regions, with a conserved order
41
and specific sizes and shapes (e.g. Starck, 1979). The strong regionalization of the vertebral
42
column is already present early in the evolution of tetrapods (Ahlberg, Clack & Blom, 2005). In
43
the early tetrapod Ichthyostega, five regions (cervical, thoracic, lumbar, sacral and caudal) can be
44
recognized as in many extant amniotes. 45
Vertebrae develop from embryonic segments (somites) that are generated from the presomitic
46
mesoderm in a sequential head-to tail order. This process involves a molecular oscillator, the
47
segmentation clock, that regulates the periodicity of segment formation (Cooke & Zeeman, 1976;
48
Palmeirim et al., 1997; Dequéant & Pourquié, 2008; Gomez & Pourquié, 2009). The duration of
49
segment formation (somitogenesis) and the speed of the segmentation clock determine the total
50
number of segments formed, and hence, the total number of vertebrae (Gomez et al., 2008, 45
Vertebrae develop from embryonic segments (somites) that are generated from the presomitic
46
mesoderm in a sequential head-to tail order. This process involves a molecular oscillator, the
47
segmentation clock, that regulates the periodicity of segment formation (Cooke & Zeeman, 1976;
48
Palmeirim et al., 1997; Dequéant & Pourquié, 2008; Gomez & Pourquié, 2009). The duration of
49
segment formation (somitogenesis) and the speed of the segmentation clock determine the total
50
number of segments formed, and hence, the total number of vertebrae (Gomez et al., 2008, PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) Manuscript to be reviewed 71
In mammals, changes in the number of cervical vertebrae are associated with deleterious 72
pleiotropic effects that lead to selection against such homeotic transformations (Galis & Metz,
73
2003; Varela-Lasheras et al., 2011). In two mammalian groups (sloths and manatees) with low
74
activity and metabolic rates, the exceptional numbers of cervical vertebrae most likely resulted
75
from the effective breaking of pleiotropic constraints due to a relaxation of stabilizing selection
76
against the pleiotropic effects (known as congenital abnormalities, Varela-Lasheras et al., 2011). 77
Furthermore, Galis et al. (2014) concluded that biomechanical problems associated with initial
78
homeotic transformations (transitional vertebrae) in fast running mammals result in strong
79
selection against changes of the presacral vertebral counts in these species. 80
In other tetrapods the number of vertebrae in different regions can be more variable. Well-known
81
examples are the variable number of cervical vertebrae in birds (Woolfenden, 1961) and the
82
variable number of presacral vertebrae in squamates (Carroll, 1997; Müller et al., 2010). In both
83
cases, these regions have a large number of vertebrae. Geoffroy St. Hilaire has postulated that as
84
a rule - structures with a large number of serially homologous repetitive elements are more
85
variable than structures with smaller numbers (Geoffroy St. Hilaire, 1932). This notion was
86
supported by Darwin (Darwin, 1860). Along the same line, Bateson (1894) concluded that series
87
containing large numbers of undifferentiated parts are more variable than series made up of a
88
few, more differentiated parts. 89
In tailed amphibians, the presacral vertebrae vary in their number but only little in shape. A
90
single, sacral vertebra is morphologically very similar to the vertebrae from the thoracic or trunk
91
region, with more robust processes for attachment of the sacral ribs which are also thicker than
92
regular thoracic ribs. There is considerable intraspecific variation in the number of thoracic
93
vertebrae in many species of salamanders (Adolphi, 1898; Gerecht, 1929; Peabody & Brodie,
94
1975; Jockusch, 1997; Litvinchuk & Borkin, 2003). Intraspecific variation originally results from
95
homeotic transformations that are subsequently maintained in the population. In salamanders, 89
In tailed amphibians, the presacral vertebrae vary in their number but only little in shape. Manuscript to be reviewed 96
transitional vertebrae at the thoraco-sacral boundary have been frequently reported (Adolphi,
97
1898; Gerecht, 1929; Highton, 1960; Worthington, 1971; Jockush, 1997; Arntzen et al., 2015). 98
Such transitional vertebrae with partial thoracic and partial sacral identity result from incomplete
99
homeotic transformations. Therefore, the frequencies of transitional vertebrae could be related to
100
the amount of variation in the number of thoracic vertebrae within species. Data on changes in
101
axial pattering and homeotic transformations in amphibians are relatively scarce and more data
102
are necessary for understanding the evolution of axial pattering in amphibians and the tetrapods. 103
In this study we explore the relationship between variation in the number of thoracic vertebrae
104
and transitional sacral vertebrae using eight species of the monophyletic genus Triturus newts as
105
a model system. Within the family Salamandridae, which is the second most diverse group of
106
tailed amphibians, Triturus newts are the most disparate in the number of thoracic vertebrae
107
(Arntzen et al., 2015). Triturus species form a morphocline from the predominantly terrestrial T. 108
marmoratus and T. pygmaeus with a short and stout body and 12 thoracic vertebrae to the
109
slender and elongated, largely aquatic T. cristatus and T. dobrogicus with 15-17 thoracic
110
vertebrae (Arntzen, 2003). Triturus species also display considerable intraspecific variation in
111
vertebral numbers (Gerecht, 1929; Crnobrnja et al., 1997; Arntzen et al., 2015). Moreover, there
112
is a well-documented, extensive hybridization in the area of sympatry of T. marmoratus (12
113
thoracic vertebrae) and T. cristatus (15 thoracic vertebrae). The hybridization of these two
114
species leads to sterile F1 hybrids with intermediate morphologies and number of thoracic
115
vertebrae (Vallée, 1959; Arntzen et al., 2009). Interspecific hybridization at contact zones also
116
occurs between other Triturus species with parapatric distributions (Mikulíček et al., 2012;
117
Arntzen, Wielstra & Wallis, 2014), providing the opportunity to investigate the relationship
118
between vertebral number and frequencies of transitional sacral vertebrae. Here, we compared
119
variation in the number of thoracic vertebrae and transitional sacral vertebrae among Triturus
120
species, T. marmoratus × T. cristatus F1 hybrids with parental species, and populations from
121
contact zones with populations away from contact zones. Manuscript to be reviewed A
90
single, sacral vertebra is morphologically very similar to the vertebrae from the thoracic or trunk
91
region, with more robust processes for attachment of the sacral ribs which are also thicker than
92
regular thoracic ribs. There is considerable intraspecific variation in the number of thoracic
93
vertebrae in many species of salamanders (Adolphi, 1898; Gerecht, 1929; Peabody & Brodie,
94
1975; Jockusch, 1997; Litvinchuk & Borkin, 2003). Intraspecific variation originally results from
95
homeotic transformations that are subsequently maintained in the population. In salamanders, PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) Manuscript to be reviewed More specifically, we explored the
122
intra- and interspecific variation in number of the thoracic vertebrae and frequencies of
123
transitional vertebrae at the thoraco-sacral boundary to test the following hypotheses:
124
1. Species with more vertebrae in the thoracic region are more variable in the number of thoracic
125
vertebrae than those with fewer vertebrae in the thoracic region. 96
transitional vertebrae at the thoraco-sacral boundary have been frequently reported (Adolphi,
97
1898; Gerecht, 1929; Highton, 1960; Worthington, 1971; Jockush, 1997; Arntzen et al., 2015). 98
Such transitional vertebrae with partial thoracic and partial sacral identity result from incomplete
99
homeotic transformations. Therefore, the frequencies of transitional vertebrae could be related to
100
the amount of variation in the number of thoracic vertebrae within species. Data on changes in
101
axial pattering and homeotic transformations in amphibians are relatively scarce and more data
102
are necessary for understanding the evolution of axial pattering in amphibians and the tetrapods. 103
In this study we explore the relationship between variation in the number of thoracic vertebrae
104
and transitional sacral vertebrae using eight species of the monophyletic genus Triturus newts as
105
a model system. Within the family Salamandridae, which is the second most diverse group of
106
tailed amphibians, Triturus newts are the most disparate in the number of thoracic vertebrae
107
(Arntzen et al., 2015). Triturus species form a morphocline from the predominantly terrestrial T. 108
marmoratus and T. pygmaeus with a short and stout body and 12 thoracic vertebrae to the
109
slender and elongated, largely aquatic T. cristatus and T. dobrogicus with 15-17 thoracic
110
vertebrae (Arntzen, 2003). Triturus species also display considerable intraspecific variation in
111
vertebral numbers (Gerecht, 1929; Crnobrnja et al., 1997; Arntzen et al., 2015). Moreover, there
112
is a well-documented, extensive hybridization in the area of sympatry of T. marmoratus (12
113
thoracic vertebrae) and T. cristatus (15 thoracic vertebrae). The hybridization of these two
114
species leads to sterile F1 hybrids with intermediate morphologies and number of thoracic
115
vertebrae (Vallée, 1959; Arntzen et al., 2009). Interspecific hybridization at contact zones also
116
occurs between other Triturus species with parapatric distributions (Mikulíček et al., 2012;
117
Arntzen, Wielstra & Wallis, 2014), providing the opportunity to investigate the relationship
118
between vertebral number and frequencies of transitional sacral vertebrae. 135
Triturus newts and their characteristics 136
The vertebral column in Triturus newts is differentiated in: the cervical region– consisting of a
137
single anterior vertebra, the atlas; the thoracic (trunk) region,–consisting of a rib-bearing
138
vertebrae; the sacral region – usually a single vertebra with well-developed stout transverse
139
processes for the attachment of sacral ribs and pelvic girdle; the caudosacral region – up to three
140
vertebrae that continue from the caudal to the sacral vertebra and are associated with the pelvic
141
musculature and cloaca and the caudal region – the remaining vertebrae in the tail (Fig. 1). The
142
body elongation in Triturus species appears to be correlated with the length of the aquatic phase
143
– more terrestrial species have a short and stout trunk with relatively longer legs compared to
144
species with a more aquatic life style which have a more elongated trunk and relatively shorter
145
legs. Body elongation involves a larger number of thoracic vertebrae. More specifically, the
146
number of thoracic vertebrae in the vertebral formulae varies from 12 in T. marmoratus and T. 147
pygmaeus, which have a short aquatic phase (T. marmoratus only two months), 13 in T. karelinii
148
and T. ivanbureschi, 14 in T. macedonicus and T. carnifex, 15 in T. cristatus to 16 or 17 in T. 149
dobrogicus, the most aquatic species which has six months long aquatic phase (Arntzen, 2003)
150
(Fig. 2). 151
The distribution of the genus Triturus is well documented (Arntzen, Wielstra & Wallis, 2014). 152
Triturus cristatus and T. marmoratus have an area of range overlap in France and can often be
153
found in syntopy (Arntzen & Wallis, 1991; Lescure & de Massary, 2012). Other Triturus species
154
contact zones are generally narrow and show a weak but significant negative relationship
155
between the level of hybridization and the genetic distance of species pairs (Arntzen, Wielstra &
156
Wallis, 2014). 151
The distribution of the genus Triturus is well documented (Arntzen, Wielstra & Wallis, 2014). 152
Triturus cristatus and T. marmoratus have an area of range overlap in France and can often be
153
found in syntopy (Arntzen & Wallis, 1991; Lescure & de Massary, 2012). Other Triturus species
154
contact zones are generally narrow and show a weak but significant negative relationship
155
between the level of hybridization and the genetic distance of species pairs (Arntzen, Wielstra &
156
Wallis, 2014). Manuscript to be reviewed 126
2. The higher the variation in the number of thoracic vertebrae, the higher the frequencies of
127
transitional vertebrae are. In hybrids, we would expect that the range of variation in the number
128
of vertebrae overlaps the ranges of parental species’ variation. Also, we would expect the higher
129
frequencies of transitional vertebrae in comparison with parental species. In species with
130
parapatric distributions we would expect the higher variation in the number of vertebrae and the
131
higher frequencies of transitional vertebrae in populations from contact zones in comparison with
132
populations away from contact zones. 133
134
Materials and Methods
135
Triturus newts and their characteristics
136
The vertebral column in Triturus newts is differentiated in: the cervical region– consisting of a
137
single anterior vertebra, the atlas; the thoracic (trunk) region,–consisting of a rib-bearing
138
vertebrae; the sacral region – usually a single vertebra with well-developed stout transverse
139
processes for the attachment of sacral ribs and pelvic girdle; the caudosacral region – up to three
140
vertebrae that continue from the caudal to the sacral vertebra and are associated with the pelvic
141
musculature and cloaca and the caudal region – the remaining vertebrae in the tail (Fig. 1). The
142
body elongation in Triturus species appears to be correlated with the length of the aquatic phase
143
– more terrestrial species have a short and stout trunk with relatively longer legs compared to
144
species with a more aquatic life style which have a more elongated trunk and relatively shorter
145
legs. Body elongation involves a larger number of thoracic vertebrae. More specifically, the
146
number of thoracic vertebrae in the vertebral formulae varies from 12 in T. marmoratus and T. 147
pygmaeus, which have a short aquatic phase (T. marmoratus only two months), 13 in T. karelinii
148
and T. ivanbureschi, 14 in T. macedonicus and T. carnifex, 15 in T. cristatus to 16 or 17 in T. 149
dobrogicus, the most aquatic species which has six months long aquatic phase (Arntzen, 2003)
150
(Fig. 2). 151
The distribution of the genus Triturus is well documented (Arntzen, Wielstra & Wallis, 2014). 152
Triturus cristatus and T. marmoratus have an area of range overlap in France and can often be
153
found in syntopy (Arntzen & Wallis, 1991; Lescure & de Massary, 2012). Manuscript to be reviewed Here, we compared
119
variation in the number of thoracic vertebrae and transitional sacral vertebrae among Triturus
120
species, T. marmoratus × T. cristatus F1 hybrids with parental species, and populations from
121
contact zones with populations away from contact zones. More specifically, we explored the
122
intra- and interspecific variation in number of the thoracic vertebrae and frequencies of
123
transitional vertebrae at the thoraco-sacral boundary to test the following hypotheses:
124
1. Species with more vertebrae in the thoracic region are more variable in the number of thoracic 124
1. Species with more vertebrae in the thoracic region are more variable in the number of thoracic
125
vertebrae than those with fewer vertebrae in the thoracic region. PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) Manuscript to be reviewed Other Triturus species
154
contact zones are generally narrow and show a weak but significant negative relationship
155
b t
th l
l f h b idi
ti
d th
ti di t
f
i
i
(A
t
Wi l t
& 126
2. The higher the variation in the number of t
127
transitional vertebrae are. In hybrids, we wou
128
of vertebrae overlaps the ranges of parental s
129
frequencies of transitional vertebrae in comp
130
parapatric distributions we would expect the
131
higher frequencies of transitional vertebrae in
132
populations away from contact zones. 133
134
Materials and Methods
135
Triturus newts and their characteristics
136
The vertebral column in Triturus newts is dif
137
single anterior vertebra, the atlas; the thoraci
138
vertebrae; the sacral region – usually a single
139
processes for the attachment of sacral ribs an
140
vertebrae that continue from the caudal to the
141
musculature and cloaca and the caudal region
142
body elongation in Triturus species appears t
143
– more terrestrial species have a short and sto
144
species with a more aquatic life style which h
145
legs. Body elongation involves a larger numb
146
number of thoracic vertebrae in the vertebral
147
pygmaeus, which have a short aquatic phase
148
and T. ivanbureschi, 14 in T. macedonicus an
149
dobrogicus, the most aquatic species which h
150
(Fig. 2). 151
The distribution of the genus Triturus is well
152
Triturus cristatus and T. marmoratus have an
153
found in syntopy (Arntzen & Wallis, 1991; L
154
contact zones are generally narrow and show 126
2. The higher the variation in the number of thoracic vertebrae, the higher the frequencies of
127
transitional vertebrae are. In hybrids, we would expect that the range of variation in the number
128
of vertebrae overlaps the ranges of parental species’ variation. Also, we would expect the higher
129
frequencies of transitional vertebrae in comparison with parental species. In species with
130
parapatric distributions we would expect the higher variation in the number of vertebrae and the
131
higher frequencies of transitional vertebrae in populations from contact zones in comparison with
132
populations away from contact zones. 133 Right side asymmetry of a sacral vertebra is
186
scored when the thoracic rib is present on the right side and the sacral rib on the left side of
187
transitional thoraco-sacral vertebra and vice versa for left side asymmetry For a 3D model of 157
158
Material analysed
159
We analysed axial skeletons of 1436 adult newts that originate from 126 populations of all eight
160
species of Triturus newts (Fig. 3). For this study we used X-ray images of good quality and
161
cleared and stained skeletons. The X-ray images were obtained on a Faxitron 43855C/D with an
162
exposure of 20–40s at 3 mA and 70kV. Other material was cleared with trypsin and KOH and
163
stained with Alizarin Red S for bone deposition (Dingerkus & Uhler, 1977) and stored in
164
glycerine. Analyzed specimens are from the batrachological collection of the Institute for
165
Biological Research “Siniša Stanković”, Belgrade, Serbia (N = 601) and from the collection of
166
the Naturalis Biodiversity Center, Leiden, The Netherlands (N = 835). Our material covers the
167
geographic, taxonomic and molecular genetic diversity of the genus. On the basis of well
168
documented species distributions (Mikulíček et al., 2012; Arntzen, Wielstra & Wallis, 2014)
169
populations were assigned as “central” or “fringe” based upon their geographical position away
170
(>= 50km) or close to (< 50 km) congeneric species. For localities and sample size per
171
population see Supplementary Table S1. 172
173
Scoring vertebral formulae and transitional thoraco-sacral vertebrae
174
We determined the vertebral formula by counting the number of cervical (C), thoracic (T) and
175
sacral vertebrae (S). The caudosacral and caudal regions are excluded from our formula as the
176
detailed inspection of cleared and stained specimens showed that a variable number of
177
caudosacral vertebrae frequently bear small, much reduced ribs which could be fused with
178
transverse processes and cannot always be distinguished on X-ray images. The number of tail
179
vertebrae was available only for a subset of specimens; in most specimens tails had been
180
removed for enzyme electrophoretic analyses or were broken or damaged. Manuscript to be reviewed 157
158
Material analysed
159
We analysed axial skeletons of 1436 adult newts that originate from 126 populations of all eight
160
species of Triturus newts (Fig. 3). For this study we used X-ray images of good quality and
161
cleared and stained skeletons. The X-ray images were obtained on a Faxitron 43855C/D with an
162
exposure of 20–40s at 3 mA and 70kV. Other material was cleared with trypsin and KOH and
163
stained with Alizarin Red S for bone deposition (Dingerkus & Uhler, 1977) and stored in
164
glycerine. Analyzed specimens are from the batrachological collection of the Institute for
165
Biological Research “Siniša Stanković”, Belgrade, Serbia (N = 601) and from the collection of
166
the Naturalis Biodiversity Center, Leiden, The Netherlands (N = 835). Our material covers the
167
geographic, taxonomic and molecular genetic diversity of the genus. On the basis of well
168
documented species distributions (Mikulíček et al., 2012; Arntzen, Wielstra & Wallis, 2014)
169
populations were assigned as “central” or “fringe” based upon their geographical position away
170
(>= 50km) or close to (< 50 km) congeneric species. For localities and sample size per
171
population see Supplementary Table S1. 172
173
Scoring vertebral formulae and transitional thoraco-sacral vertebrae
174
We determined the vertebral formula by counting the number of cervical (C), thoracic (T) and
175
sacral vertebrae (S). The caudosacral and caudal regions are excluded from our formula as the
176
detailed inspection of cleared and stained specimens showed that a variable number of
177
caudosacral vertebrae frequently bear small, much reduced ribs which could be fused with
178
transverse processes and cannot always be distinguished on X-ray images. The number of tail
179
vertebrae was available only for a subset of specimens; in most specimens tails had been
180
removed for enzyme electrophoretic analyses or were broken or damaged. 181
Homeotic transformations of thoracic vertebra into sacral vertebra, or vice versa (transitional
182
sacral vertebra having half of the identity of thoracic vertebra and half of the identity of sacral
183
vertebra) were assigned 0.5 and this score was added to the number of complete thoracic
184
vertebrae. Only complete changes of identity on one side of the vertebrae (on one side thoracic
185
and on one side sacral) were declared transitional. 135
Triturus newts and their characteristics PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) 181
Homeotic transformations of thoracic vertebra into sacral vertebra, or vice versa (transitional
182
sacral vertebra having half of the identity of thoracic vertebra and half of the identity of sacral
183
vertebra) were assigned 0.5 and this score was added to the number of complete thoracic
184
vertebrae. Only complete changes of identity on one side of the vertebrae (on one side thoracic
185
and on one side sacral) were declared transitional. Right side asymmetry of a sacral vertebra is
186
scored when the thoracic rib is present on the right side and the sacral rib on the left side of
187
transitional thoraco-sacral vertebra and vice versa for left side asymmetry For a 3D model of 215
Results 216
Vertebral formula and transitional sacral vertebra in Triturus newts
217
The most common vertebral formulae were 1C 12T 1S in T. marmoratus and T. pygmaeus, 1C
218
13T 1S in T. karelinii and T. ivanbureschi, 1C 14T 1S in T. macedonicus and T. carnifex, 1C 15T 173
Scoring vertebral formulae and transitional thoraco-sacral vertebrae 173
Scoring vertebral formulae and transitional thoraco-sacral vertebrae
174
We determined the vertebral formula by counting the number of cervical (C), thoracic (T) and
175
sacral vertebrae (S). The caudosacral and caudal regions are excluded from our formula as the
176
detailed inspection of cleared and stained specimens showed that a variable number of
177
caudosacral vertebrae frequently bear small, much reduced ribs which could be fused with
178
transverse processes and cannot always be distinguished on X-ray images. The number of tail
179
vertebrae was available only for a subset of specimens; in most specimens tails had been
180
removed for enzyme electrophoretic analyses or were broken or damaged. 173
Scoring vertebral formulae and transitional thoraco-sacral vertebrae
174
We determined the vertebral formula by counting the number of cervical (C), thoracic (T) and
175
sacral vertebrae (S). The caudosacral and caudal regions are excluded from our formula as the
176
detailed inspection of cleared and stained specimens showed that a variable number of
177
caudosacral vertebrae frequently bear small, much reduced ribs which could be fused with
178
transverse processes and cannot always be distinguished on X-ray images. The number of tail
179
vertebrae was available only for a subset of specimens; in most specimens tails had been
180
removed for enzyme electrophoretic analyses or were broken or damaged. PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) Manuscript to be reviewed 192 193
Statistical analyses
194
The Spearman correlation coefficient (rs) was used to quantify correlation between species
195
modal numbers of thoracic vertebrae (Tn) and 1) the percentage of individuals with the number
196
of complete thoracic vertebrae different from the modal number (Tvar) and 2) the range of
197
variation in the number of thoracic vertebrae (Trange) within species. The same measure was used
198
to quantify the relationship between percentages of transitional sacral vertebrae (Str) and Tvar and
199
Trange. To test for differences between hybrids and parental species across fringe and central
200
populations we used the G-test of independence. To analyse interspecific variation in a
201
phylogenetic context, we used a well resolved time-calibrated phylogeny of genus Triturus
202
(Arntzen et al., 2015) shown in Figure 2. Associations derived from common ancestry were
203
evaluated by calculating the strength of the phylogenetic signal for analysed variables (Tn, Tvar,
204
Trange and Str). The procedure involves the random permutation of the variables over the terminal
205
units of the phylogenetic tree (10,000 iterations), in which the test statistic is the total amount of
206
squared change summed over all branches of the tree. We applied the phylogenetic independent
207
contrasts approach (Felsenstein, 1985) to obtain a set of independent contrasts. The regression of
208
1) Tvar independent contrasts on Tn independent contrasts and 2) Trange independent contrasts on
209
Tn independent contrasts were used to explore the relationship between evolutionary change in
210
the number of the thoracic vertebra in vertebral formula and amount of variation in the number
211
of thoracic vertebrae. The regressions of Tvar and Trange independent contrasts on Str were used to
212
explore changes and frequencies of transitional vertebrae, taking the similarity due to shared
213
ancestry into account. Manuscript to be reviewed 188
regular and transitional thoraco-sacral vertebra obtained by CT-scanning see Supplementary
189
data. We assumed that the frequency of transitional vertebrae with a complete change of identity
190
at one side of vertebra reflects the frequency of all homeotic transformations, including more
191
gradual ones, which could not always be scored. 192
193
Statistical analyses
194
The Spearman correlation coefficient (rs) was used to quantify correlation between species
195
modal numbers of thoracic vertebrae (Tn) and 1) the percentage of individuals with the number
196
of complete thoracic vertebrae different from the modal number (Tvar) and 2) the range of
197
variation in the number of thoracic vertebrae (Trange) within species. The same measure was used
198
to quantify the relationship between percentages of transitional sacral vertebrae (Str) and Tvar and
199
Trange. To test for differences between hybrids and parental species across fringe and central
200
populations we used the G-test of independence. To analyse interspecific variation in a
201
phylogenetic context, we used a well resolved time-calibrated phylogeny of genus Triturus
202
(Arntzen et al., 2015) shown in Figure 2. Associations derived from common ancestry were
203
evaluated by calculating the strength of the phylogenetic signal for analysed variables (Tn, Tvar,
204
Trange and Str). The procedure involves the random permutation of the variables over the terminal
205
units of the phylogenetic tree (10,000 iterations), in which the test statistic is the total amount of
206
squared change summed over all branches of the tree. We applied the phylogenetic independent
207
contrasts approach (Felsenstein, 1985) to obtain a set of independent contrasts. The regression of
208
1) Tvar independent contrasts on Tn independent contrasts and 2) Trange independent contrasts on
209
Tn independent contrasts were used to explore the relationship between evolutionary change in
210
the number of the thoracic vertebra in vertebral formula and amount of variation in the number
211
of thoracic vertebrae. The regressions of Tvar and Trange independent contrasts on Str were used to
212
explore changes and frequencies of transitional vertebrae, taking the similarity due to shared
213
ancestry into account. 214 188
regular and transitional thoraco-sacral vertebra obtained by CT-scanning see Supplementary
189
data. We assumed that the frequency of transitional vertebrae with a complete change of identity
190
at one side of vertebra reflects the frequency of all homeotic transformations, including more
191
gradual ones, which could not always be scored. Manuscript to be reviewed 219
1S in T. cristatus and 1C 17T 1S in T. dobrogicus (See Fig. 2 and Table 1). The percentage of
220
individuals with a number of complete thoracic vertebrae different from the modal number (Tvar)
221
varied among species, from 3.6% in T. pygmaeus to 24.6% in T. dobrogicus. The range of
222
variation in the number of thoracic vertebrae is 12 – 13 observed in T. marmoratus and T. 223
pygmaeus, 12 – 14 in T. ivanbureschi, 12 – 14 in T. karelinii, 13 – 15 in T. macedonicus, 13 – 16
224
in T. carnifex, 13 – 16 in T. cristatus and 15 – 18 in T. dobrogicus (see Table 1). The variation in
225
the vertebrae number per population is shown in Supplementary Table S2. 219
1S in T. cristatus and 1C 17T 1S in T. dobrogicus (See Fig. 2 and Table 1). The percentage of
220
individuals with a number of complete thoracic vertebrae different from the modal number (Tvar)
221
varied among species, from 3.6% in T. pygmaeus to 24.6% in T. dobrogicus. The range of
222
variation in the number of thoracic vertebrae is 12 – 13 observed in T. marmoratus and T. 223
pygmaeus, 12 – 14 in T. ivanbureschi, 12 – 14 in T. karelinii, 13 – 15 in T. macedonicus, 13 – 16
224
in T. carnifex, 13 – 16 in T. cristatus and 15 – 18 in T. dobrogicus (see Table 1). The variation in
225
the vertebrae number per population is shown in Supplementary Table S2. 226
Frequencies of recorded homeotic transformations in Triturus newt species are listed in Table 2. 227
The least common is the homeotic transformation of cervical vertebra with the thoracic rib
228
attached to one or both sides of the vertebra, recorded six times (0.41%) and in four out of eight
229
species. Two types of transitional vertebrae at the thoraco-sacral boundary were recorded. The
230
first type involves changes of two succeeding vertebrae - transitional sacral vertebra with
231
thoracic rib at one side and sacral rib at the other side, followed by transitional vertebra having
232
sacral rib at one side (opposite than previous vertebra) and no rib attached on the other side (see
233
Figs 1 and 4). Manuscript to be reviewed The second type of transitional sacral vertebra involves transitional thoraco-sacral
234
vertebra, with a thoracic rib at one side and a sacral rib at the other, followed by regular sacral
235
vertebra. The transitional changes involving two adjacent vertebrae, thoracic and sacral (here
236
termed transitional sacral) are more frequent than transitional changes of thoracic to sacral
237
vertebra without changes of sacral vertebra. Excluding F1 hybrids, we recorded a transitional
238
sacrum in 70 out of 1368 specimens (5.1%). Both, right side and left side asymmetries were
239
recorded (Table 2). We found that Tn and Tvar are significantly positively correlated (rs = 0.75, p
240
= 0.023), indicating that species with more vertebrae in the thoracic region are more variable in
241
the number of vertebrae. A significant correlation was also found between Tn and Trange (rs =
242
0.90, p = 0.002), indicating that the range of variation was significantly higher in species with
243
more thoracic vertebrae. We did not find a correlation between Str and Tvar (rs = 0.31, p = 0.46) or
244
between Str and Trange (rs = 0.13, p = 0.76). 247
We found a statistically significant phylogenetic signal in Tn (p = 0.013) and Trange (p = 0.033)
248
and no significant phylogenetic signal in Tvar (p = 0.730) and Str (p = 0.970). The regression of 1)
249
Tvar independent contrasts on Tn independent contrasts (p = 0.018) and 2) the regression of Trange Manuscript to be reviewed 250
independent contrasts on Tn independent contrasts revealed a significant relationship between the
251
increase in the number of thoracic vertebrae and the amount of variation in the number of
252
vertebrae (p = 0.006). We found no significant relationship between Trange independent contrasts
253
and Str independent contrasts (p = 0.413). 254 250
independent contrasts on Tn independent contrasts revealed a significant relationship between the
251
increase in the number of thoracic vertebrae and the amount of variation in the number of
252
vertebrae (p = 0.006). We found no significant relationship between Trange independent contrasts
253
and Str independent contrasts (p = 0.413). 254
255
Hybridization and variation in vertebral formula
256
There were statistically significant differences in changes in vertebral formulae between central’
257
and fringe populations (G-test for independence, G = 18.61, p = 0.001). For fringe populations,
258
the observed range of variation in number of thoracic vertebrae is 12-15 in T. ivanbureschi, 13-
259
16 in T. macedonicus, 13-15 in T. carnifex, 13-17 in T. cristatus and 14-18 in T. dobrogicus. In
260
T. dobrogicus and T. ivanbureschi fringe populations differed significantly from central
261
populations in the frequencies of individuals with non-modal vertebrae formulae. For other
262
species no significant differences between central and fringe populations were found (Table 4). 263
In T. cristatus × T. marmoratus, sixty F1 hybrids (88.2%) have a vertebral formula with an
264
intermediate number of thoracic vertebrae (Table 3). Six hybrids (8.8%) possess an incomplete
265
homeotic transformation. Among these, one has an incomplete transformation of a cervical into a
266
thoracic vertebra. The most frequent incomplete homeotic transformation involves an
267
asymmetrical sacrum. The frequencies of transitional sacral vertebra in hybrids and parental
268
species are similar (G-test for independence, G = 1.07, p = 0.59). 269
270
Di
i 252
vertebrae (p = 0.006). We found no significant relationship between Trange independent contrasts
253
and Str independent contrasts (p = 0.413). 254
255
Hybridization and variation in vertebral formula
256
There were statistically significant differences in changes in vertebral formulae between central’
257
and fringe populations (G-test for independence, G = 18.61, p = 0.001). For fringe populations,
258
the observed range of variation in number of thoracic vertebrae is 12-15 in T. ivanbureschi, 13-
259
16 in T. macedonicus, 13-15 in T. carnifex, 13-17 in T. cristatus and 14-18 in T. dobrogicus. PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) Manuscript to be reviewed marmoratus, sixty F1 hybrids (88.2%) have a vertebral formula with an
264
intermediate number of thoracic vertebrae (Table 3). Six hybrids (8.8%) possess an incomplete
265
homeotic transformation. Among these, one has an incomplete transformation of a cervical into a
266
thoracic vertebra. The most frequent incomplete homeotic transformation involves an
267
asymmetrical sacrum. The frequencies of transitional sacral vertebra in hybrids and parental
268
species are similar (G-test for independence, G = 1.07, p = 0.59). 269 Manuscript to be reviewed In
260
T. dobrogicus and T. ivanbureschi fringe populations differed significantly from central
261
populations in the frequencies of individuals with non-modal vertebrae formulae. For other
262
species no significant differences between central and fringe populations were found (Table 4). 263
In T. cristatus × T. marmoratus, sixty F1 hybrids (88.2%) have a vertebral formula with an
264
intermediate number of thoracic vertebrae (Table 3). Six hybrids (8.8%) possess an incomplete
265
homeotic transformation. Among these, one has an incomplete transformation of a cervical into a
266
thoracic vertebra. The most frequent incomplete homeotic transformation involves an
267
asymmetrical sacrum. The frequencies of transitional sacral vertebra in hybrids and parental
268
species are similar (G-test for independence, G = 1.07, p = 0.59). 269
270
Discussion
271
Our study shows a substantial variation in the number of thoracic vertebrae in Triturus newts,
272
suggesting the absence of strong selection against change in the number of thoracic vertebrae. In
273
agreement with the postulations of Geoffroy St. Hilaire, Darwin and Bateson, the variation in the
274
vertebral column is positively correlated with the number of vertebrae. Using independent
275
contrasts we observed a statistically significant, correlated evolutionary change between an
276
increase in the modal number of thoracic vertebrae and variation in vertebrae number. The range
277
of variation in the number of thoracic vertebrae is also significantly higher in species with a
278
larger modal number of thoracic vertebrae. In T. marmoratus and T. pygmaeus the variation is
279
limited to one extra thoracic vertebra, while in crested newt species the variation range is up to
280
four vertebrae. Although such a pattern of intraspecific variation is in agreement with Geoffroy 255
Hybridization and variation in vertebral formula
256
There were statistically significant differences in changes in vertebral formulae between central’
257
and fringe populations (G-test for independence, G = 18.61, p = 0.001). For fringe populations,
258
the observed range of variation in number of thoracic vertebrae is 12-15 in T. ivanbureschi, 13-
259
16 in T. macedonicus, 13-15 in T. carnifex, 13-17 in T. cristatus and 14-18 in T. dobrogicus. In
260
T. dobrogicus and T. ivanbureschi fringe populations differed significantly from central
261
populations in the frequencies of individuals with non-modal vertebrae formulae. For other
262
species no significant differences between central and fringe populations were found (Table 4). 263
In T. cristatus × T. Manuscript to be reviewed We do not know the cause of the
306
constraint on the number of cervical and sacral vertebra in tailed amphibians, but further studies
307
in various amphibian groups that will consider survival rates of individuals with changes in the
308
cervical and sacral region across ontogenetic stages should provide valuable data to solve this
309
issue. 310
Although we hypothesized that frequencies of transitional vertebrae at the thoraco-sacral
311
boundary should be correlated to the range of variation in the number of thoracic vertebrae as in 281
St. Hilaire’s and Darwin’s rules, the obser
282
variation in vertebral number and differen
283
species occupy, in particular the amount o
284
environment. During the terrestrial phase
285
forward propulsion by the synchronous us
286
weight does not need to be supported by t
287
limbs tightly held against the body to redu
288
2006). Therefore, the selection pressures r
289
probably different with respect to the dura
290
variation in the number of thoracic verteb
291
species, but more data on different selectiv
292
are needed to find out whether this is the c
293
294
Frequencies of transitional vertebrae
295
In Triturus newts, the frequency of transit
296
than ten times lower than changes at the th
297
salamanders (Wake & Lawson, 1973) and
298
explained by stronger interactivity and low
299
early organogenesis, or phylotypic stage, w
300
2006). At later stages, development is incr
301
changes are expected to be associated with
302
mutations with an effect during early orga
303
a consequence to more vulnerability and m
304
strongly supported in rodents (Galis & Me
305
hypothesis is discussed by Galis, Wagner
306
constraint on the number of cervical and s
307
in various amphibian groups that will con
308
cervical and sacral region across ontogene
309
issue. 310
Although we hypothesized that frequencie
311
boundary should be correlated to the rang 281
St. Hilaire’s and Darwin’s rules, the observed pattern may also represent an association between
282
variation in vertebral number and differences in selection in different habitats that Triturus
283
species occupy, in particular the amount of time they spend in the terrestrial versus the aquatic
284
environment. During the terrestrial phase the limbs support the weight of the body and provide
285
forward propulsion by the synchronous use of diagonal limb pairs. Manuscript to be reviewed 281
St. Hilaire’s and Darwin’s rules, the observed pattern may also represent an association between
282
variation in vertebral number and differences in selection in different habitats that Triturus
283
species occupy, in particular the amount of time they spend in the terrestrial versus the aquatic
284
environment. During the terrestrial phase the limbs support the weight of the body and provide
285
forward propulsion by the synchronous use of diagonal limb pairs. In water where the body
286
weight does not need to be supported by the limbs the newts move by tail propulsion with the
287
limbs tightly held against the body to reduce hydrodynamic drag (Gvoždik & Van Damme,
288
2006). Therefore, the selection pressures related to specific biomechanical requirements are
289
probably different with respect to the duration of the terrestrial and aquatic phase. The larger
290
variation in the number of thoracic vertebrae may indicate relaxed selection in more aquatic
291
species, but more data on different selective pressures in terrestrial versus aquatic environments
292
are needed to find out whether this is the case. 293
294
Frequencies of transitional vertebrae
295
In Triturus newts, the frequency of transitional changes at the cervico-thoracic boundary is more
296
than ten times lower than changes at the thoraco-sacral boundary. This is also observed in other
297
salamanders (Wake & Lawson, 1973) and mammals (Galis et al., 2006). This pattern may be
298
explained by stronger interactivity and low modularity of developmental processes during the
299
early organogenesis, or phylotypic stage, when the cervical vertebra is determined (Galis et al.,
300
2006). At later stages, development is increasingly less interactive and more modular, such that
301
changes are expected to be associated with fewer pleiotropic effects. The hypothesis that
302
mutations with an effect during early organogenesis stage lead to more pleiotropic effects and as
303
a consequence to more vulnerability and mortality than earlier or later stages was tested and
304
strongly supported in rodents (Galis & Metz 2001). In amphibians indirect support for this
305
hypothesis is discussed by Galis, Wagner & Jockusch (2003). Manuscript to be reviewed In water where the body
286
weight does not need to be supported by the limbs the newts move by tail propulsion with the
287
limbs tightly held against the body to reduce hydrodynamic drag (Gvoždik & Van Damme,
288
2006). Therefore, the selection pressures related to specific biomechanical requirements are
289
probably different with respect to the duration of the terrestrial and aquatic phase. The larger
290
variation in the number of thoracic vertebrae may indicate relaxed selection in more aquatic
291
species, but more data on different selective pressures in terrestrial versus aquatic environments
292
are needed to find out whether this is the case. 293 281
St. Hilaire’s and Darwin’s rules, the observed pattern may also represent an association between
282
variation in vertebral number and differences in selection in different habitats that Triturus
283
species occupy, in particular the amount of time they spend in the terrestrial versus the aquatic
284
environment. During the terrestrial phase the limbs support the weight of the body and provide
285
forward propulsion by the synchronous use of diagonal limb pairs. In water where the body
286
weight does not need to be supported by the limbs the newts move by tail propulsion with the
287
limbs tightly held against the body to reduce hydrodynamic drag (Gvoždik & Van Damme,
288
2006). Therefore, the selection pressures related to specific biomechanical requirements are
289
probably different with respect to the duration of the terrestrial and aquatic phase. The larger
290
variation in the number of thoracic vertebrae may indicate relaxed selection in more aquatic
291
species, but more data on different selective pressures in terrestrial versus aquatic environments
292
are needed to find out whether this is the case. 293 294
Frequencies of transitional vertebrae
295
In Triturus newts, the frequency of transitional changes at the cervico-thoracic boundary is more
296
than ten times lower than changes at the thoraco-sacral boundary. This is also observed in other
297
salamanders (Wake & Lawson, 1973) and mammals (Galis et al., 2006). This pattern may be
298
explained by stronger interactivity and low modularity of developmental processes during the
299
early organogenesis, or phylotypic stage, when the cervical vertebra is determined (Galis et al.,
300
2006). At later stages, development is increasingly less interactive and more modular, such that
301
changes are expected to be associated with fewer pleiotropic effects. 270
Discussion 271
Our study shows a substantial variation in the number of thoracic vertebrae in Triturus newts,
272
suggesting the absence of strong selection against change in the number of thoracic vertebrae. In
273
agreement with the postulations of Geoffroy St. Hilaire, Darwin and Bateson, the variation in the
274
vertebral column is positively correlated with the number of vertebrae. Using independent
275
contrasts we observed a statistically significant, correlated evolutionary change between an
276
increase in the modal number of thoracic vertebrae and variation in vertebrae number. The range
277
of variation in the number of thoracic vertebrae is also significantly higher in species with a
278
larger modal number of thoracic vertebrae. In T. marmoratus and T. pygmaeus the variation is
279
limited to one extra thoracic vertebra, while in crested newt species the variation range is up to
280
four vertebrae. Although such a pattern of intraspecific variation is in agreement with Geoffroy 271
Our study shows a substantial variation in the number of thoracic vertebrae in Triturus newts,
272
suggesting the absence of strong selection against change in the number of thoracic vertebrae. In
273
agreement with the postulations of Geoffroy St. Hilaire, Darwin and Bateson, the variation in the
274
vertebral column is positively correlated with the number of vertebrae. Using independent
275
contrasts we observed a statistically significant, correlated evolutionary change between an
276
increase in the modal number of thoracic vertebrae and variation in vertebrae number. The range
277
of variation in the number of thoracic vertebrae is also significantly higher in species with a
278
larger modal number of thoracic vertebrae. In T. marmoratus and T. pygmaeus the variation is
279
limited to one extra thoracic vertebra, while in crested newt species the variation range is up to
280
four vertebrae. Although such a pattern of intraspecific variation is in agreement with Geoffroy PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) Manuscript to be reviewed Available literature data indicate
313
that incomplete homeotic transformation of sacral vertebrae are relatively common, with up to
314
10% across the various salamander lineages: 4.5% in Batrachoseps attenuatus (Jockusch, 1997),
315
5.7% in Rhyacotriton olympicus (Worthington, 1971), 6% in Plethodon cinereus (Highton,
316
1960), up to 9% for newt genera Lissotriton and Ichthyosaura (Arntzen et al., 2015) and between
317
1.9% and 9.0% in Triturus newts (this study). The lower than expected incidence of transitional
318
vertebrae could result from developmental mechanisms favoring complete numbers of thoracic
319
vertebrae and/or from selection against transitional sacral vertebrae due to associated problems
320
related to an asymmetric sacrum (c.f. Galis et al., 2014). Potential problems associated to
321
asymmetrical sacrum might arise due to asymmetrical muscle attachments, blood vessels and
322
innervation, or biomechanical problems during locomotion. In salamanders, the selection
323
pressures related to specific biomechanical requirements are probably different in fully aquatic
324
larvae and metamorphosed individuals that spend most of their time on land. Furthermore,
325
selection pressures may vary with respect to the duration of annual aquatic and terrestrial phase. 326
More detailed morphological and functional studies of locomotion of larval and metamorphic
327
stages could shed more light on the functional significance of variation in the axial skeleton in
328
Triturus newts. However, it is possible that our results are biased as we have not included the full
329
range of transitional vertebrae. We scored only easily identifiable transitional vertebrae with
330
complete morphological transformations of one side of the vertebra under the assumption that
331
the frequency of these transitional vertebrae reflects the total amount of homeotic
332
transformations. Nonetheless, initial mutations for homeotic transformations can lead to a whole
333
series of gradually transitional homeotic transformations; in the case of thoraco-sacral vertebrae
334
ranging from predominantly thoracic and only slightly sacral to predominantly sacral and slightly
335
thoracic. Inclusion of all transitional vertebral morphologies might change the observed
336
relationship between incomplete homeotic transformations and changes in the number of
337
thoracic vertebrae in newts. 338 Manuscript to be reviewed The hypothesis that
302
mutations with an effect during early organogenesis stage lead to more pleiotropic effects and as
303
a consequence to more vulnerability and mortality than earlier or later stages was tested and
304
strongly supported in rodents (Galis & Metz 2001). In amphibians indirect support for this
305
hypothesis is discussed by Galis, Wagner & Jockusch (2003). We do not know the cause of the
306
constraint on the number of cervical and sacral vertebra in tailed amphibians, but further studies
307
in various amphibian groups that will consider survival rates of individuals with changes in the
308
cervical and sacral region across ontogenetic stages should provide valuable data to solve this
309
issue. 310
Although we hypothesized that frequencies of transitional vertebrae at the thoraco-sacral
311
boundary should be correlated to the range of variation in the number of thoracic vertebrae as in PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) Manuscript to be reviewed 312
mammals (Ten Broek et al., 2012), no correlation was found. Available literature data indicate
313
that incomplete homeotic transformation of sacral vertebrae are relatively common, with up to
314
10% across the various salamander lineages: 4.5% in Batrachoseps attenuatus (Jockusch, 1997),
315
5.7% in Rhyacotriton olympicus (Worthington, 1971), 6% in Plethodon cinereus (Highton,
316
1960), up to 9% for newt genera Lissotriton and Ichthyosaura (Arntzen et al., 2015) and between
317
1.9% and 9.0% in Triturus newts (this study). The lower than expected incidence of transitional
318
vertebrae could result from developmental mechanisms favoring complete numbers of thoracic
319
vertebrae and/or from selection against transitional sacral vertebrae due to associated problems
320
related to an asymmetric sacrum (c.f. Galis et al., 2014). Potential problems associated to
321
asymmetrical sacrum might arise due to asymmetrical muscle attachments, blood vessels and
322
innervation, or biomechanical problems during locomotion. In salamanders, the selection
323
pressures related to specific biomechanical requirements are probably different in fully aquatic
324
larvae and metamorphosed individuals that spend most of their time on land. Furthermore,
325
selection pressures may vary with respect to the duration of annual aquatic and terrestrial phase. 326
More detailed morphological and functional studies of locomotion of larval and metamorphic
327
stages could shed more light on the functional significance of variation in the axial skeleton in
328
Triturus newts. However, it is possible that our results are biased as we have not included the full
329
range of transitional vertebrae. We scored only easily identifiable transitional vertebrae with
330
complete morphological transformations of one side of the vertebra under the assumption that
331
the frequency of these transitional vertebrae reflects the total amount of homeotic
332
transformations. Nonetheless, initial mutations for homeotic transformations can lead to a whole
333
series of gradually transitional homeotic transformations; in the case of thoraco-sacral vertebrae
334
ranging from predominantly thoracic and only slightly sacral to predominantly sacral and slightly
335
thoracic. Inclusion of all transitional vertebral morphologies might change the observed
336
relationship between incomplete homeotic transformations and changes in the number of
337
thoracic vertebrae in newts. 338
339
Hybridization, marginality and homeotic transformations 312
mammals (Ten Broek et al., 2012), no correlation was found. Manuscript to be reviewed 343
phenotypes with 13 thoracic vertebrae, an intermediate number. It is interesting to note that 13
344
thoracic vertebrae is the only number that is shared by both parental species. In T. cristatus × T. 345
marmoratus offspring there is considerable mortality and almost all of F1 hybrids (~90%) had T. 346
cristatus as mother. The marmoratus-mothered specimens were all male, due to low survival of
347
female embryos (Arntzen et al., 2009). Developmental anomalies in T. cristatus × T. marmoratus
348
crosses, including more digital anomalies compared with parental species (hybrids 16.9%,
349
parental species pooled 5.4%) (Vallée, 1959; more data in Arntzen & Wallis, 1991) are observed,
350
and therefore, the higher number of changes in the axial skeleton may be related to a generally
351
higher number of anomalies. The high mortality may also influence the incidence of the
352
variability and transitional vertebrae. 353
Significantly higher frequency of changes in vertebral formula in fringe populations of T. 354
ivanbureschi and T. dobrogicus species may well have to do with the confirmed presence of
355
hybridization in the contact zones of T. cristatus and T. dobrogicus populations (Mikulíček et al.,
356
2012), of T. carnifex and T. dobrogicus populations (Wallis & Arntzen, 1989) and of T. 357
ivanbureschi and T. macedonicus populations (Arntzen, Wielstra & Wallis, 2014). However, the
358
effect of the various genotype combinations on the survival rate and morphology of these species
359
remains to be studied. 353
Significantly higher frequency of changes in vertebral formula in fringe populations of T. 354
ivanbureschi and T. dobrogicus species may well have to do with the confirmed presence of
355
hybridization in the contact zones of T. cristatus and T. dobrogicus populations (Mikulíček et al.,
356
2012), of T. carnifex and T. dobrogicus populations (Wallis & Arntzen, 1989) and of T. 357
ivanbureschi and T. macedonicus populations (Arntzen, Wielstra & Wallis, 2014). However, the
358
effect of the various genotype combinations on the survival rate and morphology of these species
359
remains to be studied. 359
remains to be studied. 360
In conclusion, Triturus newts have a relatively large amount of variation in the number of
361
thoracic vertebrae, both with respect to the frequency of non-modal numbers and the range of
362
variation. In agreement with Geoffroy St. Hilaire’s rule, variation was larger in species with a
363
larger number of thoracic vertebrae. PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) Manuscript to be reviewed The absence of a correlation between the frequency of
364
homeotic change (transitional sacral vertebrae, Str) and variation in the number of vertebrae (Tvar,
365
Trange) could be a result of developmental mechanisms that favour complete numbers of presacral
366
vertebrae and/or selection against transitional vertebrae in this group of tailed amphibians. 367
368
Acknowledgements
369
370
We thank Hans Metz and Joost Woltering for discussions, Ben Wielstra for providing a
371
distribution map and Marieke Vinkenoog for help with X – ray imaging. 372
373
References 360
In conclusion, Triturus newts have a relatively large amount of variation in the number of
361
thoracic vertebrae, both with respect to the frequency of non-modal numbers and the range of
362
variation. In agreement with Geoffroy St. Hilaire’s rule, variation was larger in species with a
363
larger number of thoracic vertebrae. The absence of a correlation between the frequency of
364
homeotic change (transitional sacral vertebrae, Str) and variation in the number of vertebrae (Tvar,
365
Trange) could be a result of developmental mechanisms that favour complete numbers of presacral
366
vertebrae and/or selection against transitional vertebrae in this group of tailed amphibians. 367 PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) Manuscript to be reviewed Manuscript to be reviewed marmoratus in western France. Evolution 45 (4): 805–826
381
Arntzen JW. 2003. Triturus cristatus Superspezies - Kammolch-Artenkreis. (Triturus cristatus
382
(Laurenti, 1768) – Noërdlicher Kammolch, Triturus carnifex (Laurenti, 1768) –
383
Italienischer Kammolch. Triturus dobrogicus (Kiritzescu, 1903) – Donau-
384
Kammolch,Triturus karelinii (Strauch, 1870) – Suëdlicher Kammolch). In: K. 385
Grossenbacher, B. Thiesmeier, eds. Handbuch der Reptilien und Amphibien Europas,
386
Band 4/IIA: Schwanzlurche (Urodela) IIA. Wiebelsheim: Aula–Verlag, 421–514
387
Arntzen JW, Jehle R, Bardakci F, Burke T, Wallis GP. 2009. Asymmetric viability of
388
reciprocal-cross hybrids between crested and marbled newts (Triturus cristatus and T. 389
marmoratus). Evolution 63 (5): 1191–202 DOI: 10.1111/j.1558-5646.2009.00611.x
390
Arntzen JW, Wielstra B, Wallis GP. 2014. The modality of nine Triturus newt hybrid zones
391
assessed with nuclear, mitochondrial and morphological data. Biological Journal of the
392
Linnean Society 113 (2): 604–622 DOI: 10.1111/bij.12358 y
g
379
Arntzen JW, Wallis GP. 1991. Restricted gene flow in a moving hybrid zone of the newts
380
Triturus cristatus and T. marmoratus in western France. Evolution 45 (4): 805–826
381
Arntzen JW. 2003. Triturus cristatus Superspezies - Kammolch-Artenkreis. (Triturus cristatus
382
(Laurenti, 1768) – Noërdlicher Kammolch, Triturus carnifex (Laurenti, 1768) –
383
Italienischer Kammolch. Triturus dobrogicus (Kiritzescu, 1903) – Donau-
384
Kammolch,Triturus karelinii (Strauch, 1870) – Suëdlicher Kammolch). In: K. 385
Grossenbacher, B. Thiesmeier, eds. Handbuch der Reptilien und Amphibien Europas,
386
Band 4/IIA: Schwanzlurche (Urodela) IIA. Wiebelsheim: Aula–Verlag, 421–514 387
Arntzen JW, Jehle R, Bardakci F, Burke T, Wallis GP. 2009. Asymmetric viability of
388
reciprocal-cross hybrids between crested and marbled newts (Triturus cristatus and T. 389
marmoratus). Evolution 63 (5): 1191–202 DOI: 10.1111/j.1558-5646.2009.00611.x
390
Arntzen JW, Wielstra B, Wallis GP. 2014. The modality of nine Triturus newt hybrid zones
391
assessed with nuclear, mitochondrial and morphological data. Biological Journal of the
392
Linnean Society 113 (2): 604–622 DOI: 10.1111/bij.12358 393
Arntzen JW, Beukema W, Galis F, Ivanović A. 2015. Vertebral number is highly evolvable in
394
salamanders and newts (family Salamandridae) and variably associated with climatic
395
parameters. Contributions to Zoology 84 (2): 87–116 393
Arntzen JW, Beukema W, Galis F, Ivanović A. 2015. Vertebral number is highly evolvable in
394
salamanders and newts (family Salamandridae) and variably associated with climatic
395
parameters. Contributions to Zoology 84 (2): 87–116 396
Aulehla A, Pourquié O. 2010. Signaling gradients during paraxial mesoderm development. Manuscript to be reviewed 374
375
Adolphi H. 1898. Über das Wandern der Extremitätenplexus und des Sacrum bei Triton
376
taeniatus. Morphologisches Jahrbuch 25: 544–554
377
Ahlberg PE, Clack JA, Blom H. 2005. The axial skeleton of the Devonian tetrapod
378
Ichthyostega. Nature 437: 137–140 DOI:10.1038/nature.03893
379
Arntzen JW, Wallis GP. 1991. Restricted gene flow in a moving hybrid zone of the newts
380
Triturus cristatus and T. marmoratus in western France. Evolution 45 (4): 805–826
381
Arntzen JW. 2003. Triturus cristatus Superspezies - Kammolch-Artenkreis. (Triturus cristatus
382
(Laurenti, 1768) – Noërdlicher Kammolch, Triturus carnifex (Laurenti, 1768) –
383
Italienischer Kammolch. Triturus dobrogicus (Kiritzescu, 1903) – Donau-
384
Kammolch,Triturus karelinii (Strauch, 1870) – Suëdlicher Kammolch). In: K. 385
Grossenbacher, B. Thiesmeier, eds. Handbuch der Reptilien und Amphibien Europas,
386
Band 4/IIA: Schwanzlurche (Urodela) IIA. Wiebelsheim: Aula–Verlag, 421–514
387
Arntzen JW, Jehle R, Bardakci F, Burke T, Wallis GP. 2009. Asymmetric viability of
388
reciprocal-cross hybrids between crested and marbled newts (Triturus cristatus and T. 389
marmoratus). Evolution 63 (5): 1191–202 DOI: 10.1111/j.1558-5646.2009.00611.x
390
Arntzen JW, Wielstra B, Wallis GP. 2014. The modality of nine Triturus newt hybrid zones
391
assessed with nuclear, mitochondrial and morphological data. Biological Journal of the
392
Linnean Society 113 (2): 604–622 DOI: 10.1111/bij.12358
393
Arntzen JW, Beukema W, Galis F, Ivanović A. 2015. Vertebral number is highly evolvable in
394
salamanders and newts (family Salamandridae) and variably associated with climatic
395
parameters. Contributions to Zoology 84 (2): 87–116
396
Aulehla A, Pourquié O. 2010. Signaling gradients during paraxial mesoderm development. 397
Cold Spring Harbor Perspectives in Biology 2 (2): a000869
398
DOI:10.1101/cshperspect.a000869
399
Bateson W. 1894. Materials for the Study of Variation. London: MacMillan
400
Carapuço M, Nóvoa A, Bobola N, Mallo M. 2005. Hox genes specify vertebral types in the
401
presomitic mesoderm. Genes & Development 19 (18): 2116–2121
402
Carroll RL. 1997. Patterns and Processes of Vertebrate Evolution. Cambridge: Cambridge 374
375
Adolphi H. 1898. Über das Wandern der Extremitätenplexus und des Sacrum bei Triton
376
taeniatus. Morphologisches Jahrbuch 25: 544–554 377
Ahlberg PE, Clack JA, Blom H. 2005. The axial skeleton of the Devonian tetrapod
378
Ichthyostega. Nature 437: 137–140 DOI:10.1038/nature.03893 377
Ahlberg PE, Clack JA, Blom H. 2005. The axial skeleton of the Devonian tetrapod
378
Ichthyostega. Nature 437: 137–140 DOI:10.1038/nature.03893
379
Arntzen JW, Wallis GP. 1991. Restricted gene flow in a moving hybrid zone of the newts
380
Triturus cristatus and T. Manuscript to be reviewed 404
Cooke J, Zeeman ECA. 1976. Clock and wavefront model for control of the number of repeated
405
structures during animal morphogenesis. Journal of Theoretical Biology 58 (2): 455–476 404
Cooke J, Zeeman ECA. 1976. Clock and wavefront model for control of the number of repeated
405
structures during animal morphogenesis. Journal of Theoretical Biology 58 (2): 455–476
406
Crnobrnja J, Džukić G, Krstić N, Kalezić ML. 1997. Evolutionary and paleogeographic
407
effects on the distribution of the Triturus cristatus superspecies in the central Balkans. 408
Amphibia-Reptilia 18 (4): 321–332 DOI: 10.1163/156853897X00378 406
Crnobrnja J, Džukić G, Krstić N, Kalezić ML. 1997. Evolutionary and paleogeographic
407
effects on the distribution of the Triturus cristatus superspecies in the central Balkans. 408
Amphibia-Reptilia 18 (4): 321–332 DOI: 10 1163/156853897X00378 09
Darwin CR. 1860. The Origin of Species. New York: D. Appleton and Comp 410
Dequéant ML, Pourquié O. 2008. Segmental patterning of the vertebrate embryonic axis. 411
Nature Review Genetics 9: 370–382 DOI: 10.1038/nrg2320 410
Dequéant ML, Pourquié O. 2008. Segmental patterning of the vertebrate embryonic axis. 411
Nature Review Genetics 9: 370–382 DOI: 10.1038/nrg2320 412
Diezdel-Corral R, Olivera-Martinez I, Goriely A, Gale E, Maden M, Storey K. 2003. 413
Opposing FGF and retinoid pathways control ventral neural pattern, neuronal
414
differentiation, and segmentation during body axis extension. Neuron 40 (1): 65–79 415
Dingerkus G, Uhler LD. 1977. Enzyme clearing of alcian blue stained whole small vertebrates
416
for demonstration of cartilage. Stain Technology 52 (4): 229–232 417
Dubrulle J, McGrew MJ, Pourquié O. 2001. FGF signaling controls somite boundary position
418
and regulates segmentation clock control of spatiotemporal Hox gene activation. Cell 106
419
(2):219–32 420
Felsenstein J. 1985. Phylogenies and the comparative method. American Naturalist 125 (1): 1–
421
15 422
Galis F, Metz JAJ. 2001. Testing the vulnerability of the phylotypic stage: on modularity and
423
evolutionary conservation. Journal of Experimental Zoology Part B: Molecular and
424
Developmental Evolution 291 (2): 195–204 425
Galis F, Metz JAJ. 2003. Anti-cancer selection as a source of developmental and evolutionary
426
constraints. Bioessays 25 (11): 1035–1039 427
Galis F, Wagner GP, Jockusch E. 2003. Why is limb regeneration possible in amphibians but
428
not in reptiles, birds and mammals? Evolution & Development 5 (2): 208–220 429
Galis F, Van Dooren TJ, Feuth JD, Metz JA,Witkam A, Ruinard S, Steigenga MJ,
430
Wijnaendts LC. 2006. Manuscript to be reviewed 397
Cold Spring Harbor Perspectives in Biology 2 (2): a000869
398
DOI:10.1101/cshperspect.a000869 99
Bateson W. 1894. Materials for the Study of Variation. London: MacMillan 400
Carapuço M, Nóvoa A, Bobola N, Mallo M. 2005. Hox genes specify vertebral types in the
401
presomitic mesoderm. Genes & Development 19 (18): 2116–2121 402
Carroll RL. 1997. Patterns and Processes of Vertebrate Evolution. Cambridge: Cambridge
403
University Press PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) Manuscript to be reviewed 434
Proceedings of the National Academy of Sciences of the United States of America 111
435
(31): 11401–11406 DOI: 10.1073/pnas.1401392111 436
Geoffroy SHI. 1832. Histoire Générale et Particulière des Anomalies de L'organisation chez
437
L'homme et les Animaux. Paris: J.–B. Baillère, Libraire de l’ Academie Royale de
438
Médecine 436
Geoffroy SHI. 1832. Histoire Générale et Particulière des Anomalies de L'organisation chez
437
L'homme et les Animaux. Paris: J.–B. Baillère, Libraire de l’ Academie Royale de
438
Médecine
439
Gerecht A. 1929. Über die Richtung der phyletischenWanderung der Sakralregionbei Triton
440
cristatus und Triton taeniatus. Bulletin de la Société de biologie de Lettonie 1(1): 9–19
441
Gomez C, Özbudak EM, Wunderlich J, Baumann D, Lewis J, Pourquié 2008. Control of
442
segment number in vertebrate embryos. Nature 454 (7202): 335-339
443
Gomez C, Pourquié O. 2009. Developmental control of segment numbers in vertebrates. 444
Journal of Experimental Zoology Part B: Molecular and Developmental Evolution. 312
445
(6): 533–544 DOI: 10.1002/jez.b.21305
446
Gvoždik L,Van Damme R. 2006. Triturus newts defy the running-swimming dilemma. 447
Evolution 60 (10): 2110–2121 437
L'homme et les Animaux. Paris: J.–B. Baillère, Libraire de l’ Academie Royale de
438
Médecine
439
Gerecht A. 1929. Über die Richtung der phyletischenWanderung der Sakralregionbei Triton
440
cristatus und Triton taeniatus. Bulletin de la Société de biologie de Lettonie 1(1): 9–19
441
Gomez C, Özbudak EM, Wunderlich J, Baumann D, Lewis J, Pourquié 2008. Control of
442
segment number in vertebrate embryos. Nature 454 (7202): 335-339 440
cristatus und Triton taeniatus. Bulletin de la Société de biologie de Lettonie 1(1): 9–19
441
Gomez C, Özbudak EM, Wunderlich J, Baumann D, Lewis J, Pourquié 2008. Control of
442
segment number in vertebrate embryos. Nature 454 (7202): 335-339 ,
q
p
g
444
Journal of Experimental Zoology Part B: Molecular and Developmental Evolution. 312
445
(6): 533–544 DOI: 10.1002/jez.b.21305 446
Gvoždik L,Van Damme R. 2006. Triturus newts defy the running-swimming dilemma. 447
Evolution 60 (10): 2110–2121 446
Gvoždik L,Van Damme R. 2006. Triturus newts defy the running-swimming dilemma. 447
Evolution 60 (10): 2110–2121 448
Harima Y, Takashima Y, Ueda Y, Ostsuka T, Kageyama R. 2013. Accelerating the tempo of
449
the segmentation clock by reducing the number of introns in the Hes7 Gene. Cell Reports
450
3 (1): 1–7 DOI: 10.1016/j.celrep.2012.11.012 451
Highton R. 1960. Heritability of geographic variation in trunk segmentation in the red- backed
452
salamander, Plethodon cinereus. Manuscript to be reviewed Extreme selection in humans against homeotic transformations of
431
cervical vertebrae. Evolution 60 (12): 2643–2654 429
Galis F, Van Dooren TJ, Feuth JD, Metz JA,Witkam A, Ruinard S, Steigenga MJ,
430
Wijnaendts LC. 2006. Extreme selection in humans against homeotic transformations of
431
cervical vertebrae. Evolution 60 (12): 2643–2654 432
Galis F, Carrier DR, Alphen J, Van der Mije SD, Doorn TV, Metz JAJ, Ten Broek CMA. 433
2014. Fast running restricts evolutionary change of the vertebral column in mammals. . PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) Manuscript to be reviewed Evolution 14 (3): 351–360 454
vertebrae in slender salamanders, Batrachoseps (Caudata: Plethodontidae). Evolution 51
455
(6): 1966–1982 DOI:10.2307/2411017 456
Lescure J, de Massary J-C. 2012. Atlas des Amphibiens et Reptiles de France. Paris: Muséum
457
national d’Histoire naturelle 458
Litvinchuk SN, Borkin LJ. 2003.Variation in number of trunk vertebrae and in count of costal
459
grooves in salamanders of the family Hynobiidae. Contributions to Zoology 72 (4):195–
460
209 461
Mallo M, Wellik D, Deschamps J. 2010. Hox genes and regional patterning of the vertebrate
462
body plan. Developmental Biology 344 (1): 7–15 DOI: 10.1016/j.ydbio.2010.04.024
463
Mikulíček P, Horák A, Zavadil V, Kautman J, Piálek J. 2012. Hybridization between three
464
crested newt species (Triturus cristatus superspecies) in the Czech Republic and 461
Mallo M, Wellik D, Deschamps J. 2010. Hox genes and regional patterning of the vertebrate
462
body plan. Developmental Biology 344 (1): 7–15 DOI: 10.1016/j.ydbio.2010.04.024 y p
p
gy
( )
j y
463
Mikulíček P, Horák A, Zavadil V, Kautman J, Piálek J. 2012. Hybridization between three
464
crested newt species (Triturus cristatus superspecies) in the Czech Republic and PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) Manuscript to be reviewed 465
Slovakia: comparison of nuclear markers and mitochondrial DNA. Folia Zoologica 61
466
(3-4): 202–218 467
Müller J, Scheyer TM, Head JJ, Barrett PM, Werneburg I, Ericson PGP, Pol D, Sanchez-
468
Villagra MR. 2010. Homeotic effects, somitogenesis and the evolution of vertebral
469
numbers in recent and fossil amniotes. Proceedings of the National Academy of Sciences
470
of the United States of America 107 (5): 2118–2123 471
Palmeirim I, Henrique D, Ish-Horowicz D, Pourquié O. 1997. Avian hairy gene expression
472
identifies a molecular clock linked to vertebrate segmentation and somitogenesis. Cell 91
473
(5): 639–48 474
Peabody RB, Brodie ED. 1975. Effect of temperature, salinity and photoperiod on the number
475
of trunk vertebrae in Ambystoma maculatum. Copeia 1975 (4): 741–746 476
Schroeter C, Oates AC. 2010. Segment number and axial identity in a segmentaion clock
477
period mutant. Current Biology 20 (14): 1254–1258 478
Starck D. 1979. Vergleichende Anatomie der Wirbeltiere. Berlin: Springer Verlag 479
Ten Broek CM, Bakker AJ,Varela-Lasheras I, Bugiani M, Van Dongen S, Galis F. 2012. 480
Evo-Devo of the human vertebral column: on homeotic transformations, pathologies and
481
prenatal selection. Evolutionary Biology 39(4): 456–471 482
Vallée L. 1959. Recherches sur Triturus blasii de l’Isle, hybride naturel de Triturus cristatus
483
Laur. x Triturus marmoratus Latr. Mém. Bulletin de la Société zoologique de France 31: 482
Vallée L. 1959. Recherches sur Triturus blasii de l’Isle, hybride naturel de Triturus cristatus 482
Vallée L. 1959. Recherches sur Triturus blasii de l’Isle, hybride naturel de Triturus cristatus
483
Laur. x Triturus marmoratus Latr. Mém. Bulletin de la Société zoologique de France 31:
484
1–95 483
Laur. x Triturus marmoratus Latr. Mém. Bulletin de la Société zoologique de France 31:
484
1–95 485
Varela-Lasheras I, Bakker A, Mije S, Metz J, Alphen J, Galis F. 2011. Breaking
486
evolutionary and pleiotropic constraints in mammals: On sloths, manatees and homeotic
487
mutations. EvoDevo 2: 11 DOI:10.1186/2041-9139-2-11
488
Wallis GP, Arntzen JW. 1989. Mitochondrial-DNA variation in the crested newt superspecies:
489
limited cytoplasmic gene flow among species. Evolution 43 (1): 88–10 DOI:
490
10.2307/2409166
491
Wake DB, Lawson R. 1973. Developmental and adult morphology of the vertebral column in
492
the plethodontid salamander Eurycea bislineata, with comments on vertebral evolution in
493
the amphibian. Journal of Morphology 139 (3):251–299 DOI: 10.1002/jmor.1051390302
494
Woltering JM. 2012. From lizard to snake; behind the evolution of an extreme body plan. Manuscript to be reviewed 495
Current Genomics 13 (4): 289–299 DOI:10.2174/138920212800793302 485
Varela-Lasheras I, Bakker A, Mije S, Metz J, Alphen J, Galis F. 2011. Breaking
486
evolutionary and pleiotropic constraints in mammals: On sloths, manatees and homeotic 485
Varela-Lasheras I, Bakker A, Mije S, Metz J, Alphen J, Galis F. 2011. Breaking
486
evolutionary and pleiotropic constraints in mammals: On sloths, manatees and homeotic
487
mutations. EvoDevo 2: 11 DOI:10.1186/2041-9139-2-11 488
Wallis GP, Arntzen JW. 1989. Mitochondrial-DNA variation in the crested newt superspecies:
489
limited cytoplasmic gene flow among species. Evolution 43 (1): 88–10 DOI:
490
10.2307/2409166
491
Wake DB, Lawson R. 1973. Developmental and adult morphology of the vertebral column in
492
the plethodontid salamander Eurycea bislineata, with comments on vertebral evolution in
493
the amphibian. Journal of Morphology 139 (3):251–299 DOI: 10.1002/jmor.1051390302
494
Woltering JM. 2012. From lizard to snake; behind the evolution of an extreme body plan. 495
C
t G
i
13 (4) 289 299 DOI 10 2174/138920212800793302 488
Wallis GP, Arntzen JW. 1989. Mitochondrial-DNA variation in the crested newt superspecies:
489
limited cytoplasmic gene flow among species. Evolution 43 (1): 88–10 DOI:
490
10.2307/2409166 491
Wake DB, Lawson R. 1973. Developmental and adult morphology of the vertebral column in
492
the plethodontid salamander Eurycea bislineata, with comments on vertebral evolution in
493
the amphibian. Journal of Morphology 139 (3):251–299 DOI: 10.1002/jmor.1051390302 494
Woltering JM. 2012. From lizard to snake; behind the evolution of an extreme body plan. 495
Current Genomics 13 (4): 289–299 DOI:10.2174/138920212800793302 495
Current Genomics 13 (4): 289–299 DOI:10.2174/138920212800793302 PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) Manuscript to be reviewed Manuscript to be reviewed 496
Wong SFL, Agarwal W, Mansvield JH, Denans N, Schwartz MG, Prosser HM, Pourquié
497
O. 2015. Independent regulation of vertebral number and vertebral identity by
498
microRNA-196 paralogs. Proceedings of the National Academy of Sciences of the United
499
States of America 112 (35): E4884–93 DOI: 10.1073/pnas1512655112
500
Woolfenden GE. 1961. Postcranial osteology of the waterfowl. Bulletin of the Florida State
501
Museum 6: (1) –129
502
Worthington RD. 1971. Postmetamorphic changes in the vertebrae of the marbled salamander
503
Ambystoma opacum Gravenhorst (Amphibia, Caudata). El Paso: Texas Western Press
504 496
Wong SFL, Agarwal W, Mansvield JH, Denans N, Schwartz MG, Prosser HM, Pourquié
497
O. 2015. Independent regulation of vertebral number and vertebral identity by
498
microRNA-196 paralogs. Proceedings of the National Academy of Sciences of the United
499
States of America 112 (35): E4884–93 DOI: 10.1073/pnas1512655112
500
Woolfenden GE. 1961. Postcranial osteology of the waterfowl. Bulletin of the Florida State
501
Museum 6: (1) –129
502
Worthington RD. 1971. Postmetamorphic changes in the vertebrae of the marbled salamander 496
Wong SFL, Agarwal W, Mansvield JH, Denans N, Schwartz MG, Prosser HM, Pourquié
497
O. 2015. Independent regulation of vertebral number and vertebral identity by
498
microRNA-196 paralogs. Proceedings of the National Academy of Sciences of the United
499
States of America 112 (35): E4884–93 DOI: 10.1073/pnas1512655112
500
Woolfenden GE. 1961 Postcranial osteology of the waterfowl Bulletin of the Florida State 500
Woolfenden GE. 1961. Postcranial osteology of the waterfowl. Bulletin of the Florida State
501
Museum 6: (1) –129 502
Worthington RD. 1971. Postmetamorphic changes in the vertebrae of the marbled salamander
503
Ambystoma opacum Gravenhorst (Amphibia, Caudata). El Paso: Texas Western Press
504 PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) Manuscript to be reviewed 505
Figure 1. 506
The regionalization of the vertebral column in Triturus newts and schematic presentation of
507
scored homeotic transformations (example of T. marmoratus). A) vertebral column without
508
homeotic transformation and regular number of vertebrae – the first three caudal vertebrae are
509
shown; B) incomplete homeotic transformations of cervical vertebra into thoracic; C) complete
510
homeotic transformation of cervical into thoracic vertebra; D) transitional sacral vertebra with
511
thoracic rib at one side and sacral rib at the other side followed by transitional vertebra with
512
sacral rib at the one side and no rib at the other; E) transitional thoraco-sacral vertebra, with
513
thoracic rib at one side and sacral rib at the other, followed by regular sacral vertebra. 511
thoracic rib at one side and sacral rib at the other side followed by transitional vertebra with
512
sacral rib at the one side and no rib at the other; E) transitional thoraco-sacral vertebra, with
513
thoracic rib at one side and sacral rib at the other, followed by regular sacral vertebra. 514
PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) 514 514 Manuscript to be reviewed 516
Calibrated phylogeny for the genus Triturus with the modal number of vertebrae indicated by
517
colour code (after Arntzen et al., 2015). 516
Calibrated phylogeny for the genus Triturus with the modal number of vertebrae indicated by
517
colour code (after Arntzen et al., 2015). 516
Calibrated phylogeny for the genus Triturus with the modal number of vertebrae indicated by
517
colour code (after Arntzen et al., 2015). 518 519 519 519 PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) 17
colour code (after Arntzen et al., 2015). Manuscript to be reviewed 520
Figure 3. 521
Distribution of eight Triturus species across Europe and adjacent Asia. Studied populations are
522
marked by solid dots (central populations) and open dots (fringe populations). For detailed
523
information see Supplementary Table S1. 524
525 524
525 PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) Manuscript to be reviewed 539
The number of thoracic vertebrae in Triturus species (central populations only). Modal numbers
540
of thoracic vertebrae in vertebral formulae are shown in bold. Str – percentage of individuals with
541
transitional vertebrae at thoraco-sacral boundary. Tvar – percentage of individuals with the
542
complete number of thoracic vertebrae different from the modal number. 539
The number of thoracic vertebrae in Triturus species (central populations only). Modal numbers
540
of thoracic vertebrae in vertebral formulae are shown in bold. Str – percentage of individuals with
541
transitional vertebrae at thoraco-sacral boundary. Tvar – percentage of individuals with the
542
complete number of thoracic vertebrae different from the modal number. 543
Species
Sample
size
Number of thoracic vertebrae
Str (%)
Tvar (%)
12 12.5 13 13.5 14 14.5 15 15.5 16 16.5 17 17.5 18
T. marmoratus
58
46
4
8
6.9
13.8
T. pygmaeus
55
52
1
2
1.8
3.6
T. ivanbureschi
175
1
3
150
4
17
4.0
10.3
T. karelinii
43
1
40
2
2.3
4.7
T. carnifex
66
4
5
53
3
1
7.6
12.1
T. macedonicus
67
9
5
51
1
1
9.0
14.9
T. cristatus
122
1
6
1
98
5
11
4.9
14.8
T. dobrogicus
57
2
11
1
42
1
1.8
24.6 543 Manuscript to be reviewed 526
Figure 4. 527
Homeotic transformations recorded: A) complete vertebral column without homeotic
528
transformations and transitional vertebrae. B) transitional cervical vertebra (cervical into
529
thoracic) (1); C) complete homeotic transformation of cervical into thoracic vertebra (2) and
530
transitional sacral vertebra with thoracic rib on the right side and sacral rib on the other side
531
followed by sacral vertebra (3); D) Transitional sacral vertebra - first vertebra with thoracic rib
532
on the right side and sacral rib on the left side, followed by second transitional vertebra, with
533
sacral rib on the right side and no rib attached on the left side (transitional sacral vertebra with a
534
left side asymmetry) (4); Transformations are marked by arrows and showed in close-up at the
535
bottom of the Figure. 536 537 PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) Manuscript to be reviewed Manuscript to be reviewed 545
Table 2. 546
Overview of homeotic transformations observed in Triturus species. Number and percentage of
547
individuals with transitional vertebrae are given. Left and right asymmetries of transitional sacral
548
vertebra are shown separately. 546
Overview of homeotic transformations observed in Triturus species. Number and percentage of
547
individuals with transitional vertebrae are given. Left and right asymmetries of transitional sacral
548
vertebra are shown separately. 9
Cervical to thoracic
Transitional sacral
Thoracic to sacral
Species
Sample
size
Transitional
%
Complete
Incomplete
%
Left
Right
%
Incomplete
%
T. marmoratus
58
5
8.6
0
1
1.7
1
3
6.9
0
T. pygmaeus
55
1
1.8
0
0
1
0
1.8
0
T. ivanbureschi
361
25
6.9
1
3
1.1
6
12
5.0
3
0.8
T. karelinii
43
1
2.3
0
0
0
1
2.3
0
T. carnifex
123
8
6.5
0
0
6
2
6.5
0
T. macedonicus
226
14
6.2
0
1
0.4
8
4
5.3
1
0.4
T. cristatus
286
16
5.6
2
1
1.0
10
3
4.5
0
T. dobrogicus
216
13
6.0
0
0
7
6
6.0
0
Total
1368
83
6.1
3
6
0.6
39
31
4
0.3 549 PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) Manuscript to be reviewed 551
Table 3. 552
The number of thoracic vertebrae in T. cristatus × T. marmoratus F1 hybrids and parental
553
species. Modal numbers of thoracic vertebrae in vertebral formulae are shown in bold. Str –
554
percentage of individuals with transitional vertebrae at thoraco-sacral boundary. Tvar– percentage
555
of individuals with the complete number of thoracic vertebrae different from the modal number. 556
Species
Sample
size
Number of thoracic vertebrae
Str (%)
Tvar (%)
12 12.5 13 13.5 14 14.5 15 15.5 16 16.5 17 17.5 18
T. marmoratus
58
46
4
8
6.9
13.8
T. cristatus
122
1
6
1
98
5
11
4.9
14.8
Hybrids
68
2
1
39
4
16
6
7.4
35.3 552
The number of thoracic vertebrae in T. cristatus × T. marmoratus F1 hybrids and parental
553
species. Modal numbers of thoracic vertebrae in vertebral formulae are shown in bold. Str –
554
percentage of individuals with transitional vertebrae at thoraco-sacral boundary. Tvar– percentage
555
of individuals with the complete number of thoracic vertebrae different from the modal number. 556
Species
Sample
size
Number of thoracic vertebrae
Str (%)
Tvar (%)
12 12.5 13 13.5 14 14.5 15 15.5 16 16.5 17 17.5 18
T. marmoratus
58
46
4
8
6.9
13.8
T. cristatus
122
1
6
1
98
5
11
4.9
14.8
Hybrids
68
2
1
39
4
16
6
7.4
35.3 557 PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) Manuscript to be reviewed Manuscript to be reviewed 559
The number of individuals with regular and changed vertebral formulae in “central’’ and
560
“fringe” populations. 559
The number of individuals with regular and changed vertebral formulae in “central’’ and
560
“fringe” populations. 1
Number with modal vertebral
formula
Number with non-modal
vertebral formula
Species
Central
Fringe
Central
Fringe
G
p
T. ivanbureschi
150
110
25
76
18.86
***
T. carnifex
53
37
13
20
2.09
ns
T. macedonicus
51
122
16
37
0.006
ns
T. cristatus
98
130
24
34
0.032
ns
T. dobrogicus
42
67
15
92
6.65
**
Total
394
466
83
259
ns, not significant; ** p < 0.01; *** p < 0.001 561 562 Manuscript to be reviewed Manuscript to be reviewed Figure 1(on next page) Figure 1(on next page) Schematic presentation of the vertebral column in Triturus newts and homeotic
transformations scored Figure 1. The regionalization of the vertebral column in Triturus newts and schematic
presentation of scored homeotic transformations (example of T. marmoratus). A) vertebral
column without homeotic transformation and regular number of vertebrae – the first three
caudal vertebrae are shown; B) incomplete homeotic transformations of cervical vertebra
into thoracic; C) complete homeotic transformation of cervical into thoracic vertebra; D)
transitional sacral vertebra with thoracic rib at one side and sacral rib at the other side
followed by transitional vertebra with sacral rib at the one side and no rib at the other; E)
transitional thoraco-sacral vertebra, with thoracic rib at one side and sacral rib at the other,
followed by regular sacral vertebra. PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) Manuscript to be reviewed PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) Manuscript to be reviewed Manuscript to be reviewed Figure 2(on next page) Figure 2(on next page) Figure 2(on next page) Calibrated phylogeny for the genus Triturus with the modal number of vertebrae
indicated by colour code Figure 2. Calibrated phylogeny for the genus Triturus with the modal number of vertebrae
indicated by colour code (after Arntzen et al., 2015). PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) Manuscript to be reviewed Manuscript to be reviewed Figure 3(on next page) Figure 3(on next page) Distribution of Triturus species and geographic positions of populations Figure 3. Distribution of eight Triturus species across Europe and adjacent Asia. Studied
populations are marked by solid dots (central populations) and open dots (fringe
populations). For detailed information see Supplementary Table S1. PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) Manuscript to be reviewed Manuscript to be reviewed T. marmoratus
T. pygmaeus
T. karelinii
T. ivanbureschi
T. carnifex
T. macedonicus
T. cristatus
T. dobrogicus
25
20
15
10
5
30
35
Ma
12
13
14
15
17
0
PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) T. marmoratus
T. pygmaeus
T. karelinii
T. ivanbureschi
T. carnifex
T. macedonicus
T. cristatus
T. dobrogicus
25
20
15
10
5
30
35
Ma
12
13
14
15
17
0 T. marmoratus Ma PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) Manuscript to be reviewed Manuscript to be reviewed 4 X-ray images of homeotic transformations recorded Figure 4. Homeotic transformations recorded: A) complete vertebral column without
homeotic transformations and transitional vertebrae. B) transitional cervical vertebra
(cervical into thoracic) (1); C) complete homeotic transformation of cervical into thoracic
vertebra (2) and transitional sacral vertebra with thoracic rib on the right side and sacral rib
on the other side followed by sacral vertebra (3); D) Transitional sacral vertebra - first
vertebra with thoracic rib on the right side and sacral rib on the left side, followed by second
transitional vertebra, with sacral rib on the right side and no rib attached on the left side
(transitional sacral vertebra with a left side asymmetry) (4); Transformations are marked by
arrows and showed in close-up at the bottom of the Figure. PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) Manuscript to be reviewed Manuscript to be reviewed T. cristatus
T. dobrogicus
T. carnifex
T. macedonicus
T. karelinii
T. ivanbureschi
ygmaeus
marmoratus T. pygmaeus
T. marmoratus T. pygmaeus
T. marmoratus PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) Manuscript to be reviewed Manuscript to be reviewed Table 1(on next page) Table 1(on next page) The number of thoracic vertebrae in Triturus species Table 1. The number of thoracic vertebrae in Triturus species (central populations only). Modal numbers of thoracic vertebrae in vertebral formulae are shown in bold. Str –
percentage of individuals with transitional vertebrae at thoraco-sacral boundary. Tvar –
percentage of individuals with the complete number of thoracic vertebrae different from the
modal number. PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) Manuscript to be reviewed 1 2
Species
Sample
size
Number of thoracic vertebrae
Str (%)
Tvar (%)
12 12.5
13
13.5 14 14.5 15 15.5 16 16.5 17 17.5 18
T. marmoratus
58
46
4
8
6.9
13.8
T. pygmaeus
55
52
1
2
1.8
3.6
T. ivanbureschi
175
1
3
150
4
17
4.0
10.3
T. karelinii
43
1
40
2
2.3
4.7
T. carnifex
66
4
5
53
3
1
7.6
12.1
T. macedonicus
67
9
5
51
1
1
9.0
14.9
T. cristatus
122
1
6
1
98
5
11
4.9
14.8
T. dobrogicus
57
2
11
1
42
1
1.8
24.6 PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) Manuscript to be reviewed Manuscript to be reviewed Manuscript to be reviewed PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) Table 2(on next page) Table 2(on next page) Homeotic transformations observed in Triturus Table 2. Overview of homeotic transformations observed in Triturus species. Number and
percentage of individuals with transitional vertebrae are given. Left and right asymmetries of
transitional sacral vertebra are shown separately. Manuscript to be reviewed 1 2
Cervical to thoracic
Transitional sacral
Thoracic to sacral
Species
Sample
size
Transitional
%
Complete
Incomplete
%
Left
Right
%
Incomplete
%
T. marmoratus
58
5
8.6
0
1
1.7
1
3
6.9
0
T. pygmaeus
55
1
1.8
0
0
1
0
1.8
0
T. ivanbureschi
361
25
6.9
1
3
1.1
6
12
5.0
3
0.8
T. karelinii
43
1
2.3
0
0
0
1
2.3
0
T. carnifex
123
8
6.5
0
0
6
2
6.5
0
T. macedonicus
226
14
6.2
0
1
0.4
8
4
5.3
1
0.4
T. cristatus
286
16
5.6
2
1
1.0
10
3
4.5
0
T. dobrogicus
216
13
6.0
0
0
7
6
6.0
0
Total
1368
83
6.1
3
6
0.6
39
31
4
0.3 2 PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) Manuscript to be reviewed Table 3(on next page) Table 3(on next page) The number of thoracic vertebrae in T. cristatus × T. marmoratus hybrids Table 3. The number of thoracic vertebrae in T. cristatus × T. marmoratus F1 hybrids and
parental species. Modal numbers of thoracic vertebrae in vertebral formulae are shown in
bold. Str – percentage of individuals with transitional vertebrae at thoraco-sacral boundary. Tvar– percentage of individuals with the complete number of thoracic vertebrae different from
the modal number. PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) Manuscript to be reviewed Manuscript to be reviewed 1
2
Species
Sample
size
Number of thoracic vertebrae
Str (%)
Tvar (%)
12 12.5 13 13.5 14 14.5 15 15.5 16 16.5 17 17.5 18
T. marmoratus
58
46
4
8
6.9
13.8
T. cristatus
122
1
6
1
98
5
11
4.9
14.8
Hybrids
68
2
1
39
4
16
6
7.4
35.3 1 1 2 PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) Manuscript to be reviewed Manuscript to be reviewed Manuscript to be reviewed Table 4(on next page) Table 4(on next page) Table 4(on next page) The number of individuals with regular and changed vertebral formulae in central and
fringe populations of species with parapatric distributions Table 4. The number of individuals with regular and changed vertebral formulae in “central’’
and “fringe” populations. PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) Manuscript to be reviewed 1 2
Number with modal vertebral
formula
Number with non-modal
vertebral formula
Species
Central
Fringe
Central
Fringe
G
p
T. ivanbureschi
150
110
25
76
18.86
***
T. carnifex
53
37
13
20
2.09
ns
T. macedonicus
51
122
16
37
0.006
ns
T. cristatus
98
130
24
34
0.032
ns
T. dobrogicus
42
67
15
92
6.65
**
Total
394
466
83
259
ns, not significant; ** p < 0.01; *** p < 0.001 3 PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015) PeerJ reviewing PDF | (2015:08:6170:1:0:NEW 19 Oct 2015)
|
https://openalex.org/W4389835209
|
https://jsr.lib.ums.ac.id/index.php/ujp/article/download/104/87
|
Indonesian
| null |
LITERATURE REVIEW : EFEKTIVITAS BERAS HITAM (Oryza sativa L. indica) SEBAGAI ANTIDIABETES
|
Usadha Journal of Pharmacy
| 2,022
|
cc-by
| 6,177
|
Abstrak Diabetes merupakan penyakit kronis yang terjadi karena pankreas tidak dapat
menghasilkan cukup insulin atau insulin yang diproduksi tidak dapat digunakan secara
efektif oleh tubuh. Hiperglikemia atau glukosa darah yang tinggi merupakan efek utama
pada diabetes yang tidak terkontrol dan dalam jangka panjang menyebabkan kerusakan
serius pada saraf dan pembuluh darah. Pada tahun 2015 Indonesia menempati peringkat 7
sebagai negara penyandang diabetes mellitus terbanyak di dunia, dan diperkirakan akan
naik ke peringkat 6 pada tahun 2040. Saat ini tanaman obat banyak digunakan oleh
masyarakat dan pada saat yang sama banyak masyarakat berasumsi penggunaan tanaman
obat atau obat tradisional relatif lebih aman daripada obat sintetis. Beras berpigmen memiliki
potensi aktivitas antioksidan karena kandungan senyawa bioaktif yang tinggi. Beras hitam
mengandung pigmen antosianin yang termasuk komponen flavonoid, yaitu turunan
polifenol. Beberapa penelitian yang dilakukan secara in vitro, in vivo dan klinis pada manusia
menunjukkan antosianin dalam beras hitam mempunyai efek antidiabetes. Studi literatur ini
bertujuan untuk mengetahui efektivitas beras hitam (Oryza sativa L. indica) sebagai
antidiabetes. Studi literatur ini bersumber dari artikel yang ditelusuri melalui database
Google Scholar, dengan menggunakan kata kunci “black rice and antidiabetic” serta kata
kunci sekunder “beras hitam sebagai penurun kadar gula darah pada tikus” dan “beras
hitam sebagai penurun kadar gula darah pada mencit”. Kriteria inklusi yang digunakan
adalah artikel yang dipublikasi pada tahun 2011 – 2021 membahas uji in vivo beras hitam,
dan dapat diunduh full text. Dari hasil tersebut didapatkan 7 artikel yang memenuhi kriteria
dan digunakan sebagai tinjauan. Berdasarkan hasil analisis dari 7 artikel tersebut didapatkan
bahwa beras hitam memiliki efektivitas sebagai antidiabetes dengan beberapa mekanisme. Kata Kunci: beras hitam (Oryza sativa L. indica), antidiabetes, in vivo Usadha: Journal of Pharmacy. Vol. 1, No. 3, (2022). ISSN 2827-9905
Available online at: https://jsr.lib.ums.ac.id/index.php/ujp LITERATURE REVIEW: EFFECTIVENESS OF BLACK RICE (Oryza sativa L. indica)
AS ANTIDIABETIC Alya Abida1, Arini Fadhilah1*
1Laboratorium Farmakologi dan Farmasi Klinis, Fakultas Farmasi,
Universitas Muhammadiyah Surakarta, Jl A. Yani No 157, Sukoharjo, Indonesia
*E-mail: af993@ums.ac.id Abstract Diabetes is a chronic disease caused by either inability of pancreas producing insulin in
sufficient amount or the insulin cannot be utilized effectively by the body. Hyperglycemia or
high blood glucose is the main effect caused by the uncontrolled diabetes and in the long
term damages the nerves and blood vessels. In 2015 Indonesia was the world's 7th country
with the greatest number of diabetes mellitus cases and is expected to rise to the 6th by
2040. Medicinal plants currently are the most common therapy used by the community and
at the same time people consider the use of medicinal plants or traditional medicine to be
safer than synthetic drugs. Black rice has a high level of antioxidant because it contains a
high amount of bioactive compound. It also contains anthocyanin pigments that are also
component of flavonoids and polyphenol derivatives. Several research conducted in vivo, in
vitro, and clinically in humans, showed anthocyanins in the black rice have antidiabetic 375 Usadha: Journal of Pharmacy. Vol. 1, No. 3, (2022). ISSN 2827-9905
Available online at: https://jsr.lib.ums.ac.id/index.php/ujp effect. This literature study is aimed to apprehand the Effectivity of black rice (Oryza sativa
L. indica) as an antidiabetic. This literature study is based on articles browsed in the Google
Scholar database with keywords of "black rice and antidiabetic" “beras hitam sebagai
penurun kadar gula darah pada tikus” and “beras hitam sebagai penurun kadar gula darah
pada mencit”. The inclusion criteria of the literatures are articles published in 2011-2021
discussing the in vivo research on black rice and downloadable full text. Seven articles that
meet the criteria were obtained and studied. Based on the studies, it was found that black
rice is effective as an antidiabetic with several mechanisms. Keywords: black rice (Oryza sativa L. indica), antidiabetic, in vivo 375 Usadha: Journal of Pharmacy. Vol. 1, No. 3, (2022). ISSN 2827-9905
Available online at: https://jsr.lib.ums.ac.id/index.php/ujp
effect. This literature study is aimed to apprehand the Effectivity of black rice (Oryza sativa
L. indica) as an antidiabetic. This literature study is based on articles browsed in the Google
Scholar database with keywords of "black rice and antidiabetic" “beras hitam sebagai
penurun kadar gula darah pada tikus” and “beras hitam sebagai penurun kadar gula darah
pada mencit”. The inclusion criteria of the literatures are articles published in 2011-2021
discussing the in vivo research on black rice and downloadable full text. Seven articles that
meet the criteria were obtained and studied. Based on the studies, it was found that black
rice is effective as an antidiabetic with several mechanisms. PENDAHULUAN Diabetes adalah penyakit kronis yang disebabkan oleh pankreas tidak dapat
menghasilkan cukup insulin atau insulin yang diproduksi tidak dapat digunakan secara efektif
oleh tubuh. Hiperglikemia, atau glukosa darah yang tinggi adalah efek utama pada diabetes
yang tidak terkontrol dan dalam jangka panjang menyebabkan kerusakan serius pada saraf dan
pembuluh darah (Kurniawaty, 2014). Diabetes mellitus merupakan penyakit kronis yang tidak
dapat disembuhkan dan dapat menimbulkan komplikasi penyakit kronis lain. Morbiditas dan
mortalitas dari diabetes mellitus juga terus berkembang di seluruh dunia dalam beberapa
dekade terakhir ini. Orang-orang di negara berkembang rentan terkena penyakit diabetes
mellitus. Saat ini diabetes mellitus merupakan penyakit yang harus diberi perhatian ketat karena
meningkatnya angka kecacatan yang disebabkan oleh diabetes dan meningkatnya kematian
dini serta dapat menurunkan kualitas hidup seseorang (Zhao et al., 2014). Diabetes mellitus sebagai permasalahan global terus meningkat prevalensinya dari
tahun ke tahun baik di dunia maupun di Indonesia. Berdasarkan data International Diabetes
Federation (IDF) prevalensi diabetes mellitus global pada tahun 2019 diperkirakan 9,3% (463
juta orang), naik menjadi 10,2 % (578 juta orang) pada tahun 2030 dan 10,9% (700 juta orang)
pada tahun 2045. Pada tahun 2015 Indonesia menempati peringkat 7 sebagai negara
penyandang diabetes mellitus terbanyak di dunia, dan diperkirakan akan naik ke peringkat 6
pada tahun 2040 (Widiastuti, 2020). Hal ini menunjukkan peningkatan prevalensi diabetes
mellitus yang cukup berarti. Masyarakat sekarang ini banyak menggunakan tanaman obat. Bahkan dari waktu ke waktu perkembangan tanaman obat semakin meningkat, terlebih dengan
munculnya tren kembali ke alam (back to nature). Pada saat yang sama banyak masyarakat
berasumsi penggunaan tanaman obat atau obat tradisional relatif lebih aman daripada obat
sintetis (Herman dkk, 2019). Beras berpigmen memiliki potensi aktivitas antioksidan karena kandungan senyawa
bioaktif yang tinggi. Senyawa bioaktif pada beras berpigmen dapat mengurangi stres oksidatif,
penyakit kardiovaskular, mencegah kanker, komplikasi diabetes, dan lainnya (Arifin et al.,
2019). Beras hitam mengandung pigmen antosianin yang termasuk komponen flavonoid, yaitu
turunan polifenol (Wijayanti, 2016). Beberapa penelitian yang dilakukan secara in vitro, in vivo
dan klinis pada manusia menunjukkan antosianin dalam beras hitam mempunyai efek
antidiabetes. Data yang dipublikasikan menunjukkan bahwa antosianin dapat menurunkan
glukosa darah dengan mekanisme aksi meningkatkan sensitivitas insulin, melindungi sel,
meningkatkan sekresi insulin dan mengurangi pencernaan gula di usus halus. Mekanisme aksi
yang utama terkait dengan sifat antioksidannya, tetapi penghambatan enzimatik dan jalur 376 Usadha: Journal of Pharmacy. Vol. 1, No. 3, (2022). ISSN 2827-9905
Available online at: https://jsr.lib.ums.ac.id/index.php/ujp
lainnya mungkin juga relevan (Thanuja and Parimalavalli, 2018). METODE PENELITIAN Penyusunan studi literatur ini dilakukan dengan pencarian literatur secara online melalui
database Google Scholar. Pencarian artikel efektivitas antidiabetes beras hitam sebagai
antidiabetes menggunakan kata kunci “black rice and antidiabetic” serta kata kunci sekunder
“beras hitam sebagai penurun kadar gula darah pada tikus” dan “beras hitam sebagai penurun
kadar gula darah pada mencit”. Pencarian artikel dilakukan dengan menetapkan beberapa
kriteria inklusi. Kriteria inklusi yang ditetapkan adalah artikel primer yang dipublikasi pada tahun
2011-2021, membahas uji in vivo beras hitam, dan dapat diunduh full text. Skrining dilakukan
dengan membaca judul dan abstrak, kemudian artikel yang diperoleh dilakukan pengunduhan
file full text dan diidentifikasi kelayakannya secara manual. Kelayakan potensi yang diidentifikasi
dari artikel tersebut adalah yang membahas aktivitas dari beras hitam sebagai antidiabetes dan
dilakukan secara in vivo. Artikel yang membahas beras hitam tetapi menjelaskan aktivitas selain
antidiabetes atau tidak membahas penurunan glukosa darah, serta penelitian antidiabetes yang
dilakukan secara in vitro tidak termasuk ke dalam kriteria. Dari beberapa artikel tersebut
kemudian dilakukan ekstraksi data pada masing-masing artikel seperti berdasarkan autor,
bentuk sampel dari beras hitam, metode pengukuran glukosa darah, agen diabetagonik, hasil
akhir dari penelitian tersebut serta signifikansi dari hasil akhir penelitian. Artikel yang sudah di
ekstraksi tersebut dikaji mengenai efektivitas beras hitam (Oryza sativa L. indica) sebagai
antidiabetes. PENDAHULUAN Studi literatur ini bertujuan
untuk mengkaji efektivitas beras hitam (Oryza sativa L. indica) dalam mengatasi kadar gula
darah. Usadha: Journal of Pharmacy. Vol. 1, No. 3, (2022). ISSN 2827-9905
Available online at: https://jsr.lib.ums.ac.id/index.php/ujp Usadha: Journal of Pharmacy. Vol. 1, No. 3, (2022). ISSN 2827-9905
Available online at: https://jsr.lib.ums.ac.id/index.php/ujp lainnya mungkin juga relevan (Thanuja and Parimalavalli, 2018). Studi literatur ini bertujuan
untuk mengkaji efektivitas beras hitam (Oryza sativa L. indica) dalam mengatasi kadar gula
darah. HASIL DAN PEMBAHASAN Berdasarkan pencarian pada database Google Scholar dengan 3 kata kunci yaitu “black
rice and antidiabetic“, “beras hitam sebagai penurun kadar gula darah pada tikus”, dan “beras
hitam sebagai penurun kadar gula darah pada mencit” masing-masing muncul sebanyak
15.100, 1.220 dan 1.370 artikel yang mengandung kata kunci tersebut, kemudian setelah
dilakukan identifikasi berdasarkan judul dan abstraknya, seleksi berdasarkan kriteria inklusinya,
untuk kata kunci “black rice and antidiabetic“ diperoleh sebanyak 5 jurnal, lalu untuk kata kunci
“beras hitam sebagai penurun kadar gula darah pada tikus” dan “beras hitam sebagai penurun
kadar gula darah pada mencit” diperoleh masing-masing sebanyak 1 jurnal, sehingga diperoleh
7 artikel. Ringkasan dari 7 artikel yang membahas beras hitam berkhasiat sebagai antidiabetes
dijelaskan pada Tabel 1. Secara umum terdapat tiga macam beras, yaitu beras putih (Oryza sativa), beras merah
(Oryza glaberrima) dan beras hitam (Oryza sativa L. indica). Beras hitam memiliki indeks
glikemik yang rendah sehingga dapat berguna dalam pengaturan kadar gula darah dan
produksi insulin. Beras hitam juga memiliki kandungan serat 100 kali lipat dari beras putih
sehingga sangat menyehatkan (Istiarini, 2016). Selain kadar seratnya yang tinggi beras hitam
juga mengandung senyawa metabolit sekunder yaitu antosianin yang tinggi yang memiliki
aktivitas sebagai antioksidan (Yuda et al., 2020). 377 Usadha: Journal of Pharmacy. Vol. 1, No. 3, (2022). ISSN 2827-9905
Available online at: https://jsr.lib.ums.ac.id/index.php/ujp Tabel 1. Hasil tinjauan artikel tentang efektivitas beras hitam sebagai antidiabetes
No
Referensi
Sampel
Metode
Pengukuran
Glukosa
Darah
Penginduksi
Hasil
Signifikansi
1
Almundarij et al. (2020)
EEBH
100 mg/kgBB
dan 20% beras hitam
dalam pakan
Glukometer
STZ 75
mg/kgBB
KGD ↓ signifikan
dibandingkan
kelompok model
tikus DM tipe 2
P< 0.001
2
Chaiyasut et al. (2016)
EEBH berkecambah
500 dan
1000 mg/kgBB
Kolorimetri
enzimatik
STZ 45
mg/kgBB
KGD kelompok pra
perlakuan dan
pasca perlakuan ↓
signifikan pada
minggu ke 12
P<0,05
OGTT
Glukometer
Larutan
glukosa 2
g/kgBB
Sediaan uji tidak
mempengaruhi
toleransi glukosa
-
3
El-Kholie et al. (2018)
EABH dan EEBH (500
dan 1000 mg/kgBB),
TBH (2,5% dan 5%
dari ransum)
Metode kinetik
dengan kit
Aloksan 150
mg/kgBB
KGD ↓ signifikan
dibandingkan
kelompok DM
dengan penurunan
paling baik pada
EEBH dosis 1000
mg/kgbb
P<0,05
4
Tantipaiboonwong
et al. (2017)
EEBH
10, 50 dan
100 mg/kgBB
Kolorimetri
enzimatik
STZ 75
mg/kgBB
KGD ↓ signifikan
dibandingkan
kelompok DM pada
dosis 50 dan 100
mg/kgbb
P<0,05
5
Yuda et al. HASIL DAN PEMBAHASAN Beras hitam juga semakin popular di masyarakat
dan dikonsumsi sebagai pangan fungsional karena manfaatnya bagi kesehatan (Mangiri et al.,
2016). Berdasarkan hal tersebut beras hitam berpotensi sebagai agen antidiabetes, seperti
yang tertera pada Tabel 1 menunjukkan bahwa beras hitam memiliki khasiat sebagai
antidiabetes. Almundarij et al (2020) melakukan penelitian untuk mengevaluasi penggantian beras
putih dengan varietas lain dari beras berpigmen yaitu beras merah, beras hitam dan beras
coklat dalam model tikus diabetes tipe 1, dengan STZ 75 mg/kgBB sebagai penginduksi DM. Sampel yang digunakan adalah EEBH dengan dosis 100 mg/kgBB dan pakan yang
mengandung 20% beras hitam utuh. Penelitian ini menggunakan 110 ekor tikus wistar jantan
yang dilakukan selama 1 bulan. Pada uji statistik ANOVA, kelompok perlakuan EEBH dengan
dosis 100 mg/kgBB dan 20% beras hitam dalam pakan dapat menurunkan KGD yang berbeda
secara signifikan (P<0,001) dibandingkan kelompok model tikus diabetes tipe 1 dan
menunjukkan KGD paling kecil diantara beras lainnya yaitu 132, 66 mg/dL dan 137, 55 mg/dL. Chaiyasut et al. (2016) melakukan penelitian untuk mengetahui apakah ekstrak beras
hitam berkecambah dari Thailand melindungi model tikus diabetes dari DM dan konsekuensi
lain terkait diabetes dengan STZ 45 mg/kgBB sebagai penginduksi DM. Sampel yang
digunakan adalah EEBH berkecambah dari Thailand dengan dosis 500 mg/kgBB dan 1000
mg/kgBB yang dibuat dengan cara beras hitam dikecambahkan dalam air kemudian
diekstraksi dengan larutan asam asetat 4% dalam etanol 50%. Penelitian ini menggunakan
tikus wistar jantan dewasa sebanyak 72 ekor selama 12 minggu. Pada uji statistik ANOVA,
KGD berkurang selama 8 minggu percobaan pada kelompok pra-perlakuan yaitu tikus diberi
ekstrak sebelum dan sesudah diabetes, dan pada minggu ke 12 KGD berkurang secara
signifikan (P<0,05) dibandingkan dengan kelompok DM. Pada kelompok pasca perlakuan yaitu
tikus diberi ekstrak setelah diabetes, nilai KGD berkurang dibandingkan dengan kelompok DM
pada minggu ke 12 masa percobaan dan memiliki signifikansi (P<0,05). EEBH berkecambah
dosis 1000 mg/kgBB memberikan penurunan KGD yang lebih besar di kedua prosedur pra
perlakuan dan pasca perlakuan. Pada metode OGTT, EEBH berkecambah dapat menurunkan
tingkat toleransi glukosa/penurunan area under the curve (AUC). Penurunan AUC tidak
dipengaruhi oleh dosis pada kelompok pra-perlakuan, sedangkan pada kelompok pasca-
perlakuan penurunan AUC dipengaruhi oleh dosis. El-Kholie et al. (2018) melakukan penelitian untuk mengetahui efek antidiabetes beras
hitam dan ekstraknya pada tikus diabetes dengan aloksan 150 mg/kgBB sebagai penginduksi
DM. HASIL DAN PEMBAHASAN (2019)
IBH
10, 50 dan
100%
OGTT
Glukometer
Larutan
glukosa 50%
KGD ↓ paling balik
pada IBH 50%
-
6
Daeli et al. (2018)
Nasi beras hitam
14,8 g + pakan
standar COMFEED
AD II 5,2 g
GOD-PAP
STZ 50
mg/kgBB +
NA 110
mg/kg BB
KGD ↓ sebesar
62,55% pada hari
ke 28
P=0,000
7
Septiyaningtyas
and Agustini
(2021)
Yeast beras hitam
Glukometer
Pakan tinggi
lemak +
Fruktosa 55
%
KGD ↓ dan berbeda
signifikan terhadap
kontrol negatif
P= 0,021
(P<0,05)
Catatan : EEBH : Ekstrak etanol beras hitam; TBH : Tepung beras hitam; EABH : Ekstrak air beras hitam;
IBH : Infusa beras hitam; STZ : Streptozotocin; NA : Nikotinamid, KGD : Kadar glukosa darah; OGTT :
Oral glucose tolerance test; DM : Diabetes mellitus. abel 1. Hasil tinjauan artikel tentang efektivitas beras hitam sebagai antidiabetes Antosianin merupakan kelompok polifenol yang memiliki aktivitas sebagai antioksidan
yang kuat, antidiabetes, anti-inflamasi, dan mampu mengendalikan obesitas (Hani, 2020). Beras
hitam (Oryza sativa L. indica) termasuk famili Poaceae yang merupakan tanaman pangan dan 378 Usadha: Journal of Pharmacy. Vol. 1, No. 3, (2022). ISSN 2827-9905
Available online at: https://jsr.lib.ums.ac.id/index.php/ujp
sudah dikonsumsi sebagai makanan pokok terutama oleh kaum bangsawan atau kerajaan. Ekstrak beras hitam juga dimanfaatkan untuk dikonsumsi sebagai makanan sehat di China dan
negara Asia Timur lainnya (Yuda et al., 2020). Beras hitam juga semakin popular di masyarakat
dan dikonsumsi sebagai pangan fungsional karena manfaatnya bagi kesehatan (Mangiri et al.,
2016). Berdasarkan hal tersebut beras hitam berpotensi sebagai agen antidiabetes, seperti
yang tertera pada Tabel 1 menunjukkan bahwa beras hitam memiliki khasiat sebagai
antidiabetes. Usadha: Journal of Pharmacy. Vol. 1, No. 3, (2022). ISSN 2827-9905
Available online at: https://jsr.lib.ums.ac.id/index.php/ujp Usadha: Journal of Pharmacy. Vol. 1, No. 3, (2022). ISSN 2827-9905
Available online at: https://jsr.lib.ums.ac.id/index.php/ujp sudah dikonsumsi sebagai makanan pokok terutama oleh kaum bangsawan atau kerajaan. Ekstrak beras hitam juga dimanfaatkan untuk dikonsumsi sebagai makanan sehat di China dan
negara Asia Timur lainnya (Yuda et al., 2020). Beras hitam juga semakin popular di masyarakat
dan dikonsumsi sebagai pangan fungsional karena manfaatnya bagi kesehatan (Mangiri et al.,
2016). Berdasarkan hal tersebut beras hitam berpotensi sebagai agen antidiabetes, seperti
yang tertera pada Tabel 1 menunjukkan bahwa beras hitam memiliki khasiat sebagai
antidiabetes. sudah dikonsumsi sebagai makanan pokok terutama oleh kaum bangsawan atau kerajaan. Ekstrak beras hitam juga dimanfaatkan untuk dikonsumsi sebagai makanan sehat di China dan
negara Asia Timur lainnya (Yuda et al., 2020). Usadha: Journal of Pharmacy. Vol. 1, No. 3, (2022). ISSN 2827-9905
Available online at: https://jsr.lib.ums.ac.id/index.php/ujp Usadha: Journal of Pharmacy. Vol. 1, No. 3, (2022). ISSN 2827-9905
Available online at: https://jsr.lib.ums.ac.id/index.php/ujp Usadha: Journal of Pharmacy. Vol. 1, No. 3, (2022). ISSN 2827-9905
Available online at: https://jsr.lib.ums.ac.id/index.php/ujp Usadha: Journal of Pharmacy. Vol. 1, No. 3, (2022). ISSN 2827-9905
Available online at: https://jsr.lib.ums.ac.id/index.php/ujp Tantipaiboonwong et al. (2017) melakukan penelitian efek anti-hiperglikemik dan
antihiperlipidemia beras hitam dan merah pada tikus diabetes yang diinduksi streptozotocin
dengan dosis 75 mg/kgBB. Sampel yang digunakan adalah EEBH dengan dosis 10, 50 dan
100 mg/kgBB. Penelitian ini menggunakan tikus wistar albino sebanyak 54. EEBH diberikan
secara oral selama 8 minggu. Pada uji statistik ANOVA, nilai KGD kelompok EEBH dosis 50
dan 100 mg/kgBB tidak menunjukkan penurunan KGD setelah 4 minggu pengobatan, namun
pada akhir minggu ke-8 KGD menurun pada kelompok tersebut yaitu masing-masing 1035
dan 1017 mg/dL dan berbeda signifikan (P<0,05) dibandingkan dengan kelompok DM dengan
kadar glukosa darah 1205 mg/dL. Hasilnya menunjukkan bahwa konsumsi jangka panjang dari
dosis tinggi EEBH dapat menurunkan KGD pada model tikus diabetes yang diinduksi STZ dosis tinggi EEBH dapat menurunkan KGD pada model tikus diabetes yang diinduksi STZ. Yuda et al. (2019) melakukan penelitian uji aktivitas antidiabetes infusa beras hitam
(Oryza sativa L. indica) dengan metode toleransi glukosa dengan memberikan larutan glukosa
50% secara oral kepada hewan uji. Sampel yang digunakan pada penelitian ini adalah infusa
beras hitam dengan konsentrasi 10 %, 50% dan 100%. Pada penelitian ini digunakan 15 ekor
tikus putih jantan (Rattus norvegicus) galur Sprague Dawley sebagai hewan uji. Sebelum
perlakuan, KGD tikus diukur terlebih dahulu menggunakan glukometer. Selanjutnya setiap
kelompok diberi sediaan uji secara oral, yaitu kelompok kontrol diberi Na-CMC, kelompok
kontrol positif diberi glibenklamid dan 3 kelompok infusa beras hitam konsentrasi 10 %, 50%
dan 100%. Setelah 30 menit, semua tikus diberi larutan glukosa 50% secara oral, kemudian
dilakukan pengukuran kadar glukosa darah kembali pada menit ke- 30, 60, 90 dan 120 dengan
menggunakan glukometer. Pada menit ke-0, perbedaan kadar glukosa darah setiap kelompok
tidak terlalu signifikan (<170 mg/dL), namun kadar glukosa tiap kelompok meningkat ketika
diberikan larutan glukosa 50% pada menit ke-30. Peningkatan kadar glukosa darah terjadi
karena glukosa yang diberi secara oral akan diserap oleh usus halus dan masuk ke aliran
darah. Pada menit ke-60, hampir semua kelompok mengalami penurunan kadar glukosa
darah, kemudian pada menit ke-90 dan 120, kadar glukosa darah paling rendah ditunjukkan
oleh perlakuan dengan infusa beras hitam 50%. HASIL DAN PEMBAHASAN Peneliti menggunakan 3 sampel yaitu EEBH dan EABH dengan dosis 500 dan 1000
mg/kgBB serta TBH 2,5 % dan 5 % dari berat ransum. Penelitian ini dilakukan dengan desain
eksperimen 48 ekor tikus dewasa jantan albino Sprague Dawley. Masa percobaan pada
penelitian ini berlangsung selama 28 hari. Pada uji statistik ANOVA, nilai rata-rata KGD pada
kelompok EEBH, EABH dan TBH terjadi penurunan yang signifikan (P<0.05) dibandingkan
dengan kelompok DM. Hasil terbaik ditunjukkan oleh kelompok EEBH dengan dosis 1000
mg/kgBB dengan kadar glukosa darahnya yaitu 99 mg/dL. 379 Usadha: Journal of Pharmacy. Vol. 1, No. 3, (2022). ISSN 2827-9905
Available online at: https://jsr.lib.ums.ac.id/index.php/ujp Usadha: Journal of Pharmacy. Vol. 1, No. 3, (2022). ISSN 2827-9905
Available online at: https://jsr.lib.ums.ac.id/index.php/ujp Hasil ini menunjukkan bahwa infusa beras
hitam memberikan toleransi glukosa terbaik pada konsentrasi 50% dan dapat disimpulkan
bahwa infusa beras hitam mempengaruhi toleransi glukosa dengan pola yang hampir sama
dengan glibenklamid. Daeli et al. (2018) melakukan uji pengaruh pemberian nasi beras merah (Oryza nivara)
dan nasi beras hitam (Oryza sativa L. indica) terhadap perubahan kadar glukosa darah dan
trigliserida tikus wistar (Rattus norvegicus) diabetes mellitus tipe 2 dengan STZ 50 mg/kgBB
dan NA 110 mg/kgBB sebagai penginduksi DM. Peneliti menggunakan pakan sebagai sampel
uji yang dibuat dengan beras hitam dimasak terlebih dahulu hingga menjadi nasi, kemudian
dicampur dan dihomogenkan dengan pakan standar COMFEED AD II dengan jumlah nasi
beras hitam dan pakan standar sebesar 14,8 gram dan 5,2 gram. Rancangan penelitian ini
simple random sampling dengan tikus jantan galur wistar sebanyak 24 ekor yang dibagi
menjadi 4 kelompok yaitu kelompok kontrol normal, kontrol DM, perlakuan 1 (P1) (beras
merah), dan perlakuan 2 (P2) (beras hitam). Penelitian dilakukan selama 28 hari. Hasil
penelitian ini menunjukkan bahwa penurunan kadar glukosa darah terjadi lebih besar pada
kelompok P2 dengan pemberian nasi beras hitam yaitu terjadi penurunan kadar glukosa darah
puasa (GDP) sebesar 62.55% dengan kadar glukosa darahnya menjadi 98,9 mg/dL dan 380 Usadha: Journal of Pharmacy. Vol. 1, No. 3, (2022). ISSN 2827-9905
Available online at: https://jsr.lib.ums.ac.id/index.php/ujp Usadha: Journal of Pharmacy. Vol. 1, No. 3, (2022). ISSN 2827-9905
Available online at: https://jsr.lib.ums.ac.id/index.php/ujp
berdasarkan uji ANOVA yang dilakukan terdapat perbedaan rata-rata GDP yang signifikan
(P=0,000) pada hari ke 28. Usadha: Journal of Pharmacy. Vol. 1, No. 3, (2022). ISSN 2827-9905
Available online at: https://jsr.lib.ums.ac.id/index.php/ujp Usadha: Journal of Pharmacy. Vol. 1, No. 3, (2022). ISSN 2827-9905
Available online at: https://jsr.lib.ums.ac.id/index.php/ujp berdasarkan uji ANOVA yang dilakukan terdapat perbedaan rata-rata GDP yang signifikan
(P=0,000) pada hari ke 28. Septiyaningtyas dan Agustini (2021) melakukan uji pengaruh jenis yeast terhadap
kadar glukosa darah mencit (Mus musculus) diabetes mellitus tipe 2 yang diinduksi pakan
tinggi lemak dan fruktosa 55%. Sampel dalam penelitian ini menggunakan yeast beras putih,
beras merah, dan beras hitam yang dibuat dengan cara tepung dari masing-masing beras
difermentasi. Dari fermentasi tersebut dihasilkan yeast ekstrak yang kemudian disterilisasi
dengan autoklaf bersuhu 115 ˚C selama ±15 menit dan menghasilkan supernatan dan pasta
fermentasi. Penelitian ini menggunakan desain eksperimen Post test Only Control Group
Design dengan mencit jantan yang berumur ±4 minggu sebagai hewan uji. Sampel Yeast dari
masing-masing beras diberikan terhadap mencit sebanyak 1 mL/hari selama 7 hari. Usadha: Journal of Pharmacy. Vol. 1, No. 3, (2022). ISSN 2827-9905
Available online at: https://jsr.lib.ums.ac.id/index.php/ujp Setiap 7
hari dilakukan pengukuran KGD mencit. Data yang dihasilkan tersebut kemudian dianalisis
mengunakan SPSS uji non parametrik Kruskall Wallis. Didapatkan hasil P=0,002 (P<0,05) pada
kadar glukosa darah sehingga diketahui adanya pengaruh pemberian jenis yeast terhadap
kadar glukosa darah mencit. Selanjutnya dilakukan uji lanjut menggunakan Mann Whitney U
Test untuk mengetahui perbedaan pengaruh pemberian yeast beras hitam terhadap kontrol
negative (tikus DM diberi Na-CMC), didapatkan perbedaan secara signifikan P=0,021
(P<0,05). Kadar glukosa darah pada mencit yang diberi yeast beras hitam mencapai 51 mg/dL
yang termasuk ke dalam hipoglikemia. Hal seperti ini dapat disebabkan oleh beberapa faktor
diantaranya faktor stres pada mencit dan kurangnya kesesuaian dosis dalam pemberian
perlakuan. Berdasarkan hal tersebut yeast beras hitam berpengaruh terhadap penurunan
kadar glukosa darah dari mencit yang telah terindikasi diabetes melitus tipe 2. Senyawa Antidiabetes dalam Beras Hitam Aktivitas antidiabetes dalam beras hitam disebabkan oleh kandungan pigmen antosianin
yang merupakan komponen utama dalam beras hitam (Yuda et al., 2020). Antosianin adalah
senyawa kimia organik yang memiliki warna merah, biru, dan hitam biasanya terdapat pada
tanaman. Antosianin merupakan pigmen alami yang termasuk golongan flavonoid, senyawa
flavonoid mempunyai kerangka karbon yaitu dua cincin benzene tersambung dengan rantai
alifatik tiga karbon (Pasaribu et al., 2021). Antosianin utama dalam beras hitam adalah cyanidin-
3-glucoside (C3G) yang merupakan sumber antosianin penting (Mangiri et al., 2016). Senyawa
cyanidin-3-glucoside (C3G) adalah 93% dari total antosianin dalam beras hitam (El-kholie et al.,
2018). Studi literatur yang dilakukan oleh Hani (2020) menunjukkan bahwa cyanidin-3-
glucoside (C3G) merupakan senyawa antosianin yang memiliki aktivitas antidiabetes. 381 Usadha: Journal of Pharmacy. Vol. 1, No. 3, (2022). ISSN 2827-9905
Available online at: https://jsr.lib.ums.ac.id/index.php/ujp Usadha: Journal of Pharmacy. Vol. 1, No. 3, (2022). ISSN 2827-9905
Available online at: https://jsr.lib.ums.ac.id/index.php/ujp Usadha: Journal of Pharmacy. Vol. 1, No. 3, (2022). ISSN 2827-9905
Available online at: https://jsr.lib.ums.ac.id/index.php/ujp Gambar 1. Struktur senyawa cyanidin -3-glucoside (https://pubchem.ncbi.nlm.nih.gov/) Gambar 1. Struktur senyawa cyanidin -3-glucoside (https://pubchem.ncbi.nlm.nih.gov/) Antosianin bersifat sebagai antioksidan yang berefek positif bagi kesehatan (Mangiri et
al., 2016). Antioksidan merupakan senyawa yang dapat menghambat reaksi oksidasi atau zat
yang dapat menetralkan atau menangkap radikal bebas, dimana molekul tersebut memicu
kerusakan sel dan dapat meningkatkan risiko kanker dan penyakit jantung. Seseorang yang
menderita DM diketahui mengalami peningkatan stres oksidatif. Stres oksidatif dapat
menyebabkan apoptosis sel-pankreas sehingga terjadi resistensi insulin. Tingginya kadar
antosianin pada beras hitam diketahui mampu mencegah kerusakan pada sel-pankreas serta
menekan apoptosis sel-pankreas tanpa mengubah proliferasi dari sel tersebut (Daeli et al.,
2018). Mekanisme Kerja Beras Hitam Sebagai Antidiabetes
Meregenerasi Sel Beta di Pankreas Salah satu penyebab kejadian hiperglikemia pada diabetes mellitus tipe 2 (DMT2) yaitu
pada sel beta pankreas terjadi kegagalan untuk mensekresikan insulin yang cukup. Pada
perjalanan penyakit DMT2, sel beta pankreas mengalami penurunan fungsi dan terjadi
peningkatan resistensi insulin yang berlanjut sehingga terjadi hiperglikemia kronik dengan
segala dampaknya. Hiperglikemia kronik juga memperburuk disfungsi sel beta pankreas
(Decroli, 2019). Flavonoid yang tinggi dalam beras hitam merupakan agen antidiabetik yang
memiliki efek meregenerasi sel beta, dan memungkinkan untuk meningkatkan pelepasan
insulin. Stimulasi pelepasan insulin oleh flavonoid disebabkan oleh peningkatan konsentrasi
Ca2+ di sel pulau Langerhans (Almundarij et al., 2020). Menghambat Pelepasan Glukosa Selama Pencernaan Serat pangan merupakan komponen utama penyusun dinding sel tanaman seperti pada
sayur, buah, serealia, serta umbi (Daeli et al., 2018). Serat merupakan komponen dinding sel
tanaman yang tidak dapat dicerna oleh sistem pencernaan manusia. Serat dapat merangsang
alat cerna agar mendapat cukup getah cerna, membentuk volume sehingga menimbulkan rasa
kenyang dan membantu pembentukan feses (Nurul et al., 2019). Serat disebut memiliki efek
hipoglikemik karena mampu memperlambat pengosongan lambung, memperlambat difusi
glukosa, mengubah peristaltik lambung, menurunkan aktifitas amilase akibat meningkatnya
viskositas isi usus, serta menurunkan waktu transit yang mengakibatkan pendeknya absorbsi
glukosa dan berpengaruh terhadap peningkatan sekresi insulin dan pemakaian glukosa oleh 382 382 Usadha: Journal of Pharmacy. Vol. 1, No. 3, (2022). ISSN 2827-9905
Available online at: https://jsr.lib.ums.ac.id/index.php/ujp Usadha: Journal of Pharmacy. Vol. 1, No. 3, (2022). ISSN 2827-9905
Available online at: https://jsr.lib.ums.ac.id/index.php/ujp
sel hati. Nasi beras hitam memiliki kandungan serat dan antosianin yang lebih tinggi Usadha: Journal of Pharmacy. Vol. 1, No. 3, (2022). ISSN 2827-9905
Available online at: https://jsr.lib.ums.ac.id/index.php/ujp Usadha: Journal of Pharmacy. Vol. 1, No. 3, (2022). ISSN 2827-9905
Available online at: https://jsr.lib.ums.ac.id/index.php/ujp Usadha: Journal of Pharmacy. Vol. 1, No. 3, (2022). ISSN 2827-9905
Available online at: https://jsr.lib.ums.ac.id/index.php/ujp sel hati. Nasi beras hitam memiliki kandungan serat dan antosianin yang lebih tinggi
dibandingkan dengan nasi beras merah (Daeli et al., 2018). Kandungan serat yang tinggi dalam beras hitam tidak banyak memengaruhi kadar
glukosa darah karena serat dapat menghambat pelepasan glukosa ke dalam darah (Buchari,
2019). Indeks glikemik (IG) yang rendah serta kandungan serat diketahui mampu meningkatkan
sensitivitas insulin sehingga glukosa dalam darah mampu diserap masuk ke dalam sel (Daeli et
al., 2018). Meningkatkan Kadar Transporter Glukosa 1 (GLUT 1) dan 4 (GLUT 4) Karbohidrat yang masuk ke dalam tubuh kemudian dicerna dan diserap di dalam usus
dan masuk ke dalam darah dan didistribusikan ke berbagai jaringan tubuh. Proses distribusi
tersebut melibatkan Glucose transporters (GLUT) yaitu suatu famili protein transport yang
bekerja sebagai pengangkut untuk memindahkan glukosa melewati membran sel. Bentuk
transporter glukosa yang telah diketahui dengan baik adalah GLUT 1-4. GLUT 4 adalah suatu
protein spesifik yang memfasilitasi transpor glukosa. GLUT 4 bertanggung jawab atas
transportasi glukosa ke dalam otot, jantung, otak dan jaringan adiposa. GLUT 4 merupakan
transporter glukosa utama dan dapat meregulasi glukosa yang distimulasi insulin yang
disekresikan oleh sel beta sebagai sensor glukosa. Insulin berikatan dengan reseptor insulin
yang terdapat pada permukaan membran sel kemudian akan menghasilkan biosignal seluler. Signal ini akan menstimulasi GLUT 4 yang ada di dalam badan golgi untuk translokasi bergerak
menuju membran plasma. Selanjutnya GLUT 4 yang berikatan dengan membran plasma akan
menjadi pintu masuk glukosa ke dalam sel. GLUT 1 tersebar pada otak dan eritrosit. GLUT 1
sampai GLUT 3 dalam keadaan fisiologis yang normal sudah berasosiasi dengan membran
plasma sehingga mekanisme kerja ambilan glukosa tidak bergantung pada insulin (Utama and
Arjentinia, 2014). Dilaporkan bahwa peningkatan regulasi GLUT 4 pada penderita diabetes dapat
meningkatkan sensitivitas insulin (Tantipaiboonwong et al., 2017). Berbagai polifenol
mempromosikan translokasi GLUT4 di otot rangka dengan cara mengaktifkan 5′-adenosine
monophosphate-activated protein kinase (AMPK) (Kang et al., 2017). AMPK merupakan sensor
glukosa sensitif yang berkontribusi dalam meningkatkan pengambilan glukosa dan
mempertahankan homeostasis glukosa untuk menunda perkembangan resistensi insulin
(Anugrahini, 2020). Penelitian yang dilakukan oleh Apichai et al (2012) melaporkan bahwa diet
yang mengandung dedak beras hitam dapat menurunkan kondisi diabetes pada tikus diabetes
yang diinduksi STZ dengan mekanisme meningkatkan kadar GLUT 4 di otot soleus (El-kholie
et al., 2018). Boue et al (2016) juga melaporkan bahwa cyanidin-3-O-Glucoside dapat
meningkatkan level ekspresi Gen GLUT 1 dan GLUT 4 (Tantipaiboonwong et al., 2017). Mengendalikan Metabolisme Glukosa Penelitian yang dilakukan oleh Septianingtyas and Agustini (2021) melaporkan bahwa
yeast dari beras hitam dapat menurunkan glukosa darah. Pemberian yeast merupakan suatu
terapi yang dapat menurunkan glukosa darah. Yeast memiliki kandungan kromium. Kromium
trivalen (Cr3+) berperan untuk mengendalikan metabolisme glukosa dalam tubuh, sehingga
disebut juga dengan faktor pengendali kadar gula darah (Glucose Tolerance Factor)
(Septianingtyas and Agustini, 2021). Kromium dalam bentuk aktif memudahkan aktifitas insulin,
dimana ketika kromium masuk ke dalam sel dan berikatan dengan kromodulin menyebabkan
ikatan kompleks kromodulin-kromium yang berikatan dengan reseptor insulin dan menjaganya
agar tetap dalam kondisi yang aktif sehingga akan meningkatkan masuknya glukosa ke dalam
sel (Mutiarani, 2017). Bagi penderita diabetes yang disebabkan oleh hambatan metabolisme
glukosa, suplementasi Cr3+ sebagai mikronutrien sangat dibutuhkan (Septianingtyas and
Agustini, 2021). Menghambat Enzim α-glukosidase dan α-amilase Enzim α-glukosidase merupakan enzim yang memiliki peran penting pada hidrolisis
karbohidrat menjadi glukosa (Yuniarto and Selifiana, 2018). Amilase (α-amilase) adalah enzim
yang mengkatalisis hidrolisis ikatan α-1, 4- glikosidik amilosa pati menghasilkan glukosa
(Ariandi, 2016). Penghambatan pada kedua enzim ini akan memberikan dampak pada
penundaan penyerapan glukosa dan dapat menurunkan glukosa dalam darah. Mekanisme
penghambatan enzim α-amilase dan α-glukosidase yaitu senyawa dengan gugus fungsi
hidroksi-OH pada flavonoid dan fenolik membentuk ikatan hidrogen dengan asam amino 383 Usadha: Journal of Pharmacy. Vol. 1, No. 3, (2022). ISSN 2827-9905
Available online at: https://jsr.lib.ums.ac.id/index.php/ujp Usadha: Journal of Pharmacy. Vol. 1, No. 3, (2022). ISSN 2827-9905
Available online at: https://jsr.lib.ums.ac.id/index.php/ujp
spesifik di situs aktif enzim (Anugrahini, 2020). Penelitian Yuda et al (2020) melaporkan bahwa
infusa beras hitam aktif menginhibisi α-glukosidase yang bersumber dari yeast dengan nilai
IC50 sebesar 77,92 g/mL, relatif lebih baik dibandingkan dengan standar akarbose. Kadar
glukosa darah dapat dikontrol pada tahap pencernaan yaitu telah dilaporkan bahwa antosianin
dalam dedak beras hitam dan fraksi yang diperkaya proanthocyanidin dari dedak beras merah
memiliki efek penghambatan pada α-glukosidase dan α-amilase yang dapat mengganggu
pencernaan pati, sehingga penyerapan glukosa selanjutnya dapat ditekan (Tantipaiboonwong
et al., 2017). Usadha: Journal of Pharmacy. Vol. 1, No. 3, (2022). ISSN 2827-9905
Available online at: https://jsr.lib.ums.ac.id/index.php/ujp Usadha: Journal of Pharmacy. Vol. 1, No. 3, (2022). ISSN 2827-9905
Available online at: https://jsr.lib.ums.ac.id/index.php/ujp spesifik di situs aktif enzim (Anugrahini, 2020). Penelitian Yuda et al (2020) melaporkan bahwa
infusa beras hitam aktif menginhibisi α-glukosidase yang bersumber dari yeast dengan nilai
IC50 sebesar 77,92 g/mL, relatif lebih baik dibandingkan dengan standar akarbose. Kadar
glukosa darah dapat dikontrol pada tahap pencernaan yaitu telah dilaporkan bahwa antosianin
dalam dedak beras hitam dan fraksi yang diperkaya proanthocyanidin dari dedak beras merah
memiliki efek penghambatan pada α-glukosidase dan α-amilase yang dapat mengganggu
pencernaan pati, sehingga penyerapan glukosa selanjutnya dapat ditekan (Tantipaiboonwong
et al., 2017). Menghambat Stres Oksidatif Antosianin merupakan senyawa bioaktif, yang pada strukturnya terdapat susunan ikatan
rangkap terkonjugasi dan mampu memfungsikan antosianin sebagai senyawa penghancur dan
penangkal radikal bebas alami atau yang lebih dikenal sebagai senyawa antioksidan alami pada
manusia. Semakin banyak gugus hidroksil fenolik dalam struktur antosianin akan dapat
meningkatkan fungsi antioksidannya. Antosianin dapat memangsa berbagai jenis radikal bebas
turunan oksigen reaktif, seperti hidroksil (OH*), peroksil (ROO*), dan oksigen tunggal (O2*)
(Priska et al., 2018). Antosianin dapat mencegah resistensi insulin dengan menghambat stres
oksidatif, sehingga dapat meningkatkan sensitivitas insulin. (Wahyuni et al., 2020) DAFTAR PUSTAKA Almundarij T.I., Zaki A.K.A., Albarrak S.M., Alharbi Y.M., Almuzaini S.A. and Abo-Aziza F.A.M., 2020,
Evaluation of the anti-diabetic activities of colored rice varieties in streptozotocin-induced diabetic
rats, Systematic Reviews in Pharmacy, 11 (11), 1424–1433. Anugrahini C. P. H., 2020, Narrative Review : Aktivitas Antidiabetes Tanaman Tradisional Di Pulau Jawa,
Naskah Publikasi, Fakultas Farmasi Universitas Muhammadiyah Surakarta. Ariandi, 2016, Pengenalan Enzim Amilase (Alpha-Amylase) dan Reaksi Enzimatisnya Menghidrolisis
Amilosa Pati Menjadi Glukosa, Jurnal Dinamika, 07 (1), 74–82. Arifin A.S., Yuliana N.D. and Rafi M., 2019, Aktivitas Antioksidan Pada Beras Berpigmen dan Dampaknya
terhadap Kesehatan, Pangan, 28 (1), 11–22. Buchari., Utamie L., 2019, Pengaruh Penambahan Gula Terhadap Sifat Kimia Dan Tingkat Kesukaan
Tape Beras Merah, Beras Hitam, Dan Beras Ketan, Skripsi thesis, Universitas Mercu Buana
Yogyakarta. Chaiyasut, 2017, Germinated Thai Black Rice Extract Protects Experimental Diabetic Rats from
Oxidative Stress and Other Diabetes-Related Consequences, Pharmaceuticals, 10 (3), 1-16. Daeli E., Ardiaria M., Candra A., 2018, Pengaruh Pemberian Nasi Beras Merah (Oryza nivara) dan Nasi
Beras Hitam (Oryza sativa L. indica) terhadap Perubahan Kadar Gula Darah dan Trigliserida Tikus
Wistar (Rattus norvegicus) Diabetes Mellitus Tipe 2, JNH (Journal of Nutrition and Health), 6 (2),
42-56. Decroli, Eva., 2019, Diabetes Melitus Tipe 2, Padang : Penerbitan Bagian Ilmu Penyakit Dalam Fakultas
Kedokteran Universitas Andalas. El-kholie E.M., Khatib B.R.E.- and Khalil Y.A., 2018, Anti-diabetic Effect of Black Rice and their Extracts
on Alloxan-Induced Diabetic Rats, Al-majallatu Al-ilmiyah Likulliyyati At-tarbiyyah An-naw'iyyah,
551–566. Hani D.P., 2020, Tinjauan Artikel : Aktivitas Antidiabetes Antosianin, Naskah Publikasi, Universitas
Muhammadiyah Surakarta. Herman., Murniarti., Syafitri, N. A., 2019, Inventarisasi Tanaman Obat Tradisional Untuk Penderita
Diabetes Melitus Dan Hipertensi Di Desa Minang Kecamatan Bambang Kabupaten Mamasa, Jurnal
Farmasi Sandi Karsa, 5 (1), 26-32. Istiarini C.H., 2016, Pengaruh Beras Hitam Terhadap Kadar Gula Darah Pada Pasien Diabetes Melitus
Tipe Ii Di Dusun Kadirojo Ii, Sleman, Yogyakarta, Jurnal Kesehatan, 4 (2), 87–97. Kang H.W., Lim W.C., Lee J.K., Ho J.N., Lim E.J. and Cho H.Y., 2017, Germinated waxy black rice
ameliorates hyperglycemia and dyslipidemia in streptozotocin-induced diabetic rats, Biological
and Pharmaceutical Bulletin, 40 (11), 1846–1855. Kurniawaty, Evy., 2014, Diabetes Mellitus. JUKE. 4 (7), 114-119. Mangiri J., Mayulu N. and Kawengian S.E.S., 2016, Gambaran kandungan zat gizi pada beras hitam
(Oryza sativa L.) kultivar Pare Ambo Sulawesi Selatan, Jurnal e-Biomedik, 4 (1), 2–6. Mutiarani A. KESIMPULAN Berdasarkan hasil studi literatur diatas dapat disimpulkan bahwa beras hitam (Oryza
sativa L. indica) dengan berbagai bentuk memiliki khasiat menurunkan kadar glukosa darah. Beras hitam memiliki khasiat sebagai antidiabetes karena mengandung senyawa metabolit
sekunder antosianin yang merupakan komponen utama dalam beras hitam dengan aktivitas
antidiabetes meregenerasi sel beta di pankreas, meningkatkan kadar GLUT 1 dan GLUT 4,
menghambat enzim α-glukosidase dan α-amilase dan menghambat stres oksidatif. Kandungan
serat yang tinggi dalam beras hitam juga berpotensi menurunkan kadar gula darah dengan cara
menghambat pelepasan glukosa selama pencernaan. Yeast dari beras hitam memiliki khasiat
antidiabetes karena kandungan kromium yang tinggi dengan aktivitas antidiabetes
mengendalikan metabolisme glukosa darah. Dengan adanya studi literatur ini beras hitam
(Oryza sativa L. indica) dapat digunakan untuk pencegahan dan pengobatan diabetes mellitus. 384 Usadha: Journal of Pharmacy. Vol. 1, No. 3, (2022). ISSN 2827-9905
Available online at: https://jsr.lib.ums.ac.id/index.php/ujp Usadha: Journal of Pharmacy. Vol. 1, No. 3, (2022). ISSN 2827-9905
Available online at: https://jsr.lib.ums.ac.id/index.php/ujp DAFTAR PUSTAKA linda., 2017, Pengaruh Pemberian Kromium, Vitamin C, Dan Vitamin E Terhadap Gula
Darah Tikus Wistar Yang Diinduksi Aloksan, Jurnal ilmiah kedokteran, 4 (1), 39–50. Nurul et al., 2019, Asupan Karbohidrat, Asupan Lemak, Aktivitas Fisik dan Kejadian Obesitas pada
Remaja di Kota Yogyakarta, skripsi thesis, Poltekkes Kemenkes Yogyakarta. Pasaribu S.F. et al., 2021, Analisis Antosianin dan Flavonoid Ekstrak Kecambah Beras Hitam Analysis of
Anthocyanins and Flavonoids in Germinated Black Rice Extract Program Studi Ilmu Gizi , Sekolah
Pascasarjana , Universitas Sebelas Maret , Surakarta , Indonesia, Jurnal Dunia Gizi, 4 (1), 8–14. Priska M., Peni N., Carvallo L. and Ngapa Y.D., 2018, Antosianin dan Pemanfaatannya, Cakra Kimia 385 Usadha: Journal of Pharmacy. Vol. 1, No. 3, (2022). ISSN 2827-9905
Available online at: https://jsr.lib.ums.ac.id/index.php/ujp (Indonesian E-Journal of Applied Chemistry), 6 (2), 79–97. (Indonesian E-Journal of Applied Chemistry), 6 (2), 79–97. Septianingtyas E., Agustini R., 2021, Pengaruh Jenis Yeast Terhadap Kadar Glukosa Darah Mencit (Mus
musculus) Yang Terindikasi Diabetes Mellitus Tipe 2, UNESA Journal of Chemistry, 10 (2), 128-
134. Tantipaiboonwong P., Pintha K., Chaiwangyen W., Chewonarin T., Pangjit K., Chumphukam O., Kangwan
N. and Suttajit M., 2017, Anti-hyperglycaemic and anti-hyperlipidaemic effects of black and red
rice in streptozotocin-induced diabetic rats, ScienceAsia, 43 (5), 281–288. Thanuja B. and Parimalavalli R., 2018, Role of black rice in health and diseases, International journal of
Health Sciences and Research, 8 (2), 241–248. Utama I.H., Arjentinia I. P.G. Y., 2014, Ekspresi Glukosa Transporter 4(GLUT 4) Pada Berbagai Organ
Tikus Hiperglikemia, Laporan Akhir Penelitian Fundamental, Universitas Udayana Bali. Wahyuni A.S., Hakim L., Nurrochmad A. and Astuti P., 2020, The sinergistic effect of black rice bran
extract and glibenclamide on protecting renal, hepatic, and pancreatic cells in alloxan induced
rats, International Journal of Pharmaceutical Research, 12 (1), 509–517. Widiastuti, Linda., 2020, Acupressor Dan Senam Kaki Terhadap Tingkat Peripheral Arterial Disease
Pada Klien DM Tipe 2. Jurnal Keperawatan Silampari. 3 (2), 694-706. Wijayanti, Maria Intan., 2016, Kualitas Yoghurt Sinbiotik Sari Beras Hitam (Oryza sativa L.) Dengan
Variasi Susu Skim, S1 thesis, UAJY. Yuda I.P., Juniarti J., Yuhernita Y., Ferlianti R. and Taufik Nasrullah T., 2020, Uji Aktivitas Antidiabetes
Infusa Beras Hitam (Oryza sativa L. indica) dengan Metode Toleransi Glukosa dan Inhibisi α-
Glukosidase, Majalah Kesehatan Pharmamedika, 11 (2), 123–132. Yuniarto A. and Selifiana N., 2018, Aktivitas Inhibisi Enzim Alfa-glukosidase dari Ekstrak Rimpang Bangle
(Zingiber cassumunar Roxb.) secara In vitro, MPI (Media Pharmaceutica Indonesiana), 2 (1), 22–
25. Zhao, C., W. Zhao, C., W. et al., 2014, Insulin and Risk of Diabetic Retinophathy in Patients with type 2 Diabetes
Mellitus : data from meta-analysis of seven cohort studies. Diagnostic Pathology, 9 (1), 130–136. Usadha: Journal of Pharmacy. Vol. 1, No. 3, (2022). ISSN 2827-9905
Available online at: https://jsr.lib.ums.ac.id/index.php/ujp DAFTAR PUSTAKA et al., 2014, Insulin and Risk of Diabetic Retinophathy in Patients with type 2 Diabetes
Mellitus : data from meta-analysis of seven cohort studies. Diagnostic Pathology, 9 (1), 130–136. 386
|
https://openalex.org/W2083947022
|
https://dash.harvard.edu/bitstream/1/41511300/1/109854%20J.%20Biol.%20Chem.-2013-Sung-10012-23.pdf
|
English
| null |
Structural and Biochemical Analysis of DNA Helix Invasion by the Bacterial 8-Oxoguanine DNA Glycosylase MutM
|
Journal of biological chemistry/The Journal of biological chemistry
| 2,013
|
cc-by
| 11,198
|
Terms of Use This article was downloaded from Harvard University’s DASH repository, and is made available
under the terms and conditions applicable to Other Posted Material, as set forth at http://
nrs.harvard.edu/urn-3:HUL.InstRepos:dash.current.terms-of-use#LAA Citation Sung, Rou-Jia, Michael Zhang, Yan Qi, and Gregory L. Verdine. 2013. “Structural and Biochemical
Analysis of DNA Helix Invasion by the Bacterial 8-Oxoguanine DNA Glycosylase MutM.” Journal
of Biological Chemistry288 (14): 10012–23. https://doi.org/10.1074/jbc.M112.415612. Permanent link http://nrs.harvard.edu/urn-3:HUL.InstRepos:41511300 Received for publication,August 31, 2012, and in revised form, January 26, 2013 Published, JBC Papers in Press,February 12, 2013, DOI 10.1074/jbc.M112.415612 Received for publication,August 31, 2012, and in revised form, January 26, 2013 Published, JBC Papers in Press,F
Rou-Jia Sung‡§1, Michael Zhang¶1,2, Yan Qi3, and Gregory L. Verdine‡§¶**4 Rou-Jia Sung‡§1, Michael Zhang¶1,2, Yan Qi3, and Gregory L. Verdine‡§¶**4 From the Departments of ‡Molecular and Cellular Biology, ¶Chemistry and Chemical Biology, and §Stem Cell and Regenerative
Biology, Harvard University, Cambridge, Massachusetts 02138, the **Chemical Biology Initiative and Program in Cancer Chemical
Biology, Dana-Farber Cancer Institute, Boston, Massachusetts 02115, and the Graduate Program in Biophysics, Harvard Medical
School, Boston, Massachusetts 02115 nd: The role of helix-invading residues during MutM-catalyzed oxoG repair remains unclear. Background: The role of helix-invading residues during MutM-catalyzed oxoG repair remains unclear. Results: Mutation of Phe114 and Met77 to alanine destabilizes the extrahelical conformation of oxoG and decreases efficiency of
oxoG repair. Background: The role of helix-invading residues during MutM-catalyzed oxoG repair remains unclea
Results: Mutation of Phe114 and Met77 to alanine destabilizes the extrahelical conformation of oxoG an
oxoG repair. g
g
g
y
p
Results: Mutation of Phe114 and Met77 to alanine destabilizes the extrahelical conformation of oxoG and decreases efficiency of
oxoG repair. Conclusion: Helix invasion by Phe114 and Met77 is important for promoting and stabilizing extrusion of oxoG during repair. Significance: Helix invasion constitutes a critical step during the base extrusion process. MutM is a bacterial DNA glycosylase that serves as the first
line of defense against the highly mutagenic 8-oxoguanine
(oxoG) lesion, catalyzing glycosidic bond cleavage of oxoG to
initiate base excision DNA repair. Previous work has shown that
MutM actively interrogates DNA for the presence of an intrahe-
lical oxoG lesion. This interrogation process involves significant
buckling and bending of the DNA to promote extrusion of oxoG
from the duplex. Structural snapshots have revealed several dif-
ferent highly conserved residues that are prominently inserted
into the duplex in the vicinity of the target oxoG before and after
base extrusion has occurred. However, the roles of these helix-
invading residues during the lesion recognition and base extru-
sion process remain unclear. In this study, we set out to probe
the function of residues Phe114 and Met77 in oxoG recognition
and repair. Here we report a detailed biochemical and structural
characterization of MutM variants containing either a F114A or
M77A mutation, both of which showed significant decreases in
the efficiency of oxoG repair. * Thisworkwassupported,inwholeorinpart,byNationalInstitutesofHealth
Grants GM044853 and CA100742.
□
S This article contains supplemental text, Table S1, and Figs. S1–S5.
The atomic coordinates and structure factors (codes 4G4R and 4G4O) have been
deposited in the Protein Data Bank (http://wwpdb.org/).
1 These authors contributed equally to this work.
2 Supported by a grant by the Harvard College Research Program.
3 Supported by a predoctoral fellowship from the National Science
Foundation.
4 To whom correspondence should be addressed: 12 Oxford St., Cambridge,
MA 02138. Fax: 617-495-8755; E-mail: gregory_verdine@harvard.edu. Received for publication,August 31, 2012, and in revised form, January 26, 2013 Published, JBC Papers in Press,February 12, 2013, DOI 10.1074/jbc.M112.415612 These data reveal that Met77 plays
an important role in stabilizing the lesion-extruded conforma-
tion of the DNA. Phe114, on the other hand, appears to destabi-
lize the intrahelical state of the oxoG lesion, primarily by buck-
ling the target base pair. We report the observation of a
completely unexpected interaction state, in which the target
base pair is ruptured but remains fully intrahelical; this struc-
ture vividly illustrates the disruptive influence of MutM on the
target base pair. the most prevalent and deleterious forms of spontaneous
genetic change (1). Numerous DNA lesions arise from reactive
oxygen intermediate-induced damage, but that which has
received the most attention is 7,8-dihydro-8-oxoguanine (8-
oxoguanine (oxoG)),5 a nearly perfect mutagen because of its
strong preference to mispair with adenine during genome rep-
lication, which gives rise to G:C 3 T:A transversion mutations
(2, 3). In all organisms, oxoG lesions are repaired through the
action of lesion-specific DNA glycosylases that conduct surveil-
lance of the genome to locate and excise damaged nucleobases
(4, 5). oxoG repair in bacteria is carried out by the DNA repair
enzyme MutM, a bifunctional DNA glycosylase/lyase that rec-
ognizes oxoG:C base pairs, catalyzing removal of both the oxoG
nucleobase and its sugar moiety (Fig. 1B) (3). Our current
understanding of oxoG recognition and repair MutM indicates
that it takes place via four major sequential steps: 1) encounter
between MutM and an intrahelical, fully base-paired oxoG
nested within the DNA duplex; 2) extrusion of the oxoG nucle-
oside from the duplex and insertion into the extrahelical
enzyme active site; 3) excision of the N-glycosidic bond via
nucleophilic displacement by a nucleophilic proline residue on
the enzyme (Fig. 1B); and 4) conjugate elimination on both
sides of enzyme-linked abasic site to remove the sugar moiety,
leaving behind a nick bracketed by 5- and 3-phosphate groups
(6, 7). Considerable intrigue and interest surrounds the first two
steps, because oxoG differs by only two atoms from G (Fig. 1A),
because the lesion has negligible effects on duplex DNA struc-
ture and only modest effects on duplex stability (8–11), and
because oxoG residues are embedded among a million-fold
excess of undamaged G residues. Share Your Story The Harvard community has made this article openly available. Please share how this access benefits you. Submit a story . The Harvard community has made this article openly available. Please share how this access benefits you. Submit a story . Accessibility THE JOURNAL OF BIOLOGICAL CHEMISTRY VOL. 288, NO. 14, pp. 10012–10023, April 5, 2013
© 2013 by The American Society for Biochemistry and Molecular Biology, Inc. Published in the U.S.A. THE JOURNAL OF BIOLOGICAL CHEMISTRY VOL. 288, NO. 14, pp. 10012–10023, April 5, 2013
© 2013 by The American Society for Biochemistry and Molecular Biology, Inc. Published in the U.S.A. 5 The abbreviations used are: oxoG, 8-oxoguanine; RMSD, root mean square
deviation; LRC, lesion recognition complex; OCL, oxoG-capping loop; EC,
encounter complex; DNA, deoxyribonucleic acid. Received for publication,August 31, 2012, and in revised form, January 26, 2013 Published, JBC Papers in Press,February 12, 2013, DOI 10.1074/jbc.M112.415612 by guest on October 8, 2019
http://www.jbc.org/
Downloaded from Oxidative damage to DNA, caused primarily by escaped
reactive oxygen intermediates of aerobic respiration, is among The development of structure- and mechanism-based trap-
ping technologies has recently made it possible to characterize,
at atomic resolution, the interactions of MutM with lesion-
containing DNA at multiple stages of the repair cycle, including
lesion searching, encounter, recognition, and extrusion. For 10012
JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 288•NUMBER 14•APRIL 5, 2013 Helix Invasion by MutM GURE 1. Formation and repair of oxoG. A, diagram showing differences at N7 and C8 for guanine versus oxoG. B, oxoG repair by MutM. First step, encounter
etweenMutMandafullybase-pairedoxoGintheDNAduplex.Secondstep,extrusionofoxoGintotheactivesiteandcleavageoftheN-glycosydicbond.Third
ndfourthsteps,Schiffbaseformationfollowedby,-eliminationtoyieldtheendproduct-boundcomplex.Chemicalstructuresofrelevantintermediatesare
hown. C–E, positions of residues Met77, Arg112, and Phe114 relative to the target base are shown for structures of MutM bound to an intrahelical oxoG, (C) oxoG
hat has been extruded into the active site (known as the lesion recognition complex, or LRC) (D), and intrahelical G (E). The side chains are shown as cyan sticks,
nd the rest of the protein chain is shown as gray ribbon. Hydrogen bonds are indicated by the dashed lines. The OCL is indicated by a blue loop. Helix Invasion by MutM
by guest on October 8, 2019
http://www.jbc.org/
Downloaded from by guest on October 8, 2019
http://www.jbc.org/
Downloaded from FIGURE 1. Formation and repair of oxoG. A, diagram showing differences at N7 and C8 for guanine versus oxoG. B, oxoG repair by MutM. First step, encounter
betweenMutMandafullybase-pairedoxoGintheDNAduplex.Secondstep,extrusionofoxoGintotheactivesiteandcleavageoftheN-glycosydicbond.Third
andfourthsteps,Schiffbaseformationfollowedby,-eliminationtoyieldtheendproduct-boundcomplex.Chemicalstructuresofrelevantintermediatesare
shown. C–E, positions of residues Met77, Arg112, and Phe114 relative to the target base are shown for structures of MutM bound to an intrahelical oxoG, (C) oxoG
that has been extruded into the active site (known as the lesion recognition complex, or LRC) (D), and intrahelical G (E). The side chains are shown as cyan sticks,
and the rest of the protein chain is shown as gray ribbon. Hydrogen bonds are indicated by the dashed lines. The OCL is indicated by a blue loop. FIGURE 1. Formation and repair of oxoG. A, diagram showing differences at N7 and C8 for guanine versus oxoG. B, oxoG repair by MutM. First step, encounter
betweenMutMandafullybase-pairedoxoGintheDNAduplex.Secondstep,extrusionofoxoGintotheactivesiteandcleavageoftheN-glycosydicbond.Third
andfourthsteps,Schiffbaseformationfollowedby,-eliminationtoyieldtheendproduct-boundcomplex.Chemicalstructuresofrelevantintermediatesare
shown. EXPERIMENTAL PROCEDURES The clarified lysate was diluted
1:5 with buffer A (20 mM Tris, pH 7.4, and 0.1% -mercapto-
ethanol) and loaded onto a 20-ml Sepharose Fast Flow column
(GE Healthcare). The protein was purified using a linear gradi-
ent from 0.1–0.5 M NaCl over 10 column volumes, followed by
size exclusion chromatography (Superdex75; GE Healthcare)
into 1 M NaCl, 10 mM Tris, pH 7.4, 5 mM -mercaptoethanol. The unused protein was stored by adding glycerol to 20% final
concentration, flash frozen, and stored at 80 °C. by guest on October 8, 2019
http://www.jbc.org/
Downloaded from The striking changes in the arrangement of these three resi-
dues in response to the state of the target base (whether intra-
helical or extrahelical) have strong implications for their role in
the base extrusion process. Molecular dynamics simulations
have suggested a role for Arg112 in promoting base extrusion
through direct competition with oxoG for hydrogen bonding
contacts to the opposing C (14). The function of Arg112 has
been further characterized by new work observing the struc-
tural effects of an alanine substitution at Arg112 on the interac-
tion between MutM and DNA (17). The absence of Arg112
results in the entrapment of a MutM-DNA complex with pro-
tein-DNA contacts distinct from those of other intrahelical
structures. These structures suggest an additional role for
Arg112 in DNA translocation during lesion search, providing a
novel mechanism for the movement of MutM along DNA as
the enzyme probes undamaged bases in the DNA duplex. Phe114 has also been implicated in the lesion search process,
most directly by the aforementioned structural studies but also
by recent biochemical results indicating that removal of the
intercalating Phe114 side chain increases the rate of diffusion of
the protein along DNA (18). DNA Synthesis and Purification—All of the DNA substrates
were synthesized using solid phase synthesis on an ABI 392
DNA synthesizer using standard reagents and protocols
(Applied Biosystems and Glen Research). For the cross-linker-
modified oligonucleotides, the protocol was modified to
include a 3 H-phosphonate (Glen Research) in the coupling
step at the site of the backbone modification. Following cou-
pling, the disulfide tether was incorporated onto the backbone
by an oxidation step with carbon tetracholoride (Sigma-Al-
drich) and the diamine disulfide (free base). All of the oligonu-
cleotides were purified according to the same protocol. EXPERIMENTAL PROCEDURES ing accelerated extrusion of the lesion from DNA following the
initial encounter (14, 16). Protein Expression and Purification—Point mutations of Bst
MutM were made using the QuikChange site-directed
mutagenesis kit (Stratagene), and the mutations were con-
firmed by sequencing. Expression and purification of the
mutant proteins were carried using the wild-type protocol. These structures revealed that base extrusion is mediated by
three highly conserved residues: Met77, Arg112, and Phe114, that
form important contacts at the protein-DNA interface from the
minor groove side. Each of these residues occupies a distinct
position in close proximity to the target base pair. In the LRC
structure (Fig. 1D), Phe114 wedges into the DNA duplex
between the target base pair and the base pair 3 to the target
base, inducing a bend in the global DNA conformation. Arg112
invades the duplex to fill the space left by oxoG, forming hydro-
gen bonding contacts to the opposing C (Fig. 1D). Met77 is
positioned in the intervening space between the extruded base
and the resulting gap in the duplex (Fig. 1D) (12). In contrast,
these residues take on a decidedly different organization in the
presence of an intrahelical oxoG-containing substrate (Fig. 1C)
(14). The fully base-paired nature of the target base pair blocks
the entry of Arg112 and Met77 into the duplex, leaving them to
take up alternative positions in the minor groove. Phe114 occu-
pies a similar position as in the LRC state; however, insertion of
the aromatic side chain into a fully duplex substrate severely
buckles the aforementioned base pairs, significantly destabiliz-
ing the target base (Fig. 1C). Mutant MutM proteins were overexpressed and purified as
previously described (14). MutM was expressed in Escherichia
coli BL21 (DE3) plysS cells and grown at 37 °C until A600
reached 0.5–0.7. Protein expression was induced by addition of
0.5 mM isopropyl--D-thiogalactopyranoside (Invitrogen) and
0.05 mM ZnCl2. The cells were allowed to grown for 4–5 h at
30 °C, harvested by ultracentrifugation, resuspended in 50 mM
NaPO4, pH 8.0, 500 mM NaCl, and 0.1% -mercaptoethanol,
flash frozen in liquid nitrogen, and stored at 80 °C. Thawed cells were supplemented with 1 mM PMSF and 1
Complete, EDTA-free Protease inhibitor mixture tablet (Roche
Applied Science), lysed by sonication, and clarified by centrifu-
gation at 14,000 rpm for 20 min. Received for publication,August 31, 2012, and in revised form, January 26, 2013 Published, JBC Papers in Press,February 12, 2013, DOI 10.1074/jbc.M112.415612 C–E, positions of residues Met77, Arg112, and Phe114 relative to the target base are shown for structures of MutM bound to an intrahelical oxoG, (C) oxoG
that has been extruded into the active site (known as the lesion recognition complex, or LRC) (D), and intrahelical G (E). The side chains are shown as cyan sticks,
and the rest of the protein chain is shown as gray ribbon. Hydrogen bonds are indicated by the dashed lines. The OCL is indicated by a blue loop. states has revealed that MutM can detect the presence of an
oxoG lesion even at the earliest stage of initial encounter, with
the intrahelical oxoG inducing an altered DNA backbone con-
formation that encourages extrusion of the lesion from the he-
lical stack (14). Computational simulations have revealed that
this “extrudogenic conformation” indeed lowers the free energy
barrier for extrusion of oxoG versus G. It is noteworthy, more-
over, that MutM shows no thermodynamic preference to bind
DNA containing an intrahelical oxoG (14, 15). Taken together,
these data have led to formulation of a model wherein the
enzyme locates oxoG residues by a kinetic mechanism involv- example, mutation of the critical catalytic residue Glu3 enabled
structural characterization of a state poised for catalysis, specif-
ically an enzyme-DNA complex (lesion recognition complex
(LRC); Fig. 1D) having a fully extrahelical oxoG lesion inserted
into the enzyme active site (12). More recently, disulfide cross-
linking technology has been employed to stabilize and thereby
structurally elucidate ordinarily fleeting intermediates at earlier
stages of the repair pathway, including that of initial encounter
by MutM of a fully base-paired, intrahelical oxoG, and also
during interrogation of undamaged DNA (Fig. 1, C and E) (7,
12–14). Detailed comparison of the structures at these three APRIL 5, 2013•VOLUME 288•NUMBER 14 JOURNAL OF BIOLOGICAL CHEMISTRY 10013 10013 APRIL 5, 2013•VOLUME 288•NUMBER 14 EXPERIMENTAL PROCEDURES The
oligonucleotides were deprotected and cleaved from the CPG
resin using ammonium hydroxide at 55 °C for 8–10 h. The
ammonium hydroxide was removed using a SpeedVac, and the
oligonucleotides were purified using the crush and soak
method (19). The oligonucleotides were desalted using hydro-
phobic chromatography (Sep-Pak columns; Waters Chroma-
tography) and then evaporated to dryness using a SpeedVac. The oligonucleotides were dissolved in 10 mM Tris, pH 8.0, and
their appropriate masses were verified using MALDI-TOF
mass spectrometry. The lesion-containing strand was annealed
to the complementary strand in a 1:1.2 ratio in 1 NaTE (50
mM NaCl, 10 mM Tris-HCl, 1 mM EDTA, pH 8.0). The following
DNA oligomers were used for the structures discussed in this
paper, with Go designating oxoG and x representing the loca-
tion of the phosphate group bearing the nonbridging N-ethyl-
thio-tether used in generating the cross-linked complexes:
oxoG-containing strand, 5-TGCGTCCGoAGxTCTACC-3;
G-containing strand, 5-TGCGTCCGAGxTCTACC-3; and the
complementary strand, 5-AGGTAGACTCGGACGC-3. The ability to trap MutM bound to oxoG- and G-containing
DNA at different steps in the repair cycle has provided a pow-
erful tool for investigating the functions of these three residues
in promoting changes at the target base pair that facilitate base
extrusion. Here, we report the structural and biochemical char-
acterization of Bacillus stearothermophilus MutM in which the
two helix-invading side chains, those on Phe114 and Met77, have
been individually removed by mutation to Ala. These structures
provide direct evidence that MutM actively interrogates the
DNA duplex during the search for lesions, and they shed light
on the means by which MutM discourages reannealing of the
extruded oxoG lesion. 10014
JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 288•NUMBER 14•APRIL 5, 2013 RESULTS Biochemical Characterization of the M77A and F114A Vari-
ants of MutM—We expressed and purified variants of MutM
containing individually either the F114A or M77A mutation. These variants showed chromatographic behavior indistin-
guishable from that of the wild-type protein, indicating that the
mutations caused no gross conformational abnormality. To
probe more definitively the activities of these proteins, we
measured their ability to catalyze base excision DNA repair of
oxoG in a duplex oligonucleotide substrate. As shown in Fig. 2
(A and B), the rate of oxoG repair was significantly diminished
for both F114A and M77A MutM. This loss of activity could in
principle result either from a generalized defect in the protein
or from a specific defect in catalysis of glycosidic bond cleavage,
the first chemical step in the multistep reaction cascade cata-
lyzed by MutM (Fig. 1B). To resolve this ambiguity, we took
advantage of the fact that it is possible to circumvent the base
excision step and synthetically generate a downstream interme-
diate in the repair cascade—the aminal intermediate—by pro-
viding MutM with a substrate containing an abasic site (Fig. 2D). Because all steps of the entire reaction cascade are cata-
lyzed by a single active site (7), preservation of enzyme activity
on an abasic site indicates preservation of overall active site
architecture, with the catalytic defect being limited to the
upstream steps of in the repair process. When the catalytic
competence of the mutant enzymes was analyzed in this man-
ner, it was found to be virtually unaffected by either the F114A
or M77A mutation (Fig. 2C). Taken together with the oxoG
repair data, we conclude that the F114A and M77A mutations
specifically impair the early events in oxoG repair, namely
extrusion of the lesion from DNA and excision of the oxoG
glycosidic bond. p
g
p
g
(
)
Protein coordinates from the isomorphous structure of
MutM cross-linked to undamaged DNA (Protein Data Bank
code 3GPX) were used as the initial model for refinement (14). Rigid body fitting, energy minimization, and simulated anneal-
ing in PHENIX resulted in a partial model (21). The DNA
(including the disulfide tether) was built into the Fo-Fc map,
using the strong density for the cross-link to determine the
register. Manual readjustment of protein side chains and DNA
bases were done in COOT, followed by successive rounds of
simulated annealing, energy minimization, and individual
B-factor refinement (22, 23). 10014
JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 288•NUMBER 14•APRIL 5, 2013 Helix Invasion by MutM Crystallization, Data Collection, and Structure Deter-
mination—Crystallization drops were set up at 4 °C using the
hanging drop vapor diffusion method, in a 1:1 or 1:2 ratio of
protein:DNA complex to reservoir solution. The complexes
crystallized in 12–18% PEG 8000, 100 mM sodium cacodylate,
pH 7.0, and 5% glycerol. Crystals appeared with 2–3 days and
were allowed to grow to size over 1–2 weeks before briefly soak-
ing in cryoprotectant solution containing 18% PEG 8000, 100
mM sodium cacodylate, pH 7.0, and 25% glycerol before flash
freezing in liquid nitrogen. Diffraction data were collected at
the 24ID-E beamline at Argonne Photon Source and the X29
beamline at the National Synchrotron Light Source. The data
were processed using the HKL2000 program suites (20). by guest on October 8, 2019
http://www.jbc.org/
Downloaded from Helix Invasion by MutM oxoG and Abasic Site Cleavage Assays—The duplex DNA
substrate used for the cleavage assays was identical in sequence
to the oligonucleotide used for crystallization but did not con-
tain a N-ethylthio tether. The oxoG-containing strands were
5-end-labeled using T4 polynucleotide kinase (New England
Biolabs) and [-32P]ATP (PerkinElmer Life Sciences) and then
annealed with a 1.1-fold excess of the complementary strand. Single turnover cleavage reactions were carried out using 100
nM DNA duplex and 500 nM wild-type, F114A, or M77A MutM
in a standard reaction buffer of 40 mM NaCl, and 50 mM Tris-
HCl, pH 7.4, at room temperature. The data were fit using a
single exponential equation to determine the first order rate
constant as previously described (GraphPad Prism 5 Software)
(30). Abasic site substrates were prepared by pretreating a
5-32P-labeled uracil-containing 16-mer duplex DNA (5-
AGCGTCCAUGTCTACC-3) with uracil DNA glycosylase
(New England Biolabs) at 37 °C for 1 h. The abasic site cleavage
reaction was carried out using 100 nM pretreated DNA duplex
and 100 nM wild-type, F114A, or M77A MutM in a reaction
buffer of 40 mM NaCl and 50 mM Tris-HCl, pH 7.4, at room
temperature. Aliquots of the reaction were removed periodi-
cally and quenched with an equal amount of 100 mM dithiothre-
itol in 95% formamide and 1 Tris borate/EDTA buffer, sub-
jected to denaturing urea-PAGE, and then visualized on a
phosphorimaging plate. Preparative Cross-linking Reactions—The complexes were
made by cross-linking the corresponding DNA substrates to
either Q166C MutM (for the G-containing complexes) or E3Q
Q166C MutM (for the oxoG-containing complexes). Purified
protein was concentrated to roughly 400 M, and the reactions
were set up using 10 M cross-linker-containing duplex sub-
strate with 20 M protein in degassed cross-linking buffer (50
mM NaCl and 20 mM Tris, pH 7.4). The reactions were purged
with argon and kept rotating at 4 °C for 2–3 days. The reactions were purified using a MonoQ column (GE
Healthcare) with a linear gradient from 0.1–0.6 M NaCl in
buffer A (20 mM Tris, pH 7.4, degassed) over 30 column vol-
umes. The fractions were pooled, buffer-exchanged into
degassed cross-linking buffer, and concentrated to 200–250
mM for crystallization experiments. The concentration was
determined by the A260 of the DNA in the complex. RESULTS Biochemical characterization of F114A and M77A mutations on oxoG repair. A, oxoG repair assay monitoring end product formation using an
oxoG-containingsubstrate(followingthereactionpathwayinFig.1B)showingsignificantincreaseinthelengthoftimenecessaryforthereactiontocomplete. B, plot following product formation from A. Data points were measured in triplicate, with indicated error bars. The data were fit to a single exponential equation
as previously described to determine the first order rate constant (30). C, quantification of end product formation using an abasic site-containing DNA
substrate. Data points were measured in duplicate, with indicated error bars. D, alternative route toward the aminal intermediate, beginning with an abasic site
DNA substrate. complexes of MutM containing either the F114A or M77A
mutation, in complex with duplex DNA containing either a
single, centrally located oxoG lesion or an undamaged G at the
same position. All of the proteins used in these studies bore a
mutation of Glu3 to Gln (E3Q mutation), which is known to
abrogate base excision while preserving lesion recognition (12). To stabilize these otherwise ephemeral complexes and restrict
the roaming range of the mutant glycosylases to the desired
interrogation site on DNA, we employed a well established
intermolecular disulfide-cross-linking strategy, which has pre-
viously been used to capture several otherwise elusive states of
the MutM-DNA interaction (13, 14, 17, 31, 32). All complexes
utilized position 166 as the cross-linking site on the protein,
with the partner thiol on DNA being introduced as an N-ethyl-
thiol tether at phosphate 10 (p10); the target base (either oxoG
or G) was located at p8. Save for the identity of the target base
(either G or oxoG), all four structures employed an otherwise
identical DNA molecule. To determine whether the removal of these key helix-invad-
ing side chains affects the strength of DNA binding by MutM,
we analyzed this binding interaction using fluorescence polar-
ization assays. Neither mutant enzyme shows any thermody-
namic preference for binding to oxoG-containing DNA, sug-
gesting that neither forms a thermodynamically stable lesion
recognition complex. Previous work has shown that whereas
recognition of an extrahelical oxoG lesion by MutM is driven by
a thermodynamic preference to attain that state, there is no
such thermodynamic preference for MutM to bind an intrahe-
lical oxoG lesion (6, 14–16). The affinity of MutM for non-
lesion-containing DNA was unaffected by introduction of the
M77A mutation but was enhanced severalfold by F114A (sup-
plemental Fig. S1, Supplemental Methods). RESULTS After all visible protein and DNA
atoms were built and the Rfree dropped below 28%, water mol-
ecules were added to the model using automated water picking
and manual inspection of the difference map in PHENIX. Translation Liberation Screw (TLS) refinement was also
included during refinement (two groups total: the protein and
then both DNA strands were treated as one TLS group) (24, 25). Electron density for residues 217–237 (comprising the oxoG-
capping loop) was not visible, and these residues were omitted
from the model. Protein side chains were truncated in instances
where density was not observed. Data collection and refine-
ment statistics are shown in Table 1. Supplemental Table S1
lists the abbreviations used for the protein:DNA complexes
described in the manuscript. Model statistics and validation were carried out using PRO-
CHECK and Molprobity (26, 27). Analysis of the DNA confor-
mation including base step parameters, sugar puckers, and tor-
sion angles were done using 3DNA (28). The figures were made
using PyMOL v1.3 (29). APRIL 5, 2013•VOLUME 288•NUMBER 14 JOURNAL OF BIOLOGICAL CHEMISTRY 10015 10015 APRIL 5, 2013•VOLUME 288•NUMBER 14 Helix Invasion by MutM FIGURE 2. Biochemical characterization of F114A and M77A mutations on oxoG repair. A, oxoG repair assay monitoring end product formation
oxoG-containingsubstrate(followingthereactionpathwayinFig.1B)showingsignificantincreaseinthelengthoftimenecessaryforthereactiontoc
B, plot following product formation from A. Data points were measured in triplicate, with indicated error bars. The data were fit to a single exponentia
as previously described to determine the first order rate constant (30). C, quantification of end product formation using an abasic site-contain
substrate. Data points were measured in duplicate, with indicated error bars. D, alternative route toward the aminal intermediate, beginning with an a
DNA substrate. by guest on October 8, 2019
http://www.jbc.org/
Downloaded from by guest on October 8, 2019
jbc.org/ FIGURE 2. Biochemical characterization of F114A and M77A mutations on oxoG repair. A, oxoG repair assay monitoring end product formation using an
oxoG-containingsubstrate(followingthereactionpathwayinFig.1B)showingsignificantincreaseinthelengthoftimenecessaryforthereactiontocomplete. B, plot following product formation from A. Data points were measured in triplicate, with indicated error bars. The data were fit to a single exponential equation
as previously described to determine the first order rate constant (30). C, quantification of end product formation using an abasic site-containing DNA
substrate. Data points were measured in duplicate, with indicated error bars. D, alternative route toward the aminal intermediate, beginning with an abasic site
DNA substrate. FIGURE 2. Data collection and refinement statistics These unusual changes in DNA conformation cannot be
ascribed to crystal packing, because the wild-type and F114-G
mutant structures have the same packing arrangement but dif-
ferent conformations, because buckling of the target intraheli-
cal base pair has been observed in three distinct crystal forms,
and because no crystal contacts occur within 17 Å of residues 77
and 114 (supplemental Fig. S3). example, the so-called oxoG-capping loop (OCL; Fig. 1D) on
MutM contributes extensive protein-DNA contacts to the
Hoogsteen face of extrahelical oxoG; hence the OCL is consid-
ered to stabilize the extrahelical state (12). Consistent with this
notion, mutations in the OCL, such as V222P, that alter even
subtly its interaction with extrahelical oxoG abolish crystallo-
graphic observation of the extrahelical state and furnish instead
an oxoG encounter complex (EC) (hereafter referred to as
V222P-oxoG EC), in which MutM is bound to an intact albeit
severely buckled target oxoG:C base pair (Fig. 1, C and E) (14). The present studies utilized only protein constructs containing
a wild-type OCL, and therefore the structures reflect only the
influence of the residues under study, Phe114 and Met77 on the
DNA conformational state. All four complexes yielded crystals that diffracted to high
resolution (1.8–1.9 Å; Table 1), and the structures were deter-
mined using molecular replacement followed by standard
refinement. The target G in both non-lesion-containing struc-
tures is fully intrahelical, forming a Watson-Crick base pair
with the complementary C (Fig. 3A). The structure of the M77A
mutant protein bound to non-lesion-containing DNA (hereaf-
ter referred to as the M77A-G structure) was virtually identical
to the corresponding V222P-G complex (Protein Data Bank
code 2F5O), with a C RMSD between the two of 0.19 Å, thus
indicating that the Met77 thioether-bearing side chain does not
play a prominent role in directing the specific DNA conforma-
tion at the site of intrahelical interrogation by MutM. The
structure of the F114A-G complex also bore marked overall
similarity to the structure of V222P MutM bound to a target G
in DNA (V222P-G; C RMSD 0.27 Å), with one prominent
exception. Previously, all 26 intrahelical MutM-DNA com-
plexes reported to date had revealed a highly unusual and char-
acteristic buckling of the target base pair, a feature ranging in The structures of the M77A and F114A mutant proteins
interrogating a target oxoG in DNA proved particularly intrigu-
ing and informative. Data collection and refinement statistics M77A-G
M77A-oG
F114A-G
F114A-oG
Data collection
Radiation source
APS-24-IDE
APS-24-IDE
NSLS-X29
APS-24-IDE
Resolution (Å)
50-1.85
50-1.95
50-1.86
50-1.95
Unique reflections
38,803
33,477
38,633
33,409
Completeness (%)a
99.9 (100.0)
99.9 (99.9)
99.9 (99.9)
99.9 (99.9)
Redundancya
7.3 (7.3)
7.2 (7.3)
6.9 (6.9)
4.8 (4.9)
Rmerge
a,b
0.077 (0.381)
0.062 (0.408)
0.049 (0.411)
0.056 (0.405)
I/a
28.8 (5.7)
25.0 (5.4)
38.9 (3.5)
19.9 (3.6)
Space group
P212121
P212121
P212121
P212121
Unit cell dimensions
a 45.40, b 93.45, c 104.76
a 45.19, b 93.54, c 104.65
a 45.37, b 95.50, c 103.10
a 45.39, b 93.14, c 105.52
Refinement and model
Resolution (Å)
32.7-1.85
32.7-1.95
32.7-1.86
37.9-1.95
Rwork (%)a,c
16.9 (18.1)
17.5 (19.1)
17.9 (18.3)
17.2 (17.7)
Rfree (%)a,c
19.6 (23.4)
21.1 (24.4)
19.8 (18.2)
21.4 (20.9)
Mean B-factors
Protein
23.96
28.23
31.08
27.05
Water
35.98
38.32
40.95
37.06
RMSD from ideality
Bond lengths (Å)
0.007
0.007
0.007
0.007
Bond angles (°)
1.092
1.107
1.062
1.085
Ramachandran plot (%)d
Most favored
94.4
93.9
93.0
94.8
Additionally allowed
5.1
5.6
6.5
4.7
Generously allowed
0.5
0.5
0.5
0.5
Protein Data Bank code
4G4N
4G4O
4G4Q
4G4R
a The values in parentheses refer to the highest resolution shell. b Rmerge I I/I, where I is the observed intensity. c Rwork Fo Fc/Fo, where Fo and Fc are the observed and calculated structure factor amplitudes, respectively. Rfree was calculated based on 5% data randomly selected
and omitted throughout structure refinement (34). d The values were calculated using PROCHECK (27). M77A-G by guest on October 8, 2019
http://www.jbc.org/
Downloaded from b Rmerge I I/I, where I is the observed intensity. c Rwork Fo Fc/Fo, where Fo and Fc are the observed and calculated structure factor amplitudes, respectively. Rfree was calculated based on 5% data randomly selected
and omitted throughout structure refinement (34). d The values were calculated using PROCHECK (27). magnitude from 28° to 44° (buckle angles measured using
3DNA). The structure of the F114A-G complex, by contrast,
altogether lacks this otherwise invariant and distinctive feature,
having a virtually negligible buckle angle of 0.8° (versus
36.2° in the corresponding wild-type structure; Fig. 3A). This
observation provides definitive evidence supporting prior sug-
gestions that DNA intercalation by the Phe114 side chain is
responsible for buckling the target base pair (12–14, 17, 18, 32,
33). RESULTS The latter results
are consistent with prior suggestions that the insertion of the
Phe114 side chain into DNA is likely to be energetically expen-
sive (13, 14) and that this insertion is an intrinsic feature of the
MutM-DNA interaction, occurring even when the protein
slides along non-lesion-containing DNA (18). Whether a target nucleobase held in the grasp of MutM pre-
fers to remain in the intrahelical state or transition to an extra-
helical state depends upon the balance of multiple energetic
influences that stabilize or destabilize each of the two states. For Structural Characterization of DNA-bound Complexes of
F114A and M77A MutM—To gain atomic level insights into
the functions of Met77 and Phe114, we crystallized a total of four 10016 VOLUME 288•NUMBER 14•APRIL 5, 2013 10016
JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 288•NUMBER 14•APRIL 5, 2013 Helix Invasion by MutM TABLE 1 Data collection and refinement statistics Data collection and refinement statistics The
F114A-oxoG structure thus represents a complex in which the
target base pair has been disrupted, but in which the oxoG
retains its engagement in stacking interactions with the
5-neighbor and consequently remains nested within the heli-
cal stack. which the cross-linking site or the sequence was permuted
(Protein Data Bank codes 3GPY, 3GQ4, and 3JR5) (14, 31, 32);
without exception, those structures also represent extrahelical
lesion recognition complexes, lending further support to the
notion that that the truncation of the Met77 side chain alone is
sufficient to shift the energetic balance for base extrusion so as
to prefer the intrahelical state, at least in the crystal environ-
ment. Although the absolute free energies of the complexes in
our crystals cannot be assumed to equal those in solution, it is
safe to assume that free energy differences between states, or at
least the rank order of free energy differences, are the same in
the crystal and in solution. This strongly suggests that under
physiologic conditions, the Met77 side chain significantly stabi-
lizes the extrahelical state of the target oxoG. Detailed Analysis of the Role of Met77—As mentioned above,
OCL-perturbing mutations such as V222P invariably result in
crystals of the EC state (Fig. 1C) (14). Here we revealed that this
state can also be accessed by introduction of the M77A muta-
tion. Although V222P and M77A both alter the energetic land-
scape of the base extrusion pathway so as to favor the intrahe-
lical encounter state, these mutations are structurally distinct;
whereas V222P resides far from the locus of helix invasion on a
loop that is disordered in EC structures, M77A lies at the epi-
center of helix invasion. Comparison of the EC structures pro-
duced by V222P versus M77A will therefore lend insight into
the influence of the Met77 side chain on the DNA structure
in the EC state. Comparison of the M77A-oG EC structure with
the V222P-oxoG EC structure (Protein Data Bank code 3GO8)
reveals that the two EC structures are nearly identical, with a
C alignment resulting in an RMSD of 0.17 Å. Data collection and refinement statistics In the M77A structure, the target oxoG is
fully intrahelical and engaged in Watson-Crick base pair with
its complement C (Fig. 3B, middle structure), a state termed the
intrahelical lesion EC, which prior to this work had been cap-
tured only in MutM structures having OCL mutations. The
target base pair in M77A-oxoG is buckled to roughly the same
extent as in the corresponding EC structure of V222P-oxoG
(Fig. 3B, compare left and middle structures). It thus appears
that the identity of the residue at position 77 is sufficient to
convert between the two major states of lesion recognition,
with the wild-type residue Met77 crystallizing as an extrahelical
LRC (as in Fig. 1D) and the mutant form Ala77 crystallizing as an
intrahelical lesion EC; an analogous situation obviously per-
tains for single point mutations in the OCL, for example at
position 222 (14). It is worth noting that we have determined a
large number of additional structures having Met77 intact but in APRIL 5, 2013•VOLUME 288•NUMBER 14 10017 JOURNAL OF BIOLOGICAL CHEMISTRY APRIL 5, 2013•VOLUME 288•NUMBER 14 Helix Invasion by MutM FIGURE 3. The DNA conformation at the target site, with either G (A) or oxoG (B) as the target base. The corresponding V222P structure for each set is
shown on the left, the M77A structure are shown in the middle, and the F114A structures are shown on the right. The top panels show the buckle angle for the
targetbasepair(withtheexceptionoftheV222P-oGandF114A-oGstructures),whereasthebottompanelsshowaviewdownthehelicalaxisforthetargetbase
pair. All of the distances were measured in Å, and the buckle angles were calculated using 3DNA. 2Fo-Fc maps were contoured to 1.0 . by guest on October 8, 2019
http://www.jbc.org/
Downloaded from FIGURE 3. The DNA conformation at the target site, with either G (A) or oxoG (B) as the target base. The corresponding V222P structure for each set is
shown on the left, the M77A structure are shown in the middle, and the F114A structures are shown on the right. The top panels show the buckle angle for the
targetbasepair(withtheexceptionoftheV222P-oGandF114A-oGstructures),whereasthebottompanelsshowaviewdownthehelicalaxisforthetargetbase
pair. All of the distances were measured in Å, and the buckle angles were calculated using 3DNA. 2Fo-Fc maps were contoured to 1.0 . and indeed, the interaction between O8-oxoG and O2-C would
be repulsive if the two were brought much closer together. Data collection and refinement statistics In contrast, the van der Waals contact rep-
ertoire made by Met77 in the EC state is much less extensive and
limited to only the sugar moiety of the intrahelical G and the
phosphate backbone 3 to the target base compared with the
LRC state. These observations suggest that Met77 thus appears
to stabilize the extrahelical state through the establishment of
nonpolar contacts that are uniquely tailored to the extrahelical
(LRC state). Detailed Analysis of the Role of Phe114—Whereas deletion of
the Met77 side chain has no obvious effect on the structure of
the MutM complexes, deletion of the Phe114 side chain has
pronounced effects on DNA structure, and these effects pro-
vide insight into the unique function of this residue. One of the
most striking features of intrahelical DNA helix interrogation
by MutM is the pronounced buckling of the target base pair. Buckling occurs at the site of DNA duplex insertion by Phe114,
during which the base pair buckles away from Phe114 as if to
avoid it (Fig. 4). This behavior led to suggestions that Phe114
directly promotes rupture of the target base pair by putting it
under conformational and steric duress (12–14, 17, 32). The
hypothesis that Phe114 is directly and singularly responsible for
buckling of the target base pair is fully supported by the present
structural data. Precisely, with undamaged DNA (F114A-G
structure), removal of the Phe114 side chain results in complete
unbuckling of the target base pair, a 35° reduction in the buckle
angle (calculated using 3DNA (28); Fig. 3A), and restoration of
a typical Watson-Crick base pair. Apart from the degree of
buckling of the target base pair, no other significant structural
adjustment is attributable to the presence or absence of the
Phe114 side chain. by guest on October 8, 2019
http://www.jbc.org/
Downloaded from FIGURE 4. Stereo views of side chain interactions at the target site. The
sidechainsarecoloredincyan,theproteinchainisgray,thetargetoxoGisred,
and the opposing C is orange. The mutated residues are indicated in italics. Shown in the right panels, in the F114A-oG structure Arg76 shifts upward to
form two new hydrogen bonds (at distances of 2.9 and 2.7 Å) with the O8. (RMSD 0.14 Å; Fig. 4). Data collection and refinement statistics Most impor-
tantly, the three other residues that make intimate contact with
the duplex at the site of invasion: Phe114, Arg112, and Arg76,
adopt essentially identical positions in the two complexes, and
the details of DNA conformation are also fully conserved Truncation of the Phe114 side chain yields a state of MutM-
DNA interaction never before observed in the dozens of struc-
tures available for MutM and its orthologs bound to DNA. In
this state, the target oxoG is fully intrahelical and is both nearly
coaxial and coplanar with the complementary C (Fig. 3, A and B,
right panels). However, the target oxoG in the F114A-oxoG
complex is swiveled 180° about its glycosidic bond so as to
adopt the syn glycosidic configuration, as opposed to the typical
anti configuration observed in the corresponding wild-type
complex (Fig. 3B, compare right and left panels). Although the
syn-oxoG and the complementary C bear an unmistakable
resemblance to a Watson-Crick base pair, the two nucleobases
are separated by too great a distance to form hydrogen bonds, 10018 VOLUME 288•NUMBER 14•APRIL 5, 2013 10018
JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 288•NUMBER 14•APRIL 5, 2013 Helix Invasion by MutM FIGURE 4. Stereo views of side chain interactions at the target site. The
sidechainsarecoloredincyan,theproteinchainisgray,thetargetoxoGisred,
and the opposing C is orange. The mutated residues are indicated in italics. Shown in the right panels, in the F114A-oG structure Arg76 shifts upward to
form two new hydrogen bonds (at distances of 2.9 and 2.7 Å) with the O8. well be the case, our structures provide no information on the
existence of any such kinetic effect. As for thermodynamic sta-
bilization of the extrahelical state by Met77, we note that this
residue has a greatly restricted range of positions in the extra-
helical (LRC) state than intrahelical (EC) state and also that
extrusion of oxoG is accompanied by retraction of Met77 into
the space vacated by departure of oxoG (Fig. 6B). In all of the
LRC structures to date, the side chain of Met77 makes several
direct van der Waals contacts with the sugar moiety of the
extrahelical oxoG and the phosphate backbone flanking the tar-
get oxoG (Fig. 6A). Data collection and refinement statistics Intrahelical oxoG encounter has previ-
ously been found to result in a structural rearrangement of the
DNA backbone at the site of the target base, as compared with
corresponding G structure, and the nature of the rearrange-
ment was found to promote base extrusion. The driving force
behind the formation of this “extrudogenic conformation” with
oxoG is provided by a severe steric clash within the oxoG moi-
ety, between its nucleobase C8-carbonyl) and both the 2-C and
5-phosphate (14). Virtually identical rearrangements leading
to formation of the extrudogenic conformation are also
observed in comparing M77A-G with M77A-oxoG (Fig. 5A). Taken together, these structural comparisons led us to con-
clude that even though the M77A mutation perturbs the ener-
getic landscape of the protein-DNA complex, it does not per-
turb the fundamental structural nature of intrahelical oxoG
encounter. The present work reveals that the nature of the encounter
between MutM and an intrahelical oxoG is surprisingly and
dramatically dependent upon the presence or absence of the
Phe114 side chain. With Phe114 intact, MutM forms an intrahe-
lical lesion encounter complex, the hallmarks of which are
buckling of a fully intact target oxoG:C base pair and adoption
of an extrudogenic conformation in the DNA backbone (Fig. 4
and supplemental Fig. S2). Mere removal of the Phe114 side
chain drastically changes the nature of target recognition, with
the oxoG lesion being swiveled 180° about its glycosidic bond
to adopt the syn conformation and with the target nucleobase
being intrahelical and unbuckled but also unpaired. The
unusual conformation of oxoG appears to be stabilized through
a hydrogen-bonding interaction between its O8 and the side
chain of Arg76, an interaction not seen previously in any struc-
ture of MutM bound to DNA (Fig. 4). The conformation of
Arg76 required to establish this hydrogen bonding interaction
with oxoG would be expected to clash sterically with Phe114 How does Met77 stabilize the extrahelical state? High level
computational simulations will ultimately be required to
answer this question, but the x-ray structures do provide some
noteworthy hints. As a pretext to this analysis, we assume that
the x-ray structures represent the products of free equilibra-
tion, at least in solution prior to crystallization. APRIL 5, 2013•VOLUME 288•NUMBER 14 Data collection and refinement statistics Thus, although
it is tempting to speculate that Met77 acts as a steric block
toward reannealing of the oxoG lesion, and indeed that could APRIL 5, 2013•VOLUME 288•NUMBER 14 JOURNAL OF BIOLOGICAL CHEMISTRY 10019 10019 APRIL 5, 2013•VOLUME 288•NUMBER 14 FIGURE 5. Conformation of the DNA backbone at the target base for M77A (A) and F114A (B). Double-headed arrows show the distance between the O8
atom and C2 carbon when modeled into the G structures (middle panels). In the M77A-oG structure, the sugar pucker switches to C4 exo to avoid repulsive
interactions between the O8 atom and the C2 carbon, whereas the rotation of the target oxoG to syn in the F114A-oG structure allows the sugar to remain in
the ground state C2 endo pucker (the distance between C2 and the N3 atom, which is now the closest nucleobase atom to the sugar, is shown). Helix Invasion by MutM Helix Invasion by MutM by guest on October 8, 2019
http://www.jbc.org/
Downloaded from FIGURE 5. Conformation of the DNA backbone at the target base for M77A (A) and F114A (B). Double-headed arrows show the distance between the O8
atom and C2 carbon when modeled into the G structures (middle panels). In the M77A-oG structure, the sugar pucker switches to C4 exo to avoid repulsive
interactions between the O8 atom and the C2 carbon, whereas the rotation of the target oxoG to syn in the F114A-oG structure allows the sugar to remain in
the ground state C2 endo pucker (the distance between C2 and the N3 atom, which is now the closest nucleobase atom to the sugar, is shown). FIGURE 6. Role of residues Met77 and Phe114 in oxoG repair. A, van der Waals contacts between Met77 and neighboring DNA backbone in the EC and LRC states. B,theMet77sidechainsfromaCalignmentofallavailableMutM-DNAcomplexesareshownwiththeindicatedintrahelicaloxoGandintrahelicalGstructure.Atotal
ofsixintrahelicaloxoG(ProteinDataBankcodes3GO8,3GP1,3GPP,3GPU,and3GQ3andF114A-oG)andsixintrahelicalG(ProteinDataBankcodes2F5O,3GPX,3GQ5,
2F5N, and 2F5P and F114A-G) structures were used for the alignment. In the right panel, the base pair 5 to the oxoG has been removed for clarity. The black arrow
representsthepathofbaseextrusion.ThepositionofthesidechainvarieswidelyintheintrahelicaloxoGstate,whereasintheextrahelicalstatethesidechainoccupies
a firm position between the extruded oxoG and the DNA. C, modeling the Phe114 side chain into the F114A-oG structure shows a distinct steric clash between Phe114
and Arg76, which forms two stabilizing hydrogen bonding interactions with the O8 from the syn oxoG. The distances are shown in Å. FIGURE 6. Role of residues Met77 and Phe114 in oxoG repair. Data collection and refinement statistics A, van der Waals contacts between Met77 and neighboring DNA backbone in the EC and LRC states. B,theMet77sidechainsfromaCalignmentofallavailableMutM-DNAcomplexesareshownwiththeindicatedintrahelicaloxoGandintrahelicalGstructure.Atotal
ofsixintrahelicaloxoG(ProteinDataBankcodes3GO8,3GP1,3GPP,3GPU,and3GQ3andF114A-oG)andsixintrahelicalG(ProteinDataBankcodes2F5O,3GPX,3GQ5,
2F5N, and 2F5P and F114A-G) structures were used for the alignment. In the right panel, the base pair 5 to the oxoG has been removed for clarity. The black arrow
representsthepathofbaseextrusion.ThepositionofthesidechainvarieswidelyintheintrahelicaloxoGstate,whereasintheextrahelicalstatethesidechainoccupies
a firm position between the extruded oxoG and the DNA. C, modeling the Phe114 side chain into the F114A-oG structure shows a distinct steric clash between Phe114
and Arg76, which forms two stabilizing hydrogen bonding interactions with the O8 from the syn oxoG. The distances are shown in Å. FIGURE 6. Role of residues Met77 and Phe114 in oxoG repair. A, van der Waals contacts between Met77 and neighboring DNA backbone in the EC and LRC states. B,theMet77sidechainsfromaCalignmentofallavailableMutM-DNAcomplexesareshownwiththeindicatedintrahelicaloxoGandintrahelicalGstructure.Atotal
ofsixintrahelicaloxoG(ProteinDataBankcodes3GO8,3GP1,3GPP,3GPU,and3GQ3andF114A-oG)andsixintrahelicalG(ProteinDataBankcodes2F5O,3GPX,3GQ5,
2F5N, and 2F5P and F114A-G) structures were used for the alignment. In the right panel, the base pair 5 to the oxoG has been removed for clarity. The black arrow
representsthepathofbaseextrusion.ThepositionofthesidechainvarieswidelyintheintrahelicaloxoGstate,whereasintheextrahelicalstatethesidechainoccupies
a firm position between the extruded oxoG and the DNA. C, modeling the Phe114 side chain into the F114A-oG structure shows a distinct steric clash between Phe114
and Arg76, which forms two stabilizing hydrogen bonding interactions with the O8 from the syn oxoG. The distances are shown in Å. when the latter is present (Fig. 6C); however, this clash appears
as though it could be alleviated through rotation about the
Phe114 C-C bond. In an unpublished work, we have observed Arg76 taking on a similar role in structures of MutM (contain-
ing Phe114) bound to oxoG-containing DNA in which the oxoG
has been repositioned onto the opposing “complementary” 10020
JOURNAL OF BIOLOGICAL CHEMISTRY Further
investigation will be necessary to fully explore the role of this
residue in oxoG recognition and repair. by guest on October 8, 2019
http://www.jbc.org/
Downloaded from bases from DNA but is ineffective at catalyzing their excision. Computational studies have indicated that MutM lowers the
barrier for base extrusion at oxoG sites by 7 kcal/mol more
than at sequence-matched G sites, thus providing a kinetic basis
for discrimination of lesions from undamaged DNA even at the
earliest stages of encounter (14). Wresting a nucleobase from
DNA and stabilizing it in the extrahelical state does not occur
spontaneously to any appreciable extent; hence MutM must
invest energy in the form of interfacial contacts to enforce pres-
entation of an extrahelical lesion to the enzyme active site. A
priori, these contacts would be expected to consist of two broad
classes: those that serve to destabilize the intrahelical state and
those that stabilize the extrahelical state. At present, MutM is the only base excision DNA repair
enzyme for which the structural features of encounter with an
intrahelical lesion have been elucidated in detail. These EC
structures indeed provide evidence suggesting that MutM
destabilizes the target base pair, with that pair being completely
destacked from its 3-neighbor and being severely buckled
toward the 5-neigbor (Fig. 1C). The structures also provide
evidence suggesting stabilization of the extrahelical LRC state,
namely acquisition of multiple interfacial contacts to the extra-
helical oxoG (including the oxoG-capping loop, OCL), inser-
tion of Arg112 into the helical stack to hydrogen bond with the
complementary C, and establishment of van der Waals contacts
with Met77. Here and in previous work, we have used site-di-
rected mutagenesis to probe the structural and (by inference)
energetic influence of residues in MutM that prominently fig-
ure in the protein-DNA interface. The large separation of the oxoG nucleobase from its com-
plementary C is associated with a significant shift of the DNA
backbone toward MutM at the site of the lesion, one that is even
more extrudogenic than the corresponding shift seen in the EC
structures (supplemental Fig. S2B, shifting an average 2.2 Å
from the G-bound to oxoG-bound state with the F114A muta-
tion versus 0.92 Å with Phe114 intact (14)). We conclude that the
extrudogenic backbone conformation alone is not sufficient to
induce extrusion of oxoG and that indeed Phe114 plays this role
of enforcer under physiologic conditions with wild-type MutM. 10020
JOURNAL OF BIOLOGICAL CHEMISTRY 10020 10020
JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 288•NUMBER 14•APRIL 5, 2013 Helix Invasion by MutM FIGURE 7. Effects of various mutations on oxoG conformation in duplex DNA during lesion recognition by MutM. Protein Data Bank codes (clockwise
from top) were 183D, 4G4R (this work), 1R2Y, and 4G4O (this work). Helix Invasion by MutM FIGURE 7. Effects of various mutations on oxoG conformation in duplex DNA during lesion recognition by MutM. Protein Data Bank codes (clockwise
from top) were 183D, 4G4R (this work), 1R2Y, and 4G4O (this work). FIGURE 7. Effects of various mutations on oxoG conformation in duplex DNA during lesion recognition by MutM. Protein Data Bank codes (clockwise
from top) were 183D, 4G4R (this work), 1R2Y, and 4G4O (this work). FIGURE 7. Effects of various mutations on oxoG conformation in duplex DNA during lesion recognition by MutM. Protein Data Bank codes (clockwise
from top) were 183D, 4G4R (this work), 1R2Y, and 4G4O (this work). strand; in those complexes, Arg76 forms stabilizing hydrogen
bonds to a syn undamaged guanine base located on the oppos-
ing strand, directly 5 to the base occupying the position of the
opposing C in our mutant structures.6 Preliminary mutagenesis
studies into the role of Arg76 in oxoG repair show that replace-
ment of the arginine side chain with alanine significantly
impairs base excision DNA repair of a duplex oxoG-containing
substrate (supplemental Fig. S4). However, substitution to
methionine or lysine also failed to rescue oxoG repair activity,
indicating that side chain length is not the determining factor
for the function of residue 76 (supplemental Fig. S4). Further
investigation will be necessary to fully explore the role of this
residue in oxoG recognition and repair. strand; in those complexes, Arg76 forms stabilizing hydrogen
bonds to a syn undamaged guanine base located on the oppos-
ing strand, directly 5 to the base occupying the position of the
opposing C in our mutant structures.6 Preliminary mutagenesis
studies into the role of Arg76 in oxoG repair show that replace-
ment of the arginine side chain with alanine significantly
impairs base excision DNA repair of a duplex oxoG-containing
substrate (supplemental Fig. S4). However, substitution to
methionine or lysine also failed to rescue oxoG repair activity,
indicating that side chain length is not the determining factor
for the function of residue 76 (supplemental Fig. S4). 10020
JOURNAL OF BIOLOGICAL CHEMISTRY The substantial decrease in base excision activity that accom-
panies the F114A mutation (Fig. 2A) thus appears to arise from
an inability of the mutant protein to populate the extruded state
that is required for catalysis of base excision. 6 R.-J. Sung, M. Zhang, Y. Qi, and G. L. Verdine, manuscript in preparation. REFERENCES 1. Lindahl, T. (1993) Instability and decay of the primary structure of DNA. Nature 362, 709–715 2. Barnes, D. E., and Lindahl, T. (2004) Repair and genetic consequences of
endogenous DNA base damage in mammalian cells. Annu. Rev. Genet. 38,
445–476 3. Michaels, M. L., and Miller, J. H. (1992) The GO system protects organ-
isms from the mutagenic effect of the spontaneous lesion 8-hydroxygua-
nine (7,8-dihydro-8-oxoguanine). J. Bacteriol. 174, 6321–6325 4. Nash, H. M., Bruner, S. D., Schärer, O. D., Kawate, T., Addona, T. A.,
Spooner, E., Lane, W. S., and Verdine, G. L. (1996) Cloning of a yeast
8-oxoguanine DNA glycosylase reveals the existence of a base-excision
DNA-repair protein superfamily. Curr. Biol. 6, 968–980 5. van der Kemp, P. A., Thomas, D., Barbey, R., de Oliveira, R., and Boiteux,
S. (1996) Cloning and expression in Escherichia coli of the OGG1 gene of
Saccharomyces cerevisiae, which codes for a DNA glycosylase that excises
7,8-dihydro-8-oxoguanine
and
2,6-diamino-4-hydroxy-5-N-methyl-
formamidopyrimidine. Proc. Natl. Acad. Sci. U.S.A. 93, 5197–5202 6. Fedorova, O. S., Nevinsky, G. A., Koval, V. V., Ishchenko, A. A., Vasilenko,
N. L., and Douglas, K. T. (2002) Stopped-flow kinetic studies of the inter-
action between Escherichia coli Fpg protein and DNA substrates. Bio-
chemistry 41, 1520–1528 7. Fromme, J. C., and Verdine, G. L. (2002) Structural insights into lesion
recognition and repair by the bacterial 8-oxoguanine DNA glycosylase
MutM. Nat. Struct. Biol. 9, 544–552 8. Oda, Y., Uesugi, S., Ikehara, M., Nishimura, S., Kawase, Y., Ishikawa, H.,
Inoue, H., and Ohtsuka, E. (1991) NMR studies of a DNA containing
8-hydroxydeoxyguanosine. Nucleic Acids Res. 19, 1407–1412 Phe114 is inserted into DNA at the site of a pronounced DNA
bend that is localized to 3-side of the target base pair and that
completely destacks the target base pair from its 3-neighbor. This association suggests the possibility of causation, specifi-
cally that Phe114 might actually be responsible for formation of
the bend. Our results with the F114A-G complex rule out that
scenario conclusively: the bend is just as pronounced in this
complex as in other intrahelical recognition complexes,
although the target base pair is completely unbuckled. Met77 is
also inserted into the duplex at the kinked site, yet it also is not
required to produce the bend, as judged by inspection of the
M77A-G structure. Hence we conclude that interfacial con- 9. Lipscomb, L. A., Peek, M. E., Morningstar, M. L., Verghis, S. Helix Invasion by MutM tacts by residues other than Phe114 and Met77 produce the sig-
nature bend seen thus far in all MutM/DNA complexes. The
observation that the bend is retained upon removal of side
chains that invade DNA having only intact base pairs suggests
that bending is an intrinsic feature of MutM/DNA binding, in
which case the bend would be expected to migrate along with
MutM as the protein slides along DNA while searching for
lesions. By the same token, during every base pair-wise translo-
cation event, the side chain of Phe114 would have to clear the
obstruction presented by the 5-base base pair, suggesting that
Phe114 is in nearly constant repulsive contact with DNA during
the lesion search. Consistent with this notion, deleting the
Phe114 side chain modestly increases the affinity of the protein
for DNA (supplemental Fig. S1) and also increases the rate of
processive sliding (18). interaction with a lesion, one characterized by complete rup-
ture of the target base pair but failure to extrude the target oxoG
nucleobase (Fig. 7, right structure). The observation of such a
structure highlights the fact that rupture of a base pair with at
least partial preservation of stacking interactions is less ener-
getically expensive than extrusion of a nucleobase from DNA. Enforcement of extrusion requires additional repulsive interac-
tions beyond those seen in the F114A-oxoG structure, and
those repulsive interactions are provided in the wild-type
enzyme by the Phe114 side chain being jammed into the DNA
helix so forcefully as to buckle the target base pair. The very
observation of such an unusual structure attests to the extraor-
dinary ability of MutM to pry open the DNA duplex upon
encounter with an oxoG lesion. Although F114-oxoG can
hardly be considered a physiologic complex, we believe it pro-
vides insight into an exceedingly important but poorly under-
stood physiologic property of MutM. When base-paired
against A, MutM is nearly devoid of base excision activity; evo-
lution provides strong pressure against MutM processing
A:oxoG lesions, because such repair would have the pro-muta-
genic effect of accelerating C:G to T:A transversion mutations. In the A:oxoG base pair, the A and oxoG adopt a strong, stable
Hoogsteen pairing configuration, with A adopting the anti gly-
cosidic bond configuration and oxoG adopting the syn config-
uration, the same as that observed in the F114A-oxoG struc-
ture. DISCUSSION An abundance of structural, biochemical, and computational
evidence indicates that MutM actively promotes the extrusion
of oxoG lesions from DNA. This extrusion is of course a pre-
requisite to execution of a catalytic DNA repair reaction cas-
cade by MutM, because the single enzyme active site that accel-
erates all steps of the repair cascade is capable of recognizing
only a fully extrahelical nucleobase. Interestingly, evidence sug-
gests that MutM also accelerates the extrusion of undamaged Prior to the present work, two distinct classes of mutations
were known. Mutations that affect catalysis but not substrate
recognition, typified by E3Q, afford extrahelical lesion recogni-
tion complexes, LRCs (Fig. 7, left structure). Mutations that
deprive the enzyme of its ability to stabilize the extrahelical
state, typified by the OCL mutant V222P, afford intrahelical
lesion ECs (Fig. 7, middle structure). In the present work, we
have found that the M77A mutation also leads to entrapment of
an EC state; hence we conclude that Met77 stabilizes the extra-
helical state. Here we have also observed a third state of MutM 6 R.-J. Sung, M. Zhang, Y. Qi, and G. L. Verdine, manuscript in preparation. APRIL 5, 2013•VOLUME 288•NUMBER 14 JOURNAL OF BIOLOGICAL CHEMISTRY 10021 APRIL 5, 2013•VOLUME 288•NUMBER 14 Helix Invasion by MutM Although no structure is yet available for MutM bound to
an intrahelical A:oxoG base pair, the F114A-oxoG structure
provides a basis for extrapolation to A:oxoG. As shown in sup-
plemental Fig. S5, if the complementary C in the F114A-oxoG
structure were replaced with A, the A, by virtue of its larger ring
size, would span the entire gap between the two target bases and
would be perfectly positioned to Hoogsteen pair with oxoG. oxoG in turn would be hydrogen-bonded to Arg76. Taken
together, the unusual features of the F114A structure leaves the
DNA perfectly poised to accept an A:oxoG and form a stable
intrahelical recognition complex with it. This stabilization of
A:oxoG in the intrahelical state would increase the kinetic bar-
rier to extrusion of oxoG, thereby rendering A:oxoG a poor
substrate for base excision. What is difficult to predict is the
effect of Phe114 on the A:oxoG complex, but based on the avail-
able information, it seems likely that Phe114 buckles the A:oxoG
base pair, but the destabilizing effect of buckling is more than
compensated by the intrinsically high stability of the A:oxoG
base pair, the Arg76-oxoG hydrogen bond, and the other inter-
molecular contacts that stabilize the gap in the F114A-oxoG
complex. Phe114 is inserted into DNA at the site of a pronounced DNA Acknowledgments—We are grateful to staff from the 24-ID beamline
of the Argonne Photon Source and at the X29 beamline of the
National Synchrotron Light Source for assistance in data collection. We thank Kwangho Nam and Danaya Pakotiprapha for constructive
feedback during the writing of this manuscript. We also thank mem-
bers of the Verdine group for helpful advice. Acknowledgments—We are grateful to staff from the 24-ID beamline
of the Argonne Photon Source and at the X29 beamline of the
National Synchrotron Light Source for assistance in data collection. We thank Kwangho Nam and Danaya Pakotiprapha for constructive
feedback during the writing of this manuscript. We also thank mem-
bers of the Verdine group for helpful advice. by guest on October 8, 2019
http://www.jbc.org/
Downloaded from Helix Invasion by MutM 12. Fromme, J. C., and Verdine, G. L. (2003) DNA lesion recognition by the
bacterial repair enzyme MutM. J. Biol. Chem. 278, 51543–51548 a new tool for fast protein structure alignment in three dimensions. Acta
Crystallogr. D Biol. Crystallogr. 60, 2256–2268 a new tool for fast protein structure alignment in three dimensions. Acta
Crystallogr. D Biol. Crystallogr. 60, 2256–2268 13. Banerjee, A., Santos, W. L., and Verdine, G. L. (2006) Structure of a DNA
glycosylase searching for lesions. Science 311, 1153–1157 24. Painter, J., and Merritt, E. A. (2006) TLSMD web server for the generation
of multi-group TLS models. J. Appl. Crystallogr. 109–111 14. Qi, Y., Spong, M. C., Nam, K., Banerjee, A., Jiralerspong, S., Karplus, M.,
and Verdine, G. L. (2009) Encounter and extrusion of an intrahelical lesion
by a DNA repair enzyme. Nature 462, 762–766 25. Painter, J., and Merritt, E. A. (2006) Optimal description of a protein
structure in terms of multiple groups undergoing TLS motion. Acta Crys-
tallogr. D Biol. Crystallogr. 62, 439–450 y
y
15. Ishchenko, A. A., Vasilenko, N. L., Sinitsina, O. I., Yamkovoy, V. I., Fedo-
rova, O. S., Douglas, K. T., and Nevinsky, G. A. (2002) Thermodynamic,
kinetic, and structural basis for recognition and repair of 8-oxoguanine in
DNA by Fpg protein from Escherichia coli. Biochemistry 41, 7540–7548 26. Davis, I. W., Murray, L. W., Richardson, J. S., and Richardson, D. C. (2004)
MOLPROBITY. Structure validation and all-atom contact analysis for
nucleic acids and their complexes. Nucleic Acids Res. 32, W615–W619 27. Laskowski, R. A., MacArthur, M. W., Moss, D. S., and Thornton, J. M. (2011) PROCHECK: a program to check the stereochemical quality of
protein structures. J. Appl. Crystallogr. 26, 283–291 16. Nam, K., Verdine, G. L., and Karplus, M. (2009) Analysis of an anomalous
mutant of MutM DNA glycosylase leads to new insights into the catalytic
mechanism. J. Am. Chem. Soc. 131, 18208–18209 28. Lu, X.-J., and Olson, W. K. (2003) 3DNA. A software package for the
analysis, rebuilding and visualization of three-dimensional nucleic acid
structures. Nucleic Acids Res. 31, 5108–5121 17. Qi, Y., Nam, K., Spong, M. C., Banerjee, A., Sung, R.-J., Zhang, M., Karplus,
M., and Verdine, G. L. (2012) Strandwise translocation of a DNA glyco-
sylase on undamaged DNA. Proc. Natl. Acad. Sci. U.S.A. 109, 1086–1091 29. Schrodinger, LLC. (2010) The PyMOL Molecular Graphics System, Ver-
sion 1.3r1 18. Dunn, A. R., Kad, N. M., Nelson, S. REFERENCES M., Miller,
E. M., Rich, A., Essigmann, J. M., and Williams, L. D. (1995) X-ray struc-
ture of a DNA decamer containing 7,8-dihydro-8-oxoguanine. Proc. Natl. Acad. Sci. U.S.A. 92, 719–723 10. Plum, G. E., Grollman, A. P., Johnson, F., and Breslauer, K. J. (1995) Influ-
ence of the oxidatively damaged adduct 8-oxodeoxyguanosine on the con-
formation, energetics, and thermodynamic stability of a DNA duplex. Bio-
chemistry 34, 16148–16160 11. Bowman, B. R., Lee, S., Wang, S., and Verdine, G. L. (2008) Structure of the
E. coli DNA glycosylase AlkA bound to the ends of duplex DNA. A system
for the structure determination of lesion-containing DNA. Structure 16,
1166–1174 10022
JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 288•NUMBER 14•APRIL 5, 2013 VOLUME 288•NUMBER 14•APRIL 5, 2013 Rou-Jia Sung, Michael Zhang, Yan Qi and Gregory L. Verdine
8-Oxoguanine DNA Glycosylase MutM
Structural and Biochemical Analysis of DNA Helix Invasion by the Bacterial
doi: 10.1074/jbc.M112.415612 originally published online February 12, 2013
2013, 288:10012-10023.
J. Biol. Chem. Rou-Jia Sung, Michael Zhang, Yan Qi and Gregory L. Verdine
8-Oxoguanine DNA Glycosylase MutM
Structural and Biochemical Analysis of DNA Helix Invasion by the Bacterial
doi: 10.1074/jbc.M112.415612 originally published online February 12, 2013
2013, 288:10012-10023. J. Biol. Chem. Helix Invasion by MutM R., Warshaw, D. M., and Wallace, S. S. (2011) Single Qdot-labeled glycosylase molecules use a wedge amino acid
to probe for lesions while scanning along DNA. Nucleic Acids Res. 39,
7487–7498 30. Porello, S. L., Leyes, A. E., and David, S. S. (1998) Single-turnover and
pre-steady-state kinetics of the reaction of the adenine glycosylase
MutY with mismatch-containing DNA substrates. Biochemistry 37,
14756–14764 19. Maxam, A. M., and Gilbert, W. (1977) A new method for sequencing
DNA. Proc. Natl. Acad. Sci. U.S.A. 74, 560–564 31. Qi, Y., Spong, M. C., Nam, K., Karplus, M., and Verdine, G. L. (2010)
Entrapment and structure of an extrahelical guanine attempting to enter
the active site of a bacterial DNA glycosylase, MutM. J. Biol. Chem. 285,
1468–1478 by guest on October 8, 2019
http://www.jbc.org/
Downloaded from 20. Otwinowski, Z., and Minor, W. (1997) Methods in Enzymology, 276,
307–326 21. Adams, P. D., Afonine, P. V., Bunkóczi, G., Chen, V. B., Davis, I. W., Echols,
N., Headd, J. J., Hung, L.-W., Kapral, G. J., Grosse-Kunstleve, R. W., Mc-
Coy, A. J., Moriarty, N. W., Oeffner, R., Read, R. J., Richardson, D. C.,
Richardson, J. S., Terwilliger, T. C., and Zwart, P. H. (2010) PHENIX. A
comprehensive Python-based system for macromolecular structure solu-
tion. Acta Crystallogr. D Biol. Crystallogr. 66, 213–221 32. Sung, R.-J., Zhang, M., Qi, Y., and Verdine, G. L. (2012) Sequence-depen-
dent structural variation in DNA undergoing intrahelical inspection by
the DNA glycosylase MutM. J. Biol. Chem. 287, 18044–18054 by guest on October 8, 2019
http://www.jbc.org/
m 33. Koval, V. V., Kuznetsov, N. A., Ishchenko, A. A., Saparbaev, M. K., and
Fedorova, O. S. (2010) Real-time studies of conformational dynamics of
the repair enzyme E. coli formamidopyrimidine-DNA glycosylase and its
DNA complexes during catalytic cycle. Mutat. Res. 685, 3–10 22. Emsley, P., and Cowtan, K. (2004) Coot. Model-building tools for molec-
ular graphics. Acta Crystallogr. D Biol. Crystallogr. 60, 2126–2132 23. Krissinel, E., and Henrick, K. (2004) Secondary-structure matching (SSM), 34. Brunger, A. (1993) Acta Crystallogr. D 49, 24–36 APRIL 5, 2013•VOLUME 288•NUMBER 14 JOURNAL OF BIOLOGICAL CHEMISTRY 10023 Wh
i
f
hi
When this article is cited
• 10.1074/jbc.M112.415612
Access the most updated version of this article at doi: Wh
i
f
hi
When this article is cited
• When a correction for this article is posted
• to choose from all of JBC's e-mail alerts
Click here to choose from all of JBC's e-mail alerts
Click here by guest on October 8, 2019
http://www.jbc.org/
Downloaded from Supplemental material: Supplemental material:
http://www.jbc.org/content/suppl/2013/02/12/M112.415612.DC1 http://www.jbc.org/content/288/14/10012.full.html#ref-list-1
This article cites 32 references, 9 of which can be accessed free at
|
https://openalex.org/W2901186372
|
https://revistas.unab.edu.co/index.php/medunab/article/download/3394/2886, https://www.redalyc.org/journal/719/71964816009/71964816009.pdf
|
es
|
Consecuencias del amonio en la fatiga central en atletas, posible efecto neuroprotector del ejercicio
|
Medunab
| 2,018
|
cc-by
| 4,050
|
Artículo de reflexión no derivado de investigación
Vol 21(1): 115-121, abril - julio 2018
https://doi.org/1029375/01237047.3394
Consecuencias del amonio en la fatiga central en atletas,
posible efecto neuroprotector del ejercicio
Ammonium consequences in athletes’ central fatigue and its possible
neuroprotection effect thanks to physical activity
Consequências do amônio na fadiga central em atletas, possível
efeito neuroprotetor do exercício
Javier Porras-Álvarez, L. Educación Física, MSc. PhD (c)*
Resumen
Introducción. La fatiga central en el deporte está asociada
a los efectos del amonio. La principal fuente de producción
de amonio durante el ejercicio es el músculo esquelético.
El amonio se genera como consecuencia del metabolismo
energético, debido a la oxidación de aminoácidos y a
la desaminación del nucleótido de adenosin trifosfato.
Objetivo. Presentar una reflexión sobre el efecto del amonio
durante el ejercicio de alta intensidad y su relación con la
fatiga central en atletas. Discusión. Durante el ejercicio,
la concentración de amonio alcanza valores superiores a
200µM (micromolar); sin embargo, en un adulto promedio
se considera que valores superiores a 60µM en sangre
manifiestan un trastorno por hiperamonemia. El amonio
influye en la disminución del rendimiento en atletas y está
asociado con los efectos nocivos para la salud en pacientes
con encefalopatía hepática. Conclusiones. La práctica del
ejercicio físico genera neuroprotección contra las altas
concentraciones de amonio en el cerebro, pues, durante el
ejercicio con altas concentraciones de amonio, los atletas
no presentan los síntomas de pacientes con encefalopatía
hepática, lo que implica adaptaciones metabólicas que
juegan un papel importante en el metabolismo del amonio
en el cerebro. [Porras-Álvarez J. Consecuencias del amonio
en la fatiga central en atletas, posible efecto neuroprotector
del ejercicio. MedUNAB. 2018;21(1): 115-121 doi:
10.29375/01237047.3394].
Palabras
clave:
Fatiga;
Compuestos
de Amonio
* Licenciado en ciencias de la Educación Física, magíster en Ciencias de la Actividad Física y Deporte, doctorando en Medicina del Deporte, doctorando
en Ciencias Biomédicas. Docente de la Universidad Pedagógica y Tecnológica de Colombia, grupo de investigación FIMED, Tunja, Boyacá, Colombia.
Correspondencia: Javier Porras-Álvarez. Vereda Molino, municipio de Saboyá. Celular: +57 3143029381. E-mail: porrasalvarez@gmail.com; javier.
porras@uptc.edu.co.
Artículo recibido: 17 de enero de 2018.
Aceptado: 26 de marzo de 2018.
115
Javier Pórras-Álvarez, L. Educación Física., MSc., PhD (c)
Vol 21(1):115-121., abril - julio 2018
Cuaternario; Entrenamiento de Resistencia; Ejercicio;
Sistema Inmunológico; Ácido Úrico; Encefalopatía
Hepática.
Abstract
Introduction. Central fatigue in sports training is associated
with ammonium effects within the human body. The
ammonium main production source during physical training
is located in skeletal muscles and it is generated as a result
of energy metabolism. This process is caused by amino
acids oxidation and adenosine triphosphate nucleotide
deamination. Objective. This article’s objective is to present
an analysis regarding ammonium effects when high intensity
sports are performed and its relation with central fatigue in
athletes. Discussion. When high intensity sport practices
are performed, ammonium concentration levels can reach
values higher than 200 µM (micromolar). However, it is
considered that an average adult with ammonium levels
higher than 60µM evidences a hyperammonemia disorder.
Ammonium has direct influence in the decline of athletic
performance and it is associated with harmful effects for
hepatic encephalopathy patients. Conclusions. Physical
activity practice creates neuroprotection against highquantities of ammonium in the brain. Although in physical
practices athletes have high amounts of ammonium, they do
not show symptoms related to hepatic encephalopathy; thus,
this situation implies that metabolic adaptations have an
important role within ammonium metabolism in the brain.
[Porras-Álvarez J. Ammonium consequences in athletes’
central fatigue and its possible neuroprotection effect
thanks to physical activity. MedUNAB. 2018;21(1):115-121
doi: 10.29375/01237047.3394].
Key words: Fatigue;
Quaternary
Ammonium
Compounds; Resistance Training; Exercise; Immune
System; Uric Acid; Hepatic Encephalopathy.
Resumo
Introdução. A fadiga central no esporte está associada aos
efeitos do amônio. A fonte de produção do amônio mais
importante durante o exercício é o músculo esquelético.
O amônio é gerado como consequência do metabolismo
energético, devido à oxidação dos aminoácidos e à
desaminação do trifosfato de adenosina. Objetivo. Apresentar
uma reflexão sobre o efeito do amônio durante o exercício de
alta intensidade e sua relação com a fadiga central em atletas.
Discussão. Durante o exercício, a concentração do amônio
atinge valores superiores a 200µM (micromol); no entanto,
em um adulto médio, considera-se que valores superiores a
60µM no sangue mostram um distúrbio por hiperamonemia.
O amônio influencia a diminuição do desempenho em
atletas e está associado a efeitos nocivos para a saúde
em pacientes com encefalopatia hepática. Conclusões.
A prática de exercício físico gera neuroproteção contra
altas concentrações do amônio no cérebro, pois, durante
o exercício com altas concentrações do amônio, os atletas
não apresentam sintomas de encefalopatia hepática, o que
implica adaptações metabólicas que desempenham um papel
importante no metabolismo do amônio no cérebro. [PorrasÁlvarez J. Consequências do amônio na fadiga central em
atletas, possível efeito neuroprotetor do exercício. MedUNAB.
2018;21(1):115-121 doi: 10.29375/01237047.3394].
Palavras chave: Fadiga; Compostos de Amônio
Quaternário; Treinamento de Resistência; Exercício;
Sistema Imunitário; Ácido Úrico; Encefalopatia Hepática.
Introducción
El entrenamiento diario de alta intensidad al que se
someten los atletas induce cambios y adaptaciones en
los diferentes órganos y sistemas del cuerpo humano,
llevando al límite los parámetros bioquímicos y
fisiológicos. No obstante, hay barreras que impiden, por
ejemplo, ir más rápido, saltar más alto o realizar el gesto
de forma perfecta. Aunque las marcas y el nivel técnico
de los atletas mejoran continuamente, hay procesos
bioquímicos que actúan como límites o barreras que
impiden al atleta realizar marcas supra humanas. De
esta forma, la fatiga se define como una disminución
reversible de la capacidad de rendimiento. A esta se han
atribuido causas centrales y periféricas, que generan
en el atleta la incapacidad de seguir manteniendo una
velocidad dada o mejorar el rendimiento físico.
Las causas periféricas hacen referencia a aquellas que
ocurren, por ejemplo, a nivel muscular, relacionadas
con: el porcentaje ideal del tipo de fibras musculares
que requiere la modalidad deportiva; un adecuado
nivel enzimático; una disminución de sustratos
energéticos disponibles, como glucosa, ácidos grasos
e incapacidad para metabolizar oxígeno; o a altas
concentraciones de adenosín monofostato (1, 2).
Por su parte, las causas centrales se refieren a las
que ocurren en el cerebro. Sustancias o moléculas
que se producen en el músculo, como producto
del metabolismo energético, pueden causar fatiga
central, convirtiéndose en una barrera o límite en el
desempeño del atleta.
La fatiga central en los atletas se da por múltiples
factores, bien sean fisiológicos, psicológicos,
ambientales, o mecanismos neurohumorales y del
116
Consecuencias del amonio en la fatiga central en atletas, posible efecto neuroprotector del ejercicio
metabolismo cerebral. Al respecto, en 2004, Nybo
indica que la fatiga central es el último factor que
establece el punto de agotamiento, debido a cambios
metabólicos, circulatorios, de neurotransmisores,
termodinámicos, o a otros trastornos de la homeostasis
cerebral, asociados a la acumulación o al agotamiento
de diferentes sustratos en el cerebro. Estos implican
la serotonina, la dopamina, la noradrenalina, las
interleucinas, la hipoglicemia, la hipertermia y
también al amoniaco o amonio (1).
El amoniaco (NH3) es un compuesto químico cuya
molécula está formada por un átomo de nitrógeno y
tres átomos de hidrógeno. En disolución acuosa, como
el medio metabólico humano, este se transforma en ion
amonio (NH4+). El amonio es, por naturaleza, tóxico,
pues puede atravesar la barrera hematoencefálica,
ingresar al sistema nervioso central (SNC) e impedir
que las motoneuronas funcionen de manera adecuada,
ocasionando fatiga central (1, 3).
Durante el ejercicio, los atletas producen grandes
concentraciones de amonio, producto del metabolismo
energético, compuesto que puede influir en la
disminución del rendimiento atlético. Asimismo,
altas concentraciones de amonio están asociadas a
efectos nocivos para la salud, como se evidencia en los
pacientes con encefalopatía hepática (4-6). Por tanto, el
objetivo de este artículo es reflexionar acerca del efecto
del amonio durante el ejercicio de alta intensidad y su
relación con la fatiga central en atletas.
la oxidación de ciertos aminoácidos, principalmente
alanina, aspartato, glutamina, isoleucina, leucina
y valina y, por otra parte, a la desaminación del
nucleótido de adenosín trifosfato (ATP, del inglés
adenosine triphosphate), nucleótido esencial en la
producción de energía para la contracción muscular
(2, 7-10). En los pacientes con HE, el amonio se
genera, principalmente, en el intestino grueso, en la
porción del colon, aunque también se genera amonio
en otros órganos, como corazón, páncreas, estomago,
pulmones, cerebro y riñón (11). En la Figura 1 se
presentan las dos fuentes principales de producción
de amonio en atletas y pacientes con HE.
El ejercicio de alta intensidad requiere, de forma
inmediata, de grandes cantidades de energía. Para
cubrir las necesidades energéticas de la contracción
muscular, esta energía es proporcionada a través de
la hidrolisis del ATP, un nucleótido esencial como
fuente principal de energía química para la mayoría de
las funciones celulares (9, 10). El ATP está formado
por una base nitrogenada (adenina), un azúcar de
tipo pentosa (ribosa), enlazados a tres grupos fosfato.
También, es requerido para que se produzca el
acortamiento de los filamentos de actina-miosina y
Concentración y producción de amonio
durante el ejercicio
En seres humanos sanos, la concentración normal
de amonio en sangre es de alrededor de 20µM
(micromolar) hasta 60µM (1µM equivale a 1µmol/l:
micromole/litro). Por su parte, los atletas alcanzan
concentraciones de amonio superiores a 200µM
durante el ejercicio, tanto de alta intensidad como
de larga duración, según lo reportado en varias
investigaciones (4-6). Valores de amonio en sangre
superior a 60µM son considerados un trastorno
por hiperamonemia, asociada a pacientes con
encefalopatía hepática (HE).
La principal fuente de producción de amonio (NH3
y NH4+), durante el ejercicio, se sitúa en el músculo
esquelético. Este amonio se genera como consecuencia
del metabolismo energético, debido, por una parte, a
Figura 1. Producción de amonio durante el ejercicio en
atletas y en pacientes con HE.
Durante el ejercicio de alta intensidad, la producción de amonio
en los atletas se incrementa más de un 200% y se origina,
principalmente, en el músculo esquelético en contracción. En
los pacientes con HE, aproximadamente el 25% del amonio es
generado en el sistema gastrointestinal. El amonio, generado tanto
en músculo como intestinos, se metaboliza principalmente en el
hígado, pero también en el músculo esquelético y en el cerebro.
Fuente: elaborado por autor
117
Javier Pórras-Álvarez, L. Educación Física., MSc., PhD (c)
Vol 21(1):115-121., abril - julio 2018
generar la contracción muscular.
La energía se genera a partir del rompimiento de un
grupo fosfato de la molécula de ATP, con la acción
de la enzima ATPasa. El rompimiento de este enlace
produce la energía para los procesos celulares, en
este caso, para que se realice la contracción muscular.
Consecuentemente, como producto de la reacción,
se originan adenosín difosfato (ADP) y un grupo
fosfato. El ADP es una molécula de adenosina con
dos grupos fosfato, que se va acumulando a medida
que se realiza el ejercicio, como producto de la
reacción. La acumulación del ADP bloquea o inhibe
la enzima ATPasa y, consecuentemente, disminuye la
disponibilidad de energía para el músculo. Por tanto,
se requiere obtener nuevamente un grupo fosfato,
para reemplazar el fosfato liberado de la molécula
de ATP. Este grupo fosfato se obtiene de reacciones
metabólicas de otros compuestos, como fosfocreatina
(PC), glucosa, ácidos grasos y determinados
aminoácidos (2, 9-10, 12).
Durante el ejercicio, las reservas de glucógeno y los
compuestos ricos en energía, como la PC y la glucosa
presentes en las células, se agotan rápidamente, a
intensidades superiores al 80% del máximo consumo
de oxígeno (VO2). Como consecuencia, se activa la
ruta de la enzima miokinasa (MK), la cual cataliza
la reacción de dos ADP para formar un ATP. Como
resultado de la reacción, queda adenosín monofosfato
(AMP). Esto significa que las dos moléculas de ADP
suman cuatro grupos fosfato, por lo que una molécula
cede un grupo fosfato a la otra molécula. Esta es la
reacción que cataliza la MK. Consecuentemente, se
obtiene ATP y la otra molécula queda con un grupo
fosfato (AMP). De esta manera, se dispone de una
ruta metabólica alternativa para proporcionar energía
de forma inmediata.
La consecuencia de la utilización de la ruta de la MK
es la acumulación de AMP, un metabolito responsable
de causar fatiga periférica, porque la acumulación de
AMP bloquea o inhibe la acción de la enzima MK. Por
esto, también se acumula ADP, un potente inhibidor
de la ATPasa, generando así fatiga periférica (9, 10).
El AMP es un desecho metabolito, del cual el organismo
debe liberase a medida que se va produciendo (2,
9-10, 13). No obstante, la eliminación del AMP
implica el rompimiento de este nucleótido (AMP) o
su desaminación, es decir, la eliminación del grupo
α-amino de los aminoácidos, en la cual se incluyen
dos tipos de reacciones químicas: transaminación
y aminación. Este proceso produce amonio, cuya
acumulación, en consecuencia, aumenta durante el
ejercicio de alta intensidad (Figura 2). Si se agotan
las reservas de glucógeno y/o la disponibilidad de
glucosa es baja, la concentración de amonio aumenta
más rápidamente durante el ejercicio (1, 8, 14-16).
Producción de amonio en pacientes con
encefalopatía hepática
Se han evidenciado grandes concentraciones de
amonio en pacientes con HE. Esta es una grave
complicación de la cirrosis hepática, donde el daño
ocasionado en el hígado afecta el metabolismo del
amonio, causa una excesiva acumulación y alcanza
valores superiores o iguales a 60µM. Este fenómeno
es considerado hiperamonemia, produce efectos
perjudiciales para la función cerebral, el control de la
circulación, la regulación osmótica y el metabolismo
neuronal, así como lesión e inflamación de los
astrocitos. Tales factores parecen estar implicados en
la patogénesis de la manifestaciones neurológicas de
la enfermedad (17-20).
En los pacientes con HE, el amonio se genera en el
intestino, producto del catabolismo de las proteínas de
la dieta, de la desaminación de la glutamina y por la
acción de bacterias presentes en la flora intestinal (11,
21). La ureasa suministra a las bacterias el nitrógeno
(N) necesario para la síntesis de aminoácidos. Este
elemento se encuentra presente en el intestino como
producto de la ingesta de una dieta rica en proteínas,
por su parte, las bacterias intestinales producen amonio
como producto de desecho, durante la digestión de
proteínas. Consecuentemente, se produce un aumento
en la concentración de amonio que, debido a la lesión
de los hepatocitos que presentan los pacientes con HE
y a un inadecuado funcionamiento del ciclo de la urea
en el hígado, no puede ser totalmente metabolizado.
En condiciones normales, el hígado metaboliza todo
el amonio en urea, una molécula no tóxica que se
elimina fácilmente en la orina.
Metabolismo del amonio y origen de la
fatiga central
En condiciones normales, el cerebro metaboliza el
amonio producido, pues los astrocitos en el cerebro
son fuente abundante de la enzima glutamina sintasa
118
Consecuencias del amonio en la fatiga central en atletas, posible efecto neuroprotector del ejercicio
es el principal neurotransmisor excitador del SNC, por
su parte, la glutamina es precursor del ácido gammaaminobutírico (GABA) y, a su vez, el principal
neurotransmisor inhibidor del SNC. Por lo tanto, no
permite que las motoneuronas funcionen de forma
adecuada y afecta el rendimiento motor, al originar
fatiga central (1, 24). Asimismo, el amonio generado
en el músculo esquelético, al ingresar al SNC, no
permite que las motoneuronas funcionen de forma
adecuada, afectan el rendimiento motor y causan
fatiga central y periférica (13, 23). La fatiga central,
en consecuencia, limita la capacidad del organismo
para realizar actividades deportivas a velocidades o
intensidades supra humanas.
Ejercicio y sistema inmune adaptativo
en el metabolismo del amonio
Figura 2. Producción de amonio en el músculo esquelético
en ejercicio.
Durante el ejercicio de alta intensidad, se requiere de una
cantidad energía (ATP y PC) para la contracción muscular. Como
consecuencia de la reacción, se obtiene ADP, si este se acumula,
bloquea la ATPasa. Del ADP acumulado se puede obtener ATP
gracias a la acción de la enzima MK, como consecuencia de la
reacción se obtiene AMP. Si este se acumula, bloquea la MK, por
lo que la acumulación de AMP es inminente, cuya desaminación
produce amonio.
*ATP: adenosin trifosfato, PC: fosfocreatina, ATPasa: enzima
ATPasa, ADP: adenosin difosfato, MK: enzima mioquinasa,
AMP: adenosin monofosfato, NH4+: amonio, +: este símbolo
significa cantidad, │: este símbolo significa la reacción se inhibe
o bloquea.
Fuente: Elaborado por autor
(GS), que convierte el amonio más glutamato en
glutamina (glutamato + NH4+→Glutamina), una
molécula neutra no tóxica (9, 22). Posteriormente, la
glutamina es llevada hasta el hígado, donde tiene lugar
el proceso de desaminación del nitrógeno y, mediante
el eficaz ciclo de la urea, el exceso de amonio es
eliminado del organismo mediante la orina.
Sin embargo, debido a que el cerebro no tiene un
eficaz ciclo de la urea, depende únicamente de la
síntesis de glutamina para la eliminación del exceso
de amonio (23). Por ello, en el cerebro se convierte
gran parte del glutamato en glutamina. El glutamato
El análisis del transporte de amonio ha permitido
identificar, por una parte, una importante función del
músculo esquelético en la eliminación de amonio
(17) y, por otra, que el sistema inmune adaptativo está
relacionado al metabolismo del amonio, mediante las
células T. Estas son una fuente abundante de glutamato
extracelular en el cerebro, que es tomado por los astrocitos
y provoca la secreción de moléculas neuroprotectoras.
Además, estas células protegen la eliminación de
glutamato de los astrocitos, que se disminuye por el
estrés oxidativo (25). Por tanto, la interacción entre las
células T y los astrocitos juega un papel en el mecanismo
molecular, pues ocasiona neuroprotección e impide
la lesión en el SNC (25). Asimismo, el glutamato
extracelular proveniente de las células T, combinado con
NH4+ (producido por la contracción muscular), genera
glutamina como producto de la reacción y favorece así
la eliminación de glutamato de los astrocitos.
El recuento de células T o linfocitos T aumenta hasta
150% después de ejercicios de alta intensidad y corta
duración. No obstante, este aumento parece estar
más relacionado con personas sin entrenamiento que
con atletas de alto nivel competitivo (7, 26). Durante
ejercicios de resistencia de larga duración, como la
maratón, las células T disminuyen transitoriamente;
en cambio, en ejercicios de corta duración y moderada
intensidad (alrededor de una hora o menos), no tiene
efectos sobre las respuestas proliferativas de los
linfocitos T (7, 26).
De alguna manera, la práctica del ejercicio físico
regular induce adaptaciones metabólicas que hace
119
Javier Pórras-Álvarez, L. Educación Física., MSc., PhD (c)
Vol 21(1):115-121., abril - julio 2018
que las altas concentraciones de amonio, perjudiciales
para la salud, sean toleradas por los atletas. Como
se mencionó, durante el ejercicio, estos deportistas
presentan concentraciones de amonio en sangre
superiores a los niveles considerados normales. La
hiperamonemia está asociada, usualmente, con los
síntomas de pacientes con HE; sin embargo, los
atletas no presenten síntomas de dicha patología.
Conclusiones
Posiblemente, en el cerebro, el glutamato proveniente
de las células T reacciona con el amonio producido
por el musculo esquelético, durante ejercicios tanto
de alta intensidad como de larga duración, generando
glutamina como producto de la reacción, una molécula
neutra no toxica. De esta manera, disminuye la
fatiga central en atletas. No obstante, también podría
deberse a la inducción de moléculas neuroprotectoras
por parte de las células T, que estarían generando
neuroprotección e impidiendo la lesión de los astrocitos.
Consecuentemente, la práctica del ejercicio físico
regular induce adaptaciones metabólicas que juegan un
papel importante en el metabolismo del amonio en el
cerebro y, así, estaría generando neuroprotección contra
las altas concentraciones de amonio, ya que los atletas
durante el ejercicio con altas concentraciones de amonio
no presentan los síntomas de pacientes con HE. La
investigación científica referente a los efectos del amonio
inducido por el ejercicio puede ofrecer una aplicación
práctica en el tratamiento de la HE. No obstante,
este tema requiere mayor explicación bioquímica y
metabólica, así como estudios con adecuados diseños
metodológicos, que permitan estudiar mejor el tema.
Conflicto de intereses
El autor declara no tener conflicto de interés, pues es
independiente con respecto a las instituciones financiadoras
y de apoyo. Además, manifiesta que durante la ejecución
del trabajo y la redacción del manuscrito, no han incidido
intereses o valores distintos a los propios de la investigación.
Financiación
Universidad Pedagógica y Tecnológica de Colombia,
Facultad Seccional Chiquinquirá.
Referencias
1. Nybo L, Secher NH. Cerebral perturbations
provoked
by
prolonged
exercise.
Prog
Neurobiol. 2004;72(4):223-261.
doi:10.1016/j.
pneurobio.2004.03.005
2. Lopéz J, Fernández A. Fisiología del ejercicio.
Segunda edición. [Internet]. Madrid, España:
Editorial Médica Panamericana; 2006 [citado
26 de marzo de 2018]. Recuperador a partir de:
https://www.medicapanamericana.com/Libros/
Libro/3924/Fisiologia-del-Ejercicio.html
3. Entine J. Why black athletes dominate sports and
why we’re afraid to talk about it. [Internet]. New
York Times: Public Affairs. 2000 [citado 26 de
marzo de 2018]. Recuperado a partir de: https://
archive.nytimes.com/www.nytimes.com/books/
first/e/entine-taboo.html
4. Carvalho-Peixoto J, Alves RC, Cameron LC.
Glutamine and carbohydrate supplements reduce
ammonemia increase during endurance field
exercise. Appl Physiol Nutr Me. 2007;32(6):11861190. doi: 10.1139/H07-091
5. Bessa A, Nissenbaum M, Monteiro A, Gandra PG,
Nunes L, Bassini-Cameron A, et al. High-intensity
ultraendurance promotes early release of muscle
injury markers. Brit J Sport Med. 2008;42(11):889893. doi: 10.1136/bjsm.2007.043786
6. Bassini-Cameron A, Monteiro A, Gomes A,
Werneck-de-Castro JP, Cameron L. Glutamine
protects against increases in blood ammonia in
football players in an exercise intensity-dependent
way. Brit J Sport Med. 2008;42(4):260-266. doi:
10.1136/bjsm.2007.040378
7. Viru A, Viru M. Análisis y control del rendimiento
deportivo [Internet]. Barcelona: Paidotribo; 2003
[citado 26 de marzo de 2018]. Recuperado a partir de:
http://www.paidotribo.com/ficha.aspx?cod=00695
8. Hellsten Y, Richter EA, Kiens B, Bangsbo J.
AMP deamination and purine exchange in human
skeletal muscle during and after intense exercise.
J Physiol. 1999;520:909-920. doi: 10.1111/j.14697793.1999.00909.x
9. McKee T, McKee JR. Bioquímica: las bases
moleculares de la vida [Internet]. Estados
Unidos: Mc Graw Hill Education; 2009 [citado
26 de marzo de 2018]. Recuperado a partir de:
https://accessmedicina.mhmedical.com/Content.
aspx?bookid=1960§ionid=147707411
10. Nelson DL, Cox MM. Lehninger Principles of
Biochemistry [Internet]. Alemania: Grupo editorial
W. H. Freeman; 2004 [citado 26 de marzo de 2018].
Recuperado a partir de: https://onlinelibrary.wiley.
com/doi/abs/10.1002/cbf.1216
120
Consecuencias del amonio en la fatiga central en atletas, posible efecto neuroprotector del ejercicio
11. Huizenga JR, Gips CH, Tangerman A. The
contribution of various organs to ammonia formation:
a review of factors determining the arterial ammonia
concentration. Ann Clin Biochem. 1996;33(1):2330. doi: 10.1177/000456329603300103
12. Casey A, Greenhaff PL. Does dietary creatine
supplementation play a role in skeletal muscle
metabolism and performance?. Am J Clin Nutr.
[Internet]. 2000 [Acceso 26 mar 2018]; 72(2):607617. Disponible en: https://doi.org/10.1093/
ajcn/72.2.607S.
13. Bøyum A, Rønsen O, Tennfjord V, Tollefsen S,
Haugen A, Opstad P, et al. Chemiluminescence
response of granulocytes from elite athletes during
recovery from one or two intense bouts of exercise.
Eur J App Physiol. 2002;88(1-2):20-28. doi:
10.1007/s00421-002-0705-2
14. Wagenmakers AJ, Beckers EJ, Brouns F, Kuipers H,
Soeters PB, Van Der Vusse GJ, et al. Carbohydrate
supplementation, glycogen depletion, and amino
acid metabolism during exercise. Am J PhysiolEndoc M. 1991;260(6):E883-890. doi: 10.1152/
ajpendo.1991.260.6.E883
15. Banister EW, Cameron BJC. Exercise-induced
hyperammonemia: peripheral and central effects.
Int J Sport Med. 1990;11(1):29-142.
doi.
org/10.1055/s-2007-1024864
16. Richter EA, Ruderman BN. AMPK and the
biochemistry of exercise: implications for human
health and disease. Biochem J. 2009;418(2):261275. doi: 10.1042/BJ20082055
17. Butterworth RF. Pathophysiology of hepatic
encephalopathy: a new look at ammonia.
Metab Brain Dis. 2002;17(4):221-227. doi:
10.1023/A:1021989230535
18. Felipo V, Butterworth RF. Neurobiology of
ammonia. Neurobiol. 2002;67(4):259-279. doi:
10.1016/S0301-0082(02)00019-9
19. Ott P, Larsen FS. Blood–brain barrier permeability
to ammonia in liver failure: a critical reappraisal.
Neurochem Int. 2004;44(4):185-198. doi: 10.1016/
S0197-0186(03)00153-0
20. Córdoba J, Mínguez B. Hepatic encephalopathy.
Semin Liver Dis. [Internet]. 2008 [Acceso 26 mar
2018]; 28(1):70-80. Disponible en: https://doi.
org/10.1055/s-2008-1040322.
21. Romero-Gómez M. Role of phosphate-activated
glutaminase in the pathogenesis of hepatic
encephalopathy. Metab Brain Dis. [Internet]. 2005
[Acceso 26 mar 2018]; 20(4):319-325. Disponible
en: https://doi.org/10.1007/s11011-005-7913-5.
22. Boron W, Boulpaep E. Medical Physiology [Internet].
USA: Elsevier Health Sciences; 2012 [citado 26 de
marzo de 2018]. Recuperado a partir de: https://
www.elsevier.com/books/medical-physiology-2eupdated-edition/boron/978-1-4377-1753-2
23. Suarez I, Bodega G, Fernandez B. Glutamine
synthetase in brain: effect of ammonia. Neurochem
Int. 2002;41(2-3):123-142. doi: 10.1016/S01970186(02)00033-5
24. Nybo L, Dalsgaard MK, Steensberg A, Møller
K, Secher NH. Cerebral ammonia uptake and
accumulation during prolonged exercise in humans.
J Physiol. 2005; 563(1):285-290. doi: 10.1113/
jphysiol.2004.075838
25. Garg SK, Banerjee R, Kipnis J. Neuroprotective
immunity: T cell-derived glutamate endows
astrocytes with a neuroprotective phenotype. J
Immunol. 2008;180(6): 3866-3873. doi: 10.4049/
jimmunol.180.6.3866
26. Córdova MA. Serie Blanca. Inmunidad y ejercicio
físico. Fisiología del ejercicio. [Internet]. Madrid,
España: Editorial Médica Panamericana; 2006
[citado 26 de marzo de 2018]. Recuperador a partir
de: https://www.medicapanamericana.com/Libros/
Libro/3924/Fisiologia-del-Ejercicio.html
121
|
|
https://openalex.org/W4223619967
|
https://www.nature.com/articles/s41598-022-10021-5.pdf
|
English
| null |
Ultrasound and nerve stimulator guidance lumbar transforaminal epidural block for the treatment of patients with lumbosacral radicular pain
|
Scientific reports
| 2,022
|
cc-by
| 3,539
|
Ultrasound and nerve stimulator
guidance lumbar transforaminal
epidural block for the treatment
of patients with lumbosacral
radicular pain Seyed Ali Emami1, Mehdi Sanatkar1*, Ebrahim Espahbodi1 & Seyed Khalil Pestehei2 Transforaminal epidural block (TEB) is a widely accepted technique and minimally invasive procedure
for the treatment of lumbosacral radicular pain. This study aimed to evaluate the accuracy, efficacy,
and safety of ultrasound and nerve stimulator guidance lumbar transforaminal epidural block
(UNTEB) for the patients with unilateral lower lumbar radicular pain. The accuracy of this method was
evaluated by fluoroscopy. Using UNTEB via axial and the in-plane approach technique was performed
in 42 segments of 25 patients who presented with lumbosacral radicular pain to lower extremities. The contrast medium was injected to evaluate the needle tip at the intervertebral foramen under
fluoroscopic guidance. The numerical rating scale was used to assess pain before and after treatment. The success ratio of UNTEB in L3/L4 level was 100%, in L4/L5 was 95.4% and in L5/S1 was 100%. The
numerical rating scale (NRS) for lumbosacral radicular pain improved from a mean from 7.8 to 2.8
1 day after procedure (p = 0.01) and from a mean from 7.8 to 2.4 1 week after UNTEB (p = 0.01). None of
our subjects experienced any complications during and after the procedure. UNTEB with fluoroscopic
validation is an accurate, effective, and safe method for short-term pain relief of the patients with
unilateral lumbosacral radicular pain. Inflammation or irritation of a spinal nerve root due to degeneration of the intervertebral disc induces unilateral
radicular pain1. Treatment options of lumbosacral radicular pain range from conservative modalities such as
medications and physiotherapy to surgical interventions2. TEB is a widely accepted technique and minimally
invasive procedure for the treatment of lumbosacral radicular pain3. To visualization of the needle tip and
navigation to the target point during the TEB, this procedure is performed under the fluoroscopy or computed
tomography (CT)4. Many writers have sought to explain US-guided TEBbecause of the benefits of US-guided
nerve block, and they have had a lot of success in the treatment of lumbar radicular pain5,6. Based on previous
studies, fluoroscopy is the most reliable method for appropriate positioning of the needle during the TEB, but
this manner cannot stablish the exact relation with clinical symptoms6. Electrical stimulation of the nerve root,
especially a motor response at low amplitude, can cause symptoms of radicular pain during US-guided TEB;
therefore, identify the proper level involved with radicular pain7. www.nature.com/scientificreports www.nature.com/scientificreports Ultrasound and nerve stimulator
guidance lumbar transforaminal
epidural block for the treatment
of patients with lumbosacral
radicular pain We use a nerve stimulator to improve the success
of needle trajectory and correct target position during US-guided TEB. In the review of previous literatures on
US-guided lumbar TEB, one study used both ultrasound and nerve stimulator guidance (double guide) during
this procedure8. This study aims to evaluate the safety, accuracy, and pain relief of ultrasound and nerve stimula-
tor guidance lumbar transforaminal epidural block (UNTEB) with fluoroscopic validation for short-term pain
relief of patients with lumbosacral radicular pain. Materials and methodsh The ethical committee of pain research center, neuroscience institute, Tehran University of medical sciences (IR. MEDICINE.1400.878) approved this study, and we explained the associated risks of our procedure with patients,
and then informed consent was obtained. We confirm that all experiments were performed in accordance with 1Pain Research Center, Neuroscience Institute, Tehran University of Medical Sciences, Tehran, Iran. 2Department of
Anesthesiology and Critical Care, Imam Khomeini Medical Center, Tehran University of Medical Sciences, Tehran,
Iran. *email: mehdi.sanatkar@gmail.com | https://doi.org/10.1038/s41598-022-10021-5 Scientific Reports | (2022) 12:5954 www.nature.com/scientificreports/ www.nature.com/scientificreports/ Figure 1. Fluoroscope verification of needle placement at L4/L5 in anterior–posterior view. Figure 1. Fluoroscope verification of needle placement at L4/L5 in anterior–posterior view. Figure 1. Fluoroscope verification of needle placement at L4/L5 in anterior–posterior view. relevant guidelines and regulations9. We assessed 25 eligible patients with the degenerative intervertebral disc at
42 segments with chronic unilateral lumbosacral radicular pain for more than 3 months via the clinical presenta-
tion, physical examination, MRI, and nerve conduction velocity (NCV) test from April to July 2021. Fourteen
patients had disc hernia, 8 patients had canal stenosis, and 3 cases had spondylolisthesis. Our cases did not show
significant improvement of signs and symptoms following conservative therapy such as medication and physi-
otherapy for more than 6 weeks. The exclusion criteria were infections, previous block within 3 months, taking
anti-inflammatory agents, anomalies of the lumbar or sacral spine, previous lumbar spine operation and sever
cardiac, pulmonary, liver, and renal dysfunction. The procedure was performed with the patients in the prone
position and, in order to decrease lumbar lordosis, a pillow was placed below the lower abdomen. We used an
electrical nerve stimulation device with 22 G 100 mm nerve stimulation needles for all cases. The injection areas
were sterilized, and a sterile cover was placed on a curved transducer. UNTEB was performed by one experi-
enced pain specialist with 16 years of experience in all our subjects. We placed a curvilinear US probe over the
spinous process in a transverse axial scan to visualize the spinous process, lamina, facet joints, and transverse
process10. Before the procedure, injection areas were infiltrated with 25 G needle containing 1% lidocaine. A
nerve stimulation needle 22G was inserted approximately 45° into the skin with the in-plane technique while
the path of the needle was seen by ultrasound during the procedure. Materials and methodsh Electrical current was applied via nerve
stimulation needles to evoke the same lumbosacral radicular pain. When nerve stimulator needle reached the
lateral side of the lamina or medial to superior articular process, at Z joint location and the same radicular pain
evoked by nerve stimulator between 0.5 to 2.5 mA and we ensured that blood or cerebral spinal fluid would
not be detected, then 1 ml contrast medium was injected to confirm the correct location of needle in the ante-
rior–posterior and lateral view of fluoroscopy11 (Figs. 1, 2, 3, 4). We administered 2 ml of a mixture of 0.25%
ropivacaine and dexamethasone (8 mg/ml), one ml of dexamethasone and one ml ropivacaine in each segment
during UNTEB in all cases. Before being released from the hospital, all of our individuals were transported to
recovery and observed for 2 h. At the outpatient clinic visit 1 day and 1 week following the UNTEB, all patients
were assessed for lumbosacral radicular pain using the numerical rating scale (NRS). Statistical analysis. The data were analyzed using SPSS version 19 software, and the Wilcoxon signed-rank
test was used to compare NRS before and after UNTEB. Analysis of variance was used to compare the demo-
graphic characteristics of the patients, and a t-test was used for measurement data. The statistical significance
was set at p < 0.05. Ethical statement. Written informed consents were obtained from our patient. Discussion Lumbosacral radicular pain, which commonly occurs throughout life, is caused by spinal canal stenosis, interver-
tebral disc protrusion or extrusion, and intervertebral disc degeneration. TEB with fluoroscopy guidance was an
effective and well-established procedure which was associated with good results in the management of radicular
pain1. Ultrasonography was broadly used in the assessment and management of the musculoskeletal disorders. It was shown that the ultrasonography was a reliable and accurate technique in evaluating lumbar anatomy and
injection of local anesthetic and steroids in these areas for the management of the lumbosacral radicular pain12. US-guided TEB was first explained in cadavers in 200513. The first report on the ultrasound-guided selective
nerve root in the human was established in 200914, and then the first report of US-guided TEB in the human
model was described in 201315. Previous studies showed that lumbar TEB significantly explained better outcomes
compared to intralaminar epidural block in management of lumbosacral radicular pain. This could be because
the majority of TEB were distributed at the ventral epidural space and provide a high concentration of local
anesthetic and steroid at a nociceptive target such as dorsal root ganglions (DRG)16. The most challenging part
in US-guided TEB is putting the needle exactly in the correct location at the target nerve root and medications
spread in epidural space. We use a nerve stimulator to improve the success of needle trajectory and correct tar-
get position during US-guided TEB. Our study showed that a nerve stimulator could help us as a supplemental
guide for performing a successful US-guided TEB. Previous studies identified the relationship between electri-
cal stimulation and evoked nerve responses based on the needle tip distance17. Pain specialists try to determine
the exact amplitude at which an evoked response initiates during peripheral nerve block. Initiating at 0.5 mA
for selective nerve root block seems to be a very popular amplitude among physicians who apply for peripheral
nerve blocks18. The positioning of needle closer to the nerve roots needs less electrical current for nerve stimu-
lation, and the response of the nerve root would be more prominent. However, Kim et al. identified that in the
TEB, in contrast to a selective nerve root block, the needle has to bypass the nerve root for a successful block9. Therefore, it is not necessary that the amplitude at which nerve stimulation is observed have to be very small
during the TEB. Discussion We concluded that UNTEB with fluoroscopic validation is a safe, accurate and a success strategy for short- h
g
We concluded that UNTEB with fluoroscopic validation is a safe, accurate and a success strategy for short-
term pain relief of patients with unilateral lumbosacral radicular pain. Data availabilityh The data that support the findings of this study are available from the corresponding author upon reasonable
request. Received: 4 November 2021; Accepted: 31 March 2022 Received: 4 November 2021; Accepted: 31 March 2022 Received: 4 November 2021; Accepted: 31 March 2022 Resultsh The mean age was 54.8 ± 16.8 years. Sixteen of the patients were male, and nine were female. The procedures
were tolerated for all participates. Among 42 foraminal segments, the levels that underwent UNTEB were L3/L4,
in 2 cases, L4/L5, in 22 cases and L5/S1, in 18 cases. The same radicular pain was evoked by a nerve stimulator
at 0.5, 1, 1.5, 2 and more than 2 mA in 15, 11, 6, and 8 segments. No response was observed in 2 segments. The
success ratio of UNTEB in L3/L4 level was 100% (2 of 2), in L4/L5 level was 95.4% (21 of 22), and in L5/S1 level
was 100% (18 of 18). There was one failed case at the L4/L5 level during UNTEB. In this case, the needle tip
was positioned inappropriately inward to the midline and, therefore, repositioned correctly under US guidance
and then confirmed by fluoroscopy. NRS for lumbosacral radicular pain improved from a mean from 7.8 to 2.8
1 day after procedure (p = 0.01) and from a mean from 7.8 to 2.4 1 week after UNTEB (p = 0.01) in our patients. https://doi.org/10.1038/s41598-022-10021-5 Scientific Reports | (2022) 12:5954 | www.nature.com/scientificreports/ www.nature.com/scientificreports/ www.nature.com/scientificreports/ Figure 2. Fluoroscope verification of needle placement at L4/L5 in lateral view. Figure 2. Fluoroscope verification of needle placement at L4/L5 in lateral view. Figure 3. Fluoroscope verification of contrast spread at L4/L5. Figure 3. Fluoroscope verification of contrast spread at L4/L5. Figure 4. Ultrasound view of needle placement at L4/L5. Figure 4. Ultrasound view of needle placement at L4/L5. Figure 4. Ultrasound view of needle placement at L4/L5. Figure 4. Ultrasound view of needle placement at L4/L5. https://doi.org/10.1038/s41598-022-10021-5 Scientific Reports | (2022) 12:5954 | www.nature.com/scientificreports/ None of our subjects experienced any complications such as numbness of the lower extremities, dizziness, the
exacerbation of pain, headache, hemorrhage, infection, and allergic reaction during and after the procedure. None of our subjects experienced any complications such as numbness of the lower extremities, dizziness, the
exacerbation of pain, headache, hemorrhage, infection, and allergic reaction during and after the procedure. Discussion This result suggests that amplitude up to 2 mA or more could be used during a successful TEB9. In our study in 8 segments, the same lumbar radicular pain was evoked in 2 or more than 2 mA. Moreover, in
previous studies, authors have explored both axial and parasagittal approaches for US-guided TEB. Recently,
some authors used the parasagittal approach because of better vision and easier to perform the block, especially
in the upper spine and older patients6,19,20. Moreover, most of the authors performed US-guided TEB in in-plane
needle trajectory in both approaches because of needle path visualization. One study preferred the parasagittal
out-of-plane approach because of the deposition of medication close to the nerve root compared with in-plane
technique10. Some authors explained possible nerve injury during axial US-guided TEB because the needle tip
is not visible in the foramen. Gofeld et al. described a technique that placed the needle tip on the vertebral body
instead of the desired space in the foraminal area21. All previous studies on US-guided TEB have used another
standard method (fluoroscopy or CT) to confirm needle tip position because of the limited vision of the needle
tip at the foraminal area due to the bony structure of the lumbar spine6,22. We used axial approach and in-plane
needle trajectory for UNTEB and confirmed needle tip position by fluoroscopy before injection of medications. The gold standard to assess the success of UNTEB was the relief of symptoms. The success ratio of transforaminal
nerve block in our study was 95.4% in L4/L5 and 100% in L3/L4 and L5/S1 levels respectively; therefore, the drug
is able to be delivered in transforaminal and epidural spaces well. None of the previous studies identified major
complications during and after US-guided TEB23,24. We found no complications during and after UNTEB. Our
study had some limitations: first, our sample size was small, second, the long-term results were not evaluated,
third, the outcomes of the block were assessed only based on pain score, but functional tests of the lumbar spine
were not assessed; therefore, these items should be conducted in future researches and fourth, since we examined
patients with pain lasting for more than 3 months, it must be stated that this investigation monitored pain only
for 1 week. Therefore, no conclusions about long-term outcomes can be made. Author contributions Concept and design (S.A.E.); data acquisition (M.S.); data analysis/interpretation (M.S.); drafting of the manu-
script (E.E. and S.K.P.); critical revision of the manuscript (M.S.) supervision (S.A.E.); All authors read and
approved the final manuscript. Concept and design (S.A.E.); data acquisition (M.S.); data analysis/interpretation (M.S.); drafting of the manu-
script (E.E. and S.K.P.); critical revision of the manuscript (M.S.) supervision (S.A.E.); All authors read and
approved the final manuscript. Competing interests h p
g
The authors declare no competing interests. p
g
The authors declare no competing interests. www.nature.com/scientificreports/ Evaluation of lumbar transforaminal epidural injections with needle placement and contrast flow patterns:
A prospective, descriptive report. Pain Physician 7, 217–223 (2004). A prospective, descriptive report. Pain Physician 7, 217–223 (2004 y
7. Hadzic, A. et al. Nerve stimulators used for peripheral nerve blocks vary in their electrical characteristics. Anesthesiology 98
969–974 (2003). 18. Eeckelaert, J. P. et al. Supraclavicular brachial plexus block with the aid of a nerve stimulator. Acta Anaesthesiol. Belg. 35, 5–17
(1984). (
)
19. Lee, B. J., Han, J. & Park, D. A video of ultrasound-guided S1 transforaminal epidural injection using color doppler: Technical
reports. Pain Pract. 20, 396–398 (2020). p
(
)
0. Gofeld, M. et al. Ultrasound-guided lumbar transforaminal injections: Feasibility and validation study. Spine (Phila Pa 1976) 37
808–812 (2012). 21. Husseini, J. S. et al. Fluoroscopically guided lumbar spine interlaminar and transforaminal epidural injections: Inadvertent intra-
vascular injection. Acta Radiol. 61, 1534–1540 (2020).l j
22. Hashemi, M. et al. Ultrasound-guided lumbar transforaminal epidural injections; A single center fluoroscopic validation study. Bull. Emerg. Trauma 7, 251–255 (2019). j
22. Hashemi, M. et al. Ultrasound-guided lumbar transforaminal epidural injections; A single center fluoroscopic validation study
Bull. Emerg. Trauma 7, 251–255 (2019). g
3. Park, Y. J. et al. Novel method for S1 transforaminal epidural steroid injection. World Neurosurg. 133, e443–e447 (2020).il ,
J
p
j
g
,
(
)
24. Falsafi, M. et al. Examining the accuracy of ultrasound-guided lumbar transforaminal injection controlled by fluoroscopic imaging
in patients with lumbar radiculopathy: A modified technique. Turk. Neurosurg. 31, 582–586 (2021). p
j
g
24. Falsafi, M. et al. Examining the accuracy of ultrasound-guided lumbar transforaminal injection controlled by fluoroscopic imaging
in patients with lumbar radiculopathy: A modified technique. Turk. Neurosurg. 31, 582–586 (2021). www.nature.com/scientificreports/ www.nature.com/scientificreports/ 6. Soni, P. & Punj, J. Ultrasound-guided lumbar transforaminal epidural injection: A narrative review. Asian Spine J. 15, 261–270
(2021). 7. De Andrés, J. & Sala-Blanch, X. Peripheral nerve stimulation in the practice of brachial plexus anesthesia: A review. Reg. Anesth
Pain Med. 26, 478–483 (2001). 8. Sato, M. et al. Ultrasound and nerve stimulation-guided L5 nerve root block. Spine (Phila Pa 1976) 34, 2669–2673 (2009). 9. Kim, D. H. et al. Feasibility of nerve stimulator as a supplemental aid for lumbar transforaminal epidural block. Clin. Orthop. Surg
6, 324–328 (2014). 0. Wan, Q. et al. Ultrasonography-guided lumbar periradicular injections for unilateral radicular pain. Biomed. Res. Int. 2017, 8784149
(2017). (
)
1. Kim, Y. H., Park, H. J. & Moon, D. E. Ultrasound-guided pararadicular injection in the lumbar spine: A comparative study of the
paramedian sagittal and paramedian sagittal oblique approaches. Pain Pract. 15, 693–700 (2015). ,
,
,
J
,
g
p
j
p
p
y
paramedian sagittal and paramedian sagittal oblique approaches. Pain Pract. 15, 693–700 (2015). 12. Loizides, A. et al. Ultrasound guided versus CT-controlled pararadicular injections in the lumbar spine: A prospective randomized paramedian sagittal and paramedian sagittal oblique approaches. Pain Pract. 15, 693 700 (2015). 12. Loizides, A. et al. Ultrasound guided versus CT-controlled pararadicular injections in the lumbar spine: A prospective randomized
clinical trial AJNR Am J Neuroradiol 34 466 470 (2013) 2. Loizides, A. et al. Ultrasound guided versus CT-controlled pararadicular injections in the lumbar spine: A prospective randomized
clinical trial. AJNR Am. J. Neuroradiol. 34, 466–470 (2013). clinical trial. AJNR Am. J. Neuroradiol. 34, 466–470 (2013). 3. Galiano, K. et al. Real-time sonographic imaging for periradicular injections in the lumbar spine: A sonographic anatomic study
of a new technique. J. Ultrasound. Med. 24, 33–38 (2005). q
J
(
)
4. Narouze, S. N., Vydyanathan, A., Kapural, L., Sessler, D. I. & Mekhail, N. Ultrasound-guided cervical selective nerve root block
A fluoroscopy-controlled feasibility study. Reg. Anesth. Pain Med. 34(4), 343–348 (2009). q
14. Narouze, S. N., Vydyanathan, A., Kapural, L., Sessler, D. I. & Mekhail, N. Ultrasound-guided
A fluoroscopy-controlled feasibility study. Reg. Anesth. Pain Med. 34(4), 343–348 (2009). l
15. Chumnanvej, S., Kittayapirom, K. & Chumnanvej, S. Visualization of needle-tip localization by ultrasound guidance with contrast
bubble in lumbar selective nerve root block: Clinical pilot study. World Neurosurg. 111, e418–e423 (2018).l p
y
g
16. Manchikanti, L. et al. References 1. Manchikanti, L. et al. Transforaminal epidural injections in chronic lumbar disc herniation: A randomized, double-blind, active-
control trial. Pain Physician 17, E489-501 (2014). p
p
y
3. Yang, G. et al. Ultrasound-guided versus fluoroscopy-controlled lumbar transforaminal epidural injections: A prospective rand-
omized clinical trial. Clin. J. Pain 32, 103–108 (2016). 3. Yang, G. et al. Ultrasound-guided versus fluoroscopy-controlled lumbar transforaminal epidural injections: A prospective rand-
omized clinical trial. Clin. J. Pain 32, 103–108 (2016). (
)
4. Hashemi, S. M. et al. Comparison of transforaminal and parasagittal epidural steroid injections in patients with radicular low back
pain. Anesth. Pain Med. 5, e26652 (2015). 4. Hashemi, S. M. et al. Comparison of transforaminal and parasagittal epidural steroid injections in patients with radicular low back
pain. Anesth. Pain Med. 5, e26652 (2015). p
,
(
)
5. Bhatia, A., Flamer, D., Shah, P. S. & Cohen, S. P. Transforaminal epidural steroid injections for treating lumbosacral radicular pain
from herniated intervertebral discs: A systematic review and meta-analysis. Anesth. Analg. 122(3), 857–870 (2016). https://doi.org/10.1038/s41598-022-10021-5 Scientific Reports | (2022) 12:5954 | Fundingh g
The authors indicate no financial support. © The Author(s) 2022 Additional information Correspondence and requests for materials should be addressed to M.S. Correspondence and requests for materials should be addressed to M.S. Correspondence and requests for materials should be addressed to M.S. Reprints and permissions information is available at www.nature.com/reprints. Reprints and permissions information is available at www.nature.com/reprints. Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and
institutional affiliations. Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and
institutional affiliations. Open Access This article is licensed under a Creative Commons Attribution 4.0 International
License, which permits use, sharing, adaptation, distribution and reproduction in any medium or
format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the
Creative Commons licence, and indicate if changes were made. The images or other third party material in this
article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the
material. If material is not included in the article’s Creative Commons licence and your intended use is not
permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from
the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. © The Author(s) 2022 https://doi.org/10.1038/s41598-022-10021-5 Scientific Reports | (2022) 12:5954 |
|
https://openalex.org/W4387366884
|
https://www.ajol.info/index.php/emj/article/view/256439/242234
|
English
| null |
Association between medication adherence in patients with type 2 diabetes with their quality of life: A cross-sectional study
|
Ethiopian medical journal
| 2,023
|
cc-by
| 5,694
|
Abstract Abstract
Background: Assess the impact of medication adherence on quality of life (QOL) in patients with type 2 diabetes
in different setting and societies, can be considered as a great clinical importance. Therefore, this study was de-
signed for evaluation of association between medication adherence in patients with type 2 diabetes with their
QOL. Q
Methods: This cross-sectional study was conducted on 279 diabetic type 2 positive cases referred to Endocrinolo-
gy clinic in Hamadan City at the first six months of 2020. The short-form health survey SF-36 and MMAS-8 ques-
tionnaire was used for assessing health related QOL and medication adherence of patients, respectively. Correla-
tions between each QOL domain with medication adherence were tested with Pearson's r coefficient. Results: Among investigated patients, 154 (55.2%) of them were male and vast majority were urban dweller
(91.07%). The number of 101 patients (36.07%) had high treatment adherence. We found a significant and positive
correlation between medication adherence and domains of emotional wellbeing, social functioning, general health
and total score of QOL (P>0.05). While there was a significant negative correlation between medication adher-
ence and role limitations due to physical health (r=-0.14, P=0.02). p y
(
)
Conclusion: Based on the findings of this study, diabetic patients with poor QOL may have less medication adher-
ence, then such patients should be integrated e.g. in self-care education programs and reinforcement interventions
to improve QOL and medication adherence. Keywords: Medication Adherence, Quality of life, Diabetes Keywords: Medication Adherence, Quality of life, Diabetes
Citation : Borzouei S, Ayubi E, Bashirian S. Association between medication adherence in patients with type 2di
abetes with their quality of life: A cross-sectional study Ethiop Med J 61 (4) 351—357
Submission date : 6 January 2022 Accepted: 1 September 2023 Published: 24 September 2023 respectively (6). The most prevalent type of DM is
type 2 diabetes and affects about 90-95% of all
people with diabetes (7). It is one of the major
health problems in human life, because of the high
prevalence, high cost of treatment as well as due to
its severe disability (8, 9).This disease also affect
the patient and their family quality of life (QOL)
due to its multi organ involvement and chronic
nature. The lower QOL has been reported with
increased disease severity (5). Quality of life is a
multidimensional fundamental factor and a combi-
nation of cognitive (satisfaction) and emotional
(Happiness) factors. https://dx.doi.org/10.4314/emj.v61i4.7 https://dx.doi.org/10.4314/emj.v61i4.7 351 https://dx.doi.org/10.4314/emj.v61i4.7 Ethiopian Medical Journal 2023, 61(4)
Open access articles distributed in terms of
Creative Commons Attribution Licence [CC by 4.0] ISSN 0014-1755 eISSN 2415-2420 Borzouei et al. Shiva Borzouei1, Erfan Ayubi2, Saeid Bashirian2, Nasim Karimi3, Salman Khazaei4,5 1Department of Endocrinology, School of Medicine, Hamadan University of Medical Sciences,
Hamadan, Iran
2Social Determinants of Health Research Center, Hamadan University of Medical Sciences, Hamadan, Iran
3Department of Epidemiology, Hamadan University of Medical Sciences, Hamadan, Iran
4Modeling of Noncommunicable Diseases Research Center , Hamadan University of Medical Sciences, Hama-
dan, Iran
5A ti
S
t
Di
d
R
h C
t
H
d
U i
it
f M di
l S i
H
d
I sm Spectrum Disorders Research Center, Hamadan University of Medical Sciences, Hamadan, Iran Abstract Therefore, the aim of the med-
ical cares should be to improve the quality of life Data collection tools We used a researcher-made checklist for gathering
demographic variables of the patients including: gen-
der, age, marital status, location and education. g
The Iranian version of the health survey SF-36 in the
short form was used to evaluate the health related qual-
ity of life in diabetic patients (20). The validity and
reliability of this instrument was approved in previous
studies (20). This questionnaire consists of 36 items
with 8 domain including: general health (5 items),
physical functioning (10 items), role limitations due to
physical problems (4 items), social functioning (2
items), pain (2 items), energy/ fatigue (4 items), role
limitations due to emotional problems (3 items) and
emotional well-being (5 items). Each domain ranged
from 0-100 scale, with higher scores representative
better situation (21). Low adherence to diabetes drugs regimen leads to
treatment failure, damage to vital organs (such as
kidneys), imposing financial cost to the patients and
their family, lower quality of life and finally in-
creased mortality (16). It has been reported that patient’s features like age,
sex, education, income, history of disease in their
family, duration of diabetes mellitus, and prescribed
treatments could affect patient’s quality of life. Also,
some prescribed drugs can negatively affect the dia-
betes mellitus patient’s quality of life (10, 17). The
results of studies regarding the relation between qual-
ity of life of DM patients and adherence to drug treat-
ment and glycemic control was not necessarily a sig-
nificant relationship (11, 18, 19). We think some of
these discrepancies may be due to the ignoring
confounding factors and different instruments or
methods used to measure diabetic patients QOL and
adherence to drug treatment. Therefore, this study
was designed for evaluation of association between
medication adherence in patients with type 2 diabetes
with their quality of life. The Persian version of MMAS-8 was used for deter-
mine medication adherence of patients. Items 1 -7 are
binary (yes/no) and item 8 has a five-point in the Lik-
ert response scale. ‘Yes” response to questions 1-7 is
scored as 1 and “no” is scored as 0. About the item 8,
to compute a summated score, the scores (0-4) be
standardized by dividing the result by 4. Total scores
for this questionnaire is ranged from 0 to 8, with scores
of 8, 7 or 6 and <6 indicating high adherence, medium
adherence and low adherence, respectively (3). Data collection procedures The data collection process involved asking patients
questions by a trained nursing specialist before their
visit with the physician or nutritionist. Additionally,
some information was extracted from the patient's care
and treatment records. The final questionnaire check
was performed by the physician who is based at the
center. Material and Methods
Study design and setting In this cross-sectional study, a total of 279 individuals
with type 2 diabetes referred to Endocrinology clinic
affiliated to Hamadan University of Medical Scienc-
es, Hamadan City, Hamadan Province, Iran, in the
first six months of 2020 were enrolled. This is a refer-
ral center for diabetic patients in Hamedan province,
which provides diagnostic and treatment services to Data collection tools The
validity and reliability of the questionnaire to use in
Iranian context has been approved in previous studies
(22, 23). Eligibility criteria To enter the study, at least six months must have
passed since the patient was diagnosed with diabetes,
and the patient must be willing to contribute in the
study. Illiterate patients, patients with neurological
diseases or with other known co-morbidities were
omitted from the study. The clinical effect of drug therapies for diabetic
patients has been limited by low adherence behavior
(11). The World Health Organization (WHO) has
estimated that adherence behavior in diabetic patients
for long-term maintenance in less advanced countries
is only 50%, and even lower in developing countries
(12). Results of the systematic review study showed
that adherence to insulin therapy in patients with
Type 1 or Type 2 diabetes mellitus was vary widely
from 43% to 86% in different studies (13). The dis-
continuation rates of the drug use in a 30 days period
of treatment reached 42% and injectable drugs of
diabetes mellitus had the higher risk for discontinua-
tion rather than oral medications (14). Generally, the
incidence of non-adherence to drug therapies in dia-
betic patients is high due to the long length of treat-
ment duration (15). Introduction Following the industrialization of countries in the 21st
century, increasing life expectancy and lifestyle
changes, the pattern of diseases has changed from
acute to chronic such as diabetes mellitus (DM) (1). DM is the most prevalent metabolic disorder world-
wide, described by the chronic form of hyperglyce-
mia and resulted from a deficiency in insulin secre-
tion or action or both (2). There are currently more
than 415 million people lived with diabetes world-
wide (3). Diabetes is responsible for 4 million deaths
every year (9% of all deaths) (4). It is predicted that
the prevalence of DM increase by 69% in less ad-
vanced countries until 2030 (5). The prevalence of
diabetes in Iranian men and women is 8% and 10%, 352 (5). Self-management has been considered as the ba-
sis of success of long-term maintenance cares and
prevention of disease comorbidities. Proper diet,
physical activity and adherence to drug therapies
(medication
regimen)
are
among
the
self-
management behaviors of patients with diabetes (4). Adherence to the medication regimen is defines as
taking the prescribed drugs at the correct time and
amount during the prescribed period (10). patients by an endocrinology specialist. Additionally,
nutritional services are provided by a nutritionist in
this center. Since we were not aware of the population
frame of patients and the pattern of patient referrals,
we had no choice but to use non-probability sampling
methods for choosing patients. We obtained oral in-
formed consent from the patients. Statistical Methods Descriptive statistics including mean and SD for quan-
titative data and n (%) for qualitative data were used
for describe demographic features of type 2 diabetic
patients, and their medication adherence and quality of 353 life status. The normality assumption of the variables
distribution was checked with shapiro-wilk test. An
independent Student's t test was used for comparison
the means of treatment adherence and quality of life,
an analysis of variance (ANOVA) was used for com-
parison the means of medication adherence and quali-
ty of life according variables with more than two cat-
egories. Correlations between each quality of life
domain with medication adherence were tested with
Pearson's r coefficient. Data were analyzed using
Stata- 14 software. Significant level was considered
at the 5 % level (p < 0.05). cation and marital status, 132 (47.14%) had academic
education and 239 (85.36%) were married. About
half of them were in the age range of 54-60 years and
37 (13.21. %) were under 45 years old. 101 (36.07%)
of the patients had high treatment adherence and in
87 (31.07%) of them medication adherence was low. The relation between medication adherence and qual-
ity of life according demographic variables is shown
in table 1. As shown, there was not a significant asso-
ciation between patient’s demographic characteristics
and their treatment adherence status (P>0.05). While,
males had reported higher score of quality of life
compared women (P=0.008), and increasing the level
of education was associated with increasing the qual-
ity of life (P<0.001). Also there was a significant
association between marital status and quality of life
(49.58 for married vs. 45.54 for single patients,
P=0.009). With increase in the age of patients, their
quality of life has decreased significantly (P=0.03). Ethical considerations We obtained oral informed consent from the patients
for participation to the study. The study has been ap-
proved by the Ethics Committee of Hamadan Univer-
sity of Medical Sciences (Ethics code:
IR.UMSHA.REC.1398.620). Discussion
h
i
f The aim of the present study was to detect some so-
cio-demographic determinants of treatment adher-
ence and QOL in type 2 diabetes patients. Moreover,
we examined the correlation between medication
adherence and QOL. In the present study, no associa-
tion between studied socio-demographic determi-
nants and medication adherence was found while,
male patients, well-educated patients, those were
married and were younger had a higher score of
QOL. We also found a positive correlation between
medication adherence and overall QOL. Table 2. Quality of life components mean score of
the study population Table 2. Quality of life components mean score of
the study population SF36 domains
Range
Mean (SD)
Physical functioning
0, 100
69 (28.17)
Role limitations due
to physical health
2.5, 100
16.78 (26.04)
Role limitations due
to emotional problems
1.33, 100
11.79 (19.88)
Energy/ fatigue
10, 90
45.28 (19.05)
Emotional well being
8, 88
58.2 (18.17)
Social functioning
12.5, 100
79.74 (20.44)
Pain
0, 100
76.53 (28.16)
General health
0.4, 62
34.99 (14.76)
SF36, total score
13.38,
69.37
48.98 (9.22) Since the rate of non-adherence to medications is
reported lower than 50% among patients with type 2
diabetes in some previous studies in Iran (24, 25), it
is needed to pay attention to the determining factors
of medication adherence patients with type 2 diabetes
is obvious. In the present study, we found that there
was no socio-demographic difference in medication
adherence. However, other studies have demonstrat-
ed that several factors including education level, type
of medication, age, body mass index, the presence of
complications, treatment duration, perceived side
effect and negative effect of medicines, lack of fi-
nances and stage of diabetes may be associated with
medication adherence in patients with type 2 diabetes
(24, 26-30). Another point that can be mentioned is
that other type of adherence in type 2 diabetes pa-
tients e.g. exercise, blood test and foot care could be
modified by type of treatment. For example, in the
study by Zioga et al. (31), they found that the type of
treatment (diet, oral medication and insulin) is associ-
ated with adherence to blood tests. The correlation between medication adherence of
patients and domains of quality of life is shown in
table 3. There was a significant positive correlation
between medication adherence and domains of emo-
tional wellbeing, social functioning, general health
and total score of quality of life (P>0.05). Results In total 279 diabetic patents were investigated. 154
(55.2%) of them were male and vast majority of them
were urban dweller (91.07%). In regards of their edu- Table 1. Medication adherence and quality of life by demographic variables
Variable
Total (%)
Medication adherence status
(<6: poor, 6-7.9: moderate, 8: good)
Quality of life
(SF35 score)
Poor
87
(31.07)
Moder-
ate
92
(32.86)
Good
101
(36.07)
Mean
(SD)
P*
Mean
(SD)
p
Gender
Male
125 (44.8)
40
(31.75)
45
(35.71)
41
(32.54)
6.38
(1.67)
0.21
50.61
(8.96)
0.008**
Female
154 (55.2)
47
(30.52)
47
(30.52)
60
(38.96)
6.61
(1.47)
47.69
(9.24)
Educa-
tion
Illiterate
56 (20)
21 (37.5)
15
(26.79)
20
(35.71)
6.46
(1.51)
0.31
45.31
(9.92)
<0.001*
**
Diploma
and less
92 (32.86)
22
(23.91)
34
(36.96)
36
(39.13)
6.71
(1.34)
48.32
(9.22)
Academ-
ic
132 (47.14)
44
(33.33)
43
(32.58)
45
(34.09)
6.39
(1.72)
51.03
(8.37)
Marital
status
Single
41 (14.64)
15
(36.59)
13
(31.71)
13
(31.71)
6.37
(1.6)
0.54
45.54
(7.75)
0.009**
Married
239 (85.36)
72
(30.13)
79
(33.05)
88
(36.82)
6.53
(1.56)
49.58
(9.33)
Age
group
(year)
30-44.9
37 (13.21)
16
(43.24)
12
(32.43)
9
(24.32)
6.09
(1.55)
0.12
52.2
(8.06)
0.03***
45-59.9
137 (48.93)
37
(27.01)
55
(40.15)
45
(32.85)
6.47
(1.62)
49.81
(8.91)
60+
106 (37.86)
34
(32.08)
25
(23.58)
47
(44.34)
6.7
(1.48)
46.84
(9.55)
Loca-
tion
Urban
255 (91.07)
79
(30.98)
83
(32.55)
93
(36.47)
6.52
(1.57)
0.69
49.04
(8.83)
0.72**
Rural
25 (8.93)
8 (32)
9 (36)
8 (32)
6.39
(1.54)
48.34
(12.7)
*Chi- square test, ** student t-test, *** One way ANOV Table 1. Medication adherence and quality of life by demographic variables 354 The mean scores for domains of the quality of life
was ranged from 11.79 in the role limitations due to
emotional problems domain to 79.74 in the social
functioning domain. The total mean (SD) score of
quality of life for patients was 48.98 (9.22) (Table 2). *Pearson correlation coefficient Discussion
h
i
f Second, the impact of self-report response
bias on the observed results should be considered. Third, several inclusion criteria were considered
which may reduce generalizability and external valid-
ity of this study, e.g. the results may be not applied to
Illiterate patients. Forth, it was not possible to meas-
ure potential confounders in the causal chain of QOL
-medication adherence. For example, depression was
not measured in this study which is a determinant for
both QOL and medication adherence. (45, 46). Most studies used non-specific tools to measure QOL
e.g. WHO Quality of Life (WHOQOL) (11, 39, 41),
SF36 (29, 31), EQ-5D (42, 43) when evaluating the
association between QOL and medication adherence
while some used a diabetes-specific-QOL e.g. Diabe-
tes 39 instrument (5) and Diabetes Quality of Life
Clinical Trial Questionnaire (DQLCTQ) (10). In the
present study, we used SF36 for QOL assessment and
our results showed the positive statistical correlation
between medication adherence and QOL domains
including physical functioning, emotional well-being,
social functioning, pain and general health. In other
hands, the negative correlation between medication
adherence and physical role as well as no statistical
association between medication adherence and QOL
domains including physical functioning, physical role
and energy/fatigue were found. In a study by Marinho
et al. (29), they used SF36 to measure the quality of
life and their results suggested that SF-36 emotional
domain was also associated positively related to bet-
ter general medication adherence. In other work (31)
suggested that an association between all domains of
QOL and all adherence subscales including diet,
physical activity, blood sugar test, care of foots and
commendations for self-care. In another study by
Perwitasari et al (10), they used DQLCTQ a specific
tool for QOL assessment in diabetic patients and their
results showed that beliefs, recall, and beliefs about
adverse event screens as a QOL domain is related
positively to medication adherence. In the study by
Khayyat et al (44), the mean level of overall QOL
and all QOL domains among adherent’s patients with
diabetes and hypertension was higher than those in
non-adherent patients group. In conclusion, our results show an association be-
tween medication adherence and overall QOL and
some QOL domains. Based on the findings of this
study, diabetic patients with poor QOL may have less
medication adherence, then such patients should be
integrated e.g. in self-care education programs and
reinforcement interventions to improve QOL and
medication adherence. Discussion
h
i
f While there
was a significant negative correlation between medi-
cation adherence and role limitations due to physical
health (r=-0.14, P=0.02). In the line of previous studies (32-37), our results
have shown that older age patients, women, those
with lower education level and singles had a lower
score of QOL. Other important sociodemographic
characteristics that reported as determinants of QOL
and not studied in the present study were income and
occupation e.g. unemployment patients (34) and
those with lower household income per month (33)
had a lower level of overall QOL and domains of
QOL. Regardless of sociodemographic characteris-
tics, the factors including exercise, the presence of
complications, comorbidities, duration of disease,
psychological disorders and diet were also suggested
as determinants of QOL in type 2 diabetes patients
(38). Table 3. Correlation between medication adherence
and domains of quality of life Medication adherence
SF36 domains
r*
p
Physical functioning
0.04
0.54
Role limitations due to
physical health
-0.14
0.02
Role limitations due to
emotional problems
-0.09
0.14
Energy/ fatigue
0.09
0.13
Emotional well being
0.18
0.003
Social functioning
0.22
<0.001
Pain
0.2
<0.001
General health
0.16
0.006
SF36, total score
0.18
0.003
*Pearson correlation coefficient The results from previous studies showed tendency
toward a positive association (5, 10, 29, 31, 39), how-
ever, some studies reported a no association between
two aforementioned variables (11, 40). Heterogeneity
in the reported association in previous studies may be
as results of several factors e.g. type of method for
medication adherence and QOL assessment. For ex-
ample, in the studies that reported a positive associa-
tion, medication adherence was evaluated using the 355 methods such as MMAS (5) Brief Medication Ques-
tionnaire (BMQ) (10), frequency of statin use (39),
Diabetes Self-Care Activities Questionnaire (31) and
Diabetes Self-Care Activities Questionnaire (29). In
the studies that showed no association, the medica-
tion adherence was assessed using the methods such
as pill counting (11) and hemoglobin A1c (HbA1c)
(40). There are several limitations to this study. First, the
temporal relationship between medication adherence
and QOL with a 1-time measurement of them could
not be established and this is an intrinsic limitation of
a cross sectional study. For example, in one work
(44) medication adherence was studied as an inde-
pendent predictor of QOL while in another study (29)
QOL domains were introduced as important determi-
nants of medication adherence in patients with type 2
diabetes. Conflict of interest The authors declare that there is no conflict of inter-
ests. Acknowledgments The Deputy of Research and Technology from Ham-
adan University of Medical Sciences financially sup-
ported this study (Research Id: 9807305712, Ethics
code: IR.UMSHA.REC.1398.620). We would like
thanks from all patients that participate in this study. Discussion
h
i
f Further large-scale longitudi-
nal studies with a bigger dataset are recommended. 4. Dashtian M, Eftekhar Ardebili H, Karimzadeh Shirazi K, Shahmoradi M, Azam K. Predicting Factors Af-
fecting Medication Adherence and Physical Activity in Patients with Type-2 Diabetes Mellitus Based on the
Theory of Planned Behavior. Journal of School of Public Health and Institute of Public Health Research.
2017;15(2):133-46. ;
( )
5. Alfian SD, Sukandar H, Lestari K, Abdulah R. Medication adherence contributes to an improved quality of
life in type 2 diabetes mellitus patients: A cross-sectional study. Diabetes Therapy. 2016;7(4):755-64. 3. Satyawali V, Pandey S, Sharma V, Titiyal GS, Panwar N, Kumar P, et al. Assessment of level of awareness
for diabetes mellitus, its systemic and ocular complications: a questionnaire-based survey at tertiary care
Centre of Kumaon Region. Journal of evolution of medical and dental sciences-jemds. 2016;5(68):4856-60. 2. Piero M, Nzaro G, Njagi J. Diabetes mellitus-a devastating metabolic disorder. Asian journal of biomedical
and pharmaceutical sciences. 2015;5(40):1. References Glycaemic control and quality of life among ethnically
sian diabetic patients. Quality of Life Research. 2015;24(4):951-8. p
Q
y
;
( )
18. Lau CY, Qureshi AK, Scott SG. Association between glycaemic control and quality of life in diabetes melli-
tus. Journal of postgraduate medicine. 2004;50(3):189-93; discussion 94. p
g
( )
19. Pitale S, Kernan-Schroeder D, Emanuele N, Sawin C, Sacks J, Abraira C. Health-related quality of life in the
VA Feasibility Study on glycemic control and complications in type 2 diabetes mellitus. Journal of diabetes
and its complications. 2005;19(4):207-11. p
;
( )
20. Montazeri A, Goshtasebi A, Vahdaninia M, Gandek B. The Short Form Health Survey (SF-36): translation
and validation study of the Iranian version. Quality of life research. 2005;14(3):875-82. y
Q
y
( )
21. John E. SF-36 health survey: manual and interpretation guide: Health Assessment Lab; 2000. Dehghan M, Nayeri ND, Karimzadeh P, Iranmanesh S. Psychometric properties of the Persian v
Morisky medication adherence scale-8. Journal of Advances in Medicine and Medical Research. 2 22. Dehghan M, Nayeri ND, Karimzadeh P, Iranmanesh S. Psychometric properties of the Persian version of the
Morisky medication adherence scale-8. Journal of Advances in Medicine and Medical Research. 2015:1-10. 23
Moharamzad Y Saadat H Shahraki BN Rai A Saadat Z Aerab Sheibani H et al Validation of the Persian 22. Dehghan M, Nayeri ND, Karimzadeh P, Iranmanesh S. Psychometric properties of the Persian version of the
Morisky medication adherence scale-8. Journal of Advances in Medicine and Medical Research. 2015:1-10. 23. Moharamzad Y, Saadat H, Shahraki BN, Rai A, Saadat Z, Aerab-Sheibani H, et al. Validation of the Persian
version of the 8-item Morisky Medication Adherence Scale (MMAS-8) in Iranian hypertensive patients Glob- 22. Dehghan M, Nayeri ND, Karimzadeh P, Iranmanesh S. Psychometric properties of the Persian version of the
Morisky medication adherence scale-8. Journal of Advances in Medicine and Medical Research. 2015:1-10. 23. Moharamzad Y, Saadat H, Shahraki BN, Rai A, Saadat Z, Aerab-Sheibani H, et al. Validation of the Persian
version of the 8-item Morisky Medication Adherence Scale (MMAS-8) in Iranian hypertensive patients. Glob-
al journal of health science. 2015;7(4):173. 24
B
L Kh l
i A M di
i
Adh
d i
P di
i
T
2 Di b i P i
R f
i Morisky medication adherence scale-8. Journal of Advances in Medicine and Medical Research. 2015:1-10. 23. References Type 2 diabetes and quality of life. Worl
betes. 2017;8(4):120. ( )
8. Chaudhury A, Duvoor C, Reddy Dendi VS, Kraleti S, Chada A, Ravilla R, et al. Clinical review of antidiabet-
ic drugs: implications for type 2 diabetes mellitus management. Frontiers in endocrinology. 2017;8:6. g
p
yp
g
gy
9. Razavi Ns, Majlessi F, Mohebbi B, Tol A, Azam K. Assessing the effect of educational intervention on treat-
ment adherence based on aim model among patients with type 2 diabetes. Iranian Journal of Diabetes and Li-
pid Disorders. 2017;16(3):173-82. p
( )
10. Perwitasari DA, Urbayatun S. Treatment adherence and quality of life in diabetes mellitus patients in Indone-
sia. Sage Open. 2016;6(2):2158244016643748. 11
M
í
YV P d
A
il
CA R
ó
P
h
RA V ldi i M
í
JJ Q
li
f lif
i
d
i h g
p
( )
11. Martínez YV, Prado-Aguilar CA, Rascón-Pacheco RA, Valdivia-Martínez JJ. Quality of life associated with
treatment adherence in patients with type 2 diabetes: a cross-sectional study. BMC Health Services Research. 2008;8(1):164. 12. Saleh F, Mumu SJ, Ara F, Hafez MA, Ali L. Non-adherence to self-care practices & medication and health
related quality of life among patients with type 2 diabetes: a cross-sectional study. BMC public health. 2014;14(1):431. ;
( )
13. Davies M, Gagliardino JJ, Gray L, Khunti K, Mohan V, Hughes R. Real‐world factors affecting adherence to
insulin therapy in patients with Type 1 or Type 2 diabetes mellitus: a systematic review. Diabetic Medicine. 2013;30(5):512-24. ;
( )
14. Zullig LL, Gellad WF, Moaddeb J, Crowley MJ, Shrank W, Granger BB, et al. Improving diabetes medication
adherence: successful, scalable interventions. Patient preference and adherence. 2015;9:139. ,
p
;
15. Blackburn DF, Swidrovich J, Lemstra M. Non-adherence in type 2 diabetes: practical conside
preting the literature. Patient preference and adherence. 2013;7:183-9. p
DF, Swidrovich J, Lemstra M. Non-adherence in type 2 diabetes: practical considerations for inter-
iterature. Patient preference and adherence. 2013;7:183-9. p
g
p
;
16. Schoenthaler A, Montague E, Baier Manwell L, Brown R, Schwartz MD, Linzer M. Patient–physician racial/
ethnic concordance and blood pressure control: the role of trust and medication adherence. Ethnicity & health. 2014;19(5):565-78. AlMashoor SAH, Winn T. Glycaemic control and quality of life among ethnically diverse Malay-
c patients. Quality of Life Research. 2015;24(4):951-8. ;
( )
17. Daher AM, AlMashoor SAH, Winn T. References e e e ces
1. Adogu P, Ubajaka C, Emelumadu O, Alutu C. Epidemiologic transition of diseases and health-related events
in developing countries: a review. American Journal of Medicine and Medical Sciences. 2015;5(4):150-7. 2. Piero M, Nzaro G, Njagi J. Diabetes mellitus-a devastating metabolic disorder. Asian journal of biomedical
d h
ti
l
i
2015 5(40) 1 1. Adogu P, Ubajaka C, Emelumadu O, Alutu C. Epidemiologic transition of diseases and health-related events
in developing countries: a review. American Journal of Medicine and Medical Sciences. 2015;5(4):150-7. p g
( )
2. Piero M, Nzaro G, Njagi J. Diabetes mellitus-a devastating metabolic disorder. Asian journal of biomedical
and pharmaceutical sciences. 2015;5(40):1. 2. Piero M, Nzaro G, Njagi J. Diabetes mellitus-a devastating metabolic disorder. Asian journal of biomedical
and pharmaceutical sciences. 2015;5(40):1. p
(
)
3. Satyawali V, Pandey S, Sharma V, Titiyal GS, Panwar N, Kumar P, et al. Assessment of level of awareness
for diabetes mellitus, its systemic and ocular complications: a questionnaire-based survey at tertiary care
Centre of Kumaon Region. Journal of evolution of medical and dental sciences-jemds. 2016;5(68):4856-60. 3. Satyawali V, Pandey S, Sharma V, Titiyal GS, Panwar N, Kumar P, et al. Assessment of level of awareness
for diabetes mellitus, its systemic and ocular complications: a questionnaire-based survey at tertiary care
Centre of Kumaon Region. Journal of evolution of medical and dental sciences-jemds. 2016;5(68):4856-60. g
j
; (
)
4. Dashtian M, Eftekhar Ardebili H, Karimzadeh Shirazi K, Shahmoradi M, Azam K. Predicting Factors Af-
fecting Medication Adherence and Physical Activity in Patients with Type-2 Diabetes Mellitus Based on the
Theory of Planned Behavior. Journal of School of Public Health and Institute of Public Health Research. 2017;15(2):133-46. g
j
; (
)
4. Dashtian M, Eftekhar Ardebili H, Karimzadeh Shirazi K, Shahmoradi M, Azam K. Predicting Factors Af-
fecting Medication Adherence and Physical Activity in Patients with Type-2 Diabetes Mellitus Based on the
Theory of Planned Behavior. Journal of School of Public Health and Institute of Public Health Research. 2017;15(2):133 46 ;
( )
5. Alfian SD, Sukandar H, Lestari K, Abdulah R. Medication adherence contributes to an improved quality of
life in type 2 diabetes mellitus patients: A cross-sectional study. Diabetes Therapy. 2016;7(4):755-64. 356 6. Imani A, Dabirian A. Telenursing Benefits in patients with Diabetes: A Review Article. Advances in Nursing
& Midwifery. 2014;23(83):65-74. y
(
)
7. Trikkalinou A, Papazafiropoulou AK, Melidonis A. References Quality of Life in Patients with Type 2 Diabe-
tes: Application of WHOQoL-BREF Scale. Shiraz E-Med J. 2013;14(3):162-71. 34. Javanbakht M, Abolhasani F, Mashayekhi A, Baradaran HR, Jahangiri noudeh Y. Health re
life in patients with type 2 diabetes mellitus in Iran: a national survey. PloS one. 2012;7(8): 35. John R, Pise S, Chaudhari L, Deshpande PR. Evaluation of Quality of Life in Type 2 Diabetes Mellitus
Patients Using Quality of Life Instrument for Indian Diabetic Patients: A Cross-Sectional Study. Journal
of mid-life health. 2019;10(2):81-8. ;
( )
36. Khalili M, Sabouhi F, Abazari P, Aminorroaya A. Comparing the quality of life in insul
refusal patients with type 2 diabetes. Iranian journal of nursing and midwifery research. 201 refusal patients with type 2 diabetes. Iranian journal of nursing and midwifery research. 2016;21(4):351-6. 37. Zurita-Cruz JN, Manuel-Apolinar L, Arellano-Flores ML, Gutierrez-Gonzalez A, Najera-Ahumada AG,
Cisneros-González N. Health and quality of life outcomes impairment of quality of life in type 2 diabetes
mellitus: a cross-sectional study. Health and quality of life outcomes. 2018;16(1):94. 38. Jing X, Chen J, Dong Y, Han D, Zhao H, Wang X, et al. Related factors of quality of life of t
tes patients: a systematic review and meta-analysis. Health and quality of life outcomes. 2018; 39. Farhat R, Assaf J, Jabbour H, Licha H, Hajj A, Hallit S, et al. Adherence to oral glucose lowering drugs,
quality of life, treatment satisfaction and illness perception: A cross-sectional study in patients with type 2
diabetes. Saudi Pharmaceutical Journal. 2019;27(1):126-32. ;
( )
40. Kleefstra N, Ubink-Veltmaat L, Houweling S, Groenier K, Meyboom-de Jong B, Bilo H. Cross-sectional
relationship between glycaemic control, hyperglycaemic symptoms and quality of life in type 2 diabetes
(ZODIAC-2). Neth J Med. 2005;63(6):215-21. (
)
;
( )
41. Chew BH. Medication adherence on quality of life among adults with type 2 diabetes mellitus: an explora-
tory analysis on the EDDMQoL study. Quality of life research : an international journal of quality of life
aspects of treatment, care and rehabilitation. 2015;24(11):2723-31. p
(
)
42. Babenko AY, Mosikian AA, Lebedev DL, Khrabrova EA, Shlyakhto EV. Mental state, psychoemotional
status, quality of life and treatment compliance in patients with Type 2 diabetes mellitus. Journal of com-
parative effectiveness research. 2019;8(2):113-20. 43. Saleh F, Mumu SJ, Ara F, Hafez MA, Ali L. References Moharamzad Y, Saadat H, Shahraki BN, Rai A, Saadat Z, Aerab-Sheibani H, et al. Validation of the Persian
version of the 8-item Morisky Medication Adherence Scale (MMAS-8) in Iranian hypertensive patients. Glob-
al journal of health science. 2015;7(4):173. j
24. Benrazavy L, Khalooei A. Medication Adherence and its Predictors in Type 2 Diabetic Patients Referring to
Urban Primary Health Care Centers in Kerman City, Southeastern Iran. Shiraz E-Medical Journal. 2019(In
Press). )
25. Farsaei S, Sabzghabaee AM, Zargarzadeh AH, Amini M. Adherence to glyburide and metformin and associat-
ed factors in type 2 diabetes in isfahan, iran. Iran J Pharm Res. 2011;10(4):933-9. yp
( )
26. Aflakseir A. Role of illness and medication perceptions on adherence to medication in a grou
tients with type 2 diabetes. Journal of diabetes. 2012;4(3):243-7. yp
( )
. Role of illness and medication perceptions on adherence to medication in a group of Iranian pa-
ype 2 diabetes. Journal of diabetes. 2012;4(3):243-7. yp
; ( )
27. Aminde LN, Tindong M, Ngwasiri CA, Aminde JA, Njim T, Fondong AA, et al. Adherence to antidiabetic
medication and factors associated with non-adherence among patients with type-2 diabetes mellitus in two
regional hospitals in Cameroon. BMC endocrine disorders. 2019;19(1):35. 28. Lin L-K, Sun Y, Heng BH, Chew DEK, Chong P-N. Medication adherence and glycemic control among new-
ly diagnosed diabetes patients. BMJ Open Diabetes Research &amp; Care. 2017;5(1):e000429. i h
d i
i
ll
d
dh
d i
i 29. Marinho FS, Moram C, Rodrigues PC, Leite NC, Salles GF, Cardoso CR. Treatment adherence and its associ-
ated factors in patients with type 2 diabetes: results from the Rio de Janeiro type 2 diabetes cohort study. Jour-
nal of diabetes research. 2018;2018. 30. Wabe NT, Angamo MT, Hussein S. Medication adherence in diabetes mellitus and self management practices
among type-2 diabetics in Ethiopia. North American journal of medical sciences. 2011;3(9):418. 357 31. Zioga E, Kazakos K, Dimopoulos E, Koutras C, Marmara K, Marmara E-E, et al. Adherence and quality
of life in patients with type II diabetes mellitus in northern Greece. Materia socio-medica. 2016;28(4):258. p
yp
( )
32. Abedini MR, Bijari B, Miri Z, Shakhs Emampour F, Abbasi A. The quality of life of the patients with dia-
betes type 2 using EQ-5D-5 L in Birjand. Health and quality of life outcomes. 2020;18(1):18. 33. Gholami A, Azini M, Borji A, Shirazi F, Sharafi Z, Zarei E. References Non-adherence to self-care practices & medication and health
related quality of life among patients with type 2 diabetes: a cross-sectional study. BMC public health. 2014;14:431. 44. Khayyat SM, Mohamed MMA, Khayyat SMS, Hyat Alhazmi RS, Korani MF, Allugmani EB, et al. Asso-
ciation between medication adherence and quality of life of patients with diabetes and hypertension attend-
ing primary care clinics: a cross-sectional survey. Quality of life research : an international journal of qual-
ity of life aspects of treatment, care and rehabilitation. 2019;28(4):1053-61. 45. Lin EH, Katon W, Von Korff M, Rutter C, Simon GE, Oliver M, et al. Relationship of depression and
diabetes self-care, medication adherence, and preventive care. Diabetes care. 2004;27(9):2154-60. p
( )
46. Schram MT, Baan CA, Pouwer F. Depression and quality of life in patients with diabetes: a systematic
review from the European depression in diabetes (EDID) research consortium. Current diabetes reviews. 2009;5(2):112-9.
|
https://openalex.org/W4206584734
|
https://westminsterresearch.westminster.ac.uk/download/124d26335f374437415c2f634865bb0c569a020b3544d31a8e3bea98302c4bec/2554521/PIIS2452109421002487.pdf
|
English
| null |
Development and Optimization of a Machine-Learning Prediction Model for Acute Desquamation After Breast Radiation Therapy in the Multicenter REQUITE Cohort
|
Advances in radiation oncology
| 2,022
|
cc-by
| 9,456
|
Development and Optimization of a Machine-
Learning Prediction Model for Acute
Desquamation After Breast Radiation Therapy in
the Multicenter REQUITE Cohort Mahmoud Aldraimli, PhD,a Sarah Osman, PhD,b Diana Grishchuck, MSc,c
Samuel Ingram, MSc,d Robert Lyon, PhD,e Anil Mistry, MSc,f
Jorge Oliveira, PhD,g Robert Samuel, MBChB,h Leila E.A. Shelley, PhD,i
Daniele Soria, PhD,j Miriam V. Dwek, PhD,k
Miguel E. Aguado-Barrera, MD, PhD,l,m David Azria, MD,n
Jenny Chang-Claude, PhD,o,p Alison Dunning, PhD,q Alexandra Giraldo, MD,r
Sheryl Green, MD,s Sara Gutierrez-Enríquez, PhD,t Carsten Herskind, PhD,u
Hans van Hulle, MD,v Maarten Lambrecht, MD,w Laura Lozza, MD,x
Tiziana Rancati, MSc,y Victoria Reyes, MD,r Barry S. Rosenstein, PhD,z
Dirk de Ruysscher, MD,aa Maria C. de Santis, MD,x Petra Seibold, PhD,o
Elena Sperk, MD,u R. Paul Symonds, MD,bb Hilary Stobart,cc Sources of support: This research collaboration was formed by the UK Radiotherapy Machine Learning Network (RTML) funded through the
Advanced Radiotherapy Challenge+ by the Science and Technology Facilities Council (STFC). The REQUITE study received funding from the European
Union’s 7th Framework Programme for research, technological development and demonstration under grant agreement no. 601826. The research was
supported by the Quintin Hogg Trust research awards award no.165435391. The workshops were hosted by the University of Manchester and the Health
and Innovation Ecosystem at the University of Westminster. Dr Alison M. Dunning was supported by Cancer Research-UK C8197/A16565. Dr Sara
Gutierrez-Enríquez is supported by the ISCIII Miguel Servet II Program (CP16/00034). Dr Tim Rattay is currently an NIHR Clinical Lecturer (CL 2017-
11-002). He was previously funded by a National Institute of Health Research (NIHR) Doctoral Research Fellowship (DRF 2014-07-079). This publica-
tion presents independent research funded by the NIHR. The views expressed are those of the authors and not necessarily those of the NHS, the NIHR
or the Department of Health. Dr Leila Shelley reports grants from Chief Scientist Office (CSO) Scotland grant (TCS/17/26 - CSO Award). Dr Petra Sei-
bold is supported by the ERA-Net ERAPerMed / German Federal Ministry of Education and Reseach (BMBF) as well as German Federal Office for Radi-
ation Protection (BfS). Dr Elena Sperk was previously supported by the Ministry of Science and Arts of the State of Baden-W€urttemberg (2017-19)
through the Brigitte-Schlieben-Lange-Programme. Dr Ana Vega is supported by Spanish Instituto de Salud Carlos III (ISCIII) funding, an initiative of
the Spanish Ministry of Economy and Innovation partially supported by European Regional Development FEDER Funds (INT15/00070, INT16/00154,
INT17/00133, INT20/00071, PI19/01424, PI16/00046, PI13/02030, PI10/00164), and through the Autonomous Government of Galicia (Consolidation
and structuring program: IN607B). Scientific Article
Development and Optimization of a Machine-
Learning Prediction Model for Acute
Desquamation After Breast Radiation Therapy in
the Multicenter REQUITE Cohort
Mahmoud Aldraimli, PhD,a Sarah Osman, PhD,b Diana Grishchuck, MSc,c
Samuel Ingram, MSc,d Robert Lyon, PhD,e Anil Mistry, MSc,f
Jorge Oliveira, PhD,g Robert Samuel, MBChB,h Leila E.A. Shelley, PhD,i
Daniele Soria, PhD,j Miriam V. Dwek, PhD,k
Miguel E. Aguado-Barrera, MD, PhD,l,m David Azria, MD,n
Jenny Chang-Claude, PhD,o,p Alison Dunning, PhD,q Alexandra Giraldo, MD,r
Sheryl Green, MD,s Sara Gutierrez-Enríquez, PhD,t Carsten Herskind, PhD,u
Hans van Hulle, MD,v Maarten Lambrecht, MD,w Laura Lozza, MD,x
Tiziana Rancati, MSc,y Victoria Reyes, MD,r Barry S. Rosenstein, PhD,z
Dirk de Ruysscher, MD,aa Maria C. de Santis, MD,x Petra Seibold, PhD,o
Elena Sperk, MD,u R. Paul Symonds, MD,bb Hilary Stobart,cc
Advances in Radiation Oncology (2022) 7, 100890 Scientific Article
Development and Optimization of a Machine-
Learning Prediction Model for Acute
Desquamation After Breast Radiation Therapy in
the Multicenter REQUITE Cohort
Mahmoud Aldraimli, PhD,a Sarah Osman, PhD,b Diana Grishchuck, MSc,c
Samuel Ingram, MSc,d Robert Lyon, PhD,e Anil Mistry, MSc,f
Jorge Oliveira, PhD,g Robert Samuel, MBChB,h Leila E.A. Shelley, PhD,i
Daniele Soria, PhD,j Miriam V. Dwek, PhD,k
Miguel E. Aguado-Barrera, MD, PhD,l,m David Azria, MD,n
Jenny Chang-Claude, PhD,o,p Alison Dunning, PhD,q Alexandra Giraldo, MD,r
Sheryl Green, MD,s Sara Gutierrez-Enríquez, PhD,t Carsten Herskind, PhD,u
Hans van Hulle, MD,v Maarten Lambrecht, MD,w Laura Lozza, MD,x
Tiziana Rancati, MSc,y Victoria Reyes, MD,r Barry S. Rosenstein, PhD,z
Dirk de Ruysscher, MD,aa Maria C. de Santis, MD,x Petra Seibold, PhD,o
Elena Sperk, MD,u R. Paul Symonds, MD,bb Hilary Stobart,cc
Advances in Radiation Oncology (2022) 7, 100890 Advances in Radiation Oncology (2022) 7, 100890 Scientific Article ttps://do .o g/ 0. 0 6/j.ad o. 0
. 00890
2452-1094/© 2022 The Authors. Published by Elsevier Inc. on behalf of American Society for Radiation Oncology. This is an open access article under
the CC BY license (http://creativecommons.org/licenses/by/4.0/). Development and Optimization of a Machine-
Learning Prediction Model for Acute
Desquamation After Breast Radiation Therapy in
the Multicenter REQUITE Cohort Prof Catharine West is supported by Cancer Research UK (C1094/A18504, C147/A25254) and by the NIHR Man-
chester Biomedical Research Centre. Disclosures: Prof David Azria: has been involved in the creation of the start-up NovaGray in 2015. Prof Dirk de Ruysscher: none related to the cur-
rent manuscript. Outside the current manuscript: advisory board of Astra Zeneca, Bristol-Myers-Squibb, Roche/ Genentech, Merck/ Pfizer, Celgene,
Noxxon, Mologen and has received investigator initiated grants from Bristol-Myers-Squibb, Boehringer Ingelheim and Astra-Zeneca. Dr Elena Sperk:
none related to the current manuscript. Outside the current manuscript: General speakers bureau Zeiss Meditec, travel support Zeiss Meditec. The other
authors have no conflict of interests to disclose. Data sharing statement: Research data are stored in an institutional repository and will be shared upon request to the corresponding author and the
REQUITE consortium. 1 Members of the REQUITE consortium are listed in the Acknowledgments. *Corresponding author; E-mail: tr104@le.ac.uk https://doi.org/10.1016/j.adro.2021.100890 p
g
j
2452-1094/© 2022 The Authors. Published by Elsevier Inc. on behalf of American Society for Radiation Oncology. This is an open access article under
the CC BY license (http://creativecommons.org/licenses/by/4.0/). M. Aldraimli et al Advances in Radiation Oncology: May−June 2022 2 Bego~na Taboada-Valadares, MD,l,dd Christopher J. Talbot, PhD,ee
Vincent J.L. Vakaet, MD,v Ana Vega, PhD,l,m,ff Liv Veldeman, MD, PhD,v
Marlon R. Veldwijk, PhD,u Adam Webb, PhD,ee Caroline Weltens, MD,w
Catharine M. West, PhD,gg Thierry J. Chaussalet, PhD,a and
Tim Rattay, MBChB, PhDbb,*, REQUITE consortium1 aHealth Innovation Ecosystem, University of Westminster, London, United Kingdom; bPatrick G. Johnston Centre for Cancer
Research, Queen’s University Belfast, Belfast, United Kingdom; cImperial College Healthcare NHS Trust, London, United
Kingdom; dDivision of Cancer Sciences, Faculty of Biology, Medicine and Health, University of Manchester, Manchester,
United Kingdom; eDepartment of Computer Science, Edge Hill University, Ormskirk, Lancashire, United Kingdom; fGuy’s
and St. Thomas’ NHS Foundation Trust, London, United Kingdom; gMirada Medical, Oxford, United Kingdom; hUniversity
of Leeds, Leeds Cancer Centre, St. James’s University Hospital, Leeds, United Kingdom; iEdinburgh Cancer Centre, Western
General Hospital, Edinburgh, United Kingdom; jSchool of Computing, University of Kent, Canterbury, United Kingdom;
kSchool of Life Sciences, University of Westminster, London, United Kingdom; lFundacion Publica Galega Medicina
Xenomica, Santiago de Compostela, Spain; mInstituto de Investigacion Sanitaria de Santiago (IDIS), Servicio Galego de
Saude (SERGAS), Santiago de Compostela, Spain; nUniversity of Montpellier, Montpellier, France; oDivision of Cancer
Epidemiology, German Cancer Research Center (DKFZ), Heidelberg, Germany; pUKE University Cancer Center Hamburg,
University Medical Center Hamburg-Eppendorf, Hamburg, Germany; qCentre for Cancer Genetic Epidemiology, University
of Cambridge, Strangeways Research Laboratory, Worts Causeway, Cambridge, United Kingdom; rRadiation Oncology
Department, Vall d’Hebron Hospital Universitari, Vall d’Hebron Hospital Campus, Barcelona, Spain; sDepartment of
Radiation Oncology, Icahn School of Medicine at Mount Sinai, New York, New York; tHereditary Cancer Genetics Group,
Vall d'Hebron Institute of Oncology (VHIO), Vall d’Hebron Hospital Campus, Barcelona, Spain; uDepartment of Radiation
Oncology, Universit€atsmedizin Mannheim, Medical Faculty Mannheim, Heidelberg University, Mannheim, Germany;
vDepartment of Human Structure and Repair, Ghent University, Ghent, Belgium; wDepartment of Radiation Oncology,
University Hospital, Leuven, Belgium; xDepartment of Radiation Oncology 1, Fondazione IRCCS Istituto Nazionale dei
Tumori, Milan, Italy; yProstate Cancer Program, Fondazione IRCCS Istituto Nazionale dei Tumori, Milan, Italy; zIcahn
School of Medicine at Mount Sinai, New York, New York; aaMaastricht University Medical Center, Department of Radiation
Oncology (Maastro), GROW, Maastricht, The Netherlands; bbCancer Research Centre, University of Leicester, Leicester,
United Kingdom; ccIndependent Cancer Patients’ Voice, London, United Kingdom; ddDepartment of Radiation Oncology,
Complexo Hospitalario Universitario de Santiago, Servicio Galego de Saude (SERGAS), Santiago de Compostela, Spain;
eeDepartment of Genetics and Genome Biology, University of Leicester, Leicester, United Kingdom; ffBiomedical Network on
Rare Diseases (CIBERER), Madrid, Spain; ggUniversity of Manchester, Christie Hospital, Manchester, United Kingdom Received October 11, 2021; accepted December 6, 2021 Abstract
P Purpose: Some patients with breast cancer treated by surgery and radiation therapy experience clinically significant toxicity, which
may adversely affect cosmesis and quality of life. There is a paucity of validated clinical prediction models for radiation toxicity. We
used machine learning (ML) algorithms to develop and optimise a clinical prediction model for acute breast desquamation after whole
breast external beam radiation therapy in the prospective multicenter REQUITE cohort study. Methods and Materials: Using demographic and treatment-related features (m = 122) from patients (n = 2058) at 26 centers, we
trained 8 ML algorithms with 10-fold cross-validation in a 50:50 random-split data set with class stratification to predict acute breast
desquamation. Based on performance in the validation data set, the logistic model tree, random forest, and naïve Bayes models were
taken forward to cost-sensitive learning optimisation. g
Results: One hundred and ninety-two patients experienced acute desquamation. Resampling and cost-sensitive learning optimisation
facilitated an improvement in classification performance. Based on maximising sensitivity (true positives), the “hero” model was the cost-
sensitive random forest algorithm with a false-negative: false-positive misclassification penalty of 90:1 containing m = 114 predictive
features. Model sensitivity and specificity were 0.77 and 0.66, respectively, with an area under the curve of 0.77 in the validation cohort. Conclusions: ML algorithms with resampling and cost-sensitive learning generated clinically valid prediction models for acute
desquamation using patient demographic and treatment features. Further external validation and inclusion of genomic markers in ML
d
d l
h h l
d
f
d
k
f
h
b
fif g
Results: One hundred and ninety-two patients experienced acute desquamation. Resampling and cost-sensitive learning optimisation
facilitated an improvement in classification performance. Based on maximising sensitivity (true positives), the “hero” model was the cost-
sensitive random forest algorithm with a false-negative: false-positive misclassification penalty of 90:1 containing m = 114 predictive
features. Model sensitivity and specificity were 0.77 and 0.66, respectively, with an area under the curve of 0.77 in the validation cohort. Conclusions: ML algorithms with resampling and cost-sensitive learning generated clinically valid prediction models for acute
desquamation using patient demographic and treatment features. Further external validation and inclusion of genomic markers in ML
prediction models are worthwhile, to identify patients at increased risk of toxicity who may benefit from supportive intervention or
even a change in treatment plan. © 2022 The Authors. Published by Elsevier Inc. on behalf of American Society for Radiation Oncology. Abstract
P This is an open access
article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). ML Prediction Model for RT Acute Desquamation Advances in Radiation Oncology: May−June 2022 3 Study cohort and participants Several studies have examined the association between
acute breast radiation toxicity and clinical or treatment risk
factors.9-18 Nevertheless, statistical models have had limited
success to date in predicting individual patient toxicity
risk,19 and there is a paucity of validated prediction models
for acute breast radiation toxicity. It is hypothesized that
earlier prediction models failed to validate because they did
not include sufficient variables to capture the variety of sce-
narios that occur among individual patients and individual
treatment settings. Recent studies have demonstrated the
capability of machine learning (ML) to develop predictive
models for radiation toxicities in different cancers,20,21
including a thermal image-based random forest (RF) classi-
fier for radiation dermatitis (skin erythema) after the first
week of radiation therapy.22 Another recently published
abstract describes how RF, gradient boosted decision tree,
and logistic regression models were trained and validated
on treatment planning and patient data comprising 230
variables including toxicity symptoms from patients at 5
collaborating U.S. centers to predict moist desquamation
and Common Terminology Criteria for Adverse Events
(CTCAE) grade ≥2 radiation dermatitis.23 REQUITE is an international, prospective cohort
study that recruited patients with cancer before radiation
therapy in 26 hospitals from 8 countries between April
2014 and March 2017 with unified standardized data
collection.26 Patient baseline characteristics and method-
ology have been described in detail elsewhere.27 The
present study used data from the breast cancer cohort
(n = 2069). All patients were treated with breast-conserv-
ing surgery followed by whole breast EBRT according to
local protocol. Partial breast irradiation and brachyther-
apy were excluded. Patients were assessed at the start
and end of radiation therapy, and annually thereafter. Data collected at the start and at the end of radiation
treatment were used to document acute toxicity. All
patients gave written informed consent. The study was
approved by local ethics committees in participating
countries and registered at www.controlled-trials.com
(ISRCTN 98496463). Introduction intervention or treatment de-escalation. To that extent, a
logistic regression model for acute breast desquamation
after external beam radiation therapy (EBRT) recently
developed in 3 combined radiation therapy cohorts failed
to validate externally in the multicenter international
REQUITE cohort.24 Therefore, the aim of this study was
to use ML algorithms to develop and optimise a prediction
model for acute breast desquamation after EBRT in the
REQUITE breast cancer cohort. Radiation therapy is recommended for all patients with
breast cancer who have a local excision and after mastectomy
in high-risk patients.1 Over 70% of patients with breast can-
cer receive radiation therapy, which reduces local recurrence
rates and increases long-term survival.2 As survival from
breast cancer continues to improve,3 quality of life and survi-
vorship have become increasingly important research priori-
ties.4 Risk of radiation toxicity can be estimated from
empirical dosimetric models based on the dose to the target
organ and surrounding tissue.5 However, there is consider-
able variation between individual patient normal tissue reac-
tion to radiation therapy and the extent to which they
develop toxicity.6 Acute toxicity (<90 days from starting
treatment) includes breast erythema and desquamation (skin
loss).7 In a minority of patients, desquamation can cause sub-
stantial patient morbidity, worsen the cosmetic outcome after
surgery, and affect quality of life.8 It can even result in the
interruption of radiation therapy or a dose reduction, poten-
tially increasing the risk of local recurrence. Methods and Materials This was a TRIPOD (transparent reporting of a multi-
variable prediction model for individual prognosis or
diagnosis) type 2a study using a single data set with a ran-
dom split sample for development and validation.25 The
full study design is shown in Figure 1. Endpoint definition For cancers with generally good local tumor control
such as breast cancer, it is hypothesized that if a patient’s
individual risk of radiation toxicity could be estimated at
the time of diagnosis, this could inform discussions about
risks and benefits and allow treatment plans to be person-
alized for high-risk patients to minimize toxicity. Clinicians
are particularly interested in models that include readily
available clinical and treatment variables, which would
allow toxicity risk to be estimated before treatment is
planned. It is also important to predict toxicities that are
sufficiently significant to warrant increased supportive Toxicity in REQUITE was scored by treating physi-
cians using CTCAE v4.0.28 CTCAE v4.0 has separate
scales for radiation dermatitis (erythema) and skin ulcera-
tion (skin loss). The primary endpoint of this study was
acute desquamation (skin loss or moist desquamation)
occurring by the end of radiation therapy, defined as
either CTCAE grade ≥3 radiation dermatitis (moist des-
quamation) or CTCAE grade ≥1 skin ulceration, implying
that skin integrity was broken over the breast or in the
inframammary fold. Patients with high baseline scores Advances in Radiation Oncology: May−June 2022 M. Aldraimli et al 4 Fig. 1
Diagram depicting the overall study design showing data preprocessing, splitting, and imputation at the top, and
model development and optimization at the bottom. Abbreviations: ANN = artificial neural network; BVA = boundary
value analysis; C4.5 = C4.5 decision tree; CS = cost sensitive optimisation; DMI = decision-tree-based missing-value impu-
tation; ECP = equivalence class partitioning; ITD = imbalanced training data set; KNN = K-nearest neighbor;
LMT = logistic model tree; LR = logistic regression; NB = naïve Bayes; RF = random forest; ROS = random over-sampling;
RUS
random under sampling SMOTE
synthetic minority oversampling technique SVM
support vector mecha Fig. 1
Diagram depicting the overall study design showing data preprocessing, splitting, and imputation at the top, and
model development and optimization at the bottom. Abbreviations: ANN = artificial neural network; BVA = boundary
value analysis; C4.5 = C4.5 decision tree; CS = cost sensitive optimisation; DMI = decision-tree-based missing-value impu-
tation; ECP = equivalence class partitioning; ITD = imbalanced training data set; KNN = K-nearest neighbor;
LMT = logistic model tree; LR = logistic regression; NB = naïve Bayes; RF = random forest; ROS = random over-sampling;
RUS = random under-sampling; SMOTE = synthetic minority oversampling technique; SVM = support vector mecha-
nism; VD = validation data set. Endpoint definition Fig. 1
Diagram depicting the overall study design showing data preprocessing, splitting, and imputation at the top, and
model development and optimization at the bottom. Abbreviations: ANN = artificial neural network; BVA = boundary
value analysis; C4.5 = C4.5 decision tree; CS = cost sensitive optimisation; DMI = decision-tree-based missing-value impu-
tation; ECP = equivalence class partitioning; ITD = imbalanced training data set; KNN = K-nearest neighbor;
LMT = logistic model tree; LR = logistic regression; NB = naïve Bayes; RF = random forest; ROS = random over-sampling;
RUS = random under-sampling; SMOTE = synthetic minority oversampling technique; SVM = support vector mecha-
nism; VD = validation data set. (defined as CTCAE grade >1 radiation dermatitis or
CTCAE grade >0 skin ulceration) were excluded from the
analysis, as this would not be attributable to the effect of
radiation therapy. therapy physicists, and m = 136 variables remained. Boundary value analysis and equivalence class partition-
ing techniques29 were used for correcting or removing
corrupt or inaccurate records from the data set. After vari-
able-dropping (m = 13 with >37% missing values at ran-
dom compared with observed values in the remaining
variables) and case-wise deletion (n = 11 with missing
class endpoint observations),30 the final data set for
modeling had m = 123 variables including the endpoint
variable and n = 2058 patient records. Modeling Eight different ML algorithms were used to build
binary classification models to predict acute desquama-
tion in patients undergoing breast-conserving surgery and
adjuvant whole breast EBRT. They were trained in the
ITD (imbalanced modeling, Fig. 1) as well as in the 3
resampled data sets (RUS, ROS, SMOTE; data-bias
modeling, Fig. 1) with 10-fold cross-validation to reduce
overfitting,38 and then each was tested in the VD (see
Fig. 1). The ML alogrithms were discretized naïve Bayes
(NB), logistic regression with ridge estimator,39 artificial
neural networks with a multilayer perceptron architec-
ture,40 support vector machine with polynomial kernel
and
logistic
calibrator,
K-nearest
neighbour41
with
K = 1,3,5,7,9, decision trees (C4.5),32 logistic model tree
(LMT),42 and RF.43 The final set of m = 123 features consisted of 106 raw
variables. Sixteen additional features were constructed to
account for the vast number of possible combinations of
chemotherapeutic agents received by some patients before
radiation therapy. The adjuvant chemotherapy regimens
were binarized based on their generic drug names
(not shown). To adjust for different radiation therapy reg-
imens, dose was calculated as the biologically effective
dose (BED). BED is the product of the number of frac-
tions (n), dose per fraction (d), and a factor determined
by the dose and a/b ratio (10 Gy) for desquamation
(acute toxicity): Model performance was assessed using the area under
the curve (AUC). The models with the highest AUC in
the VD were taken forward for cost-sensitive learning
optimisation. Cost-sensitive classification addresses the
issue of class imbalance by imposing penalties (costs) for
the misclassification of the positive cases (ie, making a
false negative [FN] prediction). In this study, the cost for
a FN prediction was not linked to a monetary value,
instead a 10-step incremental inverse class distribution
cost was used.44 The ITD has a 96:933 ffi1:10 ratio of
examples in the positive class to examples in the negative
class. This ratio is inverted to penalize FN with a 10-step
incrementation at an initial cost x : 1 of 10:1 increasing to
100:1. The cost is applied in the form of Charles Elkan’s
explicit cost matrix notation45: BED ¼ n d
1 þ d
a=b
The endpoint definition (acute desquamation = Desq)
was used to label the patients to create a binary class vari-
able. Variable selection, imputation, and
preprocessing The raw REQUITE data set (n = 2069) contained
m = 204 variables (features) relating to patient baseline
characteristics, comorbidities, cotreatments, and radiation
therapy. Variables were initially checked for plausibility
using domain expertise by physicians and radiation The final data set was randomly shuffled and split
50:50 into training and test sets with class stratification,
yielding imbalanced training (ITD, n = 1029) and valida-
tion (VD, n = 1029) data sets. Each data set was imputed ML Prediction Model for RT Acute Desquamation Advances in Radiation Oncology: May−June 2022 5 independently with the ML decision-tree-based missing-
value imputation technique31 to enhance best expecta-
tions of missing values. By carrying out the imputation of
the data sets separately, the ITD and VD remained
completely independent and perfectly isolated. Informa-
tion levels were monitored in each data set pre- and post-
imputation with information gain attribute evaluation32
(see Fig. E4). The information gain of a feature is defined
as the expected reduction of entropy (uncertainty within
the data set) when partitioning the data; in other words,
by how much the prediction of the endpoint/class would
improve if the data were split using just that feature. The
more plausible the pattern of information gain among
data sets, the less bias is introduced in modeling. The eval-
uation of information worth is affected by the number of
records and the 50:50 training-test split allowed for a fair
information bias comparison between training and vali-
dation data sets. independently with the ML decision-tree-based missing-
value imputation technique31 to enhance best expecta-
tions of missing values. By carrying out the imputation of
the data sets separately, the ITD and VD remained
completely independent and perfectly isolated. Informa-
tion levels were monitored in each data set pre- and post-
imputation with information gain attribute evaluation32
(see Fig. E4). The information gain of a feature is defined
as the expected reduction of entropy (uncertainty within
the data set) when partitioning the data; in other words,
by how much the prediction of the endpoint/class would
improve if the data were split using just that feature. The
more plausible the pattern of information gain among
data sets, the less bias is introduced in modeling. The eval-
uation of information worth is affected by the number of
records and the 50:50 training-test split allowed for a fair
information bias comparison between training and vali-
dation data sets. Variable selection, imputation, and
preprocessing projection analysis into a 3-dimensional point cloud. Adap-
tive projection analysis37 is a multidimensional tool to visu-
alise the classes that can be separated, any outliers or
sources of error in the classification algorithms, and the
existence of clusters in the data (see Fig. E5). Modeling All numeric features (m = 63) were normalised with
z score standardization.33 Cost Matrix combinations FP 1
ð Þ
TN 0
ð Þ
TP 0
ð Þ
FN x
ð Þ
¼
1
0
0
10
;
1
0
0
20
;
1
0
0
30
; ⋯;
1
0
0
100
Resampling patients from the REQUITE patient cohort
REQUITE breast cancer cohort
Eligible patients
2059
Location
Western Europe, United States
Study design
Prospective cohort
Recruitment year
(range)
2014-2016
Treatment year
(range)
2014-2016
Toxicity assessment
scale
CTCAE v4.0
Toxicity assessment
time points
Start-of-RT
End-of-RT
Age (median, range)
58 (23-90)
Whole breast dose
(Gy, median, range)
50 (28.5-56)
Whole breast frac-
tions (median,
range)
25 (5-31)
Hypofractionated
regimen (propor-
tion of patients)
47.9%
IMRT, simple field-
in-field
39.7%
IMRT, complex
modulated
9.8%
RT to axilla
11.9%
RT to supraclavicular
fossa
12.8%
Boost
67.8%
BMI ≥25
54.0%
Smoker (current or
previous)
42.7%
Chemotherapy
31.0%
Diabetes
6.1%
Hypertension
28.0%
Cardiovascular
disease
6.9%
Toxicity (end of
treatment)
Ulceration
Grade 0
1868 (91.2%)
Grade ≥1
181 (8.8%)
Dermatitis
Grade 0
257 (12.5%)
Grade 1
1288 (62.6%)
(continued on next page) Abbreviations: BMI = body mass index; CTCAE = Common Termi-
nology Criteria for Adverse Events; IMRT = intensity modulated
radiation therapy; RT = radiation therapy. analysis 3.8.3 (with the default models' parameters
settings),47 with the C4.5 decision tree using the J4848
implementation,
K-nearest
neighbor
using
the
IBK
(instance-bases learning with parameter k) implementa-
tion, and support vector machine using the SMO (sequen-
tial minimal optimization)49 implementation. Resampling Although it is a clinically significant side effect from
breast radiation therapy, only a small proportion of
patients suffer from acute desquamation, an issue known
as “class imbalance.”9,18 Both ITD and VD in this study
were equally imbalanced (Desqþ ¼ 96; Desq ¼ 933). To address the issue of class imbalance, 3 resampling tech-
niques were applied to the training data to obtain equal
proportions of records in each class: random under-sam-
pling (RUS) (n = 192, Desqþ ¼ 96; Desq ¼ 96),34 ran-
dom over-sampling (ROS) (n
¼ 1866; Desqþ ¼ 933;
Desq ¼ 933),35 and the synthetic minority oversampling
technique (SMOTE) (n ¼ 1866; Desqþ ¼ 933; Desq ¼
933).36 The effect of resampling techniques on the training
data set was monitored with a multidimensional adaptive AUC, sensitivity (true positive rate [TPR]), and speci-
ficity (true negative rate [TNR]) were used to compare
and interpret the final models’ performance including
those developed in the resampled data sets (see bottom
half of Fig. 1). Final model selection was based on perfor-
mance in the VD in terms of AUC and the clinicians’
trade-off maximizing both TPR and TNR. The selected
model was further optimized using the mean decrease
impurity entropy filter to select fewer features and sim-
plify the “hero” model.46 All ML algorithms were imple-
mented in the Waikato environment for knowledge Advances in Radiation Oncology: May−June 2022 6 M. Aldraimli et al 6 Table
1
Summary
study
characteristics
of
eligible
patients from the REQUITE patient cohort Table 1 (Continued)
REQUITE breast cancer cohort
Grade 2
462 (22.4%)
Grade 3
28 (1.4%)
Acute desquamation
Ulceration ≥G1 or
dermatitis ≥G3
192 (9.3%)
Abbreviations: BMI = body mass index; CTCAE = Common Termi-
nology Criteria for Adverse Events; IMRT = intensity modulated
radiation therapy; RT = radiation therapy. Results Table 1 shows the main patient and treatment demo-
graphics for eligible patients. Median patient age was
58 years (range, 23-80 years). Patients were treated with a
median breast dose of 50 Gy (28.5-56 Gy) in 25 fractions
(range, 5-31) according to local protocol. In terms of
important demographic features, 54.0% of patients had a
body mass index ≥25, 42.7% were previous or current
smokers, 31.0% had also undergone chemotherapy, 6.1%
had diabetes, and 28.0% and 6.9% had hypertension and
cardiovascular disease, respectively. About half of the
patients were treated with intensity modulated radiation
therapy, with a lower proportion in France and none at
Italian or U.S. centers. The majority of patients received a
tumor-bed boost (64%), ranging from less than 20% at
the French, Italian, and Spanish centers to over 80% at
the Belgian centers, given either simultaneously (n = 257)
or sequentially (n = 1138). Patients with invasive breast
cancer in Belgium and the United Kingdom were treated
using the Standardisation of Breast Radiation therapy
Trial B (START-B) hypofractionated regimen (40 Gy in
15 fractions). In terms of regional nodal irradiation, axil-
lary nodes were treated in 11.9% and the supraclavicular
fossa was treated in 12.8% of patients, respectively. Detailed characteristics of the REQUITE patient cohorts
have previously been described elsewhere.27 Table 2 lists the performance of 12 ML classifiers using
8 different algorithms in terms of each model’s AUC,
TPR (sensitivity), and TNR (specificity). Results and TNR (the clinicians’ trade-off), with lower and
upper threshold values of 0.63 and 0.70, respectively,
which
were
crossed
by
5
classifiers:
cost-sensitive
RF
(CS-RF)
with
an
FN:false
positive
(FP)
90:1
penalty (TNR = 0.65, TPR = 0.77, AUC = 0.76); RUS-RF
(TNR = 0.65, TPR = 0.74, AUC = 0.74); cost-sensitive
NB
with
an
FN:FP
60:1
penalty
(TNR
=
0.64,
TPR = 0.70, AUC = 0.72); CS-RF with an FN:FP 80:1
penalty (TNR = 0.70, TPR = 0.65, AUC = 0.75); and
cost-sensitive
NB
with
an
FN:FP
20:1
penalty
(TNR = 0.70, TPR = 0.63, AUC = 0.73). As maximizing
sensitivity (TPR) was most important, the best perform-
ing “hero” model was the CS-RF classifier with an FN:FP
penalty of 90:1. This model exceeded others for sensitiv-
ity and AUC performance while maintaining moderate
specificity. the ITD. The 3 best-performing classifiers in terms of
AUC in the VD were LMT, RF, and NB with 0.75, 0.74,
and 0.74, respectively. These were selected for cost-sensi-
tive learning optimisation with incremental penalty rising
in 10 steps from 10 to 100. All 12 ML classifiers listed in
Table 1 were also applied to the 3 resampled training data
sets (RUS, ROS, and SMOTE). Figure 2 shows radar charts plotting sensitivity (TPR)
and specificity (TNR) in the VD for a total of 66 models
in the resampled training data (Fig. 2A-C) and after
applying cost-sensitive penalties to the 3 best performing
classifiers (Fig. 2D-F). Resampling improved sensitivity
across all classifiers, with RUS (Fig. 2A) achieving the least
variance between specificity and sensitivity on validation. For the cost-sensitive classifiers, the incremental penalty
skewed the correct classification toward the true positives
and models with higher penalty showed higher sensitivity
(TPR). NB model sensitivity ranged from 0.50 in the
unpenalized model to 0.77 for a penalty of 100. The larg-
est improvement in sensitivity was achieved for the RF
classifier, ranging from 0.01 for the unpenalized model to
0.79 at penalty of 100. LMT sensitivity improved from
0.04 without a penalty to 0.65 with a penalty of 100. Speci-
ficity (TNR) decreased for all 3 cost-sensitive classifiers
because the number of predicted false-positives increased
with each incremental penalty. p
y
The hero CS-RF (90:1) model had m = 122 features. Results Eight features were estimated to have zero importance
including features about presence/absence of systemic
lupus erythematosus and other collagen vascular diseases
and use of pertuzumab, eribulin, and amiodarone therapy. In a final step, these features were removed and the model
was rebuilt and revalidated in the VD. Table 3 lists the
features included in the final hero CS-RF classifier by
order of importance. In descending order, the top 10 fea-
tures were duration of other lipid-lowering drug use, type
of surgery (wide local excision vs quadrantectomy), use of
radiation therapy bolus, use of chemotherapy, use of
boost, radiation therapy photon dose (MV), use of epiru-
bicin therapy, hypertension, bra band size, and side of
radiation therapy. Performance of the optimized hero
final model in the VD improved slightly in terms of Results Accuracy was
biased strongly
toward
the majority
negative
class
(DesqÞ as shown by consistently high TNRs and low
TPRs across all models, likely due to class imbalance in ML Prediction Model for RT Acute Desquamation Advances in Radiation Oncology: May−June 2022 7 Table 2
Model performance with imputed imbalanced training data set DMI(ITD) and validation data set DMI(VD)
Training in ITD (n = 1029)
Validation in VD (n = 1029)
Classifier
Specificity (TNR)
Sensitivity (TPR)
AUC
Specificity (TNR)
Sensitivity (TPR)
AUC
Rank
(K = 1) NN
0.908
0.167
0.548
0.923
0.292
0.607
9
(K = 3) NN
0.975
0.094
0.601
0.979
0.125
0.627
8
(K = 5) NN
0.985
0.042
0.624
0.989
0.063
0.651
6
(K = 7) NN
0.996
0.031
0.648
0.998
0.052
0.644
7
(K = 9) NN
0.999
0.031
0.660
0.999
0.042
0.665
5
ANN
0.945
0.198
0.694
0.953
0.177
0.676
4
C4.5
0.985
0.083
0.575
0.979
0.125
0.496
12
LMT
0.996
0.010
0.578
0.995
0.042
0.746
1
LR
0.910
0.188
0.567
0.959
0.135
0.596
10
NB
0.810
0.438
0.697
0.833
0.500
0.737
3
SVM
0.966
0.156
0.561
0.976
0.146
0.561
11
RF
0.998
0.021
0.725
0.999
0.010
0.742
2
Abbreviations: ANN = artificial neural network; AUC = area under the curve; C4.5 = decision tree; DMI = decision-tree based missing value imputa-
tion; ITD = imbalanced training; KNN = K-nearest neighbor; LMT = logistic model tree; LR = logistic regression; NB = naïve Bayes; RF = random
forest; SVM = support vector machine; TNR = true negative rate; TPR = true positive rate; VD = validation. Table 2
Model performance with imputed imbalanced training data set DMI(ITD) and validation data set DMI(VD) performance with imputed imbalanced training data set DMI(ITD) and validation data set DMI(VD Table 2
Model performance with imputed imbalanced training data set DMI(ITD) and validation Validation in VD (n = 1029) Training in ITD (n = 1029) Abbreviations: ANN = artificial neural network; AUC = area under the curve; C4.5 = decision tree; DMI = decision-tree based missing value imputa-
tion; ITD = imbalanced training; KNN = K-nearest neighbor; LMT = logistic model tree; LR = logistic regression; NB = naïve Bayes; RF = random
forest; SVM = support vector machine; TNR = true negative rate; TPR = true positive rate; VD = validation. Model selection and feature filtering Figure 3 shows 2 conditions (finishing lines) selected
to maximize accuracy, that is, maximizing both TPR Advances in Radiation Oncology: May−June 2022 Advances in Radiation Oncology: May−June 2022 8 M. Aldraimli et al g. 2
Radar charts plotting sensitivity (TPR) and specificity (TNR) in the validation data set for all ML models dev
with the RUS, ROS, and SMOTE resampled training data and after applying cost-sensitive learning to the 3 best-pe
ng ML models (RF, NB, and LMT). Abbreviations: LMT = logistic model tree; ML = machine learning; NB = naïve
F = random forest; ROS = random over-sampling; RUS = random under-sampling; SMOTE = synthetic minority
ampling technique; TNR = true negative rate; TPR = true positive rate. gy
y Fig. 2
Radar charts plotting sensitivity (TPR) and specificity (TNR) in the validation data set for all ML models developed
with the RUS, ROS, and SMOTE resampled training data and after applying cost-sensitive learning to the 3 best-perform-
ing ML models (RF, NB, and LMT). Abbreviations: LMT = logistic model tree; ML = machine learning; NB = naïve Bayes;
RF = random forest; ROS = random over-sampling; RUS = random under-sampling; SMOTE = synthetic minority over-
sampling technique; TNR = true negative rate; TPR = true positive rate. Advances in Radiation Oncology: May−June 2022 ML Prediction Model for RT Acute Desquamation 9 Fig. 3
Trade-off threshold lines are shown for sensitivity (TPR) and specificity (TNR) at 0.63 and 0.70, respectively. Five
models cross both threshold lines and their TPR, TNR, and AUC values are shown at the bottom. Two out of 5 models
have a higher TNR than TPR and 3 out of the 5 models have a higher TPR than TNR. The “hero” model (no. 1) was the
cost-sensitive random forest algorithm with a penalty of 90:1. Abbreviations: AUC = area under the curve; TNR = true
negative rate; TPR = true positive rate. Fig. 3
Trade-off threshold lines are shown for sensitivity (TPR) and specificity (TNR) at 0.63 and 0.70, respectively. Five
models cross both threshold lines and their TPR, TNR, and AUC values are shown at the bottom. Two out of 5 models
have a higher TNR than TPR and 3 out of the 5 models have a higher TPR than TNR. The “hero” model (no. 1) was the
cost-sensitive random forest algorithm with a penalty of 90:1. Model selection and feature filtering Abbreviations: AUC = area under the curve; TNR = true
negative rate; TPR = true positive rate. specificity (TNR = 0.66) and AUC (0.77) while sensitivity
(TPR) remained unchanged. specificity (TNR = 0.66) and AUC (0.77) while sensitivity
(TPR) remained unchanged. Predicting cases of clinically significant radiation toxic-
ity such as acute desquamation remains challenging for
both parametric statistical and ML models due to the
issue of class imbalance leading to high FN rates, that is,
poor sensitivity. In this study, a combination of resam-
pling techniques and cost-sensitive learning was used to
try and improve predictive performance. RUS and cost-
sensitive optimisation contributed the most to optimal
performance across the different ML algorithms. Of 66
models tested, 5 fulfilled prespecified criteria for maximiz-
ing both TPR and TNR. On the basis of highest TPR, the
hero model was the CS-RF, with an FN:FP misclassifica-
tion penalty of 90:1. Given that our modeling used some-
what fewer features and had a multicenter patient sample
with diverse radiation treatment regimens, it is reassuring
that its AUC of 0.77 in the VD is similar to the range of
AUCs reported in the abstract by Reddy et al.23 Discussion A recently published logistic regression model for
acute breast desquamation after adjuvant external beam
breast radiation therapy developed in 3 combined external
breast radiation therapy cohorts failed to validate in the
multicenter REQUITE cohort.24 The aim of this study
was to use ML algorithms to develop and optimise a pre-
diction model for acute desquamation in the REQUITE
breast cancer cohort . ML techniques have previously
been used to predict acute skin toxicity during breast radi-
ation therapy.22,23 We elected to predict the occurrence of
acute desquamation rather than dermatitis (skin ery-
thema) because it can cause clinically significant patient
morbidity and can worsen the cosmetic outcome after
breast surgery. This accounts for the lower proportion of
cases with skin toxicity reported in our study versus the
study by Saednia et al22 (0.09 vs 0.38), although the pro-
portion of cases was similar to those with moist desqua-
mation in the abstract published by Reddy et al.23 Our initial models for acute desquamation included 122
features. Information gain (IG) represents the amount of
information gained about a random variable or signal
from observing another random variable. After the ran-
domized and stratified training/validation data split, only a
few variables in the VD had a different IG to discriminate
between the positive and negative cases compared with the
ITD. Zero IG does not negate the feature’s worth as this Advances in Radiation Oncology: May−June 2022 Advances in Radiation Oncology: May−June 2022 10 M. Discussion Aldraimli et al Table 3
Features in the “hero” optimized cost-sensitive RF classifier ranked by importance
Model’s feature
MDI
Model's feature
MDI
other_lipid_lowering_drugs_duration_yrs
0.52
alcohol_current_consumption
0.2
surgery_type
0.41
smoking_time_since_quitting_yrs
0.2
radio_bolus
0.4
radio_imrt
0.19
chemotherapy
0.36
radio_photon_boostdose_Gy
0.19
boost
0.35
other_antihypertensive_drug
0.19
radio_photon_dose_MV
0.34
household_members
0.19
epirubicin_chemo_drug
0.34
radio_breast_fractions_dose_per_fraction_Gy
0.19
blood_pressure
0.33
radio_elec_boost_field_y_cm
0.19
Bra_band_size
0.3
radio_photon_2nd
0.19
radio_treated_breast
0.3
bra_cup_size
0.19
tumour_size_mm
0.29
radio_breast_fractions
0.19
paclitaxel_chemo_drug
0.29
n_stage
0.18
grade_invasive
0.28
hypertension_duration_yrs
0.18
breast_separation
0.28
radio_supraclavicular_fossa
0.18
smoking
0.27
education_profession
0.18
radio_elec_energy_MeV
0.27
radio_axillary_levels
0.18
BED_boost
0.27
hypertension
0.18
docetaxel_chemo_drug
0.27
radio_photon_boost_fractions_per_week
0.17
BED_Total
0.27
smoker
0.17
radio_elec_boost_dose_Gy
0.27
depression
0.17
On_tamoxifen
0.26
menopausal_status
0.17
radio_heart_mean_dose_Gy
0.26
radio_boost_diameter_cm
0.16
t_stage
0.26
5-fluorouracil (5-FU)_chemo_drug
0.16
radio_hot_spots_107
0.25
radio_photon_boost_dose_per_fraction_Gy
0.16
BED_Breast
0.25
antidepressant_duration_yrs
0.16
tobacco_products_per_day
0.25
radio_breast_fractions_per_week
0.15
age_at_radiotherapy_start_yrs
0.25
radio_boost_type
0.15
radio_breast_ct_volume_cm3
0.25
Carboplatin_chemo_drug
0.15
hormone_replacement_therapy
0.24
radio_boost_sequence
0.15
radio_photon_boost_volume_cm3
0.24
radio_photon_boost_fractions
0.15
antidepressant
0.24
household_income
0.15
height_cm
0.24
methotrexate_chemo_drug
0.15
radio_photon_2nd_energy_MV
0.24
other_lipid_lowering_drugs
0.14
radio_ipsilateral_lung_mean_Gy
0.24
radio_photon_energy_MV or kV
0.14
alcohol_previous_consumption
0.24
ace_inhibitor
0.13
radio_photon_2nd_dose_fractions_per_week
0.23
analgesics_duration_yrs
0.13
radio_skin_max_dose_Gy
0.23
radio_photon_2nd_dose_per_fraction_Gy
0.13
histology
0.23
antidiabetic_duration_yrs
0.13
monopause_age_yrs
0.23
depression_duration_yrs
0.13
other_antihypertensive_drug_duration_yrs
0.23
on_statin_duration_yrs
0.12
weight_at_cancer_diagnosis_kg
0.23
antidiabetic
0.12 he “hero” optimized cost-sensitive RF classifier ranked by importance Table 3
Features in the “hero” optimized cost-sensitive RF classifier ranked by importance assifier ranked by importance
Model's feature
MDI
alcohol_current_consumption
0.2
smoking_time_since_quitting_yrs
0.2
radio_imrt
0.19
radio_photon_boostdose_Gy
0.19
other_antihypertensive_drug
0.19
household_members
0.19
radio_breast_fractions_dose_per_fraction_Gy
0.19
radio_elec_boost_field_y_cm
0.19
radio_photon_2nd
0.19
bra_cup_size
0.19
radio_breast_fractions
0.19
n_stage
0.18
hypertension_duration_yrs
0.18
radio_supraclavicular_fossa
0.18
education_profession
0.18
radio_axillary_levels
0.18
hypertension
0.18
radio_photon_boost_fractions_per_week
0.17
smoker
0.17
depression
0.17
menopausal_status
0.17
radio_boost_diameter_cm
0.16
5-fluorouracil (5-FU)_chemo_drug
0.16
radio_photon_boost_dose_per_fraction_Gy
0.16
antidepressant_duration_yrs
0.16
radio_breast_fractions_per_week
0.15
radio_boost_type
0.15
Carboplatin_chemo_drug
0.15
radio_boost_sequence
0.15
radio_photon_boost_fractions
0.15
household_income
0.15
methotrexate_chemo_drug
0.15
other_lipid_lowering_drugs
0.14
radio_photon_energy_MV or kV
0.14
ace_inhibitor
0.13
analgesics_duration_yrs
0.13
radio_photon_2nd_dose_per_fraction_Gy
0.13
antidiabetic_duration_yrs
0.13
depression_duration_yrs
0.13
on_statin_duration_yrs
0.12
antidiabetic
0.12
(continued on next page Table 3
Features in the hero optimized cost sensitive RF
Model’s feature
MDI
other_lipid_lowering_drugs_duration_yrs
0.52
surgery_type
0.41
radio_bolus
0.4
chemotherapy
0.36
boost
0.35
radio_photon_dose_MV
0.34
epirubicin_chemo_drug
0.34
blood_pressure
0.33
Bra_band_size
0.3
radio_treated_breast
0.3
tumour_size_mm
0.29
paclitaxel_chemo_drug
0.29
grade_invasive
0.28
breast_separation
0.28
smoking
0.27
radio_elec_energy_MeV
0.27
BED_boost
0.27
docetaxel_chemo_drug
0.27
BED_Total
0.27
radio_elec_boost_dose_Gy
0.27
On_tamoxifen
0.26
radio_heart_mean_dose_Gy
0.26
t_stage
0.26
radio_hot_spots_107
0.25
BED_Breast
0.25
tobacco_products_per_day
0.25
age_at_radiotherapy_start_yrs
0.25
radio_breast_ct_volume_cm3
0.25
hormone_replacement_therapy
0.24
radio_photon_boost_volume_cm3
0.24
antidepressant
0.24
height_cm
0.24
radio_photon_2nd_energy_MV
0.24
radio_ipsilateral_lung_mean_Gy
0.24
alcohol_previous_consumption
0.24
radio_photon_2nd_dose_fractions_per_week
0.23
radio_skin_max_dose_Gy
0.23
histology
0.23
monopause_age_yrs
0.23
other_antihypertensive_drug_duration_yrs
0.23
weight_at_cancer_diagnosis_kg
0.23 ML Prediction Model for RT Acute Desquamation Advances in Radiation Oncology: May−June 2022 11 Table 3 (Continued)
Model’s feature
MDI
Model's feature
MDI
tobacco_product
0.23
diabetes
0.11
cyclophosphamide_chemo_drug
0.22
ace_inhibitor_duration_yrs
0.11
combined_chemo_drugs
0.22
on_statin
0.11
boost_frac
0.22
doxorubicin_chemo_drug
0.11
analgesics
0.22
history_of_heart_disease
0.09
breast_cancer_family_history_1st_degree
0.22
radio_axillary_other
0.09
smoking_duration_yrs
0.21
ethnicity
0.09
radio_photon_boostdose_precise_Gy
0.21
radio_interrupted
0.08
radio_elec_boost_field_x_cm
0.21
pegfilgrastim_chemo_drug
0.07
radio_photon_2nd_fractions
0.21
history_of_heart_disease_duration_yrs
0.06
radio_boost_fractions
0.21
radiotherapy_toxicity_family_history
0.06
alcohol_intake
0.21
diabetes_duration_yrs
0.05
radio_type_imrt
0.21
radio_interrupted_days
0.05
radio_treatment_pos
0.21
trastuzumab_chemo_drug
0.04
radio_breast_dose_Gy
0.2
other_collagen_vascular_disease
0.03
rheumatoid arthritis_duration_yrs
0.2
rheumatoid arthritis
0.02
Abbreviations: BED = biologically effective dose; IMRT = intensity modulated radiation therapy; MDI = mean decrease impurity; MeV = mega elec-
tron volt; MV = mega volt; RF = random forest. Discussion Feature importance is calculated as the decrease in node impurity weighted by the probability of reaching that node. The node probability can be cal-
culated by the number of samples that reach the node, divided by the total number of samples. The higher the value, the more important the feature. after the end of treatment. Although we incorporated dif-
ferences in radiation therapy techniques by including all
available recorded treatment parameters in the analysis,
this may not fully account for variability in treatment
plans between participating centers or treating physicians. Similarly, variable transformation or feature engineering
(eg, calculating the BED, binarization of chemotherapy
drugs) could have led to the creation of a new feature that
is less powerful and suppresses important information
inferred by its raw components. In modeling the radiation
therapy dose variable, alternatives such as a categorical
variable divided by type of radiation therapy regimen
could have been used (eg, hypo- vs standard fraction-
ation). Variable aggregation could have led to model over-
fitting due to misleading combined features and may
show false significance or insignificance in the analysis.51
Although the resampling techniques used in this study
have advantages in their simplicity and transportability,
other remedies to address imbalanced data, such as
ensemble learning (which is implemented at the algorith-
mic level), could be used to improve model perfor-
mance.52 Cost-sensitive learning was selected to penalize
false negatives. However, its application depends on the
clinical situation. For example, if a model was designed to
allocate patients to a toxicity-lowering radiation therapy
regimen that might affect tumor control, then FPs may
need to have a higher cost than FNs. This study used the depends on the ML algorithm used, and any given feature
could climb up the ranking in terms of IG if additional
observations were added to the same data set. Hence, we
included all 122 features in the modeling process. The 10
most important features in the final hero model included
some that might be expected to predict breast radiation
toxicity, such as use of radiation therapy bolus, chemother-
apy, boost, radiation therapy dose, and bra size. Discussion Interest-
ingly, the most important feature (use of lipid-lowering
drugs) is not usually included in parametric statistical
models for radiation toxicity, although HMG-CoA reduc-
tase inhibitors (statins) have previously been proposed as
radioprotective agents.50 Yet unlike traditional statistical
probability modeling, feature importance should only be
interpreted within the context of the ML prediction model
but not outside. Although the resampling techniques used in this study
have advantages in their simplicity and transportability,
other remedies to address imbalanced data, such as
ensemble learning (which is implemented at the algorith-
mic level), could be used to improve model perfor-
mance.52 Cost-sensitive learning was selected to penalize
false negatives. However, its application depends on the
clinical situation. For example, if a model was designed to
allocate patients to a toxicity-lowering radiation therapy
regimen that might affect tumor control, then FPs may
need to have a higher cost than FNs. This study used the Acknowledgments 4. National Cancer Institute. Priorities for Research on Cancer Survi-
vorship. London: Department of Health; 2010. 4. National Cancer Institute. Priorities for Research on Cancer Survi-
vorship. London: Department of Health; 2010. We sincerely thank all patients who participated in the
REQUITE study and all REQUITE staff involved at the
following hospitals: 5. Emami B, Lyman J, Brown A, et al. Tolerance of normal tissue to
therapeutic irradiation. Int J Radiat Oncol Biol Phys. 1991;21:
109–122. Belgium: Prof. Wilfried de Neve, Dr. Christel Monten,
Dr. Pieter Deseyne, Giselle Post and Renee Bultijnck at
Ghent University Hospital; Gilles Defraene, Rita Aerts,
Soumia Arredouani, Maarten Lambrecht at KU Leuven. 6. Bentzen SM, Overgaard J. Patient-to-patient variability in the
expression of radiation-induced normal tissue injury. Sem Radiat
Oncol. 1994;4:68–80. 7. Knobf MT, Sun Y. A longitudinal study of symptoms and self-care
activities in women treated with primary radiotherapy for breast
cancer. Cancer Nurs. 2005;28:210–218. France: ICM Montpellier, CHU Nîmes (Department of
Radiation Oncology, CHU Nîmes, Nîmes, France). France: ICM Montpellier, CHU Nîmes (Department of
Radiation Oncology, CHU Nîmes, Nîmes, France). 8. Rochlin DH, Jeong AR, Goldberg L, et al. Postmastectomy radiation
therapy and immediate autologous breast reconstruction: Integrat-
ing perspectives from surgical oncology, radiation oncology, and
plastic and reconstructive surgery. J Surg Oncol. 2015;111:251–257. Germany: Praxis f€ur Strahlentherapie an der Stadtkli-
nik Baden-Baden (Dr. Thomas Blaschke); Klinikum
Darmstadt GmbH (PD Dr. Christian Weiß); Zentrum f€ur
Strahlentherapie Freiburg (Dr. Petra Stegmaier); ViDia
Christliche Kliniken Karlsruhe (Prof. Dr. Johannes Cla-
ßen); Klinikum der Stadt Ludwigshafen gGmbH (PD Dr. Thomas
Schnabel);
Universit€atsklinikum
Mannheim;
Strahlentherapie Speyer (Dr. J€org Sch€afer). The research-
ers at DKFZ also thank Anusha M€uller, Irmgard Helm-
bold and Dr Sabine Behrens. Germany: Praxis f€ur Strahlentherapie an der Stadtkli-
nik Baden-Baden (Dr. Thomas Blaschke); Klinikum
Darmstadt GmbH (PD Dr. Christian Weiß); Zentrum f€ur
Strahlentherapie Freiburg (Dr. Petra Stegmaier); ViDia
Christliche Kliniken Karlsruhe (Prof. Dr. Johannes Cla-
ßen); Klinikum der Stadt Ludwigshafen gGmbH (PD Dr. Thomas
Schnabel);
Universit€atsklinikum
Mannheim;
Strahlentherapie Speyer (Dr. J€org Sch€afer). The research-
ers at DKFZ also thank Anusha M€uller, Irmgard Helm-
bold and Dr Sabine Behrens. 9. Twardella D, Popanda O, Helmbold I, et al. Personal characteristics,
therapy modalities and individual DNA repair capacity as predictive
factors of acute skin toxicity in an unselected cohort of breast cancer
patients receiving radiotherapy. Radiother Oncol. 2003;69:145–153. 10. Back M. References 1. Early and Locally Advanced Breast Cancer: Diagnosis and Manage-
ment. London: National Institute for Health and Clinical Excellence;
2018. 2. Darby S, McGale P, Correa C, et al. Effect of radiotherapy after
breast-conserving surgery on 10-year recurrence and 15-year breast
cancer death: Meta-analysis of individual patient data for 10 801
women in 17 randomised trials. The Lancet. 2011;378:1707–1716. 3. Office for National Statistics. Cancer Survival in England: Patients
Diagnosed Between 2010 and 2014 and Followed Up to 2015. Lon-
don: Office for National Statistics; 2016. Study limitations Despite the rigorous error detection in the data prepro-
cessing phase, we cannot exclude errors occurring due to
manual recording during data collection. According to
the REQUITE study protocol, patients were assessed at
the start and end of treatment and annually thereafter. This may have missed cases of acute desquamation as
acute radiation toxicity is known to peak up to 2 weeks Advances in Radiation Oncology: May−June 2022 M. Aldraimli et al 12 Monika Kaushik, Kelly V Lambert, Simon M Pilgrim,
Sheila Shokuhi, Kalliope Valassiadou, Luis Aznar-Garcia,
Kerstie Johnson, Kiran Kancherla, and Kufre Sampson);
support from the HOPE Clinical Trials Unit and Leicester
Experimental Cancer Medicine Centre (ECMC), Theresa
Beaver, Kaitlin Duckworth, and Sara Barrows. impurity-based ranking mean decrease impurity filter to
simplify the final model with a known performance, but it
is important to keep in mind that feature selection based
on impurity reduction is generally biased toward prefer-
ring variables with more categories.53 Monika Kaushik, Kelly V Lambert, Simon M Pilgrim,
Sheila Shokuhi, Kalliope Valassiadou, Luis Aznar-Garcia,
Kerstie Johnson, Kiran Kancherla, and Kufre Sampson);
support from the HOPE Clinical Trials Unit and Leicester
Experimental Cancer Medicine Centre (ECMC), Theresa
Beaver, Kaitlin Duckworth, and Sara Barrows. USA: Mount Sinai Hospital, New York. Conclusion Application of ML algorithms with resampling and
cost-sensitive learning resulted in valid prediction models
for acute desquamation after whole breast EBRT using
clinical and treatment features. After optimisation, the
best model was able to classify patients with acceptable
performance in the validation cohort (AUC = 0.77). Before they can be used in clinical practice, further opti-
mization of ML prediction models, including genomic
markers, is required, and the models should be validated
in external cohorts. This approach could help identify
breast cancer patients at increased risk of toxicity to
inform discussions about risks and benefits and allow
treatment plans to be personalized with the aim of mini-
mizing toxicity or offering the patient increased support-
ive management during treatment. Supplementary materials Supplementary material associated with this article
can be found in the online version at doi:10.1016/j. adro.2021.100890. Advances in Radiation Oncology: May−June 2022 Bielza C, Larra~naga P. Discrete Bayesian network classifiers: A sur-
vey. ACM Comput Surv. 2014;47:1–43. 40. Ruck DW, Rogers SK, Kabrisky M, Oxley ME, Suter BW. The multi-
layer perceptron as an approximation to a Bayes optimal discrimi-
nant function. IEEE Trans Neural Net. 1990;1:296–298. 21. Lee S, Kerns S, Ostrer H, Rosenstein B, Deasy JO, Oh JH. Machine
learning on a genome-wide association study to predict late genito-
urinary toxicity after prostate radiation therapy. Int J Radiat Oncol
Biol Phys. 2018;101:128–135. 41. Cunningham P, Delany SJ. K-nearest neighbour classifiers. arXiv
preprint arXiv:200404523. 2020. 42. Landwehr N, Hall M, Frank E. Logistic model trees. Mach Learn. 2005;59:161–205. 22. Saednia K, Tabbarah S, Lagree A, et al. Quantitative thermal imag-
ing biomarkers to detect acute skin toxicity from breast radiation
therapy using supervised machine learning. Int J Radiat Oncol Biol
Phys. 2020;106:1071–1083. 43. Ho TK. Random decision forests. In: Proc Third Int Conf Doc Analy-
sis and Recog. 1995278–282. 23. Reddy J, Lindsay WD, Berlind CG, Ahern CA, Smith BD. Applying a
machine learning approach to predict acute toxicities during radiation
for breast cancer patients. Int J Radiat Oncol Biol Phys. 2018;102:S59. 44. Brownlee J. Imbalanced Classification with Python: Better Metrics,
Balance Skewed Classes, Cost-Sensitive Learning. Machine Learning
Mastery; 2020. 45. Elkan C. The foundations of cost-sensitive learning. Int Joint Conf
Artif Intell. 2001973–978. 24. Rattay T, Seibold P, Aguado-Barrera ME, et al. External validation of
a predictive model for acute skin radiation toxicity in the REQUITE
breast cohort. Front Oncol. 2020;10:575909. 46. Louppe G, Wehenkel L, Sutera A, Geurts P. Understanding variable
importances in forests of randomized trees. Adv Neur Inform Process
Syst. 2013;26:431–439. 25. Collins GS, Reitsma JB, Altman DG, Moons KG. Transparent
reporting of a multivariable prediction model for individual progno-
sis or diagnosis (TRIPOD): The TRIPOD statement. Br J Surg. 2015;102:148–158. 47. Garner SR. Weka: The waikato environment for knowledge
analysis. In: Proc New Zeal Comput Sci Res Student Conf. 1995
57–64. 26. West C, Azria D, Chang-Claude J, et al. The REQUITE project: vali-
dating predictive models and biomarkers of radiotherapy toxicity to
reduce side-effects and improve quality of life in cancer survivors. Clin Oncol (R Coll Radiol). 2014;26:739–742. 48. Drazin S, Montag M. Decision tree analysis using weka. Machine
Learning-Project II. 2012:1–3. 49. Keerthi SS, Shevade SK, Bhattacharyya C, Murthy KRK. Improve-
ments to Platt's SMO algorithm for SVM classifier design. Neur
Comput. 2001;13:637–649. 27. Advances in Radiation Oncology: May−June 2022 Advances in Radiation Oncology: May−June 2022 ML Prediction Model for RT Acute Desquamation 13 31. Rahman G, Islam Z. A decision tree-based missing value imputation
technique for data pre-processing. Proc Ninth Australasian Data
Min Conf. 2011;121:41–50. cancer patients receiving radiotherapy after breast conserving sur-
gery. Radiother Oncol. 2012;103:199–205. cancer patients receiving radiotherapy after breast conserving sur-
gery. Radiother Oncol. 2012;103:199–205. cancer patients receiving radiotherapy after breast conserving sur-
gery. Radiother Oncol. 2012;103:199–205. 15. Sharp L, Johansson H, Hatschek T, Bergenmar M. Smoking as an
independent risk factor for severe skin reactions due to adjuvant
radiotherapy for breast cancer. The Breast. 2013;22:634–638. 32. Quinlan JR. Improved use of continuous attributes in C4. 5. J Artif
Intell Res. 1996;4:77–90. 33. Raschka S. About feature scaling and normalization and the effect of
standardization for machine learning algorithms. Political Leg
Anthropology Rev. 2014;30:67–89. 16. Tortorelli GDML, Barbarino R, Cicchetti S, et al. Standard or hypo-
fractionated radiotherapy in the postoperative treatment of breast
cancer: A retrospective analysis of acute skin toxicity and dose inho-
mogeneities. BMC Cancer. 2013;13:230. 34. Chawla NV. Data mining for imbalanced datasets: An overview. Data Mining And Knowledge Discovery Handbook. Boston, MA:
Springer; 2009:875–886. 17. Ciammella P, Podgornii A, Galeandro M, et al. Toxicity and cos-
metic outcome of hypofractionated whole-breast radiotherapy: Pre-
dictive clinical and dosimetric factors. Radiat Oncol. 2014;9:97. 35. Branco P, Torgo L, Ribeiro RP. A survey of predictive modeling on
imbalanced domains. ACM Comput Surv. 2016;49:1–50. 18. De Langhe S, Mulliez T, Veldeman L, et al. Factors modifying the
risk for developing acute skin toxicity after whole-breast intensity
modulated radiotherapy. BMC Cancer. 2014;14:711. 36. Chawla NV, Bowyer KW, Hall LO, Kegelmeyer WP. SMOTE: Syn-
thetic
minority
over-sampling
technique. J
Artif
Intell
Res. 2002;16:321–357. 19. Mbah C, Thierens H, Thas O, et al. Pitfalls in prediction modeling for
normal tissue toxicity in radiation therapy: An illustration with the
individual radiation sensitivity and mammary carcinoma risk factor
investigation cohorts. Int J Radiat Oncol Biol Phys. 2016;95:1466–1476. 37. Faith J, Brockway M. Targeted projection pursuit tool for gene
expression visualisation. J Integrat Bioinform. 2006;3:264–273. 38. Fushiki T. Estimation of prediction error by using K-fold cross-vali-
dation. Stat Comput. 2011;21:137–146. 20. Dean J, Wong K, Gay H, et al. Incorporating spatial dose metrics in
machine learning-based normal tissue complication probability
(NTCP) models of severe acute dysphagia resulting from head and
neck radiotherapy. Clin Transl Radiat Oncol. 2018;8:27–39. 39. Acknowledgments Impact of radiation therapy on acute toxicity in breast con-
servation therapy for early breast cancer. Clin Oncol. 2004;16:
12–16. 11. Deantonio L, Gambaro G, Beldi D, et al. Hypofractionated radio-
therapy after conservative surgery for breast cancer: Analysis of
acute and late toxicity. Radiat Oncol. 2010;5:112. Italy:
Fondazione
IRCCS
Istituto
Nazionale
dei
Tumori, Milano. Italy:
Fondazione
IRCCS
Istituto
Nazionale
dei
Tumori, Milano. 12. Barnett GC, Wilkinson JS, Moody AM, et al. The Cambridge Breast
Intensity-Modulated Radiotherapy trial: Patient- and treatment-related
factors that influence late toxicity. Clin Oncol. 2011;23:662–673. Spain: Santiago: Complexo Hospitalario Universitario
de Santiago. Barcelona: The authors acknowledge the Cel-
lex Foundation for providing research facilities and equip-
ment and thank the CERCA Programme/ Generalitat de
Catalunya for institutional support. Spain: Santiago: Complexo Hospitalario Universitario
de Santiago. Barcelona: The authors acknowledge the Cel-
lex Foundation for providing research facilities and equip-
ment and thank the CERCA Programme/ Generalitat de
Catalunya for institutional support. 13. Kraus-Tiefenbacher U, Sfintizky A, Welzel G, et al. Factors of influ-
ence on acute skin toxicity of breast cancer patients treated with
standard three-dimensional conformal radiotherapy (3D-CRT) after
breast conserving surgery (BCS). Radiat Oncol. 2012;7:217. UK:
Leicester:
University
Hospitals
Leicester
(Drs. Donna Appleton, Frances Kenny, Jaroslaw Krupa, UK:
Leicester:
University
Hospitals
Leicester
(Drs. Donna Appleton, Frances Kenny, Jaroslaw Krupa, 14. Terrazzino S, Mattina PL, Masini L, et al. Common variants of
eNOS and XRCC1 genes may predict acute skin toxicity in breast Advances in Radiation Oncology: May−June 2022 Seibold P, Webb A, Aguado-Barrera ME, Azria D, Bourgier C,
Brengues M, et al. REQUITE. A prospective multicentre cohort
study of patients undergoing radiotherapy for breast, lung or pros-
tate cancer. Radiotherapy and oncology : journal of the European
Society for Therapeutic Radiology and Oncology. 2019;138:59–67. 50. Fritz G, Henninger C, Huelsenbeck J. Potential use of HMG-CoA
reductase inhibitors (statins) as radioprotective agents. Br Med Bull. 2011;97:17–26. 51. Pyle D. Data Preparation for Data Mining. Burlington, MA: Morgan
Kaufmann; 1999. 28. National Cancer Institute. Common Terminology Criteria for
Adverse Events V4.0. National Cancer Institute: NIH; 2009. 52. More A. Survey of resampling techniques for improving classification
performance in unbalanced datasets. arXiv preprint arXiv:160806048. 2016. 29. Arnicane V. Complexity of equivalence class and boundary value
testing methods. Int J Comput Sci Inform Techn. 2009;751:80–101. 30. Liu WZ, White AP, Thompson SG, Bramer MA. Techniques for
dealing with missing values in classification. In: Liu X, Cohen P,
Berthold M, eds. International Symposium on Intelligent Data Anal-
ysis. Springer; 1997:527–536. 53. Strobl C, Boulesteix AL, Zeileis A, Hothorn T. Bias in random forest
variable importance measures: Illustrations, sources and a solution. BMC Bioinformat. 2007;8:25.
|
https://openalex.org/W3120463100
|
https://www.frontiersin.org/articles/10.3389/fpls.2020.638134/pdf
|
English
| null |
Corrigendum: Editorial: Wild Plants as Source of New Crops
|
Frontiers in plant science
| 2,021
|
cc-by
| 526
|
Eric von Wettberg 1,2, Thomas M. Davis 3 and Petr Smýkal 4* 1 Plant and Soil Science and Gund Institute for the Environment, University of Vermont, Burlington, VT, United States,
2 Institute for Applied Mathematics and Mechanics, Peter the Great St. Petersburg Polytechnic University, St. Petersburg,
Russia, 3 Department of Agriculture, Nutrition, and Food Systems, University of New Hampshire, Durham, NH, United State
4 Department of Botany, Faculty of Science, Palacky University, Olomouc, Czechia Keywords: crop wild relatives, domestication, genetic diversity, neo-domestication, breeding, ethnobotany Copyright © 2021 von Wettberg, Davis and Smýkal. This is an open-access article distributed under the terms of the Creative
Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original
author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with
accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. CORRECTION
published: 14 January 2021
doi: 10.3389/fpls.2020.638134 *Correspondence:
Petr Smýkal *Correspondence:
Petr Smýkal Additionally, in the original article, we neglected to include the funder USDA National Institute
of Food and Agriculture Hatch Project, NH00678 to Thomas M. Davis. A correction has therefore
been made to the Funding statement. petr.smykal@upol.cz Specialty section:
This article was submitted to
Plant Breeding,
a section of the journal
Frontiers in Plant Science Specialty section:
This article was submitted to
Plant Breeding,
a section of the journal
Frontiers in Plant Science
Received: 05 December 2020
Accepted: 21 December 2020
Published: 14 January 2021
Citation:
von Wettberg E, Davis TM and
Smýkal P (2021) Corrigendum:
Editorial: Wild Plants as Source of
New Crops. Front. Plant Sci. 11:638134. doi: 10.3389/fpls.2020.638134 “PS is supported by the Grant Agency of the Czech Republic and Palacký University grant
Agency [IGA-2020_003]. EW is supported by the USDA Hatch program through the Vermont State
Agricultural Experimental Station and is supported to work on diversity of mungbeans by Russian
Scientific Fund Project No. 18-46-08001 on the basis of a unique scientific installation “Collection
of plant genetic resources VIR”. TD is supported in part by the USDA National Institute of Food
and Agriculture Hatch Project NH00678 (accession number 1019990).” Received: 05 December 2020
Accepted: 21 December 2020 Received: 05 December 2020 Accepted: 21 December 2020 Published: 14 January 2021 The authors apologize for these errors and state that this does not change the scientific
conclusions of the article in any way. The original article has been updated. Frontiers in Plant Science | www.frontiersin.org A Corrigendum on Edited and reviewed by:
Inaki Hormaza,
Institute of Subtropical and
Mediterranean Horticulture La
Mayora, Spain Editorial: Wild Plants as Source of New Crops
by von Wettberg, E., Davis, T. M., and Smýkal, P. (2020). Front. Plant Sci. 11:591554. doi: 10.3389/fpls.2020.591554 Editorial: Wild Plants as Source of New Crops
by von Wettberg, E., Davis, T. M., and Smýkal, P. (2020). Front. Plant Sci. 11:591554. doi: 10.3389/fpls.2020.591554 In the published article, there was an error regarding the affiliation(s) for Eric von Wettberg. As
well as having affiliation(s) 1, they should also have 2. Citation: January 2021 | Volume 11 | Article 638134 Frontiers in Plant Science | www.frontiersin.org
|
https://openalex.org/W2126064873
|
http://www.iapress.org/index.php/soic/article/download/20140606/208
|
Latin
| null |
Recurrence Relations for Moments of Order Statistics from the Lindley Distribution with General Multiply Type-II Censored Sample
|
Statistics, optimization & information computing
| 2,014
|
cc-by
| 5,197
|
STATISTICS,OPTIMIZATION AND INFORMATION COMPUTING
Stat., Optim. Inf. Comput., Vol. 2, June 2014, pp 147–160.
Published online in International Academic Press (www.IAPress.org) STATISTICS,OPTIMIZATION AND INFORMATION COMPUTING
Stat., Optim. Inf. Comput., Vol. 2, June 2014, pp 147–160. Published online in International Academic Press (www.IAPress.org) STATISTICS,OPTIMIZATION AND INFORMATION COMPUTING
Stat., Optim. Inf. Comput., Vol. 2, June 2014, pp 147–160. Published online in International Academic Press (www.IAPress.org) Recurrence relations for moments of multiply type-II
censored order statistics from Lindley distribution
with applications to inference Bander Al-Zahrani ∗and M. A. Ali Department of Statistics, King Abdulaziz University, Saudi Arabia
Received 24 December 2013; Accepted 28 April 2014 Department of Statistics, King Abdulaziz University, Saudi Arabia
Received 24 December 2013; Accepted 28 April 2014
Editor: Anwar Joarder Department of Statistics, King Abdulaziz University, Saudi Arabia Received 24 December 2013; Accepted 28 April 2014
Editor: Anwar Joarder Received 24 December 2013; Accepted 28 April 2014
Editor: Anwar Joarder Abstract
In this paper, we derive the recurrence relations for the moments of function
of single and two order statistics from Lindley distribution. We also consider the
maximum likelihood estimation (MLE) of the parameter of the distribution based on
multiply type-II censoring. The maximum likelihood estimator is comupted numerically
because it does not have an explicit form for the parameter. Then, a Monte Carlo
simulation study is carried out to evaluate the performance of the MLE obtained from
multiply type-II censored sample. Keywords Order statistics, recurrence relations, Lindley distribution,
multiply type-II censoring, Monte Carlo simulation. DOI: 10.19139/soic.v2i2.55 DOI: 10.19139/soic.v2i2.55 ISSN 2310-5070 (online) ISSN 2311-004X (print)
Copyright c⃝2014 International Academic Press BANDER AL-ZAHRANI AND M. A. ALI BANDER AL-ZAHRANI AND M. A. ALI The Lindley distribution gives a better model where the exponential distribution
is not good fit. Since the equation F(x) = u, where u ∼U(0, 1), cannot be solved
explicitly in terms of x, the inversion method for generation random data from the
Lindley distribution fails. However, one can use the fact that the distribution is a
special mixture of Exponential(θ) and Gamma(2, θ) distributions as f(x) = pf1(x) + (1 −p)f2(x); x > 0, θ > 0, where p = θ/(1 + θ), f1(x) = θe−θx and f2(x) = θ2xe−θx. where p = θ/(1 + θ), f1(x) = θe−θx and f2(x) = θ2xe−θx. To generate random data, Xi, i = 1, 2, · · · , n from the Lindley distribution with
parameter θ one may follow the acceptance-rejection method which can be given
by the following algorithm: p
/(
), f1( )
f2( )
To generate random data, Xi, i = 1, 2, · · · , n from the Lindley distribution with
parameter θ one may follow the acceptance-rejection method which can be given
by the following algorithm: i. Generate Ui ∼U(0, 1), i = 1, 2, · · · , n
ii. Generate Ei ∼Exp(θ), i = 1, 2, · · · , n
iii. Generate Gi ∼Gamma(2, θ), i = 1, 2, · · · , n i. Generate Ui ∼U(0, 1), i = 1, 2, · · · , n ii. Generate Ei ∼Exp(θ), i = 1, 2, · · · , n iii. Generate Gi ∼Gamma(2, θ), i = 1, 2, · · · , n iv. If Ui ≤p, then set Xi = Gi, otherwise, set Xi = Ei, i = 1, 2, · · · , n, where
p is as before. Ghitany, et al. (2008) developed different properties of Lindley distribution. The
main aim of this paper is to develop recurrence relations of moments of order
statistics for the function of single and two order statistics. Also develop a
maximum likelihood estimation procedure of the parameter θ by Monte Carlo
simulation from multiply type-II censored sample. Then a comparison study will
be made between maximum likelihood estimates (Ghitany, et al., 2008) and MLE
from Monte Carlo study from multiply type-II censored sample. 1. Introduction A random variable X is said to have Lindley distribution if its probability density
function (pdf) is given by f(x) =
θ2
1 + θ(1 + x) e−θx;
x > 0, θ > 0,
(1) (1) and it was introduced by Lindley (1952). The corresponding cumulative
distribution function (cdf) is given by F(x) = 1 −1 + θ + θx
1 + θ
e−θx;
x > 0, θ > 0. (2) (2) ∗Correspondence to: Department of Statistics, King Abdulaziz University, Jeddah 21589, Saudi
Arabia. E-mail: bmalzahrani@kau.edu.sa ISSN 2310-5070 (online) ISSN 2311-004X (print)
Copyright c⃝2014 International Academic Press ISSN 2310-5070 (online) ISSN 2311-004X (print)
Copyright c⃝2014 International Academic Press 148 Stat., Optim. Inf. Comput.
Vol. 2, June 2014. Stat., Optim. Inf. Comput.
Vol. 2, June 2014. BANDER AL-ZAHRANI AND M. A. ALI Let X1, X2, · · · , Xn be a random sample of size n from the pdf (1)
corresponding to the cdf (2). Then X1:n ≤X2:n ≤· · · ≤Xn:n denote ordered
statistics for the above sample. Assume that n items are put on a life test, but only
r1-th, · · · , rk-th failures are observed, the rest are unobserved, where r1, · · · , rk
are considered to be fixed. This is the multiply Type-II censoring, for more details
see e.g. Jang et al. (2001) and Schenk et al. (2011). Multiply Type-II censoring
is a generalization of Type-II censoring where only the first k failure times are
observed. In this paper, we let 0 ≤Xr1:n ≤Xr2:n ≤· · · ≤Xrk:n < ∞to be a
multiply Type-II censored sample from a population with pdf (1) and cdf (2) for
θ ∈Rq, where, 1 ≤r1 < r2 < · · · < rn ≤n. The motivation behind using multiply type-II censoring is made clear in the
particularl case if one fails to record the failure time of every subject, only several
failure times and the number of failures between them are recorded, see Kong and
Fei (1996). RECURRENCE RELATIONS FOR MOMENTS OF MULTIPLY TYPE-II CENSORED ORDER STATISTICS 149 149 2. Recurrence relation for Moments from function of single order statistic The pdf of r-th order statistic Xr:n, (1 ≤r ≤n) is given by The pdf of r-th order statistic Xr:n, (1 ≤r ≤n) is given by der statistic Xr:n, (1 ≤r ≤n) is given by fr:n(x) =
n! (r −1)! (n −r)![F(x)]r−1[1 −F(x)]n−rf(x),
x ∈(0, ∞)
(3) (3) Let g(x) be a Borel measurable function of x in the interval x ∈(0, ∞), then E [g (Xr:n)]
=
Cr:n
∫∞
0
g(x) [F(x)]r−1[1 −F(x)]n−rf(x) dx. =
pθCr:n
∫∞
0
g(x)
[
1 −(1 + px)e−θx]r−1 [1 + px]n−r
× (1 + x)e−(n−r+1)θx dx,
( (4) where Cr:n = n!/((r −1)! (n −r)!). Theorem 1
For 1 ≤r ≤n; n = 1, 2, · · · where Cr:n = n!/((r −1)! (n −r)!). where Cr:n = n!/((r −1)! (n −r)!). Th
1 Theorem 1
For 1 ≤r ≤n; n = 1, 2, · · · Theorem 1
For 1 ≤r ≤n; n = 1, 2, · · · E [g (Xr:n)] −E [g (Xr−1:n−1)]
=
(n −1
r −1
) r−1
∑
i=0
(−1)i−1−i(r −1
i
)
×
∫∞
0
g
′(x)(1 + px)n−ie−(n−i)θx dx,
(5) E [g (Xr:n)] −E [g (Xr−1:n−1)]
=
(n −1
r −1
) r−1
∑
i=0
(−1)i−1−i(r −1
i
) ×
∫∞
0
g
′(x)(1 + px)n−ie−(n−i)θx dx,
(5) (5) where p = θ/(1 + θ) and q = 1/(1 + θ). Proof. From (4), we have where p = θ/(1 + θ) and q = 1/(1 + θ). Proof. From (4), we have E [g (Xr:n)] −E [g (Xr−1:n−1)]
=
(n −1
r −1
)
θp
∫∞
0
g(x)[
1 −(1 + px)e−θx]r−1
× [1 + px]n−r
[ n −r + 1 −n (1 + px) e−θx
1 −(1 + px)e−θx
]
× (1 + x)e−(n−r+1)θxdx. Let q(x) = −
[
1 −(1 + px)e−θx]r−1 [1 + px]n−r+1e−(n−r+1)θx, then we have Let q(x) = −
[
1 −(1 + px)e−θx]r−1 [1 + px]n−r+1e−(n−r+1)θx, then we have
E [g (Xr:n)] −E [g (Xr−1:n−1)] =
(n −1
r −1
) ∫∞
0
g(x) q′(x)dx, E [g (Xr:n)] −E [g (Xr−1:n−1)] =
(n −1
r −1
) ∫∞
0
g(x) q′(x)dx, which on integration by parts gives which on integration by parts gives which on integration by parts gives E [g (Xr:n)] −E [g (Xr−1:n−1)]
=
(
n −1
r −1
) ∫∞
0
g′(x)
[
1 −(1 + px)e−θx]r−1
× [1 + px]n−r+1(1 + x)e−(n−r+2)θxdx. Proof. Putting g(x) = xk in Theorem 1, to get Proof. Putting g(x) = xk in Theorem 1, to get µ(k)
r:n −µ(k)
r−1:n−1
=
k
(n −1
r −1
) r−1
∑
i=0
(−1)r−1−i
(r −1
i
)
×
∫∞
0
xk−1(1 + px)n−ie−(n−i)θx dx. =
k
(n −1
r −1
) r−1
∑
r=0
n−i
∑
j=0
(−1)r−1−i
(r −1
i
)(n −j
j
)
pn−i−j
×
∫∞
0
xn+k−i−j−1e−(n−i)θx dx, =
k
(n −1
r −1
) r−1
∑
r=0
n−i
∑
j=0
(−1)r−1−i
(r −1
i
)(n −j
j
)
pn−i−j
×
∫∞
0
xn+k−i−j−1e−(n−i)θx dx, using the gamma function to the integrand to get the require result. using the gamma function to the integrand to get the require result. using the gamma function to the integrand to get the require result. Corollary 2
f ( )
k f
h using the gamma function to the integrand to get the require result. Corollary 2
( )
k y
If g(x) = xk, for 1 ≤r ≤n; n = 1, 2, · · · , then y
If g(x) = xk, for 1 ≤r ≤n; n = 1, 2, · · · , then µ(k)
r:n −µ(k)
r−1:n = C (constant), where C = k
(
n −1
r −2
) r−1
∑
i=0
n−i
∑
j=0
(−1)r−1−i
(
r −1
i
) (
n −i
j
)
pn−i−j Γ(n + k −i −j)
[(n −i) θ]n+k−i−j . Proof. Putting g(x) = xk in Theorem 2 and the rest is similar to Corollary 1. Proof. Putting g(x) = Corollary 3 Corollary 3
If g(x) = xk, for 1 ≤r ≤n; n = 1, 2, · · · , then
µ(k)
r−1:n−1 −µ(k)
r−1:n = C (constant), y
If g(x) = xk, for 1 ≤r ≤n; n = 1, 2, · · · , then y
If g(x) = xk, for 1 ≤r ≤n; n = 1, 2, · · · , then
µ(k)
r−1:n−1 −µ(k)
r−1:n = C ( µ(k)
r−1:n−1 −µ(k)
r−1:n = C (constant), where C = k
(
n −1
r −2
) r−1
∑
i=0
n−i
∑
j=0
(−1)r−1−i
(
r −1
i
) (
n −i
j
)
pn−i−j Γ(n + k −i −j)
[(n −i) θ]n+k−i−j . Proof. Putting g(x) = xk in Theorem 3 and the rest is similar to Corollary 1. Proof. Putting g(x) = xk in Theorem 3 and the rest is similar to Corollary 1. 2. Recurrence relation for Moments from function of single order statistic (6) Stat., Optim. Inf. Comput. Vol. 2, June 2014. Stat., Optim. Inf. Comput. Vol. 2, June 2014. BANDER AL-ZAHRANI AND M. A. ALI 150 Now expanding
[
1 −(1 + px)e−θx]r−1 binomially to get,
[
1 −(1 + px)e−θx]r−1 =
r−1
∑
i=0
(−1)r−1−i
(
r −1
i
)
(1 + px)r−1−ie−(r−1−i)θx. Now expanding
[
1 −(1 + px)e−θx]r−1 binomially to get, Now expanding
[
1 −(1 + px)e−θx]r−1 b [
1 −(1 + px)e−θx]r−1 =
r−1
∑
i=0
(−1)r−1−i
(
r −1
i
)
(1 + px)r−1−ie−(r−1−i)θx. Putting this result in (6) and on algebraic simplification gives the required proof
of the result in (5). Theorem 2 heorem 2
or 1 ≤r ≤n; n = 1, 2, · · · For 1 ≤r ≤n; n = 1, 2, · · · E [g (Xr:n)] −E [g (Xr−1:n)]
=
(
n
r −1
) r−1
∑
i=0
(−1)r−1−i
(r −1
i
)
×
∫∞
0
g′(x)(1 + px)n−ie−(n−i)θxdx. Proof. The proof of this theorem is same as that of Theorem 1. Also, for further
details we refer to Ali and Khan (1998a). Theorem 3 For 1 ≤r ≤n; n = 1, 2, · · · For 1 ≤r ≤n; n = 1, 2, · · · [E [g (Xr−1:n−1)] −E [g (Xr−1:n)]
=
(n −1
r −2
) r−1
∑
i=0
(−1)r−1−i
(r −1
i
)
×
∫∞
0
g′(x)(1 + px)n−ie−(n−i)θx dx. Proof. The proof is same as that of Theorem 2 Also, we refer to Ali and Khan
(1998a). It is important to note that the above theorems lead to establish the well-known
relation given in David and Nagaraja (2003), pp. 45. Corollary 1
If g(x) = xk, for 1 ≤r ≤n; n = 1, 2, · · · , then Corollary 1 y
If g(x) = xk, for 1 ≤r ≤n; n = 1, 2, · · · , then If g(x) = xk, for 1 ≤r ≤n; n = 1, 2, · · · , then µ(k)
r:n −µ(k)
r−1:n−1 = C (constant), where where C = k
(
n −1
r −1
) r−1
∑
i=0
n−i
∑
j=0
(−1)r−1−i
(
r −1
i
) (
n −i
j
)
pn−i−j Γ(n + k −i −j)
[(n −i) θ]n+k−i−j . Stat., Optim. Inf. Comput. Vol. 2, June 2014. Stat., Optim. Inf. Comput. Vol. 2, June 2014. 2. Recurrence relation for Moments from function of single order statistic RECURRENCE RELATIONS FOR MOMENTS OF MULTIPLY TYPE-II CENSORED ORDER STATISTICS 151 Proof. Putting g(x) = xk in Theorem 1, to get 3. Recurrence relation for Moments from the function of two order statistics The joint pdf of Xr:n = x and Xs:n = y, 1 ≤r ≤s ≤n, is given by
fr,s:n(x, y) = Cr,s:n[F(x)]r−1[F(y) −F(x)]s−r−1[1 −F(y)]n−sf(x)f(y), The joint pdf of Xr:n = x and Xs:n = y, 1 ≤r ≤s ≤n, is given by
fr,s:n(x, y) = Cr,s:n[F(x)]r−1[F(y) −F(x)]s−r−1[1 −F(y)]n−sf(x)f(y), Stat., Optim. Inf. Comput. Vol. 2, June 2014. Stat., Optim. Inf. Comput. Vol. 2, June 2014. Stat., Optim. Inf. Comput. Vol. 2, June 2014. 152 BANDER AL-ZAHRANI AND M. A. ALI where 0 ≤x < y < ∞, Cr,s:n = n!/((r −1)! (s −r −1)! (n −s)!). If g is a Borel measurable function from R2 to R, then where 0 ≤x < y < ∞, Cr,s:n = n!/((r −1)! (s −r −1)! (n −s)!). If g is a Borel measurable function from R2 to R, then where 0 ≤x < y E [g (Xr:n, Xs:n)]
=
Cr,s:n
∫∫
0≤x<y<∞
{
g(x, y)[F(x)]r−1[F(y) −F(x)]s−r−1
× [1 −F(y)]n−sf(x)f(y)
}
dxdy. E [g (Xr:n, Xs:n)]
=
Cr,s:n
∫∫
0≤x<y<∞
{
g(x, y)[F(x)]r−1[F(y) −F(x)]s−r−1 E [g (Xr:n, Xs:n)]
=
Cr,s:n
∫∫
0≤x<y<∞
{
g(x, y)[F(x)]r−1[F(y) −F(x)]s−r−1
× [1 −F(y)]n−sf(x)f(y)
}
dxdy. =
Cr,s:n θ2p2
∫∫
0≤x<y<∞
{ r−1
∑
i=0
(−1)r−1−i
(r −1
i
)
g(x, y)(1 + p
×
[
e−θx −e−θy + pxe−θx −pye−θy]s−r−1[1 + py]n−s × [1 −F(y)]n−sf(x)f(y)
}
dxdy. =
Cr,s:n θ2p2
∫∫
0≤x<y<∞
{ r−1
∑
i=0
(−1)r−1−i
(r −1
i
)
g(x, y)(1 + px)r−1−i =
Cr,s:n θ2p2
∫∫
0≤x<y<∞
{ r−1
∑
i=0
(−1)r−1−i
(r −1
i
)
g(x, y)(1 + px)r−1−i
×
[
e−θx −e−θy + pxe−θx −pye−θy]s−r−1[1 + py]n−s
× (1 + x) (1 + y) e−θ[(r−i)x+(n−s+1)y]
}
dxdy. (7) ×
[
e−θx −e−θy + pxe−θx −pye−θy]s−r−1[1 + py]n−s
× (1 + x) (1 + y) e−θ[(r−i)x+(n−s+1)y]
}
dxdy ×
[
e−θx −e−θy + pxe−θx −pye−θy]s−r−1[1 + py]n−s (7) Theorem 4 For 1 ≤r ≤n; n = 1, 2, · · · E [g (Xr:n, Xs:n)] −E [g (Xr:n, Xs−1:n)] E [g (Xr:n, Xs:n)] −E [g (Xr:n, Xs−1:n)]
=
Cr,s:n
(n −s + 1)
∫∫
0≤x<y<∞
{ r−1
∑
i=0
(−1)r−1−i
(r −1
i
)
g′(x, y)(1 + px)r−1−i
×
[
e−θx −e−θy + pxe−θx −pye−θy]s−r−1[1 + py]n−s+1 (1 + x) (1 + y)
} ×
e−θ[(r−i)x+(n−s+2)y]dxdy
}
,
(8) (8) where g′(x, y) =
d
dy g(x, y). where g′(x, y) =
d
dy g(x, y). where g′(x, y) =
d
dy g(x, y). 3. Recurrence relation for Moments from the function of two order statistics Set s = r + 1 and g(x, y) = xiyj in Theorem 4 and on algebraic
simplification in the same line of Corollary 1 to get the required proof. Proof. Set s = r + 1 and g(x, y) = xiyj in Theorem 4 and on algebraic
simplification in the same line of Corollary 1 to get the required proof. 3. Recurrence relation for Moments from the function of two order statistics where g′(x, y) =
d
dy g(x, y). Proof. From (7), we have Proof. From (7), we have E [g (Xr:n, Xs:n)] −E [g (Xr:n, Xs−1:n)] E [g (Xr:n, Xs:n)] −E [g (Xr:n, Xs−1:n)]
=
Cr,s:n
(n −s + 1)
∫∫
0≤x<y<∞
r−1
∑
i=0
(−1)r−1−i(r −1
i
)
g(x, y)(1 + px)r−1−i
[
e−θx −e−θy
+pxe−θx −pye−θys−r−2
]
[1 + py]n−s (1 + x) (1 + y) e−θ[(r−i)x+(n−s+1)y]
[(n −r) {
1 −(1 + py) e−θy}
−(n −s + 1) θp (1 + x) e−θx −(n −s + 1)]dx dy. (9) Let K(x, y) = −
[
e−θx −e−θy + pxe−θx −pye−θy]s−r−1
[1 + py]n−s+1e−(n−s+1)θy. K(x, y) = −
[
e−θx −e−θy + pxe−θx −pye−θy]s−r−1
[1 + py]n−s+1e−(n−s+1)θy. Stat., Optim. Inf. Comput. Vol. 2, June 2014. RECURRENCE RELATIONS FOR MOMENTS OF MULTIPLY TYPE-II CENSORED ORDER STATISTICS 153 θp Cr,s:n
(n −s + 1)
∫∫
0≤x<y<∞
r−1
∑
i=0
(−1)r−1−i
( r −1
i
)
g(x, y)(1 + px)r−1−ie−θ(r−1)xK′(x, y)dx dy where, K′(x, y) =
d
dyg(x, y) and hence the theorem (Ali and Khan, 1998). where, K′(x, y) =
d
dyg(x, y) and hence the theorem (Ali and Khan, 1998). where, K′(x, y) =
d
dyg(x, y) and hence the theorem (Ali and Khan, 1998). dy
The usual technique is establishing the recurrence relations will be to express
[1 −F(x)] or F(x) as a function of x and f(x) and then obtain the general form
relations. dy
The usual technique is establishing the recurrence relations will be to express
[1 −F(x)] or F(x) as a function of x and f(x) and then obtain the general form
relations. Corollary 4 Corollary 4
If g(x, y) = xiyj and 1 ≤r ≤n; n = 1, 2, 3, · · · Corollary 4
If g(x, y) = xiyj and 1 ≤r ≤n; n = 1, 2, 3, · · · µ(i,j)
r,r+1:n −µ(i+j)
r:n
=
r−1
∑
m=0
r−1−m
∑
t=0
n−r
∑
u=0
(−1)r−1−m
(r −1
m
)(r −1 −m
t
)(n −r
u
)
pn−m−t−u−1
×
∫∫
0≤x<y∞
xr−1−m−t+iyn−r−u+j−i (1 + x) e−(r−m)θxe−(n−r)θydxdy. µr,r+1:n
µr:n
=
r−1
∑
m=0
r−1−m
∑
t=0
n−r
∑
u=0
(−1)r−1−m
(r −1
m
)(r −1 −m
t
)(n −r
u
)
pn−m−t−u−1
×
∫∫
xr−1−m−t+iyn−r−u+j−i (1 + x) e−(r−m)θxe−(n−r)θydxdy Proof. 4. MLE for θ from multiply type-II censored sample The likelihood function of θ based on the multiply type-II censored sample as L (x; θ)
=
L (xr1:n ≤xr2:n ≤· · · ≤xrk:n; θ)
=
C[F (xr1:n; θ)]r1−1
k
∏
i=2
[
F (xri:n; θ) −F
(
xri−1:n; θ
)]ri−ri−1−1
× [1 −F (xrk:n; θ)]n−rk
k
∏
i=1
f (xri:n; θ)
(10) )
=
L (xr1:n ≤xr2:n ≤· · · ≤xrk:n; θ)
=
C[F (xr1:n; θ)]r1−1
k
∏
i=2
[
F (xri:n; θ) −F
(
xri−1:n; θ
)]ri−ri−1−1 × [1 −F (xrk:n; θ)]n−rk
k
∏
i=1
f (xri:n; θ)
(10) (10) Stat., Optim. Inf. Comput. Vol. 2, June 2014. 154 BANDER AL-ZAHRANI AND M. A. ALI Where, C is a constant free from θ. Taking log on both sides to get, log L (x; θ) log L (x; θ)
=
constant + (r1 −1) log [F (xr1:n; θ)]
+
k
∑
i=2
(ri −ri−1 −1) log
[
F (xri:n; θ) −F
(
xri−1:n; θ
)] + (n −rk) log [1 −F (xrk:n; θ) +
k
∑
i=1
log f (xri:n; θ)
(11) (11) Differentiating (11) with respect to θ and equate it to zero yields the likelihood
equation for θ h (θ)
=
δlog L
δθ
= (r1 −1) f (xr1:n; θ)
F (xr1:n; θ)
δ
δθf (xr1:n; θ)
+
k
∑
i=2
(ri −ri−1 −1)
[
f (xri:n; θ) −f
(
xri−1:n; θ
)]
[
F (xri:n; θ) −F
(
xri−1:n; θ
)]
[ δ
δθf (xri:n; θ) −δ
δθf
(
xri−1:n; θ
)]
+ (n −rk)
[−f (xrk:n; θ)
[1 −F (xrk:n; θ)
δ
δθf (xrk:n; θ) +
k
∑
i=1
δ
δθf (x
f (xr + (n −rk)
[−f (xrk:n; θ)
[1 −F (xrk:n; θ)
δ
δθf (xrk:n; θ) +
k
∑
i=1
δ
δθf (xrk:n; θ)
f (xri:n; θ) (12) =
0, =
0, where δ
δθf (x; θ) =
(2
θ −x −
1
θ + 1
)
f (x; θ) . The solution of (12) will be consistent, asymptotically normal and asymptotically
efficient under some conditions. For the multiply type-II censored data, let the gap between xri−1:n and xri:n is
(ri −ri−1 −1) and it is equal to the number of unobserved failures. Let maximum
gap g, where g = maxi(ri −ri−1 −1). Stat., Optim. Inf. Comput.
Vol. 2, June 2014. 4. MLE for θ from multiply type-II censored sample Condition 1: For all most all x, the derivatives ∂ilogf(x, θ)
∂θi
, i = 1, 2 and ∂i+1logf(x, θ)
∂x ∂θi
, i = 1, 2, 3 exists, and are piecewise continuous for all θ belongs to a non-degenerate
interval I and x ∈[0, ∞). Condition 2: There exist nonnegative numbers a1, a2, λij, i = 1, 2, j =
1, 2, · · · , 5, such that when θ is in some neighborhood of true value θ0, and Stat., Optim. Inf. Comput. Vol. 2, June 2014. Stat., Optim. Inf. Comput. Vol. 2, June 2014. RECURRENCE RELATIONS FOR MOMENTS OF MULTIPLY TYPE-II CENSORED ORDER STATISTICS 155 155 x is large enough, x is large enough, x is large enough, ∂ilogf(x, θ)
∂θi
≤a1xλ1i, i = 1, 2, ∂i+1logf(x, θ)
∂x ∂θi
≤a1xλ1(i+2) i = 1, 2, 3, and when x is small enough and close enough to zero, ∂ilogf(x, θ)
∂θi
≤a2x−λ2i, i = 1, 2,
∂i+1logf(x, θ)
∂x ∂θi
≤a2x−λ2(i+2), i = 1, 2, 3. ∂ilogf(x, θ)
∂θi
≤a2x−λ2i, i = 1, 2, ∂i+1logf(x, θ)
∂x ∂θi
≤a2x−λ2(i+2), i = 1, 2, 3. Also assume that there exists a function G(x) for each θ ∈R, Also assume that there exists a function G(x) for each θ ∈R, ∂3logf(x, θ)
∂θ3
≤G(x), for −∞< x < ∞ and there exists K independent of θ such that and there exists K independent of θ such that ∫∞
−∞
G(x)f(x, θ)dx ≤K < ∞. Condition 3: There exist positive numbers δ1 and δ2 for large enough x such
that f(x, θ) ≥δ1[1 −F(x, θ)]δ2 There exist positive numbers δ3 and δ4 for small enough x such that There exist positive numbers δ3 and δ4 for small enough x such that f(x, θ) ≥δ3[F(x, θ)]δ4. Condition 4: For each θ in I, the integral, Condition 4: For each θ in I, the integral, 0 <
∫∞
−∞
(∂log f (x, θ)
θ
)2
f (x, θ) dx < ∞. The conditions may be modified for multiple parameter case. Thus if the maximum
gap g is always bounded the likelihood equation (12) has a solution converging
in probability to the true value θ0 as n →∞, i.e., lim
n→∞θn
p→θ0. Stat., Optim. Inf. Comput.
Vol. 2, June 2014. 4. MLE for θ from multiply type-II censored sample If the maximum
i
l
b
d d th
l ti
f (12) i
t ti
ll
l
d gap g is always bounded the solution of (12) is an asymptotically normal and
asymptotically efficient estimate of the true value θ0. Under the conditions mentioned above and let limn→∞n e−(n/g)ε →0 for any
ε > 0. We assume ri+1−ri−1
ri−1
on the left tail or ri+1−ri−1
n−ri+1−1 on the right tail bounded
at two tails of order statistics then the MLE is consistent, see Kong and Fei (1996). Stat., Optim. Inf. Comput. Vol. 2, June 2014. 156 BANDER AL-ZAHRANI AND M. A. ALI 0.5
1.0
1.5
2.0
2.5
0
1
2
3
4
θ
MLE
MLE
Multiplied Censored MLE
Figure 1. Simulated MLE curves for n = 30 and m = 5. 0.5
1.0
1.5
2.0
2.5
0
1
2
3
4
MLE
MLE
Multiplied Censored MLE MLE Figure 1. Simulated MLE curves for n = 30 and m = 5. 5. Simulation Study Usual algebraic solution for the equation (12) is not working due to the
properties of transcendental equation. Therefore, fixed point iteration can be
used to solve the above equation. For an initial value θ(1), the (i + 1)th iterate
θ(i+1) can be obtained from the ith iterate θ(i+1) using θ(i+1) = h(θ(i)). The
iterative procedure can be stopped if |θ(i+1) −θ(i)| < ε, where ε is a pre-assigned
small positive number. The procedure of estimation is repeated 5000 times for
each value of θ with sample size, n = 30, multiply type-II censored sample
size, m = 5, 7, 9, 12, 15, 20 are presented in Table 5.2 and with n = 100, m =
5, 7, 9, 12, 15, 20 are presented in Table 5.3. A graphical comparison between
MLE obtained from the complete case and MLE obtained from multiply censored
sample are presented in Figure 1 and Figure 2 for sample size n = 30 and 100
respectively. From Figure 1, it is observed that MLE and MLE from multiply censored
sample of θ are biased except for θ = 1.0. For θ < 1.0, both estimates are
positively biased where as MLE from censored sample is highly biased than MLE. Stat., Optim. Inf. Comput. Vol. 2, June 2014. 157 RECURRENCE RELATIONS FOR MOMENTS OF MULTIPLY TYPE-II CENSORED ORDER STATISTICS 15 0.5
1.0
1.5
2.0
2.5
0.0
0.5
1.0
1.5
2.0
θ
MLE
MLE
Multiplied Censored MLE
Figure 2. Simulated MLE curves for n = 100 and m = 5. 0.5
1.0
1.5
2.0
2.5
0.0
0.5
1.0
1.5
2.0
θ
MLE
MLE
Multiplied Censored MLE
Figure 2. Simulated MLE curves for n = 100 and m = 5. Figure 2. Simulated MLE curves for n = 100 and m = 5. On the other hand when θ > 1.0, both estimators are negatively biased with almost
same value. For θ > 2.0, the MLE from multiplied censored sample admits the
better performance than MLE in complete case. From Figure 2, it is also observed
that MLE and MLE from multiply censored sample of θ are biased except for
θ = 1.0. For θ < 1.0, both estimates are positively biased where as MLE from
censored sample is moderately higher biased than MLE. On the other hand when
θ > 0.7, both estimators gives the same estimate and gradually goes to unbiased
at θ = 1.0. Stat., Optim. Inf. Comput.
Vol. 2, June 2014. Stat., Optim. Inf. Comput.
Vol. 2, June 2014. 5. Simulation Study After then increase the negatively bias. Performance of MLE from
multiply censored sample goes to superior than MLE as increase the sample size
n(= 100). Stat., Optim. Inf. Comput. Vol. 2, June 2014. BANDER AL-ZAHRANI AND M. A. ALI 158 Table I. Estimated values of θ for n = 30 and m = 5, 7, 9, 12, 15, 20. n
θ
MLE (bθn)
S.E. (bθn)
m
MLE (bθm)
S.E. (bθm)
30
0.5
1.74424
0.23863
5
4.16115
108.29687
7
5.19627
125.29512
0.6
1.52288
0.21213
5
2.46926
142.30428
7
3.09672
151.48994
9
3.68743
168.32375
12
4.07592
174.96544
0.7
1.35711
0.18284
5
1.72878
122.41613
7
2.00455
131.08574
9
2.46257
145.54077
12
2.81153
159.97457
0.8
1.24109
0.17249
5
1.38062
097.49470
7
1.49723
105.62920
9
1.66574
118.70461
12
2.00986
147.58269
15
2.24177
150.52720
20
2.27611
152.30640
1.0
1.04838
0.15706
5
1.04366
73.009340
7
1.04757
73.213190
9
1.07673
73.741700
12
1.09524
78.071560
15
1.14241
84.400380
20
1.15221
85.118260
1.5
0.76252
0.13409
5
0.70500
48.633980
7
0.67805
46.252740
9
0.61697
44.903430
12
0.57415
39.684760
15
0.50341
34.864700
20
0.45744
24.723070
2.0
0.59196
0.11506
5
0.57714
35.824450
7
0.53536
34.860380
9
0.48571
33.270660
12
0.42460
29.088130
15
0.35889
24.479550
20
0.29724
18.663180
2.5
0.47623
0.10041
5
0.50720
26.022670
7
0.48070
24.274150
9
0.40256
23.434160
12
0.34855
21.704760
15
0.29850
18.815920
20
0.22257
14.651580 Table I. Estimated values of θ for n = 30 and m = 5, 7, 9, 12, 15, 20. RECURRENCE RELATIONS FOR MOMENTS OF MULTIPLY TYPE-II CENSORED ORDER STATISTICS 159 Table II. Estimated values of θ for n = 30 and m = 5, 7, 9, 12, 15, 20. Table II. Estimated values of θ for n = 30 and m = 5, 7, 9, 12, 15, 20. n
θ
MLE (bθn)
S.E. (bθn)
m
MLE (bθm)
S.E. Stat., Optim. Inf. Comput.
Vol. 2, June 2014. Stat., Optim. Inf. Comput.
Vol. 2, June 2014. Acknowledgements The authors are grateful to the Editor and the two referees for their valuable and
helpful suggestions which have substantially improved the presentation of the
paper. 5. Simulation Study (bθm)
100
0.5
1.72839
0.13082
5
1.94828
135.31219
7
2.11653
146.56967
0.6
1.49979
0.11160
5
1.58527
110.17368
7
1.65039
114.19110
9
1.73228
119.50743
12
1.87486
128.15120
0.7
1.34292
0.09419
5
1.36471
95.611090
7
1.39678
98.057820
9
1.42720
99.721370
12
1.49337
103.41175
0.8
1.22417
0.09121
5
1.21048
85.17485
7
1.22851
86.56835
9
1.24142
87.13272
12
1.26858
88.58547
15
1.29882
90.13348
20
1.35520
93.59640
1.0
1.03487
0.08526
5
1.00874
70.41334
7
1.00868
70.26557
9
1.00869
70.30250
12
1.01051
69.96724
15
1.00845
69.70745
20
1.01126
69.59897
1.5
0.74844
0.06983
5
0.73956
51.45035
7
0.73035
50.81887
9
0.72242
50.19667
12
0.70901
49.10274
15
0.69596
47.99910
20
0.67256
46.18182
2.0
0.58144
0.06097
5
0.59953
41.67866
7
0.58909
41.34870
9
0.58000
40.57883
12
0.56468
39.30672
15
0.55145
38.23132
20
0.52870
36.21664
2.5
0.46687
0.05465
5
0.50991
27.84197
7
0.49940
34.61033
9
0.49127
33.90684
12
0.47541
32.56191
15
0.46303
31.67288
20
0.44056
29.90937 160 BANDER AL-ZAHRANI AND M. A. ALI Stat., Optim. Inf. Comput.
Vol. 2, June 2014. REFERENCES 1. Ali, M.A. and Khan, A.H. (1998a). Recurrence relations for the expectations of a function
of single order statistic from general class of distributions, Journal of the Indian Statistical
Association, Vol. 35, pp. 1–9. pp
2. Ali, M.A. and Khan, A.H. (1998b). Recurrence relations for expected values of certain
functions of two order statistics, Metron, Vol. LVI, n.1-2, pp. 107–119. pp
3. David, H.A. and Nagaraja, H.N. (2003). Order Statistics, 3rd. ed. Wiley, N.Y., USA. 4. Ghitany, M.E., Atieh, B. and Nadarajah, S. (2008). Lindley distribution and its application,
Mathematics and computers in simulation, Vol. 78, pp. 493–506. 5. Jang, D., Park, J. and Kim, C. (2011). Estimation of the scale parameter of the half-logistic
distribution with multiply type-II censored sample, Journal of the Korean Statistical Society,
Vol. 40, pp. 291–301. 6. Kong, F. and Fei, H. (1996). Limit theorems for the maximum likelihood estimate under
general multiply type-II censoring, Annals of the Institute of Statistical Mathematics, vol. 48, pp. 731–755. pp
7. Lindley, D.V. (1952). Fiducial distributions and Bayes’ theorem, Journal of the Royal
Statistical Society, Series B, Vol. 20, pp. 102–107. 8. Schenk, N.; Burkschat, M.; Cramer, E. and Kamps, U.(2011). Bayesian estimation and
prediction with multiply Type-II censored samples of sequential order statistics from one-
and two-parameter exponential distributions, Journal of Statistical Planning and Inference,
Vol. 141, pp. 1575–1587.
|
https://openalex.org/W2792311422
|
https://link.springer.com/content/pdf/10.1007%2Fs11063-018-9816-6.pdf
|
English
| null |
AIFD Based 2D Image Registration to Multi-View Stereo Mapped 3D Models
|
Neural processing letters/Neural Processing Letters
| 2,018
|
cc-by
| 9,048
|
Neural Process Lett (2018) 48:1261–1279
https://doi.org/10.1007/s11063-018-9816-6 Neural Process Lett (2018) 48:1261–1279
https://doi.org/10.1007/s11063-018-9816-6 AIFD Based 2D Image Registration to Multi-View Stereo
Mapped 3D Models Biao Zhao1 Published online: 14 March 2018
© The Author(s) 2018 Published online: 14 March 2018
© The Author(s) 2018 Abstract Multi-view stereo (MVS) map based 3D range reconstruction is to generate 3D
ranges by analyzing the surrounding snapshots from different perspectives. Different to the
traditional method which employing the expensive and difficult maintaining laser range
devices to calibrate the range of the real 3D objects, MVS has achieved its success by seek-
ing the geometrical correlations between the correspondences from the snapshot of different
perspectives. The concerning of MVS keeps rising thanks to the fast development of digital
maps and 3D printing. Several algorithms with regard to MVS has been well developed and
achieved their success with regard to reconstruction of 3D ranges. Meanwhile, most of the
algorithms were mainly focusing on the fusion and merging of different scenes and surface
refinement. Less capability of the feature matching algorithms on the affine invariant images
renders the current MVS algorithms need huge amount of images with tiny perspective differ-
ences. In this paper, we will propose a new MVS algorithm, deploying our previous published
Affine Invariant Feature Descriptor (AIFD) to detect and match the correspondences from
different perspectives and applying Homograph matrix and segmentation to define the planes
of the objects. Thanks to the AIFD and Homograph based projection model, our proposed
MVS algorithm outperform other MVS algorithms in terms of speed and efficiency. Keywords AIFD · Feature matching · Homograph · Registration · Camera model ·
Multi-view Stereo Keywords AIFD · Feature matching · Homograph · Registration · Camera model ·
Multi-view Stereo 1
University of Lincoln, Lincoln, Brayford Way, Brayford Pool, Lincoln LN6 7TS, UK B Biao Zhao
biao_zhao@yahoo.com 1 Introduction Reconstructing 3D ranges from some snapshots of different perspective is a classical com-
puter vision problem with ever existed concerning. Its applications range from 3D mapping,
navigation to 3D printing, computational photography, video games, or heritage archival. 123 B. Zhao 1262 B. Zhao Only recently have these techniques matured enough to provide industrial scale robustness,
accuracy and scalability. The target of an image-based 3D reconstruction algorithm can be
described as estimating the most likely 3D models by the given set of images under a proper
assumption of material, viewpoints and light conditions. A general MVS pipeline includes: – Image collection. – Image collection. – Calibration for the difference of the camera setting of each image. – Correspondences detection among images. – Correspondences detection among images. – Reconstruction the 3D ranges according to the geometrical correspondences. – Optionally reconstruct the materials of the scene. Most MVS algorithms focuse on the 2D scenes fusion, merging and refinement, to achieve
a dense and accurate 3D ranges estimation. Meanwhile the DoG based feature matching
algorithms, which are less capable of detecting and matching the features under an affine
invariant environment, are making the 3D ranges reconstruction less efficient until granted
with a great amount of snapshots from all the possible perspectives. An accurate and dense correspondence matching plays an important role in the MVS: an
accurate correspondence matching makes the protagonist in the camera matrix calibration,
coherent the images patch of different coordinates; a dense correspondence matching can
establish a density depth clouds, rendering it more accurate and easier to smooth and merge
the 3D ranges surface. The matches to construct the 3D ranges are determined by the images
registration. 3D ranges to 2D images registration is largely depends on an accurate 2D camera
calibration with respect to the acquired geometric 3D model. An accurate camera position
estimation can largely determine the quality of 3D range’s construction. Thus the registration
issue can be simplified as how to conduct the camera matrix in the framework of the projection
model [13]. A large part of the recent success of MVS is due to the success of the underlying
Structure from Motion algorithms that compute the camera parameters. Camera calibration is the fundamental of MVS registration. 1 Introduction It refers to a set of values
describing a camera configuration, that is, camera pose information consisting of location
and orientation, and camera intrinsic properties such as focal length and pixel sensor size. There are many different ways or models to parametrize this camera configuration. There exist
many cues that can be used to calculate the camera parameters from images including: stereo
correspondence, pre-settled devices, snapshot calibration etc. The method the algorithms
applied can largely restrict the range of its application: a controlled MVS capture use diffuse
lights and a turn table to collect the images, outdoor capture can capture series of images
aroundasmall-scalescenes,andcrowd-sourcingfromonlinephoto-sharingwebsites.General
speaking, the algorithm which is capable of tacking arbitrarily snapshots are more desirable. In this situation, a robust, dense and accurate correspondence detecting and matching schema
becomes quite necessary for the MVS. Different to the traditional DoG based feature marching algorithm, we propose a novel
MVS method utilizing our previous published AIFD to detect and match the correspondence
from images to images and introducing the Homograph model to define the smooth planes
from the 3D objects. AIFD is a feature detector and descriptor method, provided a more
improved resilience to affine and scale invariance. It borrowed some ideas of SIFT [14],
like scale space and pyramid structure, etc, but it is more capable to dealing with the image
content of different view points, which is suits the special requirement of MVS. Scale invariant feature detector, like SIFT, SURF [6], ALP [4], etc, has achieved its
success on a lot of applications, including content-based visual retrieval, robotic navigation,
image registration, etc. However, its sensitivity to the view point changes greatly restrict its
applications to a larger range, such as 3D registration for instance for a long time. Borrowed 123 AIFD Based 2D Image Registration to Multi-View Stereo... 1263 the basic principles of SIFT, we have successfully proposed Affine Invariant Feature Detector
(AIFD), which has a better resilience to affine transformations. Equipped with this more
advanced affine invariant feature detector, we can now seek the connections between images
for 3D ranges reconstruction by detecting the matched features. the basic principles of SIFT, we have successfully proposed Affine Invariant Feature Detector
(AIFD), which has a better resilience to affine transformations. Equipped with this more
advanced affine invariant feature detector, we can now seek the connections between images
for 3D ranges reconstruction by detecting the matched features. 1 Introduction g
y
g
A correct detected correspondences between two images constitute a stereo system, which
can provide the depth information. A set of points defined by the depth information out-
shapes the structure of the 3D ranges. In practice, most scenes or partials of the object will
be covered by more than 2 images, which can help to calibrate a more dense and accurate
spatial information. The origins of multi-view stereo can be traced back to human stereopsis
and the first attempts to solve the stereoscopic matching problem as a computation problem
[4]. Until today, two-view stereo algorithms have been a very active and fruitful research area
. The multi-view version of stereo originated as a natural improvement to the two-view case. Instead of capturing two images from different perspectives, multi-view stereo would capture
more viewpoints in-between to increase robustness, e.g. to image noise or surface texture
and viewpoint. What started as a way to improve two-view stereo has nowadays evolved into
a different type of problem. Only equipped with sufficient amount of correspondences from different images, are we
able to be more approach to an accurate camera matrix estimation by DLT. The knowledge
of the registration from 3D ranges to 2D images can improve to map the 3D textures and
in advance can be treated as a homograph reference to be apply to some other 2D images
[3]. With this iterative registration-mapping method, the 3D ranges can be more precisely
registered to an un-calibrated and arbitrarily snapshot [1]. Based on our proposed pipeline, a progressive mapping and registration 3D to 2D images
registration method can be formed. The experiments in the below section can prove the perfor-
mance of our proposed registration method outperforms against the traditional edge/corner
based method [2]. Though our registration proposal, the stereo mapped 3D model can be
introduced to more applications for its efficiency and simplicity [5]. 2 Stereo Visual Based 3D Range Methods 3D range model is referring to a collection of points that presenting the distance in a scene
from a specified viewpoint, which is normally associated with some type of sensor, like the
Laser deceives [7]. To a well formed range model, its pixel value reflects the corresponding
distance to a certain view plain[8]. If the sensor that is used to produce the range is properly
calibrated, the pixel values can directly give the distance in physical units, like meter [9]. The sensor device that is used for producing the range model is sometimes referred to as a
range camera. Range cameras can operate according to a number of different techniques [10],
including Stereo triangulation, Sheet of light triangulation, Time-of-flight, Structured light,
Interferometry, and Coded aperture. Sheet of light triangulation is achieved by changing of
the scene illuminated with a sheet of light this creates a reflected line as seen from the light
source. From any point out of the plane of the sheet the line will typically appear as a curve,
the exact shape of which depends both on the distance between the observer and the light
source, and the distance between the light source and the reflected points. By observing the
reflected sheet of light using a camera (often a high resolution camera) and knowing the
positions and orientations of both camera and light source, it is possible to determine the
distances between the reflected points and the light source or camera. By illuminating the
scene with a specially designed light pattern, structured light [11], depth can be determined
using only a single image of the reflected light. The structured light can be in the form 123 1264 B. Zhao B. Zhao of horizontal and vertical lines, points or checker board patterns. The depth can also be
measured using the standard time-of-flight (ToF) technique, more or less like a radar, in that
a range image similar to a radar image is produced, except that a light pulse is used instead
of an RF pulse. By illuminating points with coherent light and measuring the phase shift
of the reflected light relative to the light source it is possible to determine depth. Under the
assumption that the true range image is a more or less continuous function of the image
coordinates, the correct depth can be obtained using a technique called phase-unwrapping. 2 Stereo Visual Based 3D Range Methods Depth information may be partially or wholly inferred alongside intensity through reverse
convolution of an image captured with a specially designed coded aperture pattern with a
specific complex arrangement of holes through which the incoming light is either allowed
through or blocked. Among all of these techniques, Stereo triangulation is the most popular and widely applied
technique for 3D ranges detections, where the depth data are determined by the data acquired
by stereo or multiple-camera system. This way it is can determine the depth of a certain points
in the scene, for example, from the enter point of the line between their focal points. In order to
solve the depth measurement by employing a stereo camera system, it is necessary to detect
the corresponding points from the different images. A well solution correctly specifying
the correspondences from different images is one of the main task by applying this type
of technique. For instance, it is difficult to detect the correspondence for the image points
that lie inside the regions of homogeneous intensity or color. As a consequence, 3D range
based stereo triangulation can produce reliable depth estimation only for a subset of all points
visible from a multiple-view cameras. The advantage of this technique is that it can guarantee that the measurement is passive
and does not require a special requirement for detecting scene illumination. Other mentioned
technique do not have to tangle the correspondence but are depending on some particular
scene illuminations. To a stereo image, a pixel value is not to store the grey scale level at a certain position, but
to identify the depth information. Stereo cameras captures two images with tiny difference
on the scene. In Fig. 1 the point A on the right light from is transmitted through the entry
points of a pinhole cameras at B and D, onto image screens at E and H. In the attached
diagram the distance between the centers of the two camera lens is BD = BC + C D. The
triangles are similar, AC B and BFE and AC D and DGH. 2 Stereo Visual Based 3D Range Methods Therefore displacement d = E F + GH
= BC + C D
AC
= BD
AC
= k
z , where k = BD, z = AC
(1) (1) = k
z , where k = BD, z = AC Assuming the cameras are from the same level, and image plane is flat on the same plane,
the displacement in the Y axis between the same pixel in the two images is d = k
z , Where k
is the distance between the two cameras times the distance from the lens to the image. The
depth component in the two images are z1, z1 given by, z2(x, y) = min
v : v = z1
x, y −
k
z1(x, y)
z1(x, y) = min
v : v = z2
x, y +
k
z2(x, y)
(2) (2) 1265 AIFD Based 2D Image Registration to Multi-View Stereo... Fig. 1 Relationship of image displacement to depth with stereoscopic images, assuming flat co-planar images Fig. 1 Relationship of image displacement to depth with stereoscopic images, assuming flat co-planar images Because the computation is squaring of the points in SSD, many implementations use Sum of
Absolute Difference (SAD) as the basis for computing the measurement. Other methods use
normalized cross correlation (NCC). The least squares measure can be deployed to measure
the information content of the stereoscopic images, given the depth information at each point
z(x, y). The information needed to represent the image is called Im. Image rectification is
required to adjust the images as if they were co-planar. This may be achieved by a linear
transformation. The images also need to be rectified to make each image equivalent to the
one taken from a pinhole camera as if they are projected to a flat plane. The normal distribution is, The normal distribution is, P(x, μ, σ) =
1
σ
√
2π
e−(x−μ)2
2σ2
(3) (3) Probability is related to information content described by message length L, Probability is related to information content described by message length L, P(x) = 2−L(x)
L(x) = −log2 P(x)
(4) (4) 1266 B. Zhao B. Zhao So, L(x, μ, σ) = log2(σ
√
2π) + (x −μ)2
2σ 2
log2 e
(5) (5) The least squares measure may be used to measure the information content of the stereo-
scopic images [], given depths at each point z(x, y). 2 Stereo Visual Based 3D Range Methods Firstly the information needed to express
one image in terms of the other is derived. This is called Im. A color difference function can be used to fairly measure the difference between colors. The color difference function is written cd in the following. The measure of the information
needed to record the color matching between the two images is, Im(z1, z2) = 1
σ 2m
x,y
cd
color1
x, y +
k
z1(x, y)
, color2(x, y)
2
(6) (6) An assumption is made about the smoothness of the image. Assuming that two pixels are
more likely to be of the same color, the closer the voxels they represent are. This measure
is intended to favor colors that are similar being grouped at the same depth. For example, if
an object in front occludes an area of sky behind, the measure of smoothness favors the blue
pixels all being grouped together at the same depth. T he total measure of smoothness uses
the distance between voxels as an estimate of the expected standard deviation of the color
difference, Is(z1, z2) =
1
2σ 2
h
i:{1,2}
x1,y1
x2,y2
cd(colori(x1, y1), colori(x2, y2))2
(
)2
(
)2
( (
)
(
))2
(7) Is(z1, z2) =
1
2σ 2
h
i:{1,2}
x1,y1
x2,y2
(7) (7) cd(colori(x1, y1), colori(x2, y2))2
(x1 −x2)2 + (y1 −y2)2 + (zi(x1, y1) −zi(x2, y2))2
(7 The total information content is then summed as, It(z1, z2) = Im(z1, z2) + Is(z1, z2)
(8) (8) The z component of each pixel must be chosen to provide the minimum value for the infor-
mation content. This will give the most likely depths at each pixel. The minimum total
information measure is, Imin = min {i : i = It(z1, z2)}}
(9) (9) The depth functions for the left and right images are the pair, The depth functions for the left and right images are the pair, (z1, z2) ∈{(z1, z2) : It(z1, z2) = Imin}
(10) (10) For each point v = (vx, vy, vz), the transformed point would be Pv =
⎡
⎣
1
0
0
0
1
0
0
0
0
⎤
⎦
⎡
⎣
vx
vy
vz
⎤
⎦=
⎡
⎣
vx
vy
0
⎤
⎦
(12) (12) Often, it is more useful to use homogeneous coordinates. The transformation above can be
represented for homogeneous coordinates as P =
⎡
⎢⎢⎣
1
0
0
0
0
1
0
0
0
0
0
0
0
0
0
1
⎤
⎥⎥⎦
(13) (13) For each homogeneous vector v = (vx, vy, vz, 1), the transformed vector would be For each homogeneous vector v = (vx, vy, vz, 1), the transformed vector would be Pv =
⎡
⎢⎢⎣
1
0
0
0
0
1
0
0
0
0
0
0
0
0
0
1
⎤
⎥⎥⎦
⎡
⎢⎢⎣
vx
vy
vz
1
⎤
⎥⎥⎦=
⎡
⎢⎢⎣
vx
vy
0
1
⎤
⎥⎥⎦
(14) (14) 3 Orthogonal Projection Orthographic projection (sometimes orthogonal projection), is a means of representing three-
dimensional objects in two dimensions. It is a form of parallel projection, in which all the
projection lines are orthogonal to the projection plane,[1] resulting in every plane of the scene
appearing in affine transformation on the viewing surface. The obverse of an orthographic
projection is an oblique projection, which is a parallel projection in which the projection
lines are not orthogonal to the projection plane. The term orthographic is sometimes reserved specifically for depictions of objects where
the principal axes or planes of the object are also parallel with the projection plane, [1]
but these are better known as multi-view projections. When the principal planes or axes of 123 AIFD Based 2D Image Registration to Multi-View Stereo... 1267 an object are not parallel with the projection plane, but are rather tilted to reveal multiple
sides of the object, the projection is called an axonometric projection. Sub-types of multi-
view projection include plans, elevations and sections. Sub-types of axonometric projection
include isometric, dimetric and trimetric projections. an object are not parallel with the projection plane, but are rather tilted to reveal multiple
sides of the object, the projection is called an axonometric projection. Sub-types of multi-
view projection include plans, elevations and sections. Sub-types of axonometric projection
include isometric, dimetric and trimetric projections. A simple orthographic projection onto the plane z = 0 can be defined by the following
matrix:
⎡1
0
0⎤ P =
⎡
⎣
1
0
0
0
1
0
0
0
0
⎤
⎦
(11) (11) For each point v = (vx, vy, vz), the transformed point would be For each point v = (vx, vy, vz), the transformed point would be 4 AIFD Based Feature Matching Our previously published feature descriptor AIFD achieves its resilience to affine and scale
changes by reshaping the multi-scale image representation and local extrema detection in
order to maintain a linear relationship under the changes of affine transformation. Instead of
relying on the image simulations, AIFD achieves its affine and scale invariance completely
based on its internal mechanisms when dealing with the transformed visual content. Thus it
is more brief, reliable and feasible to more applications and has more potentials for future
research. It is based on our previous proposed affine scale space. For a given image I (x, y), its scale
space is given by a family of pre-smoothed images L(x, y, σ), where the scale parameter is
pre-defined according to its kernel size: g(x, y; σ) =
1
2πσ 2 e−x2+y2
2σ2
such as,
L(x, y; σ) = g(x, y; σ) ∗I (x, y)
(15) (15) Based on this definition, a structure more adaptable to the affine transformation is defined
below: 1268 B. Zhao B. Zhao Fig. 2 The pipeline of AIFD. The parameters in the dashed box is pre-settled and only determined by the
pyramid structure Fig. 2 The pipeline of AIFD. The parameters in the dashed box is pre-settled and only determined by the
pyramid structure Fig. 2 The pipeline of AIFD. The parameters in the dashed box is pre-settled and only determined by the
pyramid structure Fig. 2 The pipeline of AIFD. The parameters in the dashed box is pre-settled and only determined by the
pyramid structure g(η, Σs)A =
1
2π√detΣs e−ηT Σ−1
s
η
2
. (16) (16) where, Σs = Aσ 2 AT . In this formula, A represents the affine transformation, a 2 × 2 matrix. σ is the scale. This
deformed Gaussian kernel is specialized to generate affine scale space which can maintain
linear relationship regardless the change of view point. Based on this structure, the images
from any view points can be well represented from multi-scales. From the definition of affine
scale space, conventional isotropic scale space can be deemed as a spacial case, whose affine
transformation equals to the 2 × 2 identity matrix (Fig. 2). The similarity of two visual content is largely depends on the matched features detected
from the scale space. 4 AIFD Based Feature Matching To the conventional scale space, several approaches to detect the local
maximum or minimum from derivatives have been proposed [12], and local LoG extrema
detectionoutperformallothers,concerningtheaccuracyandefficiencyofamethodinpractice
[17]. The Laplacian of Gaussian (LoG ) scale space can be mathematically expressed as: ∇2L = Lxx + L yy
(17) (17) In this formula, L represents the scale space. The local maximum or minimum over the
Laplacian can then be selected as the feature candidates. In this formula, L represents the scale space. The local maximum or minimum over the
Laplacian can then be selected as the feature candidates. Borrowing the idea of LoG, we have also proposed an affine LoG, with the purpose to
promote the feature candidates detection over affine scale space. Instead of a direct laplacian
operation, we have proposed a feasible implementation based on our proposed pyramid
structure to efficiently generate affine LoG. By this implementation, the affine Gaussian and
LoG scale space can be simultaneously generated. More information about affine scale space
and affine LoG can be found [18]. 123 123 AIFD Based 2D Image Registration to Multi-View Stereo... AIFD Based 2D Image Registration to Multi-View Stereo... 1 1269 Once obtaining the affine scale space, we can then detect the local extrema by comparing
its Harris and Hessian matrix. If we suppose, A =
⎡
⎣
∂2f
∂x2
∂2f
∂x∂y
∂2f
∂x∂y
∂2f
∂y2
⎤
⎦
B =
⎡
⎢⎣
∂f
∂x
2
∂f
∂x
∂f
∂y
∂f
∂x
∂f
∂y
∂f
∂y
2
⎤
⎥⎦
(18) (18) Nominating the eigenvalues of A as ψ1 and ψ2 and the eigenvalues of A as ψ1 and ψ2, we
can then detect the features by analysing, Nominating the eigenvalues of A as ψ1 and ψ2 and the eigenvalues of A as ψ1 and ψ2, we
can then detect the features by analysing, 1
4 min{ψ1, ψ2}2 > max{ν1, ν2}
(19) 1
4 min{ψ1, ψ2}2 > max{ν1, ν2}
(19) (19) With this way to find the feature candidates, we can apply the LoG first and second
derivative filters to form the Harris and Hessian matrices. With this way to find the feature candidates, we can apply the LoG first and second
derivative filters to form the Harris and Hessian matrices. 4 AIFD Based Feature Matching l1 =A−1η 1
πσ 6 e−ηT (AAT )−1η
2σ2
2 −ηT (AAT )−1η
2σ 2
laxx = 1
πσ 6
ηT (AAT )η
2σ 2
−3
( Ma(1, 1)
σ 2
−1) −1
layy = 1
πσ 6
ηT (AAT )η
2σ 2
−3
( Ma(2, 2)
σ 2
−1) −1
laxy = 1
πσ 6
ηT (AAT )η
2σ 2
−3
Ma(1, 2)
σ 2
(20) (20) Applying the same structure as depicted in Fig. 3 to generate the LoG derivative scale
space to detect the local extrema. Applying the same structure as depicted in Fig. 3 to generate the LoG derivative scale
space to detect the local extrema. Borrowing the idea of CDVS [16], we have proposed a polynomial expression to represent
the affine scale space (also affine LoG and affine LoG derivatives) with a continual scale from. Supposing the the continual scale space can be approached by a cubic polynomial, which
can be written as: L(σ) = a · σ 3 + b · σ 2 + c · σ + d
(21) (21) where, the parameters a, b, c, d are of the scale space’s image size. σ represents a scale value
within the Octave range. L(σ) represents the image of a specified scale, it can be one image
of Gaussian scale space, LoG scale space or LoG derivatives of any scale. This expression
can be deployed as the continual form to represent the scale space of Gaussian, LoG or LoG
derivatives. Apparently, σ can be any value within the Octave range, and L(σ) represents the
corresponding scale space. According to the expression above, the image of any specified
scale can be calculated given the parameters a, b, c, d, which are related with the input image. Thanks to the pyramid structure, 4 scale space within each Octave can easily be generated. The equations at the 4 parameters can be written as: ⎡
⎢⎢⎣
a(x, y)
b(x, y)
c(x, y)
d(x, y)
⎤
⎥⎥⎦= M
⎡
⎢⎢⎣
L(σ1)
L(σ2)
L(σ3)
L(σ4)
⎤
⎥⎥⎦,
(22) ⎡
⎢⎢⎣
a(x, y)
b(x, y)
c(x, y)
d(x, y)
⎤
⎥⎥⎦= M
⎡
⎢⎢⎣
L(σ1)
L(σ2)
L(σ3)
L(σ4)
⎤
⎥⎥⎦,
(22) (22) where M is a 4 × 4 matrices and a, b, c, d are linear combination of L(σ1), L(σ2), L(σ3)
L(σ4). 4 AIFD Based Feature Matching Matrix M is a simple 4 × 4 matrix and it is determined under the certain affine
transformation. Thanks to the linear property of the polynomial expression, the matrix M 123 B. Zhao 1270 B. Zhao Fig. 3 Proposed pyramid structure to speed up the generation of Gaussian, LoG and also LoG derivatives
scale space Fig. 3 Proposed pyramid structure to speed up the generation of Gaussian, LoG and also LoG derivatives
scale space Fig. 3 Proposed pyramid structure to speed up the generation of Gaussian, LoG and also LoG derivatives
scale space will not be affected under a different view point, which can be demonstrated, if supposing
the affine transformation is A. will not be affected under a different view point, which can be demonstrated, if supposing
the affine transformation is A. L(A, σ) =
σ 3 σ 2 σ 1
M
⎡
⎢⎢⎣
L(A, σ1)
L(A, σ2)
L(A, σ3)
L(A, σ4)
⎤
⎥⎥⎦
(23) (23) The candidates, with the Harris-Hessian matrices subtraction smaller than 0.001 will
also be rejected to guarantee the extrema is larger upto a certain level compared with the
surrounding points by R = Tr(H)2/Det(H), which equals to (γ + 1)2/γ . Since L(x + x) is the local extreme, its derivative equals 0. By taking the derivative on
both sides of the equation, we can have: x = −{D2L(x)}−1DL(x)
(24) (24) If the offset is larger than 0.5 in any dimension, it implies the real extreme location is closer
to a different pixel sample. Considering the local extreme scale for each pixel sample can be
quite diverse, the feature located at the specified integral position becomes no more adequate. The offset ˆx will be summed up with the detected integral position to approach the local
extreme location to sub-pixel’s precision, according to the formula Eq. 31. The Hessian
matrix and local gradient of the pixel sample can be obtained by the corresponding LoG
derivative polynomial expressions. To cope with an affine gradient-based feature descriptor, we will introduce the affine
gradient filter and the related scale space in this section. At the very beginning, the definition
of image gradient can be given by: 123 AIFD Based 2D Image Registration to Multi-View Stereo... AIFD Based 2D Image Registration to Multi-View Stereo... 1271 1271 ∇I =
∂I
∂x , ∂I
∂y
(25) (25) It is equivalent to the first order of image derivatives. 4 AIFD Based Feature Matching The traditional method to calculate
the image gradient is by taking the subtraction from two image neighbouring pixels in the
form:
1 ∇I = 1
2 (I (x −1, y) −I (x + 1, y), I (x, y −1) −I (x, y + 1))
(26) (26) For every point of image gradient, its direction and magnitude can be given by: For every point of image gradient, its direction and magnitude can be given by: m(x, y) =
∂I
∂x
2
+
∂I
∂y
2
,
θ = arctan
⎛
⎜⎜⎜⎝
∂I
∂x
∂I
∂y
⎞
⎟⎟⎟⎠. (27) (27) where I is the original image, ∗is convolution operation, g is Gaussian filter. Thus, the
derivative of the Gaussian blurred image of the corresponding scale space can be equivalently
obtained though filtering the image with the Gaussian derivative filters, which can be derived
by: where I is the original image, ∗is convolution operation, g is Gaussian filter. Thus, the
derivative of the Gaussian blurred image of the corresponding scale space can be equivalently
obtained though filtering the image with the Gaussian derivative filters, which can be derived
by: gx
gy
=
⎡
⎢⎢⎣
∂g
∂x
∂g
∂y
⎤
⎥⎥⎦=
⎡
⎢⎢⎣
x
2πσ 4e−x2+y2
2σ2
y
2πσ 4e−x2+y2
2σ2
⎤
⎥⎥⎦=
1
2πσ 2 e−x2+y2
2σ2
⎡
⎢⎢⎣
x
σ 2
y
σ 2
⎤
⎥⎥⎦
(28) (28) Thus, the Gaussian derivatives scale space can easily be obtained by multiplying the corre-
sponding Gaussian scale space with x/σ 2 and y/σ 2. Borrowing the idea of affine LoG, we can have the affine Gaussian derivative filters as: Borrowing the idea of affine LoG, we can have the affine Gaussian derivative filters as: gA
x
gA
y
= gA
η = A−1η
2πσ 4 e−ηT (AAT )−1η
2σ2
(29) (29) where A is a 2×2 matrix, indicating the affine transformation, η = A
x
y
. Like the isotropic
Gaussian derivative scale space, the affine Gaussian derivative scale space can also be gen-
erated by multiplying the corresponding Gaussian scale space with A−1η/σ 2. where A is a 2×2 matrix, indicating the affine transformation, η = A
x
y
. Like the isotropic
Gaussian derivative scale space, the affine Gaussian derivative scale space can also be gen-
erated by multiplying the corresponding Gaussian scale space with A−1η/σ 2. 4 AIFD Based Feature Matching Gradients from affine transformed images are restrained by the affine matrices between
different viewpoints. Around each feature, the relocated gradients, according to the affine
transformation, can then form a histogram as the feature descriptor (Figs. 4, 5). The gradient relocation can be done in the form of: The gradient relocation can be done in the form of: x′
y′
= A
x
y
(30) (30) where x, y is the index of the gradient around the detected features from 1 to the descriptor,
pre-defined by the scale of the features. x′ and y′ are the indexes accounting for the affine
transformation. The relocation of the gradient can then be collected according to the new
calculated index. The interpolation of the gradients may also be applied if the new calculated
indexes are not integers. Assigning an orientation to each feature, the feature descriptor can be represented relative
to this orientation and achieve its invariance to image rotation. To calibrate the orientation 123 123 1272 B. Zhao Fig. 4 Gradient relocated from the specified area around the detected features. The index of the gradient can
be calculated according to the affine transformations Fig. 4 Gradient relocated from the specified area around the detected features. The index of the gradient can
be calculated according to the affine transformations Fig. 5 The gradient around the feature will be added to the gradient histogram weighted by its magnitude
and a Gaussian-weighted circular window [14] Fig. 5 The gradient around the feature will be added to the gradient histogram weighted by its magnitude
and a Gaussian-weighted circular window [14] of a feature, an area of scale space gradient around the feature will first be formed, after
proceeding our proposed gradient relocation, eliminating the effect of the affine distortion. The area of gradient to be collected is in a square shape with its size equal to 3 times that of
the feature scale. Then, the orientation of each sample of scale space gradient can be added
to the orientation histogram weighted by its gradient magnitude and by a Gaussian-weighted
circular window with 1.5 times the scale [14]. Then, the orientation histogram will be subdivided into 36 bins covering the 360◦range of
orientations and filled the corresponding accumulated magnitude. The peak of the histogram
points to its main direction. 4 AIFD Based Feature Matching The image rotation can also be equalized as a special case of affine
transformation in the form of: R(θ) =
cos θ −sin θ
sin θ
cos θ
(34) (34) Combining the affine transformation, the total transformation can be synthesized as: A′ =
cos θ −sin θ
sin θ
cos θ
· A
(35) (35) By applying the gradient relocation depicted in Fig. 6, a square patch of gradient can be
obtained (Fig. 7). By applying the gradient relocation depicted in Fig. 6, a square patch of gradient can be
obtained (Fig. 7). As depicted in Fig. 8, affine descriptor can then be generated. It has an exact same form
of SIFT descriptor and can be directly applied in a lot of pre-SIFT applications 4 AIFD Based Feature Matching Any other local peak that is within 80% of the highest peak and
higher than the average of its two neighbors will be assigned with different orientations. The
features with multiple peaks will be respectively created at the same location with the same
scale but different orientations [14]. In advance, the main directions of the feature can also promote its accuracy by utilizing
the second order of Taylor expansion in the form of: x = −{D2L(x)}−1DL(x)
(31) (31) Supposing the detected main direction bin is M and its two neighbor direction bins are M−
and M+, the above equation can be implemented in the form of: Supposing the detected main direction bin is M and its two neighbor direction bins are M−
and M+, the above equation can be implemented in the form of: M = −
1
2(M+ −M−)
1
2(M+ + M−) −M
(32) (32) 123 123 123 1273 AIFD Based 2D Image Registration to Multi-View Stereo... 0
5
10
15
20
25
30
35
orientation bin
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
accumulated magnitude
Fig. 6 Orientation histogram, created by subdividing the surrounding gradient into 36 bins according to the
gradient orientations and accumulating weighted magnitude. The largest bin will be selected as the main
direction of the descriptor 0
5
10
15
20
25
30
35
orientation bin
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
accumulated magnitude Fig. 6 Orientation histogram, created by subdividing the surrounding gradient into 36 bins according to the
gradient orientations and accumulating weighted magnitude. The largest bin will be selected as the main
direction of the descriptor Then, the main direction is: Then, the main direction is: θ = (M +
M) · 360
36
(33) (33) The offset larger than 0.5 will be autocratically rejected. The offset larger than 0.5 will be autocratically rejected. The offset larger than 0.5 will be autocratically rejected. Assigning an orientation to each feature descriptor is to gain the invariance to image
rotation by indicating the relations between the representation and the feature itself. Gradient
is introduced with the purpose of avoiding the effects of the illumination changes. The
patch to collect the gradient will then steer to the main direction to generate the traditional
SIFT descriptor [14]. 5 Camera Matrix Calculation Via the Image-Image Feature
Correspondences With the matched correspondences, we can then apply the DLT method to calculate the
camera matrix to define the 2D image to 3D models position mapping (Figs. 8, 9, 10). 1274 B. Zhao B. Zhao Fig. 7 Gradient relocation according to the synthesized transformation matrix. After the relocation, we will
obtain a set of gradients eliminating the affine transformation and pointing to the main direction Fig. 7 Gradient relocation according to the synthesized transformation matrix. After the relocation, we will
obtain a set of gradients eliminating the affine transformation and pointing to the main direction Fig. 8 Borrowing the idea of SIFT descriptor, affine descriptor can be generated from the relocated gradient Fig. 8 Borrowing the idea of SIFT descriptor, affine descriptor can be generated from the relocated gradi Fig. 8 Borrowing the idea of SIFT descriptor, affine descriptor can be generated from the relocated gradient The mapping from the coordinates of a 3D point P to the 2D image coordinates of the
point’s projection onto the image plane, according to the pinhole camera model is given by y1
y2
= f
x3
x1
x2
(36) (36) where (x1, x2, x3) are the 3D coordinates of P relative to a camera centered coordinate
system, (y1, y2) are the resulting image coordinates, and f is the camera’s focal length for
which we assume f > 0. Furthermore, we also assume that x3 > 0. To derive the camera
matrix this expression is rewritten in terms of homogeneous coordinates. Instead of the 2D
vector (y1, y2) we consider the projective element (a 3D vector)y = (y1, y2, 1) and instead
of equality we consider equality up to scaling by a non-zero number, denoted ∼. First, we
write the homogeneous image coordinates as expressions in the usual 3D coordinates. where (x1, x2, x3) are the 3D coordinates of P relative to a camera centered coordinate
system, (y1, y2) are the resulting image coordinates, and f is the camera’s focal length for
which we assume f > 0. Furthermore, we also assume that x3 > 0. To derive the camera
matrix this expression is rewritten in terms of homogeneous coordinates. Instead of the 2D
vector (y1, y2) we consider the projective element (a 3D vector)y = (y1, y2, 1) and instead
of equality we consider equality up to scaling by a non-zero number, denoted ∼. 5 Camera Matrix Calculation Via the Image-Image Feature
Correspondences First, we
write the homogeneous image coordinates as expressions in the usual 3D coordinates. ⎛
⎝
y1
y2
1
⎞
⎠= f
x3
⎛
⎝
x1
x2
x3
f
⎞
⎠∼
⎛
⎝
x1
x2
x3
f
⎞
⎠
(37) (37) 123 AIFD Based 2D Image Registration to Multi-View Stereo... 1275 Fig. 9 A typical example of our proposed method. We will compare the performance our proposed method
with some other most frequently applied algorithm on this special case [15] Fig. 9 A typical example of our proposed method. We will compare the performance our proposed method
with some other most frequently applied algorithm on this special case [15] Fig. 10 Typical 3D and 2D images with fiducial markers: a CBCT axial slice (a), a 2D-MAX LAT (b) and
2D-MAX AP (c) images. The arrows indicate the locations of the fiducial markers [15] Fig. 10 Typical 3D and 2D images with fiducial markers: a CBCT axial slice (a), a 2D-MAX LAT (b) and
2D-MAX AP (c) images. The arrows indicate the locations of the fiducial markers [15] Finally, also the 3D coordinates are expressed in a homogeneous representation x and this
is how the camera matrix appears: ⎛
⎝
y1
y2
1
⎞
⎠∼
⎛
⎝
1 0 0 0
0 1 0 0
0 0 1
f 0
⎞
⎠
⎛
⎜⎜⎝
x1
x2
x3
1
⎞
⎟⎟⎠
(38) ⎛
⎝
y1
y2
1
⎞
⎠∼
⎛
⎝
1 0 0 0
0 1 0 0
0 0 1
f 0
⎞
⎠
⎛
⎜⎜⎝
x1
x2
x3
1
⎞
⎟⎟⎠
(38) (38) Or, Or,
y ∼C x
(39) Or,
y ∼C x
(39) y ∼C x
(39) Or,
y ∼C x
(3 (39) where C is the camera matrix, which here is given by C =
⎛
⎝
1 0 0 0
0 1 0 0
0 0 1
f 0
⎞
⎠ C =
⎛
⎝
1 0 0 0
0 1 0 0
0 0 1
f 0
⎞
⎠
(40) (40) 1276 B. Zhao B. 5 Camera Matrix Calculation Via the Image-Image Feature
Correspondences Zhao and the corresponding camera matrix now becomes and the corresponding camera matrix now becomes and the corresponding camera matrix now becomes and the corresponding camera matrix now becomes and the corresponding camera matrix now becomes C =
⎛
⎝
1 0 0 0
0 1 0 0
0 0 1
f 0
⎞
⎠∼
⎛
⎝
f 0 0 0
0 f 0 0
0 0 1 0
⎞
⎠
(41) (41) The last step is a consequence of C itself being a projective element. The last step is a consequence of C itself being a projective element. The camera matrix derived here may appear trivial in the sense that it contains very few
non-zero elements. This depends to a large extent on the particular coordinate systems which
have been chosen for the 3D and 2D points. In practice, however, other forms of camera
matrices are common, as will be shown below. The camera matrix C derived in the previous section has a null space which is spanned
by the vector n =
⎛
⎜⎜⎝
0
0
0
1
⎞
⎟⎟⎠
(42) (42) This is also the homogeneous representation of the 3D point which has coordinates (0,0,0),
that is, the “camera center” (aka the entrance pupil; the position of the pinhole of a pinhole
camera) is at O. This means that the camera center (and only this point) cannot be mapped to
a point in the image plane by the camera (or equivalently, it maps to all points on the image
as every ray on the image goes through this point). For any other 3D point with x3 = 0, the result y ∼C is well-defined and has the form
y = (y1 y2 0)⊤. This corresponds to a point at infinity in the projective image plane (even
though, if the image plane is taken to be a Euclidean plane, no corresponding intersection
point exists). In this way, can we achieve the 2D images registered to the 3D stereo visual based mode 6 Experiment Detecting the features across the images of different perspective is the fundamental for
acquiring accurate depth information. An accurate depth information is also the fundamental
for an accurate 3D to 2D image model registration. Based on this situation, we have listed
our feature matching upon the images of different perspectives for 3D-2D registration in the
Table 1. In the Table, we present the performance of our proposed algorithm comparing with some
most widely applied feature matching algorithms in the 3D-2D image registration, including
SIFT, SURF and ALP. By comparison, our proposed method outperforms agains other feature
matching algorithms specially on the 2D registration images. General speaking, our proposed
3D-2D registration method is largely depending on an accurate feature matching algorithm
which can uttermost detect the potential correlations between different perspectives of images
(Tables 2, 3). The clinical image database was used to quantitatively evaluate the performances of our
proposed AIFD based method comparing with three state-of-the-art 3D-2D registration meth-
ods Selection of the state-of- the-art methods was limited to methods that are well established
in the field of 3D-2D registrations, and that are capable of registering a 3D image either to
one 2D view or to multiple 2D views simultaneously. There were about 14,000 centerline
points per 2D image for which the distance transform was precomputed so as to speed 123 AIFD Based 2D Image Registration to Multi-View Stereo... 6 Experiment 1277 Table 1 The average matching performance upon the images from the different perspectives (dataset QUBR)
Skewing
SIFT
ALP
SURF
Proposed AIFD
Number
Ratio
Number
Ratio
Number
Ratio
Number
Ratio
1
423.345
0.305
454.521
0.408
371.667
0.380
483.479
0.623
2
93.243
0.058
102.356
0.095
89.032
0.047
436.138
0.577
3
81.543
0.052
97.456
0.085
72.546
0.038
381.59
0.496
4
24.456
0.017
37.056
0.031
22.453
0.015
328.897
0.431
5
16.657
0.010
21.802
0.0155
15.234
0.00967
289.805
0.378
6
13.234
0.0078
18.307
0.0135
12.342
0.00503
235.845
0.322
7
5.213
0.0017
6.234
0.0029
3.434
0.00093
183.395
0.289
8
1.367
0.0013
2.921
0.00192
1.412
0.00107
157.281
0.228
9
0.572
0.0003
0.560
0.0005
0.442
0.0003
127.362
0.220
The row of skewing level refers to the level of view point difference ble 1 The average matching performance upon the images from the different perspectives (dataset QUBR) Table 2 Comparison of our proposed method and some state-of-the-art algorithms
View
Method
MEAN±STD
SR (%)
CR (mm)
Time (s)
DSA
MAX
DSA
MAX
DSA
MAX
LAT
MIP-MI
0.32± 0.21
0.56±0.53
72.32
34.23
4
3
76.3
ICP
0.44±0.23
42.02
1
1.1
BGB
0.40± 0.37
0.41± 0.36
52.38
48.43
3
2
13.4
MGP
0.61± 0.37
0.63± 0.39
73.23
69.98
5
3
0.9
MGP+BGB
0.26±0.23
0.29±0.27
73.21
72.21
5
3
12.8
AIFD
0.62±0.25
0.58±0.31
70.31
68.23
5
4
5.7
AP
MIP-MI
0.27±0.32
0.68±0.45
91.78
32.87
9
3
65.2
ICP
0.32±0.25
72.48
1
0.4
BGB
0.32±0.35
0.44±0.33
58.32
52.13
3
5
13.8
MGP
0.53±0.27
0.63±0.33
92.43
85.68
10
9
0.9
MGP+BGB
0.28±0.17
0.39±0.27
95.45
85.3
11
8
10.5
AIFD
0.45±0.33
0.56±0.21
72.34
63.21
4
4
6.8
Tested on dataset I and II. Mean and std of mtre values of successful registrations, success rates (sr), capture
ranges (cr) and mean execution times averaged over all clinical datasets Table 2 Comparison of our proposed method and some state-of-the-art algorithms Tested on dataset I and II. Mean and std of mtre values of successful registrations, success rates (sr), capture
ranges (cr) and mean execution times averaged over all clinical datasets up the nearest neighbor search to the projected 3D centerline points. In the BGB method
the 3D intensity gradients were using the Canny edge detector, which resulted in about
17,000 edge points. The 2D intensity gradients were computed by the central difference
kernel. up the nearest neighbor search to the projected 3D centerline points. 6 Experiment In the BGB method
the 3D intensity gradients were using the Canny edge detector, which resulted in about
17,000 edge points. The 2D intensity gradients were computed by the central difference
kernel. Parameters of the state-of-the-art 3D-2D registration methods were experimentally set
to obtain the best registration performances on the clinical image dataset. For MIP-MI
method, the sampling step along the projection rays was 0.375 mm and the intensities were
discretized in 64 bins to compute the MI histograms. The ICP had no user-controlled param-
eters, while in the BGB method the sensitivity of the angle weighting function, was set to
n = 4. 123 B. Zhao 1278 B. Zhao Table 3 Comparison of our proposed method and some state-of-the-art algorithms
View
Method
MEAN±STD
SR (%)
CR(mm)
Time (s)
DSA
MAX
DSA
MAX
DSA
MAX
LAT
MIP-MI
0.23±0.22
0.57±0.46
67.32
32.12
3
3
116.3
ICP
0.48±0.33
41.57
2
1.1
BGB
0.38±0.32
0.38±0.31
51.32
42.41
2
2
11.4
MGP
0.61±0.37
0.63±0.39
73.23
69.98
5
2
1.8
MGP+BGB
0.24±0.17
0.26±0.21
72.29
71.87
4
2
18.7
AIFD
0.45±0.34
0.23±0.46
73.21
61.32
4
3
8.2
Tested on dataset III. Mean and std of mtre values of successful registrations, success rates (sr), capture ranges
(cr) and mean execution times averaged over all clinical datasets Table 3 Comparison of our proposed method and some state-of-the-art algorithms Tested on dataset III. Mean and std of mtre values of successful registrations, success rates (sr), capture ranges
(cr) and mean execution times averaged over all clinical datasets 7 Conclusion In this paper, we presented a novel method for 3D-2D rigid registration based on our previous
proposed feature matching algorithm AIFD, which it more able to detect the correspondences
across the images of different perspectives. The main advantage of the proposed method is
it is more robust to viewpoint difference, resulting a less number of snapshot around. By
experiment, it can be proven that our proposed method performs best among all the most
applied registration method, and the overall execution time is also quite fast. Translation of any 3D-2D registration method into clinical practice requires extensive and
rigorous evaluations on real-patient image databases. Therefore, we acquired a clinical image
database representative of cerebral-EIGI and established a highly accurate gold standard reg-
istration that enables objective quantitative evaluation of 3D-2D rigid registration methods. The quantitative and comparative evaluation of three state-of-the-art methods showed that
the performance of the proposed method best met the demands of cerebral EIGI. Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 Interna-
tional License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and
reproduction in any medium, provided you give appropriate credit to the original author(s) and the source,
provide a link to the Creative Commons license, and indicate if changes were made. 1. Datta R, Joshi D, Li J, Wang JZ (2008) Image retrieval: ideas, influences, and trends of the new age. ACM
Comput Surv 40(2):5 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 Interna-
tional License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and
reproduction in any medium, provided you give appropriate credit to the original author(s) and the source,
provide a link to the Creative Commons license, and indicate if changes were made. p
2. Du S, Guo Y, Sanroma G, Ni D, Wu G, Shen D (2015) Building dynamic population graph for accurate
correspondence detection. Medical Image Analysis g
y
3. Du S, Liu J, Zhang C, Zhu J, Li K (2015) Probability iterative closest point algorithm for m-d point set
registration with noise. Neurocomputing 157:187–198 6. Förstner W, Gülch E (1987) A fast operator for detection and precise location of distinct points, corners
and centres of circular features. In: Proceeding ISPRS intercommission conference on fast processing of
photogrammetric data, pp. 281–305 5. Florack L, Maas R, Niessen W (1999) Pseudo-linear scale-space theory. Int J Comput Vision 31(2–
3):247–259 4. Fischler MA, Bolles RC (1981) Random sample consensus: a paradigm for model fitting with applications
to image analysis and automated cartography. Commun ACM 24(6):381–395 References p
2. Du S, Guo Y, Sanroma G, Ni D, Wu G, Shen D (2015) Building dynamic population graph for accur
correspondence detection. Medical Image Analysis 2. Du S, Guo Y, Sanroma G, Ni D, Wu G, Shen D (2015)
correspondence detection. Medical Image Analysis 3. Du S, Liu J, Zhang C, Zhu J, Li K (2015) Probability iterative closest point algorithm for m-d point
registration with noise. Neurocomputing 157:187–198 3. Du S, Liu J, Zhang C, Zhu J, Li K (2015) Probability ite
registration with noise. Neurocomputing 157:187–198 4. Fischler MA, Bolles RC (1981) Random sample consensus: a paradigm for model fitting with applications
to image analysis and automated cartography. Commun ACM 24(6):381–395 5. Florack L, Maas R, Niessen W (1999) Pseudo-linear scale-space theory. Int J Comput Vision 31
3):247–259 6. Förstner W, Gülch E (1987) A fast operator for detection and precise location of distinct points, corners
and centres of circular features. In: Proceeding ISPRS intercommission conference on fast processing of
photogrammetric data, pp. 281–305 123 1279 AIFD Based 2D Image Registration to Multi-View Stereo... 7. Gao Y, Ji R, Cui P, Dai Q, Hua G (2014) Hyperspectral image classification through bilayer graph-ba
learning. IEEE Trans Image Process 23(7):2769–2778 g
g
8. Gao Y, Ji R, Liu W, Dai Q, Hua G (2014) Weakly supervised visual dictionary learning by harnessing
image attributes. IEEE Trans Image Process 23(12):5400–5411 g
g
9. Gao Y, Wang M, Tao D, Ji R, Dai Q (2012) 3-d object retrieval and recognition with hypergraph analysis. IEEE Trans Image Process 21(9):4290–4303 g
g
9. Gao Y, Wang M, Tao D, Ji R, Dai Q (2012) 3-d
IEEE Trans Image Process 21(9):4290–4303 g
10. Gao Y, Wang M, Zha ZJ, Shen J, Li X, Wu X (2013) Visual-textual joint relevance learning for tag-based
social image search. IEEE Trans Image Process 22(1):363–376 11. Gonzalez RC, Woods RE, Eddins SL (2004) Digital image processing using matlab. Pearson Prentice
Hall, Upper Saddle River g
(
)
p
g g
( )
13. Lindeberg T (2013) Generalized axiomatic scale-space theory. Adv Imaging Electron Phys 178:1 14. Lowe DG (2004) Distinctive image features from scale-invariant keypoints. Int J Comput Vis 60(2):9
110 15. Mitrovic U, Špiclin Ž, Likar B, Pernuš F (2013) 3D–2D registration of cerebral angiograms: a method
and evaluation on clinical images. IEEE Trans. Med. Imaging 32(8):1550–1563 16. References Paschalakis S, Francini G (2014) Test model 12: compact descriptors for visual search. testmodel ISO/IEC
JTC1/SC29/WG11/N14961, MPEG, Strasbourg, France 16. Paschalakis S, Francini G (2014) Test model 12: compact d
JTC1/SC29/WG11/N14961, MPEG, Strasbourg, France g
17. Tuytelaars T, Mikolajczyk K (2008) Local invariant feature detectors: a survey. Found Trends Comput
Gr Vis 3(3):177–280 18. Zhao B, Lepsoy S, Magli E (2015) Affine scale space for viewpoint invariant keypoint detection. In:
Multimedia Signal Processing (MMSP), 2015 IEEE 17th International Workshop on. pp 1–6 123
|
https://openalex.org/W4311361634
|
https://jfootankleres.biomedcentral.com/counter/pdf/10.1186/s13047-022-00594-9
|
English
| null |
Experiences of using a digital tool, the D‐foot, in the screening of risk factors for diabetic foot ulcers
|
Journal of foot and ankle research
| 2,022
|
cc-by
| 8,802
|
Abstract Background: Individuals living with diabetes run an increased risk of developing diabetic foot ulcers (DFUs), lead-
ing to high costs to society and reduced quality of life for the individual. Regular screening is important to avoid
complications. Aim: To evaluate patients’ and clinicians’ experiences of using a digital tool, the D-Foot, in the screening of risk factors
for developing DFUs. The secondary aims were to investigate whether patients had had their feet examined by a
nurse or doctor during the past year, had been referred to podiatry and whether patients had received information
about self-care. Methods: A prospective study was carried out, comprising 90 patients with diabetes visiting a Department of Pros-
thetics and Orthotics (DPO). Two Certified Prosthetists and Orthotists (CPOs) were included, and they assessed foot
status and the risk of developing DFUs with the D-Foot software, prior to prescribing footwear. The quality of services
at the DPO was assessed by the patients using the Orthotics and Prosthetics Users’ Survey (OPUS). The CPOs answered
the System Usability Scale (SUS) before and after the study to assess the usability of the D-Foot. Results: No patient had risk grade 1. One (1%) patient had risk grade 2, 78 (87%) patients had risk grade 3 and 11
(12%) patients had risk grade 4. Patients reported high levels of satisfaction on eight of ten OPUS items and the two
items with lower scores were not related to the use of the D-Foot. The two CPOs reported levels above the mean
regarding usability both before (77.5 and 90) and after (70 and 97.5) using the D-Foot. Results: No patient had risk grade 1. One (1%) patient had risk grade 2, 78 (87%) patients had risk grade 3 and 11
(12%) patients had risk grade 4. Patients reported high levels of satisfaction on eight of ten OPUS items and the two
items with lower scores were not related to the use of the D-Foot. The two CPOs reported levels above the mean
regarding usability both before (77.5 and 90) and after (70 and 97.5) using the D-Foot. Conclusions: Patients expressed a high level of satisfaction with the services when their feet were examined with
the D-Foot prior to the provision of footwear. The CPOs found that the D-Foot system was usable. Several comments
were made by patients and CPOs and will support the future development and testing of the D-Foot. Zügner et al. Journal of Foot and Ankle Research (2022) 15:90
https://doi.org/10.1186/s13047-022-00594-9 Zügner et al. Journal of Foot and Ankle Research (2022) 15:90
https://doi.org/10.1186/s13047-022-00594-9 Open Access Abstract There is a need
to increase referrals for preventive podiatry and improve information on self-care for patients at risk of DFUs. Trial registration: ClinicalTrials gov ID: NCT04054804 etic foot, Foot ulcer, eHealth, Orthotics, Foot orthotics, Insole, Shoes, Implementation, diabetes, foot Keywords: Diabetic foot, Foot ulcer, eHealth, Orthotics, Foot orthotics, Insole, Shoes, Implementation, diabetes, foot
deformities © The Author(s) 2022. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which
permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the
original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or
other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line
to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory
regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this
licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativeco
mmons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Roland Zügner1,2* , Gustav Jarl3,4, Leif Sundberg5 and Ulla Hellstrand Tang1,6 Roland Zügner1,2* , Gustav Jarl3,4, Leif Sundberg5 and Ulla Hellstrand Tang1,6 Background Individuals living with diabetes run an increased risk of
developing complications such as diabetic foot ulcers
(DFUs) and this complication might lead to an amputa-
tion [1]. In 2021, the total number of patients at risk of
developing DFUs was estimated at 270 million globally *Correspondence: roland.zugner@gu.se 2 Forskningsenhet Ortopedi, Göteborgsvägen 31, SE‑431 80 Mölndal, Sweden
Full list of author information is available at the end of the article 2 Forskningsenhet Ortopedi, Göteborgsvägen 31, SE‑431 80 Mölndal, Sweden
Full list of author information is available at the end of the article 2 Forskningsenhet Ortopedi, Göteborgsvägen 31, SE‑431 80 Mölndal, Sweden
Full list of author information is available at the end of the article Study design and procedureh This prospective study was carried out in 2019 in a
cohort of patients with diabetes referred to the DPO at
Sahlgrenska University Hospital in Gothenburg, Sweden. The purpose of the visit to the DPO was to provide the
patient with pressure-relieving footwear, insoles together
with shoes, to counteract the occurrence of DFU. The
referred patients were contacted by phone and patients
interested in participating in the study were informed
by the principal investigator and received an invitation
letter to the DPO, Fig. 1. The letter included informa-
tion on how to answer the patient questionnaire in the
D-Foot, Additional file 1. The second way for the patients
to answer this questionnaire was at the DPO, where the
survey was visualised on a tablet (Samsung Galaxy Tab. A 10.1). The patients’ answers were registered before the
clinical examination began. If needed, the investigator
helped the patients to register their answers on the tab-
let. The clinical examinations were assessed by one of two
CPOs, at the DPO, following the routine in the D-Foot. The daily clinical work process in the prescription is as
follows: a) the CPO evaluated the patients’ need for pre-
scribed footwear, b) suggested and discussed a care plan
with the patient, c) gave self-care advice and d) provided
the patient with footwear as described in clinical guide-
lines [35]. The information, advice and provision of foot-
wear were based on the patient’s risk of developing DFUs
according to the D-Foot examination. Finally, the D-Foot
assessment summary was printed out as a PDF report
and given to the patient. The footwear was delivered
either on the first visit or on a second visit to the DPO,
two to 8 weeks later, Additional file 9.h The opportunities for introducing new clinical exami-
nation methods, such as a digital tool, need to consider
how the healthcare system is organised and the laws, reg-
ulations and guidelines when examining and prescribing
assistive devices to patients with diabetes at risk of devel-
oping DFUs [20]. © The Author(s) 2022. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which
permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the
original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or
other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line
to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory
regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this
licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativeco
mmons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Page 2 of 10 Zügner et al. Journal of Foot and Ankle Research (2022) 15:90 on the users’ experiences to develop and implement
improvements are necessary [32, 33].h [2] and 250,000 in Sweden [3–5], given that 50% of the
patients have peripheral neuropathy [6, 7]. DFUs are
associated with poor quality of life [8] and high costs;
the annual cost/DFU has been estimated at 19,000 US$
(around 110,000 SEK) [9–11]. The prevention of DFUs
is essential to preserve the quality of life of patients and
reduce healthcare costs to society [12–14]. The preven-
tion of DFUs should include: 1) regular screening for
risk factors for developing DFUs, 2) intervention with
prescribed footwear (shoes and/or insoles), 3) podiatry
and 4) patient self-care education [15–18]. As stated by
the World Health Organisation, therapeutic footwear
is necessary for persons with diabetes with peripheral
neuropathy and risk factors for developing DFU. Pre-
scribed footwear is one of 50 assistive devices that should
be available for those in need, as recommended by the
World Health Organisation in the Priority Assistive
Products List [19].h The primary aim of the study was to evaluate the users’
experiences of using the D-Foot in the screening of risk
factors for developing DFUs. The secondary aims were to
investigate whether patients had had their feet examined
by a nurse or doctor during the past year, had received
podiatry and whether the patient had been given infor-
mation about self-care. Study sample – CPOs
d Two experienced CPOs at the DPO at Sahlgrenska Uni-
versity Hospital agreed to participate in the study. They
were skilled in prescribing footwear to patients at risk of
developing DFUs with assistive devices and had worked
as CPOs for nine and 20 years respectively. Furthermore, in the current version, the order of the
assessment was reorganised to fit more effectively into a
more clinical flow with the patient being examined first
in a sitting position, followed by standing assessments. Before the start of the study, each CPO completed a
D-Foot introductory course which lasted for 2 h on two
separate occasions. Moreover, the CPOs participated in
a workshop with the purpose of becoming familiar with
the D-Foot examination routine. Based on the patients’ answers in the survey and the
assessment made by the CPO, a risk classification (1–4)
was generated according to the national risk classifica-
tion system [26, 38], Additional file 3. In this risk clas-
sification, the symptoms that are related to each of the
risk categories are presented: peripheral neuropathy/
angiopathy, foot deformities, skin pathologies, previ-
ous DFU/amputation, active DFU/Charcot deformity
and/or severe pain syndrome. The risk classification,
together with the recommended interventions, such as
podiatry, footwear, and regular foot examinations, was
automatically displayed on the screen of the tablet and
was also included in the PDF report that was given to
the patient. Finally, the CPO copied the results from the
D-Foot and pasted them into the patient’s EMR and, in
addition, all the data were stored in a separate D-Foot
database. The reliability and validity of the D-Foot Study design and procedureh 1 A visualisation of the different steps in the study. Note: OPUS; Orthotics and Prosthetics Users’ Survey (OPUS) [34] has been presented in a previous study by Hellstrand
Tang et al. [25]. Based on the results from the previ-
ous study, the users (eight CPOs) [25] suggested refine-
ments as compared with the first version. The following
improvements were therefore made: 1) adding a test of
signs of peripheral angiopathy with the question “Has a
healthcare professional confirmed that you have periph-
eral angiopathy?”, 2) excluding the navicular drop test
[37], 3) excluding the test entitled “Can you extend and
flex you toes”, 4) excluding the question of whether the
foot had areas of excessive pressure with calluses, 5)
excluding the assessment “Gait deviation, affected from
hip/knee joint” and 6) splitting the assessment of Char-
cot foot (acute Charcot foot and manifest deformed
Charcot foot) [25]. (patients and CPOs) comments as they used the D-Foot
at the DPO, Fig. 1. The method used was the think-aloud
method, a method frequently used in usability testing
where the spontaneous comments from the users are
registered, in this case the patients and CPOs [36]. The
registration was made by the principal investigator. The CPOs’ expectations and experiences of using the
D-Foot were evaluated using the System Usability Scale
(SUS), before and after the study period. Study design and procedureh Some of the laws that form care in Swe-
den are: a) the patient law that states that care should be
patient centred and that healthcare professionals should
have a dialogue with the patient regarding examination,
treatment and a care plan [21] and b) the law relating to
the need to consider the integrity of the patient when
data are registered in the medical record systems [22, 23]. These laws aim to promote the patient’s integrity, self-
determination and participation and must be followed by
all healthcare professionals [22, 23]. When implementing
new eHealth systems, the European regulations regarding
the development and maintenance of a medical product
need to be followed [24]. Starting in 2010, the D-Foot was developed by health-
care professionals, patients and researchers in Sweden
with the aim of: 1) facilitating foot screening, 2) gener-
ating an objective risk stratification (1 = no risk, 2 = low
risk, 3 = medium risk and 4 = ongoing DFU) and 3) indi-
vidualising care and advice regarding self-care with the
aim of promoting good foot health and preventing DFUs. A detailed description of the D-Foot is presented in the
method section. The D-Foot was designed to be used pri-
marily by clinicians working at DPOs [25] and secondar-
ily by other healthcare professionals to improve the care
of patients at risk of developing DFUs [26, 27]. i
The CPOs documented the assessments, tests, care
plan decisions and the information they gave to patients
in the local EMR system, Pilot, used at the DPO situated
in Region Västra Götaland, Sweden.h The patients’ experiences of the visit to the DPO and
the D-Foot examination were evaluated using the Orthot-
ics and Prosthetics Users’ Survey (OPUS) [34]. The sur-
vey was filled in after the patients had been provided
with their prescribed footwear. The users’ experiences of
using the D-Foot were explored by registering the users’ It has been suggested that digital tools, such as the
D-Foot, improve DFU management in care [28–31] and
stored data can be useful when auditing the care of DFU
patients to support clinical improvements, as presented
by Leese et al. [30]. Continuous improvements based Zügner et al. Journal of Foot and Ankle Research (2022) 15:90 Page 3 of 10 Fig. 1 A visualisation of the different steps in the study. Note: OPUS; Orthotics and Prosthetics Users’ Survey (OPUS) [34] ig. software have previously been presented by Hellstrand
et al. [25]. footwear (< 6 min., 6–10 min., 11–15 min., 16–20 min.),
Additional file 8. Study sample – patients Patients on the waiting list, n = 107, at the DPO, Sahlg-
renska University Hospital, with a referral to prevent foot
complications in diabetes, were invited to participate in
the study. The inclusion criteria were being diagnosed
with diabetes, age ≥ 18 years and understanding the
Swedish language. Equipmenth Descriptive statistics were used to present patient demo-
graphics. The mean and standard deviation were calcu-
lated for continuous variables. Differences between the
included patients and the non-included patients were
compared with an independent t-test for age and Pear-
son’s chi-square test for gender. IBM SPSS Statistics for
PC, Version 25, was used for all statistical calculations. The CPOs recorded their findings on a laptop. Patients
answered the patient questionnaire using a smart phone,
tablet, or personal computer. A goniometer was used to
measure passive maximum dorsal flexion at the metatar-
sal phalangeal joint and passive dorsiflexion at the ankle
joint, as previously described by Hellstrand Tang et al. [25]. A foot calliper was used to measure foot length and
foot width, while toe height was measured with a ruler
[25], Additional file 4. Questionnaire – patientsh The patients answered the Swedish version of the Client
Satisfaction with Services module of the OPUS after they
had received their footwear. The OPUS module assesses
patients’ experiences of the service quality and has been
validated with Swedish patients at DPOs [34]. The 10
items are answered on a three-level Likert scale (disa-
gree = 1, agree = 2 and strongly agree = 3) and were ana-
lysed by calculating the mean and standard deviation for
each question. The OPUS questionnaire is recommended
by the Swedish Orthotic and Prosthetic industry advisory
council to evaluate patient satisfaction with the services
at DPOs located in Sweden [39]. The studied group finally consisted of 90 patients, 53
men and 37 women, with a mean age of 68 ± 13 years. The patients who were not included (n = 17), nine men
and eight women, had a mean age of 64 ± 15 years and
did not differ significantly from the included participants
according to gender (p = 0.65) and age (p = 0.27). Of the
entire group, 57 patients had retired and 17 were work-
ing, Table 1. The D‑foot softwareh The D-Foot software has seven functions: 1) a booking
system to schedule patients’ appointments, 2) a ques-
tionnaire to be answered by patients, 3) examinations
to be completed by a healthcare professional, Addi-
tional files 2, 4) a system for generating a risk classifica-
tion for each patient, Additional files 3, 5) a summary
as a PDF report, 6) a database containing all registered
information and 7) an administration part licensing
the CPOs to obtain access to the program [25]. A brief
description of some different components included in
the D-Foot is also included in Additional file 4 and a
summary in Additional file 10. A detailed description Page 4 of 10 Zügner et al. Journal of Foot and Ankle Research (2022) 15:90 footwear (< 6 min., 6–10 min., 11–15 min., 16–20 min.),
Additional file 8. footwear (< 6 min., 6–10 min., 11–15 min., 16–20 min.),
Additional file 8. software have previously been presented by Hellstrand
et al. [25]. Questionnaire – CPOs In connection with the start of the study, a Swedish ver-
sion of the System Usability Scale (SUS) was answered
by the CPOs to capture their expectations of using the
D-Foot in clinical work, Additional file 5. The SUS con-
sists of 10 questions answered on a five-level Likert scale
(from strongly disagree = 1, to agree completely = 5) and
is a reliable tool for measuring usability [40]. At the end
of the study, the CPOs answered the SUS a second time
to assess their experiences of using the D-Foot, Addi-
tional file 6. These answers were transformed to a 0–100
scale, where a higher score indicates a higher level of
usability. Table 1 Descriptive data of included patients
Categorical and continuous variables
n (%)
Mean (SD)
Women
37 (41)
Type 1 diabetes
33 (37)
Type 2 diabetes
55 (61)
Diabetes, other types
2 (2)
Duration, yrs
21 (16)
Age, yrs
68 (13)
HbA1c, mmol/mol, self-reported
62 (19)
Height, m
1.72 (0.9)
Weight, kg
83 (20)
BMI, kg/m2
28 (6)
Medication for high blood pressure
61 (68)
Medication for heart disease
45 (50)
Smoker
6 (7)
Use snuff
9 (10)
Occupation
Working
17 (19)
Students
0 (0)
Retired
57 (63)
Other
6 (7)
Missing value
10 (11) Table 1 Descriptive data of included patients Table 1 Descriptive data of included patients In addition, at the end of study, the CPOs answered a
study-specific questionnaire on how they had experi-
enced the digital foot examination, what information
they had given to the patients and how long it took for
the CPOs to execute different sequences of the exami-
nation, Additional file 7. The questions on time esti-
mations were related to how long it took to: 1) finalise
the patient appointment and foot assessment when
the D-Foot routine was used (< 30 min., 31–45 min.,
46–60 min., > 60 min), 2) make the medical recording
based on the D-Foot examination (< 6 min., 6–10 min.,
11–15 min., 16–20 min.) and 3) order their prescribed Page 5 of 10 Zügner et al. Journal of Foot and Ankle Research (2022) 15:90 The examination showed that no patient had risk level
1, one (1%) patient had risk level 2, 78 (87%) patients had
risk level 3 and 11 (12%) patients had risk level 4. At the
examination, the patients were asked about the interven-
tions to prevent DFU they had received. Orthotics and prosthetics users’ survey‑OPUS p
y
Eighty-two (91%) of 90 patients answered OPUS. The
mean value of the questions ranged from 2.3 to 2.9,
where 2 = agree and 3 = strongly agree, Table 2. Satisfac-
tion with services was highest for the questions related
to how the CPOs showed courtesy and respect, how the
patients were informed about the choice of footwear and
the opportunity to express their own concerns. Satisfac-
tion was lowest for the waiting time for the appointment
and the co-ordination of services. Other suggested improvements, coming from both
patients and CPOs, were to design the PDF report in
an easy-to-understand format with larger text, linking
the risk grade to each of the specific risk factors and, in
addition, creating customised information and advice
about self-care of the feet for patients. Requests were
also made to be able to post the report in digital format
in addition to the printable version. The individual risk
factors should be linked to the recommended treatment
(e.g. recommendation to use a high toe box in the pres-
ence of hammer toes). Moreover, digital referrals to other
healthcare professionals, based on the results, should be
included in upcoming versions. System usability scale One CPO had an SUS score of 77.5 before using the
D-Foot and 70.0 after the study was completed. The
second CPO had an SUS score of 90.0 before using the
D-Foot and 97.5 after the study was completed, Table 3. The CPOs requested integration with co-working sys-
tems, such as the population register, the EMR used by
other healthcare professionals and the EMR used at the
DPO. Loss of wi-fi, breakdown of the EMR at the DPO
and problems registering findings in the D-Foot system
were registered. Eleven patients were not booked in the
D-Foot system and these patients were therefore unable
to answer the patients’ questions at home, prior to the
visit. Questionnaire – CPOs Seventy-eight
(87%) patients had had their feet examined by a nurse or
doctor during the past year, 59 (66%) answered that they
had received podiatry and 59 patients (66%) had been
given information about the self-care of their feet. you had a foot ulcer? and “Are your feet less sweaty
now compared with recent years?” difficult to answer. Patients suggested specifying the question of “Have
you previously had hard-to-heal ulcers?” and removing
the question about sweaty feet, because some patients
said that they had never had sweaty feet. Further, they
suggested expanding the question on the sensation of
tingling or numbness to include sensations of cramp,
walking on pillows and feelings such as “it feels like a
band around my foot”. The question about the presence
of pain should be more specific, such as “How much
pain have you experienced in your feet during the past
week”. Comments and feedback from patients and CPOs Comments and feedback from patients and CPOs
In total, 147 comments were registered, (n = 50 from
patients and n = 97 from CPOs), Table 4. Overall, 55
comments were related to improvements (n = 26 from
patients and n = 29 from CPOs). Patients suggested
refinements to the patient questionnaire regarding the
perceptions of neuropathy; pain, previous ulcers, sweat
and HbA1c. Patients found questions such as “Have Table 2 Results from the orthotic and prosthetics users’ survey (n = 82) Table 2 Results from the orthotic and prosthetics users’ survey (n = 82) Table 2 Results from the orthotic and prosthetics users’ survey (n = 82)
Note: Rating scale: disagree =1, agree = 2 and strongly agree = 3
Item
Mean (SD)
Not
applicable,
n (%)
Missing, n (%)
I received an appointment with a prosthetist/orthotist within a reasonable amount of time. 2.3 (0.8)
0
0
I was shown the proper level of courtesy and respect by the staff. 2.9 (0.3)
0
0
I waited a reasonable amount of time to be seen. 2.8 (0.4)
0
1 (1)
Clinic staff fully informed me about equipment choices. 2.9 (0.3)
7 (9)
0
The prosthetist/orthotist gave me the opportunity to express my concerns regarding my equipment. 2.9 (0.4)
1 (1)
0
The prosthetist/orthotist was responsive to my concerns and questions. 2.8 (0.4)
7 (9)
0
I am satisfied with the training I received in the use and maintenance of my prosthesis/orthosis. 2.7 (0.5)
19 (23)
0
The prosthetist/orthotist discussed problems I might encounter with my equipment. 2.7 (0.5)
14 (17)
0
The staff co-ordinated their services with my therapists and doctors. 2.3 (0.6)
39 (48)
0
I was a partner in decision-making with clinic staff regarding my care and equipment. 2.8 (0.4)
3 (4)
1 (1) Zügner et al. Journal of Foot and Ankle Research (2022) 15:90 Page 6 of 10 Table 3 Results from the System Usability Scale (SUS) answered by two CPOs
Questions, before the study start The Swedish response format was: 1) Håller absolut inte med, 2) Håller
inte med, 3) Håller varken med eller inte med, 4) Håller med and 5) Håller fullkomligt med Table 4 Summarised comments (n = 51) from the patients and the CPOs registered as they used the D-Foot
Note: DPO, departments of prosthetics and orthotics
CPO Certified Prosthetist and Orthotist
Category of comments
Frequency comments
Patients (n = 50)
CPO (n = 97)
D-Foot improvements
26
29
Medical record system not functioning
1
14
Patient was not booked in D-Foot alongside with the booking in the electronic medical
recording system at the DPO
6
11
New way of working
1
5
New way to access care
7
1
No access to patient survey from home
2
4
Patients identity and integrity
0
7
Problem with the tablet
0
3
Difficult to change booking in the D-Foot
0
2
Services at DPO
4
1
Wi-Fi not functioning
0
2
Technical equipment
0
4
Technical error with CPOs’ laptop
0
2
Technical problems
0
2
Double-booked in the D-Foot
0
2
Patient not familiar with the Swedish language
2
1
Bad ergonomics to work with a small laptop
0
1
A complicated booking system in the D-Foot
0
4
Problems with the D-Foot web program
1
0
Problems with the study setting
0
2 sed comments (n = 51) from the patients and the CPOs registered as they used the D-Foot Table 3 Results from the System Usability Scale (SUS) answered by two CPOs
Questions, before the study start The sum was based on the converted numbers
multiplied by 2.5 and a summary score for all 10 questions, ranging from 0 to 100, was obtained [40]. The 10 questions in the SUS have been translated into Swedish, from the original English text. The Swedish response format was: 1) Håller absolut inte med, 2) Håller The original data are presented in the table. Thereafter, the total SUS score was calculated as follows: each question was answered with a score (strongly disagree = 1
to strongly agree = 5). From the answers with an odd number, “one” was subtracted. Thereafter, from the answers with an even number, the number was subtracted
from “five”. The value obtained per person and question ranged from 0 to 4, with four as the highest response. The sum was based on the converted numbers
multiplied by 2.5 and a summary score for all 10 questions, ranging from 0 to 100, was obtained [40]. The 10 questions in the SUS have been translated into Swedish, from the original English text. Table 3 Results from the System Usability Scale (SUS) answered by two CPOs
Questions, before the study start Table 3 Results from the System Usability Scale (SUS) answered by two CPOs
Note: The questions that are presented in this table were answered by the CPOs prior to study start The questions in the SUS that were answered after the study was
Questions, before the study start
CPO 1 before (after)
CPO 2
before
(after)
1. I think that I would like to make a digital assessment. 4 (4)
5 (5)
2. I think I will find that a digital foot assessment is unnecessarily complex. 2 (1)
1 (1)
3. I think I will find that a digital foot assessment is easy to perform. 4 (4)
5 (5)
4. I think that I will need the support of a technical person to be able to make a digital assessment. 4 (3)
5 (2)
5. I think I will find that the various functions for performing digital assessments are well integrated. 4 (3)
5 (5)
6. I think I will find that there is too much inconsistency in the digital assessment. 2 (2)
1 (1)
7. I think that most people will quickly learn how to make a digital assessment. 4 (4)
4 (5)
8. I think I will find making digital assessments very complicated. 2 (2)
1 (1)
9. I think I will feel very confident about making a digital assessment. 4 (4)
5 (5)
10. I think I will need to learn a lot of things before I can make a digital assessment. 2 (3)
4 (1)
Total SUS score per observer
77.5
(70.0)
90.0
(97.5) Total SUS score per observer e presented in this table were answered by the CPOs prior to study start. The questions in the SUS that were answered after the study wa
n Hellstrand et al. 2017 [31] Low scores to the questions with an even number should be interpreted as positive, as the user somewhat disagrees (strongly disagree = 1 to strongly agree = 5)
The original data are presented in the table. Thereafter, the total SUS score was calculated as follows: each question was answered with a score (strongly disagree = 1
to strongly agree = 5). From the answers with an odd number, “one” was subtracted. Thereafter, from the answers with an even number, the number was subtracted
from “five”. The value obtained per person and question ranged from 0 to 4, with four as the highest response. CPOs’ experiences of the digital foot examination
and information to the patienth The CPOs felt that they had provided “complete” infor-
mation regarding a) how patients should perform
self-care for their feet, b) where patients could obtain
additional information, c) potential risks when using
the assistive device and d) warning signs relating to
health and the assistive device. Regarding the ques-
tions: a) whether the examination took place in privacy,
b) information on the D-Foot results and c) whether the
patient had received oral and written information about
the assistive device, they were judged to have been per-
formed “completely”. Two questions were judged to be
less well fulfilled by one of the CPOs: a) the discussion
of the health condition and b) whether the results of the
D-Foot survey were explained in a way the patient under-
stood, Table 5.hf The usability, according to the SUS, was generally high
both before and after using the D-Foot, as scores above
68 can be regarded as “above average” [40, 41]. For one
CPO, the SUS score improved by 7.5 points after hav-
ing used the D-Foot, indicating that this CPO found the
D-Foot’s usability better than he had expected prior to
using it. For the other CPO, the SUS score deteriorated
by 7.5 points after having used the D-Foot. This is most
likely explained by technical problems with the wi-fi sys-
tem. However, on individual items, most ratings were
similar before and after using the D-Foot. Only two rat-
ings by one of the CPOs changed by more than one step
on the rating scale. One expectation changed in both the
positive and negative direction after the CPOs had used
the D-Foot. One CPO responded more positively (from
5 = strongly agree to 2 = do not agree) regarding “I think
that I will need the support of a technical person to be
able make a digital assessment”, indicating that, by using
the D-Foot, the need for technical support decreased. Several other questions were scored positively both prior
to and after use, such as the D-Foot was easy to use, it
was quick to learn, and the user felt confident about
using the D-Foot. The estimated time required for different sequences
when meeting the patients was assessed equally by the
two CPOs. The time required for the patient appoint-
ment was 40–60 minutes. Writing the medical record
took six to 10 minutes and ordering shoes and materials
took six to 10 minutes. Category of comments Page 7 of 10 Zügner et al. Journal of Foot and Ankle Research (2022) 15:90 Of the 90 patients, 18 (20%) answered the patient sur-
vey prior to the visit to the DPO and 72 (80%) answered
the survey on a tablet at the visit. More than 50% (n = 44)
were helped at the DPO to answer the survey. patients reported that they had not received podiatry
or received information about foot self-care, which is
concerning.h The patients generally expressed a high level of satisfac-
tion with the services at the DPO when using the D-Foot
for the foot examination. It is worth noting that not all
the OPUS items relate to the foot examination and the
two items with the lowest scores (waiting times and co-
ordination of services) relate to the items that were not
related to the foot examination and the use of the D-Foot.h CPOs’ experiences of the digital foot examination
and information to the patienth The patient’s visit was finalised
within a total of 52–80 minutes. Discussionhi This is the first implementation study of a digital screen-
ing tool, the D-Foot, and we found that patients were
generally satisfied with the services when being routinely
examined using the D-Foot. The CPOs found the D-Foot
software usable. It is worth noting that one third of the A variety of comments were collected from the CPOs
and these comments from users are useful in the con-
tinuous improvement of digital tools such as the D-Foot Table 5 Answers from the certified prosthetists and orthotists
Questions
1. Not at all
2. 3. 4. Yes,
completely
5. Not
applicable
Did the patient receive enough information about how to make self-care of the feet? 2
Was the illness/health condition discussed? 1
1
Did you leave information about where to go if the patient needed help or had additional questions after
the visit? 2
Did the patient receive enough information about possible risks with using the assistive device? 2
Did the patient receive enough information about warning signs to be aware of regarding his/her illness/
health condition or your assistive device? 2
Was there enough privacy when you and the patient discussed his/her condition or treatment? 2
Did you explain the results of the D-Foot survey in a way that the patient understood? 1
1
Did you give the patient oral user information about the assistive devices? 2
Did you give the patient written information about the assistive devices? 2 Zügner et al. Journal of Foot and Ankle Research (2022) 15:90 Zügner et al. Journal of Foot and Ankle Research (2022) 15:90 Page 8 of 10 software [32]. In future versions, the authors suggest that
improvements should be made. For example, to specify
the question regarding pain as “How much pain as a
mean value did you perceive last week in your right foot”,
followed by a Likert scale. Furthermore, a more easy-
to-understand report, with custom-made information
regarding footwear and self-care, would make it easier for
persons with diabetes to perform self-care based on their
individual risk factors. The report should be available in
both a printed and a digital version, with the option of
being sent to the patient. One limitation of the study was that no control group
of patients who received traditional examination of the
feet was used. Supplementary Information Unstable wi-fi is a concern, as presented in this study. In addition to developing functioning software, the
D-Foot, the DPOs need to assure that wi-fi is in place. The online version contains supplementary material available at https://doi.
org/10.1186/s13047-022-00594-9. The online version contains supplementary material available at https://doi.
org/10.1186/s13047-022-00594-9. Additional file 1. Patient survey included and programmed in the
D-Foot. Additional file 2. Examples of different steps in the D-Foot examination. Additional file 3. Prevention and multidisciplinary service (MDS) of foot
complications in diabetes. Additional file 4. Additional file 5. System Usability Scale answered prior to the study. Additional file 6. System Usability Scale answered after the study. Additional file 7. Questions related to the information the certified pros-
thetist and orthotist gave to the patients and questions related to integrity
and time to finalise the visit. Additional file 8. Additional file 9. Additional file 10. Additional file 1. Patient survey included and programmed in the
D-Foot. Additional file 2. Examples of different steps in the D-Foot examination. Additional file 3. Prevention and multidisciplinary service (MDS) of foot
complications in diabetes. Additional file 4. Additional file 5. System Usability Scale answered prior to the study. Additional file 6. System Usability Scale answered after the study. Additional file 7. Questions related to the information the certified pros-
thetist and orthotist gave to the patients and questions related to integrity
and time to finalise the visit. Additional file 8. Additional file 9. Additional file 10. i
Secondly, the use of the stored D-Foot is useful for
audit and follow-up at local, regional and national level
and fills a gap identified in national guidelines [20]. Additional file 1. Patient survey included and programmed in the
D-Foot. Additional file 2. Examples of different steps in the D-Foot examination. Additional file 3. Prevention and multidisciplinary service (MDS) of foot
complications in diabetes. Additional file 4. Additional file 5. System Usability Scale answered prior to the study. Additional file 6. System Usability Scale answered after the study. Additional file 7. Questions related to the information the certified pros-
thetist and orthotist gave to the patients and questions related to integrity
and time to finalise the visit. Additional file 8. Additional file 9. Additional file 10. Additional file 1. Patient survey included and programmed in the
D-Foot. Additional file 1. Patient survey included and programmed in the
D-Foot. Abbreviations DFUs: Diabetic foot ulcers; DPO: Department of Prosthetics and Orthotics;
CPOs: Certified Prosthetists and Orthotists; OPUS: Orthotics and Prosthetics
Users’ Survey; SUS: System Usability Scale; EMR: Electric medical recording;
US$: US dollar (USD); SEK: Swedish krona. Discussionhi As a result, it is not possible to determine
the extent to which the patients’ high level of satisfac-
tion with the services was due to the use of the D-Foot
or to the general appreciation of the services at the DPO. Only two CPOs were included, limiting the interpreta-
tions of the CPOs’ answers regarding the SUS. A larger
sample is recommended in future studies. The two CPOs
only answered a study-specific questionnaire at the end
of the study. This evaluation, Appendix 4, should prefer-
ably be carried out after each D-Foot assessment made
for each of the participants. This could not be carried out
after each patient, due to the daily CPO-scheduled work. Finally, the study was conducted at a single DPO, which
limits the generalisability of the results, and future stud-
ies should include CPOs at different locations. Continuous improvements, based on the needs of dif-
ferent healthcare users, are a necessity, as described by
Koltveit et al., when implementing digital tools in the
care of DFUs [33]. The authors suggest that digital tools,
such as the D-Foot, are useful when implementing seam-
less and person-centred care for persons living with dia-
betes with a risk of foot complications, as recommended
by the Swedish Association of Local Authorities and
Regions [42]. Conclusion Patients with diabetes expressed a high level of satisfac-
tion with the services when their feet were examined fol-
lowing the routine in the D-Foot software prior to the
provision of footwear. The CPOs felt that the D-Foot sys-
tem was usable. Several comments were made by patients
and CPOs and will support the future development of
and improvements to the D-Foot software. The current
study revealed that not all patients in need had access to
preventive interventions such as podiatry and informa-
tion about self-care. Several improvements were suggested by the patients
and CPOs, such as clarifying and differentiating the
patients’ questions and facilitating the booking system by
connecting the D-Foot to the population register. Patient
security and integrity will definitely be improved by inte-
gration with the population register. This would ensure
that the right patient with the right identity is booked
in the D-Foot system. Integration with the EMR used by
other healthcare professionals would lead to data being
available where the patient is treated: in primary care, in
specialist care or at the DPO. Moreover, with integration
between systems, digital referrals and responses to refer-
rals would improve communication, thereby improving
the quality of care. To summarise, the D-Foot is useful
for CPOs and would improve the quality of care without
requiring more time for each visit. Consent for publication
Not applicable. Not applicable. 18. Bus SA, Lavery LA, Monteiro-Soares M, Rasmussen A, Raspovic A, Sacco
ICN, et al. Guidelines on the prevention of foot ulcers in persons with
diabetes (IWGDF 2019 update). Diabetes Metab Res Rev. 2020;36(Suppl
1):e3269. Declarations 13. Ragnarson Tennvall G, Apelqvist J. Health-related quality of life in
patients with diabetes mellitus and foot ulcers. J Diabetes Complicat. 2000;14:235–41. 13. Ragnarson Tennvall G, Apelqvist J. Health-related quality of life in
patients with diabetes mellitus and foot ulcers. J Diabetes Complicat. 2000;14:235–41. Funding Open access funding provided by University of Gothenburg. This research
was supported by Familjen Knut & Ragnvi Jacobssons stiftelse, Innovation-
splattformen in the Västra Götaland Region, Stiftelsen Promobilia, Stiftelsen
Skobranschens Utvecklingsfond, the Research and Development Council
of the County of Göteborg and Södra Bohuslän, the Health & Medical Care
Committee of the Västra Götaland Region, Stiftelsen Felix Neubergh, Stiftelsen
Gunnar Holmgrens Minne, IngaBritt & Arne Lundbergs Forskningsstiftelse,
Adlerbertska forskningsstiftelsen, Diabetesfonden, the Gothenburg Diabetes
Association (Inger Hultman med fleras fond and Utvecklingsfonden), Sveriges
Ortopedingenjörers Förening, Greta och Einar Askers Stiftelse, Hans Dahlbergs
stiftelse för miljö och hälsa and Vinnova. 8. Khunkaew S, Fernandez R, Sim J. Health-related quality of life among
adults living with diabetic foot ulcers: a meta-analysis. Qual Life Res. 2019;28:1413–27. 9. Hellstrand S, Sundberg L, Karlsson J, Zugner R, Tranberg R, Hellstrand
Tang U. Measuring sustainability in healthcare: an analysis of two
systems providing insoles to patients with diabetes. Environ Dev Sustain. 2020;23:1–15. 10. Ragnarson Tennvall G, Apelqvist J. Prevention of diabetes-related foot
ulcers and amputations: a cost-utility analysis based on Markov model
simulations. Diabetologia. 2001;44:2077–87. 10. Ragnarson Tennvall G, Apelqvist J. Prevention of diabetes-related foot
ulcers and amputations: a cost-utility analysis based on Markov model
simulations. Diabetologia. 2001;44:2077–87. Availability of data and materials 11. Ragnarson Tennvall G, Apelqvist J, Eneroth M. Costs of deep foot
infections in patients with diabetes mellitus. Pharmacoeconomics. 2000;18:225–38. 11. Ragnarson Tennvall G, Apelqvist J, Eneroth M. Costs of deep foot
infections in patients with diabetes mellitus. Pharmacoeconomics. 2000;18:225–38. The datasets used and/or analysed during the current study are available from
the corresponding author in response to a reasonable request, if the appropri-
ate permits are obtained from the correct authorities. 12. Ribu L, Hanestad BR, Moum T, Birkeland K, Rustoen T. A comparison of the
health-related quality of life in patients with diabetic foot ulcers, with a
diabetes group and a nondiabetes group from the general population. Qual Life Res. 2007;16:179–89. Supplementary Information ii
A default value of “no risk factor” was suggested by one
CPO as a way of facilitating the CPOs’ assessment. If a
risk factor is present, the CPO registers this specific risk
factor. Patient satisfaction with the service at the DPO
revealed relatively high scores according to the results
from OPUS. Two questions were answered as “not appli-
cable” and the reason might be that patients do not
regard footwear (insoles and shoes) as assistive devices. This semantic question needs to be clarified and specified
as shoes and insoles, to avoid misunderstanding. Additional file 8. Additional file 9. Additional file 10. Page 9 of 10 Page 9 of 10 Zügner et al. Journal of Foot and Ankle Research (2022) 15:90 Competing interests GJ is a consultant for Novo Nordisk. The other authors declare no competing
interests. 19. Priority Assistive Products List. https://www.adcet.edu.au/resource/9364/
who-priority-assistive-products-list#:~:text=The%20World%20Health%
20Organisation%20%28WHO%29%20Priority%20Assistive%20Products
,to%20live%20a%20healthy%2C%20productive%20and%20dignified%
20life. Acknowledgements 3. Hellstrand S, Hellstrand Tang U. A new approach to quantifying the sus-
tainability effects of healthcare: applied to the diabetic foot. Foot Ankle
Online J. 2019;12(3):5. We are most grateful for all the support from the patients and staff at the
Department of Prosthetics and Orthotics at Sahlgrenska University Hospital. Statistician Christer Johansson contributed his skills, experience and knowl-
edge and the group are most thankful for all support. 4. Hellstrand Tang U. The diabetic foot - assessment and assistive devices. Diss (sammanfattning) Göteborg: Göteborgs universitet; 2017. Hellstrand Tang U. The diabetic foot - assessment and assistive devi 5. The Swedish National Diabetes Register: Annual report 2021. 2021. Avail-
able at: https://www.ndr.nu/#/arsrapport. Accessed 12 Dec 2022. 24. EU: Medical device regulation 2017/745. 2017. 25. Hellstrand Tang U, Tranberg R, Zügner R, Karlsson J, Lisovskaja V, Siegent-
haler J, et al. The D-foot, for prosthetists and orthotists, a new eHealth
tool useful in useful in risk classification and foot assessment in diabetes. Foot and Ankle Online J. 2017;10(2):4. Authors’ contributions RZ and UHT conceived this study. UHT undertook the statistical analysis. RZ
and UHT interpreted the data. RZ, UHT, GJ and LS developed the manuscript. RZ, UHT, LS and GJ critically revised the manuscript and approved the final
manuscript for publication. 6. Kärvestedt L, Mårtensson E, Grill V, Elofsson S, Von Wendt G, Hamsten A,
et al. The prevalence of peripheral neuropathy in a population-based
study of patients with type 2 diabetes in Sweden. J Diabetes Complicat. 2011;25:97–106. 7. Lu B, Yang Z, Wang M, Yang Z, Gong W, Yang Y, et al. High prevalence
of diabetic neuropathy in population-based patients diagnosed with
type 2 diabetes in the Shanghai downtown. Diabetes Res Clin Pract. 2010;88:289–94. 1.
Edmonds M, Manu C, Vas P. The current burden of diabetic foot disease.
J Clin Orthop Trauma. 2021;17:88–93. Ethical approval and consent to participate The study was approved by the ethical review board in Gothenburg, reference
number: 2019–01039. All the patients and the CPOs received oral and written
information about the study and their right to withdraw from the study at any
time without explanation. The patients and CPOs signed a written consent
after they had received verbal and written information. The study was carried
out in accordance with the Code of Ethics of the World Medical Association,
the Declaration of Helsinki, for experiments involving humans [43]. 14. International Diabetes Federation: Clinical practice recommendations on
the diabetic foot – 2017. 2017. 15. Handbook for prescribing personal aids. https://www.vgregion.se/halsa-
och-vard/vardgivarwebben/vardriktlinjer/hjalpmedel/handbok-for-forsk
rivig-av-personliga-hjalpmedel/. 16. IWGDF Guidelines. https://iwgdfguidelines.org/guidelines/. 17. Bus SA, Armstrong DG, Gooday C, Jarl G, Caravaggi C, Viswanathan V, et al
Guidelines on offloading foot ulcers in persons with diabetes (IWGDF
2019 update). Diabetes Metab Res Rev. 2020;36:e3274. Author details 1 Department of Orthopedics, Institute of Clinical Sciences, Sahlgrenska Acad-
emy, University of Gothenburg, Sahlgrenska University, SE‑413 45 Göteborg,
Sweden. 2 Forskningsenhet Ortopedi, Göteborgsvägen 31, SE‑431 80 Mölndal,
Sweden. 3 Department of Prosthetics and Orthotics, Faculty of Medicine
and Health, Örebro University, Örebro, Sweden. 4 University Health Care
Research Centre, Faculty of Medicine and Health, Örebro University, Örebro,
Sweden. 5 Gothenburg Diabetes Association, Gothenburg, Sweden. 6 Depart-
ment of Prosthetics and Orthotics, Sahlgrenska University Hospital, Gothen-
burg, Sweden. 20. Sweden’s Municipalities and Regions: Person-centred and coherent care
course Diabetes with a high risk of foot ulcers. 2022. 21. Socialdepartementet. Patient law (SFS 2014:821): Socialdepartementet;
2014. 22. Ministry of Social Affairs: Patient Data Law (SFS 2008:355). 200 23. Information management and record keeping in health care (SOSFS
2008:14). http://www.socialstyrelsen.se/Lists/Artikelkatalog/Attachments/
8560/2008-10-14.pdf. Received: 19 August 2022 Accepted: 25 November 2022 24. EU: Medical device regulation 2017/745. 2017. 2.
International Diabetes Federation: IDF diabetes atlas 2021. 2022. References World Medical Association. World medical association declaration of
Helsinki: ethical principles for medical research involving human subjects. JAMA. 2013;310:2191–4. References 26. Sweden’s Municipalities and Regions: Foot examination in diabetes
- National care program for the prevention of foot complications in dia-
betes. (Östensson C-G, Johansson K, Insulander L, Jonsson LV, Löndahl M,
Sergu-Bogdan C, Jansson S, Gustafsson S, Hellstrand Tang U eds.); 2018. Page 10 of 10 Zügner et al. Journal of Foot and Ankle Research (2022) 15:90 27. National Board of Health and Welfare. National guidelines for diabetes
care: Indicators. Bilaga; 2018. 28. Praxel TA, Ford TJ, Vanderboom EW. Improving the efficiency and
effectiveness of performing the diabetic foot exam. Am J Med Qual. 2011;26:193–9. 29. Leese GP, Stang D. When and how to audit a diabetic foot service. Diabe-
tes Metab Res Rev. 2016;32(Suppl 1):311–7. 30. Leese GP, Stang D, McKnight JA. A national strategic approach to diabetic
foot disease in Scotland: changing a culture. Br J Diabetes Vasc Dis. 2011;11:69–73. 31. Greenhalgh T. How to improve success of technology projects in health
and social care. Public Health Res Pract. 2018;28(3):2831815. 32. Scandurra I, Hagglund M, Persson A, Ahlfeldt RM. Disturbing or facilitat-
ing? - on the usability of Swedish eHealth systems 2013. E Health Conti-
nuity of Care. 2014;205:221–5. y
33. Kolltveit B-CH, Gjengedal E, Graue M, Iversen MM, Thorne S, Kirkevold M. Conditions for success in introducing telemedicine in diabetes foot care:
a qualitative inquiry. BMC Nurs. 2017;16:2–2. y
34. Jarl GM, Heinemann AW, Norling Hermansson LM. Validity evidence for
a modified version of the orthotics and prosthetics Users’ survey. Disabil
Rehabil Assist Technol. 2012;7:469–78. 35. National Board of Health and Welfare: Prescription of assistive devices. 2016. 35. National Board of Health and Welfare: Prescription of assistive devices. 2016. 36. Payne JW. Thinking aloud: insights into information processing. Psychol
Sci. 1994;5:241–8. 36. Payne JW. Thinking aloud: insights into information processing. Psychol
Sci. 1994;5:241–8. 37. Nielsen RG, Rathleff MS, Simonsen OH, Langberg H. Determination of
normal values for navicular drop during walking: a new model correcting
for foot length and gender. J Foot Ankle Res. 2009;2:12. 38. National Diabetes Registry: Annu Report 2019. 2020 g
y
39. The Orthopedic Industry Council. https://ot-branschen.se/. 39. The Orthopedic Industry Council. https://ot-branschen.se/. 40. Bangor A, Kortum PT, Miller JT. An empirical evaluation of the system
usability scale. Int J Hum Comput Interact. 2008;24:574–94. 41. Measuring usability with the system usability scale. https://measuringu.
com/sus/. 42. National clinical knowledge support. https://kunskapsstyrningvard.se/
kunskapsstyrningvard/kunskapsstod/omvarakunskapsstod/nationelltklini
sktkunskapsstod.44269.html. 42. National clinical knowledge support. https://kunskapsstyrningvard.se/
kunskapsstyrningvard/kunskapsstod/omvarakunskapsstod/nationelltklini
sktkunskapsstod.44269.html. 43. Zügner et al. Journal of Foot and Ankle Research (2022) 15:90 Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in pub-
lished maps and institutional affiliations. •
fast, convenient online submission
•
thorough peer review by experienced researchers in your field
•
rapid publication on acceptance
•
support for research data, including large and complex data types
•
gold Open Access which fosters wider collaboration and increased citations
maximum visibility for your research: over 100M website views per year
•
At BMC, research is always in progress. Learn more biomedcentral.com/submissions
Ready to submit your research
Ready to submit your research ? Choose BMC and benefit from:
? Choose BMC and benefit from: •
fast, convenient online submission
•
thorough peer review by experienced researchers in your field
•
rapid publication on acceptance
•
support for research data, including large and complex data types
•
gold Open Access which fosters wider collaboration and increased citations
maximum visibility for your research: over 100M website views per year
•
At BMC, research is always in progress. Learn more biomedcentral.com/submissions
Ready to submit your research
Ready to submit your research ? Choose BMC and benefit from:
? Choose BMC and benefit from: •
fast, convenient online submission
•
thorough peer review by experienced researchers in your field
•
rapid publication on acceptance
•
support for research data, including large and complex data types
•
gold Open Access which fosters wider collaboration and increased citations
maximum visibility for your research: over 100M website views per year
•
At BMC, research is always in progress. Learn more biomedcentral.com/submissions
Ready to submit your research
Ready to submit your research ? Choose BMC and benefit from:
? Choose BMC and benefit from: •
fast, convenient online submission
•
thorough peer review by experienced researchers in your field
•
rapid publication on acceptance
•
support for research data, including large and complex data types
•
gold Open Access which fosters wider collaboration and increased citations
maximum visibility for your research: over 100M website views per year
•
At BMC, research is always in progress. Learn more biomedcentral.com/submissions
Ready to submit your research
Ready to submit your research ? Choose BMC and benefit from:
? Choose BMC and benefit from:
|
https://openalex.org/W2766549289
|
https://digilib.phil.muni.cz/bitstream/handle/11222.digilib/137168/1_EtudesRomanesDeBrno_47-2017-2_6.pdf?sequence=1, https://dialnet.unirioja.es/descarga/articulo/6314046.pdf
|
es
|
La confección del Glosario de aritmética y álgebra en el Renacimiento hispano como instrumento para el estudio del léxico matemático del siglo XVI
|
Études romanes de Brno
| 2,017
|
cc-by-sa
| 6,722
|
Molina Sangüesa, Itziar
La confección del Glosario de aritmética y álgebra en el Renacimiento hispano
como instrumento para el estudio del léxico matemático del siglo XVI
Études romanes de Brno. 2017, vol. 38, iss. 2, pp. 51-68
ISSN 1803-7399 (print); ISSN 2336-4416 (online)
Stable URL (DOI): https://doi.org/10.5817/ERB2017-2-4
Stable URL (handle): https://hdl.handle.net/11222.digilib/137168
Access Date: 17. 02. 2024
Version: 20220831
Terms of use: Digital Library of the Faculty of Arts, Masaryk University provides access to
digitized documents strictly for personal use, unless otherwise specified.
Digital Library of the Faculty of Arts,
Masaryk University
digilib.phil.muni.cz
études romanes de Brno 38 / 2017 / 2
DOI: 10.5817/ERB2017-2-4
The Preparation of the Glossary of Arithmetic and Algebra
in the Spanish Renaissance as an Instrument for the Study
of Mathematical Vocabulary of the Sixteenth Century
Itziar Molina Sangüesa [itziarmolina@usal.es]
Universidad de Salamanca, España
Resumen
38 / 2017 / 2
El objetivo de este trabajo es la presentación de un glosario de especialidad sobre la aritmética y el álgebra del
siglo XVI, con el cual hemos llevado a cabo, por un lado, una modesta contribución a la lexicografía histórico-científica y, por otro, la investigación de una parcela del léxico hispánico hasta el momento desatendida
(y, en buena parte, desconocida) por los historiadores de la lengua española y filólogos en general. El estudio
lexicológico que emana del inventario lexicográfíco confeccionado, como pondremos de manifiesto, nos ha
permitido obtener un mejor y más adecuado conocimiento del tecnolecto matemático renacentista, como
la procedencia de sus voces, los procedimientos morfológicos empleados en la formación de los tecnicismos
y los aspectos semánticos que lo caracterizan. Para la confección de este inventario léxico especializado nos
hemos servido del corpus del Diccionario de la Ciencia y de la Técnica del Renacimiento (DICTER) en el que
se integra. Durante el proceso hemos empleado ciertas herramientas informáticas, cuyo manejo, igualmente,
pretendemos mostrar en esta contribución, así como las principales dificultades surgidas de esta investigación interdisciplinar. En suma, consideramos que este diccionario y, concretamente, el apartado aritméticoalgebraico en el que nos centramos, puede ser un instrumento útil y provechoso para el estudio de la lengua
y literatura científica hispánicas, dado que ponen al alcance del usuario aspectos fundamentales de una etapa
crucial del devenir de ambas disciplinas y de la propia lengua española.
Palabras clave
diccionarios electrónicos; lexicografía histórica; lexicología histórica; matemáticas; renacimiento
Abstract
The objective of this paper is to present a specialized glossary about the arithmetic and algebra of the sixteenth
century, with which we have carried out, on the one hand, a contribution to the historical-scientific lexicography and, secondly, the investigation of a Hispanic vocabulary area until unattended (and largely unknown) by
historians of the Spanish language and philologists in general. This lexical study that emanates from the lexicographical inventory drawn up, as we will show, has allowed us to get a better knowledge of the Renaissance mathematician technolect, as the relevance of their voices, the morphological processes involved in the formation
51
DOSSIER thématique
La confección del Glosario de aritmética y álgebra
en el Renacimiento hispano como instrumento
para el estudio del léxico matemático del siglo XVI
Itziar Molina Sangüesa
La confección del Glosario de aritmética y álgebra en el Renacimiento hispano como instrumento para el estudio ...
Keywords
electronic dictionaries; historical lexicography; historical lexicology; mathematics; renaissance
recibido 2016–09–30; aceptado 2016–12–30
El desarrollo de esta investigación ha sido posible gracias a la ayuda predoctoral (FPU), concedida en 2011
por el Ministerio de Educación, Cultura y Deporte (Ref.: AP2010–3663).
1. Presentación
El objetivo de esta comunicación es la presentación de un glosario de especialidad sobre la aritmética y el álgebra del siglo XVI, con el cual hemos llevado a cabo, por un lado, una modesta contribución a la lexicografía histórico-científica y, por otro, la investigación de una parcela del léxico
hispánico hasta el momento desatendida (y, en buena parte, desconocida) por los historiadores de
la lengua española y filólogos en general.
2. Glosario de aritmética y álgebra en el Renacimiento hispano
Como se deduce del título fijado, este glosario recopila el vocabulario de un par de disciplinas
esenciales en el saber matemático de una época clave tanto para el asentamiento de las bases
y desarrollo de estas ciencias exactas como para la conformación de la terminología aritméticoalgebraica que, por vez primera, empleará la lengua española como vehículo de expresión lingüística de estos conocimientos científicos.
Para la confección de este inventario léxico especializado nos hemos servido del corpus —constituido por 74 obras— (Mancho y Quirós 2005)1 del Diccionario de la Ciencia y de la Técnica del
Renacimiento (DICTER)2 en el que se integra. Entre las mismas, seis son especialmente relevantes3:
1
Digitalizados y accesibles en: < http://DICTER.usal.es/?idContent=elenco_obras>.
2
Proyecto desarrollado, bajo la dirección de la Dra. Dña. María Jesús Mancho, por el “Grupo de Investigaciones lexicográficas y lexicológicas del español moderno y contemporáneo”, reconocido como grupo de investigación de Excelencia
“GR. 56. Grupo interuniversitario para el estudio de la Historia de la Lengua Española (GIEHLE)” por la Junta de Castilla
y León, en el Departamento de Lengua Española de la Facultad de Filología de la Universidad de Salamanca, concretamente, en el Instituto de Estudios Medievales y Renacentistas (IEMYR). Véase: <http://DICTER.usal.es/>.
3
No obstante, el carácter aplicado de la aritmética, especialmente, el léxico referido a los paradigmas numerales y las
proporciones, así como el cálculo constituido por las dos operaciones básicas de adición y sustracción, trae consigo la
52
38 / 2017 / 2
DOSSIER thématique
of the technical terms and the semantic aspects that characterize it. For the preparation of this specialized lexical
inventory, we have used the corpus of the Dictionary of Science and Technology of the Renaissance (DICTER)
in which it is integrated. During the process, we have used certain computer tools, whose management, also,
we intend to show in this conference, as well as the problems encountered in this interdisciplinary research. In
short, we believe that this dictionary, specifically the-algebraic arithmetic area in which we focus, can be a useful
and helpful tool for the study of Spanish language and scientific literature, since put within reach of the user key
aspects of a crucial stage of the evolution about both disciplines and the Spanish language.
Itziar Molina Sangüesa
La confección del Glosario de aritmética y álgebra en el Renacimiento hispano como instrumento para el estudio ...
Juan de Ortega (1512): Conpusición de la arte de la Arismética y de Geometría. Lyon: Maistro Nicolau de
Benedictis (por Joannes Trinxer).
Marco Aurel (1552): Libro primero de Arithmética algebrática. Valencia: Joán de Mey.
Juan Pérez de Moya (1562): Arithmética práctica y speculativa. Salamanca: Mathías Gast.
Juan Pérez de Moya (1589): Manual de contadores. Madrid: Pedro Madrigal.
Juan de Herrera (1584): Institución de la Academia Real Mathemática. Madrid: Guillermo Droy.
2.1. Selección del léxico
Por lo que respecta al establecimiento de los términos técnicos aritmético-algebraicos, procedimos, en primer lugar, a la minuciosa lectura de los textos científico-técnicos renacentistas. De ahí
partimos a la delimitación, selección y establecimiento de las unidades léxicas especializadas del
tecnolecto matemático. Como apunta Mancho Duque en la planta del DICTER (2014), conviene
tener presente
la dificultad real para distinguir las voces especializadas de las generales en un periodo en el
que se inician tímidamente las terminologías y se asiste a la eclosión de la neología de sentido. La
detección de acepciones especializadas a partir del vocabulario común es una de las tareas que
consideramos prioritaria en una época que está surgiendo un vocabulario técnico y el trasvase
entre ambos niveles léxicos es constante4.
38 / 2017 / 2
2.1.1. Herramientas informáticas y metodología
Para facilitar la detección de estas acepciones especializadas y del léxico matemático en general
nos servimos de un programa informático de búsquedas complejas, denominado Ultraedit, el cual
nos permitió localizar las voces, así como precisar los sentidos de las mismas, en concordancias
con el resto de unidades léxicas y/o gramaticales que integran el corpus.
Posteriormente, procedimos a la lematización del conjunto léxico en una base de datos, creada
mediante la herramienta informática FileMaker Pro 11.0, que constituye la ficha lexicográfica. Esta
ficha, en la actualidad, comprende diversos campos que, a su vez, se subdividen en nueve bloques
temáticos —enumerados en las páginas siguientes— en los que recogemos la información lexicográfica, concretamente:
• Datos relativos a la lematización.
• Información histórica, etimológica y morfológica.
documentación de gran parte de estos vocablos en textos tecno-científicos de diversa índole (astronomía y cosmografía,
náutica, geometría, arte militar, etc.).
4
Etinger pone de manifiesto (1982: 391) que las “relaciones entre el vocabulario común y el técnico son muy variadas
y casi siempre se han formado diacrónicamente, por lo que no es nada fácil marcar los límites entre ambos”.
53
DOSSIER thématique
Pedro Núñez Salaciense (1567): Libro de Álgebra en Arithmética y Geometría. Anvers: Herederos de Arnoldo Birckman.
Itziar Molina Sangüesa
La confección del Glosario de aritmética y álgebra en el Renacimiento hispano como instrumento para el estudio ...
Estos bloques temáticos se encuentran organizados en cuatro pestañas: ‘forma e historia’, ‘clasificación semántica’, ‘ejemplos e información’ e ‘imagen’, como puede comprobarse en las figuras5
siguientes:
Las relaciones entre lingüística e informática en el quehacer de los lexicógrafos son indisolubles, así pues, apostamos por la aplicación de herramientas informáticas en la metodología
de trabajo en aras de optimizar el almacenamiento y la gestión de la información (cf. Clavería/
Mancho 2006). Además, con respecto a la lexicografía terminológica, declararon Arntz y Pitch
(1995: 219), “se utiliza en creciente medida el tratamiento electrónico de los datos”, por lo que
podemos concluir que el empleo de recursos informáticos en investigación lexicográfica resulta
imprescindible.
2.2. Estructura
2.2.1. Macroestructura
La macroestructura de este glosario matemático está constituida por un número determinado
de entradas, concretamente, 2353, que se distribuyen del siguiente modo: un total de 846 fichas
encabezan el artículo lexicográfico, 813 son remisiones a variantes gráficas y 220 remisiones a unidades pluriverbales. Las 474 fichas restantes corresponden a subentradas en la microestructura, en
las que encontramos 230 acepciones y un total de 244 compuestos sintagmáticos y locuciones.
2.2.2. Microestructura
De acuerdo con Haensch (1997: 41), entendemos por microestructura “la ordenación de los elementos que componen el artículo lexicográfico”. En cuanto a la de nuestro inventario, como puede
visualizarse en el siguiente artículo lexicográfico, son múltiples informaciones las que la componen:
lema, variantes gráficas, etimología, etc. que, a continuación, exponemos con detalle: la categoría
gramatical, el género y el número, la primera documentación, la marcación temática y la definición.
Asimismo, aparecen un máximo de tres ejemplos —los más ilustrativos—, además de información
sobre las relaciones semánticas y genéticas (sinonimia, antonimia, familias léxicas) e información
complementaria (denominada‚ ‘información enciclopédica’) sobre el concepto definido.
5
Un análisis detallado sobre el almacenamiento y rellenado de las fichas lexicográficas del glosario sobre aritmética
y álgebra del Renacimiento hispano puede leerse en Molina Sangüesa (2014a).
54
38 / 2017 / 2
DOSSIER thématique
• Clasificación semántica.
• Definición e información lexicográfica.
• Ejemplos.
• Información enciclopédica.
• Observaciones semánticas.
• Otras observaciones.
• Imagen (referencia e información sobre la misma).
Itziar Molina Sangüesa
DOSSIER thématique
La confección del Glosario de aritmética y álgebra en el Renacimiento hispano como instrumento para el estudio ...
38 / 2017 / 2
Figura 1. Organización de la información en la base de datos: ‘Forma e historia’.
Figura 2. Organización de la información en la base de datos: ‘Clasificación semántica’.
55
Itziar Molina Sangüesa
DOSSIER thématique
La confección del Glosario de aritmética y álgebra en el Renacimiento hispano como instrumento para el estudio ...
38 / 2017 / 2
Figura 3. Organización de la información en la base de datos: ‘Ejemplos e información’.
Figura 4. Organización de la información en la base de datos: ‘Imagen’
56
Itziar Molina Sangüesa
38 / 2017 / 2
DOSSIER thématique
La confección del Glosario de aritmética y álgebra en el Renacimiento hispano como instrumento para el estudio ...
Figura 5. Modelo de artículo lexicográfico.
57
Itziar Molina Sangüesa
La confección del Glosario de aritmética y álgebra en el Renacimiento hispano como instrumento para el estudio ...
Dado que el glosario confeccionado recopila el léxico matemático hispano del periodo renacentista, documentamos para cada uno de nuestros lemas la proliferación de numerosas variantes
gráficas6 correspondientes al denominado ‘reajuste fonológico del Siglo de Oro’, ya que, como es
sabido, el siglo XVI se trata de una época de crisis, en la que se está fraguando la norma del español moderno.
Esta información sobre las variantes gráficas aparece, asimismo, en la macroestructura del glosario, en forma de remisiones a la forma canónica lematizada, mediante la abreviatura V. (véase),
seguida de la palabra destino de esa referencia. En este sentido, de acuerdo con Morales (1998: 21),
consideramos que la remisión en un diccionario histórico se convierte, además, en una información lexicográfica más acerca de la historia del lema.
2.2.2.2. Etimología y primera documentación
En un glosario de tipo histórico, como es el Glosario de aritmética y álgebra del Renacimiento
hispano, resulta esencial la información etimológica de las voces lematizadas, motivo por el que
hemos dedicado un apartado, que va entre corchetes, en cada uno de los artículos lexicográficos
que configuran nuestro inventario a estos aspectos diacrónicos del léxico matemático.
Para ello, nos hemos servido del Diccionario crítico etimológico castellano e hispánico (DCECH)
de Joan Corominas y José Antonio Pascual. Únicamente para aquellos términos no recopilados
en este repertorio, o cuyo origen no estaba claro, hemos recurrido a la consulta del Oxford latin
dictionary (OLD), el Thesaurus Linguae Latinae (TLL), A Latin Dictionary: Founded on Andrews’
Edition of Freund’s Latin Dictionary de Charlton Lewis y Charles Short (Lewis-Short), Laterculi
Vocum latinarum de Gradenwitz (Laterculi), Mediae latinitatis lexicon minus de Niermeyer, Glosarium infimae et medievae latinitatis de Du Cange, Le Dictionnaire latin-français de Gaffiot, así
como el Trésor de la langue française informatisé (TLFi), el Tesoro della lingua italiana delle origini
(TLIO), Dizionario etimologico della lingua italiana de Cortelazzo y Zolli (DELI) o el Diccionari
Català-Valencià-Balear de Alcover y Moll, entre otros muchos.
La información etimológica se complementa con el establecimiento de la primera de la documentación de las voces aritmético-algebraicas registradas. De nuevo hemos recurrido al DCECH
y, en los casos en que este repertorio no la aportaba, hemos consultado el CORDE y CNDHE
académicos e incluso los propios testimonios textuales del corpus analizado, ya que muchos términos del registro matemático manifiestan en él su primera aparición; en este sentido, debemos
puntualizar y dejar constancia de que estos datos se presentan con una prudente provisionalidad.
6
Hemos tratado las variantes gráfico-fonéticas que atañen a los numerales cardinales en Molina Sangüesa (2014b).
58
38 / 2017 / 2
DOSSIER thématique
2.2.2.1. Variantes gráficas
Itziar Molina Sangüesa
La confección del Glosario de aritmética y álgebra en el Renacimiento hispano como instrumento para el estudio ...
2.2.2.3. Categoría e información gramatical
adj.
adv.
sust.
v.
comp. sintag.
loc. adv.
loc. prep.
loc. v.
adjetivo
adverbio
sustantivo
verbo
compuesto sintagmático
locución adverbial
locución prepositiva
locución verbal
Para las cuales ofrecemos, además, una serie complementaria de abreviaturas que indican la
subcategorización (Porto Dapena 2002: 176–177) de sustantivos y verbos:
38 / 2017 / 2
f.
m.
u. t. c. f.
sg.
pl.
u. t. c. pl.
u. m. en pl.
tr.
intr.
prnl.
u. t. c. tr.
u. t. c. intr.
u. t. c. prnl.
femenino
masculino
usado también como femenino
singular
plural
usado también como plural
usado más en plural
transitivo
intransitivo
pronominal
usado también como transitivo
usado también como intransitivo
usado también como pronominal
Finalmente, damos cuenta de las frecuentes recategorizaciones que se dan entre adjetivos
y sustantivos, así como de otros trasvases entre categorías gramaticales7; aspectos que se ponen
de manifiesto mediante el siguiente código de abreviaturas:
u. t. c. adj.
u. t. c. adv.
u. t. c. sust.
u. t. c. pron.
usado también como adjetivo
usado también como adverbio
usado también como sustantivo
usado también como pronombre
7
Un estudio sobre las recategorizaciones y trasvases entre las categorías que constituyen los paradigmas numerales
renacentistas puede leerse en Molina Sangüesa (en prensa c).
59
DOSSIER thématique
A continuación, damos cuenta, mediante una serie de abreviaturas, de la categoría a la que pertenecen los tecnicismos contenidos en el glosario reducidos a la forma canónica de su paradigma:
Itziar Molina Sangüesa
La confección del Glosario de aritmética y álgebra en el Renacimiento hispano como instrumento para el estudio ...
Como era de esperar, en un repertorio de léxico especializado como el que presentamos, la distinción de la ciencia o la técnica a la que pertenece cada uno de sus lemas se convierte en una tarea
de vital importancia.
De acuerdo con la práctica tradicional de la lexicografía, mediante una serie de abreviaturas,
ofrecemos, en una posición previa a la definición, la marca que informa de la pertenencia a un
tecnolecto o ámbito del saber específico a cada una de las voces especializadas del glosario
confeccionado8, las cuales ascienden a un total de 1163 voces marcadas9. Estas se dividen en las
siguientes áreas temáticas: Aritmética (855), Álgebra (141), Matemáticas (29), Geometría (12),
Astronomía (15), Filosofía de la ciencia (139), Cronometría (6), Metrología (9), Música (16),
representadas mediante las abreviaturas Arit.; Álg.; Mat.; Geom.; Astr.; Fil.; Cron., Metr. y Mús.
respectivamente10.
Por último, cabe señalar que hemos atestiguado un par de arcaísmos, que hemos etiquetado con
la marca diacrónica ‘anticuada’ (representada por la abreviatura ant.): el numeral cardinal sece y el
fraccionario secén que, además, son dos dialectalismos, pues provienen del aragonés (cf. DCECH,
s. v. seis) y aparecen documentados exclusivamente en el tratado matemático más antiguo del corpus textual manejado, el del dominico Juan de Ortega (1512). Asimismo, llevan la marca diatópica
‘arag.’ las voces numerales centenal, cinquén, cuatrén, decén, docén, novén, ochosén, oncén, quincén,
setén, trecén y treintén.
2.2.2.5. Definición
Para la expresión del significado nos servimos de definiciones. En este cometido, de acuerdo con
las directrices de la redacción del DICTER, partimos de las fuentes lexicográficas existentes en el
tradición hispánica, principalmente de la lexicografía académica; por lo que las definiciones del
glosario que presentamos, se fundamentan en definiciones del Diccionario de la Lengua Española (DLE23) —siempre que estas sean claras, sistemáticas, informativas (Gutiérrez Cuadrado 2008
y 2010: 135) y oportunas o apropiadas al contexto del lema a definir—, esto es, siempre que ha
sido posible hallar una definición que se adecuara a los usos detectados en los textos, la hemos
aprovechado.
8
No obstante, tal y como arguyen Quirós (2010: 329–330) y Mancho Duque (planta del DICTER, 2014), distinguir la
especificación del ámbito temático de uso del léxico especializado frente al léxico común no es una labor sencilla “no es
una tarea fácil distinguir las voces propias de un ámbito técnico-científico de las generales, por el continuo trasvase que se
produce de una esfera a la otra, en ambas direcciones, máxime en una época como esta en la que se constata un incipiente
establecimiento de terminologías, donde el recurso a la incrustación de acepciones especializadas es frecuentísimo y de
delicada detección”.
9
En contrapartida, documentamos 150 cabeceras sin marcar.
10 Como puede apreciarse, postulamos para la caracterización de este repertorio léxico marcaciones jerarquizadas,
donde además de matemáticas se da geometría, aritmética, álgebra, etc. (vid. Fajardo 1996–1997), que corresponden con
subdivisiones de ciertos dominios lingüísticos (Azorín 1992: 448) o subáreas correspondientes a las disciplinas que las
integran, en este caso, las ciencias exactas.
60
38 / 2017 / 2
DOSSIER thématique
2.2.2.4. Marca diatécnica
Itziar Molina Sangüesa
Pero, cuando este repertorio no nos ha resultado convincente, por suponer, entre otras causas,
por ejemplo, un claro anacronismo, nos hemos apoyado en otros diccionarios históricos, como los
fascículos del Diccionario Histórico, el Tesoro de la lengua castellana o española de Covarrubias, el
Diccionario de Autoridades, el Nuevo tesoro lexicográfico de la lengua española de la RAE o el Diccionario castellano con las voces de las ciencias y las artes de Terreros, así como en el especializado
Vocabulario matemático-etimológico de Picatoste.
En todos los casos, dejamos entre paréntesis tras la definición adoptada constancia explícita
de la procedencia de la misma. Sin embargo, ante la ausencia de los términos en los diccionarios
manejados, en ciertas ocasiones hemos tenido que proceder a la creación de definiciones propias.
En este difícil quehacer, se ha procurado, ante todo, la mayor sencillez y transparencia. Para ello,
hemos seguido los criterios establecidos en los estudios teóricos (Bosque 1982; Ahumada 1989:
102–255; Porto Dapena 2002: 277–320 y 2014; Medina Guerra 2003).
2.2.2.6. Acepciones
En el glosario aritmético-algebraico que presentamos las distintas acepciones van introducidas
y numeradas, como es habitual, a excepción de la primera, mediante números arábigos correlativamente y distribuidas en párrafos distintos. Por otro lado, el establecimiento de acepciones, junto
con la elaboración de las definiciones, es una de las labores más arduas y complicadas a las que
inexcusablemente debe enfrentarse el lexicógrafo. En palabras de Pascual y García (2007: 97):
38 / 2017 / 2
Las acepciones tratan de distinguir diversos sentidos dentro del hilo conductor que constituye la
rama de la semántica general. Sabemos que la separación entre ellas y su redacción es una tarea
difícil sobre la que la lexicografía no ha establecido criterios únicos e indiscutibles.
La complejidad radica, esencialmente, en la detección y determinación de subacepciones y en
su ordenación; máxime en un repertorio lexicográfico de tipo histórico como el confeccionado. En nuestro inventario disponemos únicamente de fuentes documentales, lo que requiere un
exhaustivo trabajo de interpretación de los textos. Incluso en un ámbito como el científico-técnico
—en el que teóricamente se postula la correspondencia unívoca entre significante y significado—,
reconocemos la polisemia en multitud de los vocablos para la designación de los distintos referentes propios de las ciencias exactas de la época renacentista. Así, en la macroestructura del Glosario
de aritmética y álgebra del Renacimiento hispano, contabilizamos un total de 1076 acepciones, de
las cuales 846 son cabeceras y 230 subacepciones.
Con respecto a la ordenación de las 230 subacepciones que contabilizamos, entre los diversos criterios11 que se han llevado a cabo, hemos adoptado un criterio lógico (Lara 1998–199912):
11 Werner (1982: 315) enumera los siguientes: criterio de la cronología, criterio etimológico, criterio lógico, criterio
de la conciencia lingüística, criterio de la frecuencia, criterio de la posición dentro del sistema colectivo y criterio de la
distribución semántica.
12 “La base principal de todo orden lógico de acepciones es la noción de significado principal, pues el orden se establece
a partir de un punto de referencia que, en este caso, debe ser necesariamente semántico [...]. El criterio llamado lógico indica que para los lexicógrafos hay un orden de acepciones que permite reconocer, primero, la existencia de un significado
61
DOSSIER thématique
La confección del Glosario de aritmética y álgebra en el Renacimiento hispano como instrumento para el estudio ...
Itziar Molina Sangüesa
La confección del Glosario de aritmética y álgebra en el Renacimiento hispano como instrumento para el estudio ...
situamos en primer lugar la acepción más general o difundida dentro del área de especialidad
para, a partir de ahí, continuar por las más particulares hasta llegar a las unidades pluriverbales.
Los ejemplos constituyen un elemento esencial en la microestructura de un diccionario de corte
histórico, pues estos, como afirma Porto Dapena, “autorizan y atestiguan la existencia o presencia
del vocablo a lo largo del tiempo” (2002: 106). Por ello, cada voz, acepción o unidad pluriverbal
del glosario que nos ocupa aparece avalada por hasta tres ejemplos, obtenidos del expurgo de las
obras pertenecientes al corpus del DICTER. En principio, siempre y cuando ha sido lo suficientemente ilustrativo, hemos intentado recoger la primera aparición, es decir, el ejemplo más antiguo,
de cada acepción de un término. En el caso de no cumplirse esta condición, se ha desechado la
primera datación y se ha procedido a la selección de otro ejemplo, también temprano, que refleje
de modo más transparente en sentido de la voz detectada. Los ejemplos restantes aparecen ordenados en una secuencia cronológica ascendente.
2.2.2.8. Relaciones semánticas y genéticas
Después de los ejemplos, hemos dispuesto, en primer lugar, la información relativa a las relaciones
semánticas que contraen algunas de las palabras recopiladas en el glosario. De modo que hemos
indicado todos los casos de antonimia y sinonimia detectados en los textos aritmético-algebraicos13, precedidos por las abreviaturas ant. y sin. respectivamente, escritos en letra cursiva negrita
y por orden alfabético.
El léxico matemático renacentista se caracteriza por una fuerte presencia de voces sinónimas
(las cuales ascienden a un total de 678). Esta proliferación de sinónimos estaría provocada por
los tanteos designativos de los primeros cultivadores de estas disciplinas (cf. Mancho 2014).
Aunque en menor medida, detectamos, asimismo, en el corpus manejado un conjunto formado exactamente por 239 voces que presentan uno o más antónimos. En último lugar, tras los
antónimos (si los hubiere), y seguidas de la marca fam., hemos hecho explícitas las relaciones
genéticas y formales de las palabras lematizadas en nuestro catálogo de voces matemáticas, es
decir, las denominadas familias léxicas, “una de las formas en que se manifiesta de un modo más
acabado la estructuración14 de léxico de una lengua” (Pena y Campos 2009: 21). En este aparprincipal de cada vocablo y, después, una serie de significados emparentados con el principal en alguna forma, que corresponde, o bien a una posible estructura de sememas de cada vocablo, o bien a un orden que se puede formular mediante
criterios ampliamente aceptables, como la relación general/particular o viceversa” (Lara 1998–1999: 628).
13 En este quehacer nos hemos apoyado en el Diccionario general de sinónimos y antónimos (1999) dirigido por Blecua
y en el Diccionario de sinónimos y antónimos (publicado en 2004 por Espasa Calpe), entre otros.
14 Según Morales (1998: 37), “la evolución de la lengua desde un punto de vista léxico se definiría como un engranaje
de piezas que van encajando formas, acepciones y usos en su continuo devenir de la historia de cada sociedad o cultura. El
diccionario histórico debe trazar esa evolución, sucesión de formas y acepciones, para todos y cada uno de los elementos
que componen la lengua; tarea que solo es posible si se considera la lengua como un sistema en el que los elementos y sus
interrelaciones con los demás elementos forman un todo”.
62
38 / 2017 / 2
DOSSIER thématique
2.2.2.7. Ejemplos
Itziar Molina Sangüesa
La confección del Glosario de aritmética y álgebra en el Renacimiento hispano como instrumento para el estudio ...
los diccionarios etimológicos nos aportan la concepción genética de la familia como conjunto de
palabras agrupadas en torno a una raíz, étimo o palabra primitiva, como punto de arranque de las
demás palabras emparentadas.
2.2.2.9. Información enciclopédica e ilustraciones
38 / 2017 / 2
En algunas ocasiones, hemos considerado necesario proporcionar información adicional o complementaria de las voces lematizadas, con la finalidad de que estos datos ayudaran a contextualizar
histórica y culturalmente la definición y a interpretar mejor los ejemplos.
La información se ha extraído bien de los textos del propio corpus o bien de diferentes diccionarios, como Autoridades (álgebra, aritmética, contador, quipos), Terreros (áureo número, septunce, sextario) o vocabularios especializados, primordialmente del redactado por Felipe Picatoste y Rodríguez (algoritmo, axioma, común denominador, finito, igualación, infinito, matemáticas,
numeración, número racional, número triangular, problema, prueba), así como de sus Principios
y ejercicios de Aritmética y Geometría (multiplicar en cruz, regla de compañía) y del Diccionario razonado de matemáticas de A. Wausfel (medio armónico, en infinito). También de obras de carácter
histórico sobre ciencia y técnica (González Urbaneja 2007; Meavilla Seguí 2001; Moreno Castillo 2004) o de trabajos especializados en aritmética (Salavert Fabiani 1994) y álgebra (Franci
1988 y 2010; Bashmakova y Smirnova 2000).
Asimismo, las ilustraciones sirven, en algunos casos, de apoyo a la óptima comprensión de la
información lexicográfica. Por esta causa, hemos estimado oportuno dar cabida a elementos no
lingüísticos, pero que, en este contexto de lenguas especializadas al que se adscribe el glosario
confeccionado son, como certifica Lerat (1997: 177), de suma importancia.
3. El tecnolecto matemático renacentista
Una vez confeccionado el glosario procedimos a la explotación de los datos contenidos en el
mismo, a través de un estudio lexicológico (Molina Sangüesa 2015) que nos ha permitido obtener un mejor y más adecuado conocimiento del tecnolecto matemático renacentista del que,
a continuación, esquematizamos las principales conclusiones.
63
DOSSIER thématique
tado hemos englobado voces patrimoniales y cultas con una misma ascendencia etimológica
—surgidas a partir de un mismo núcleo o base— y las ramas léxicas creadas por los procedimientos de derivación y composición en lengua española.
Para establecer estos subgrupos de palabras en el ámbito de las ciencias exactas, nos hemos
servido principalmente del DCECH, puesto que, como afirman Pena y Campos (2009: 32),
Itziar Molina Sangüesa
La confección del Glosario de aritmética y álgebra en el Renacimiento hispano como instrumento para el estudio ...
Los datos recopilados en el glosario ponen de manifiesto que uno de los mecanismos más empleados para suplir las necesidades designativas que suponían el hecho de verbalizar por vez primera
ideas, objetos o conceptos matemáticos fue recurrir a la adopción de préstamos léxicos —o neologismos externos— de otras lenguas de cultura, tales como el latín o el griego, el francés, el italiano,
etc. Así pues, de 846 voces que constituyen la selección léxica estudiada 371 son préstamos.
Como puede apreciarse en el gráfico que presentamos, la aportación fundamental es del latín,
tanto como punto de partida del caudal hereditario o tradicional (que supone un 28% del conjunto total, con 136 voces patrimoniales y, por otro lado, 7 semicultismos) como de los cultismos
(constituye el grupo más cuantioso, que asciende a un 68%, 347 voces doctas).
Gráfico 1. Procedencia de las voces matemáticas
Aspectos que revelan la importancia que el componente latino adquiere en la conformación del
vocabulario matemático y el marcado carácter culto del mismo, frente a otras ciencias o técnicas
de la época estudiada.
3.2. Procedimientos morfológicos y semánticos empleados
en la formación de los tecnicismos
Igualmente, la selección de voces integradas en el glosario, nos ha permitido concluir que, junto
con los préstamos, el tecnolecto matemático se nutre de mecanismos neológicos internos a la
propia lengua de tipo formal. Uno de los procedimientos morfológicos más productivos en la
formación de la terminología aritmético-algebraica es la derivación. Así, en nuestro corpus léxico
documentamos un total de 251 derivados, generados a partir de procesos de prefijación, sufijación
y parasíntesis. Entre los mismos, la sufijación es el mecanismo neológico más empleado: contabilizamos un conjunto de 232 voces nuevas entre sufijados nominales (39 casos, 17%), adjetivales
(132, un 57%, de los cuales 68 son denumerales), verbales (6 términos, 3%) y adverbiales (que
ascienden a 35 voces, 15%) —frente a una voz parasintética y 18 prefijadas—. Como puede observarse:
64
38 / 2017 / 2
DOSSIER thématique
3.1. Procedencia de las voces aritmético-algebraicas
Itziar Molina Sangüesa
Gráfico 2. Procedimientos morfológicos empleados en la formación de las voces matemáticas: derivación
38 / 2017 / 2
De manera análoga, la composición léxica —crucial para la creación de numerales complejos— y, especialmente, la composición sintagmática designan más de dos centenares (215) de
conceptos relativos a este par de ciencias exactas.
Por otro lado, aunque en menor medida, constatamos algunos procesos neológicos de corte
semántico. Por ejemplo, detectamos un par de préstamos semánticos cantidad (del italiano,
como ‘segunda incógnita de una ecuación’) y nombre (del catalán, y este, a su vez, del francés,
como ‘número’, en la aritmética redactada por Juan de Ortega); la transferencia de una serie de
términos de un campo especializado o disciplina a otra, como sucede con los numerales fraccionarios sufijados en -avo: cincoavo, seisavo, sieteavo y ochavo, los cuales pasan de la aritmética
a la geometría, más concretamente, al léxico de los canteros, para designar polígonos regulares
‘que constan de cinco, seis, siete u ocho lados iguales’; cambios semánticos en voces de la lengua común que, tras procesos neológicos en los que se les atribuye un nuevo sema, pasan al
léxico especializado aritmético-algebraico, este es el caso de las palabras: carácter, figura, letra,
especie, obra, obrar, conjugar y conjugación. Con todo, la acuñación de metáforas (antropomórficas, espaciales, etc.) se revela como el mecanismo semántico más fructífero e interesante de la
creación léxica del tecnolecto matemático renacentista (por ejemplo, número/raíz sorda, número
cuadrado o número triangular).
4. A modo de conclusión
En síntesis, hemos procurado que en este glosario encuentren acogida todas las palabras relativas
a la aritmética y el álgebra que se han podido rescatar del pasado y sus empleos, que nos permitan
estudiar las conexiones entre el léxico del Siglo de Oro y el actual; motivo por el que consideramos
que este diccionario (DICTER) y, más concretamente, el apartado matemático en el que nos hemos centrado, puede ser un instrumento útil y provechoso para el estudio de la lengua y literatura
científica hispánicas, dado que ponen al alcance del usuario aspectos fundamentales de una etapa
crucial del devenir de ambas disciplinas y de la propia lengua española.
65
DOSSIER thématique
La confección del Glosario de aritmética y álgebra en el Renacimiento hispano como instrumento para el estudio ...
Itziar Molina Sangüesa
La confección del Glosario de aritmética y álgebra en el Renacimiento hispano como instrumento para el estudio ...
Ahumada, I. (1989). Aspectos de lexicografía teórica. Granada: Universidad de Granada.
[DCVB]: Alcover, A. Mª, & Moll, F. B. (1935). Diccionari català-valencià-balear. Palma de Mallorca: Imprenta Mn. Alcover
Arntz, R., & Picht, H. (1995). Introducción a la terminología. Madrid: Fundación Germán Sánchez Ruipérez,
Pirámide.
Aurel, M. (1552). Libro primero de Arithmética algebrática. Valencia: Joán de Mey.
Azorín Fernández, D. (1992). El Diccionario general de la lengua frente a los vocabularios científicos y técnicos. In Actas del IV Congreso Internacional de EURALEX (pp. 445–453). Barcelona: Biblograf.
Bashmakova, I., & Smirnova, G. (2000). The beginnings and evolution of Algebra. Washington, D.C.: The
Mathematical Association of America.
[TLIO]: Beltrami, P. (dir.) (1997). Tesoro della lingua italiana delle origini. <http://tlio.ovi.cnr.it/TLIO>
Blecua, J. M. (dir.) (1999). Diccionario general de sinónimos y antónimos. Barcelona: VOX Universidad.
Bosque, I. (1982). Sobre la teoría de la definición lexicográfica. Verba, 9, 105–123.
Clavería Nadal, G., & Mancho Duque, Mª J. (Eds.) (2006). Estudio del léxico y bases de datos. Barcelona:
Universitat Autònoma de Barcelona, Servei de Publicacions.
[DCECH]: Corominas, J., & Pascual, J. A. (1980–1991). Diccionario crítico etimológico castellano e hispánico.
Madrid: Gredos.
[DCECH]: Corominas, J., & Pascual, J. A. (2012). Diccionario Crítico Etimológico Castellano e Hispánico
(Edición CD-ROM). Madrid: Gredos.
[DELI]: Cortelazzo, M., & Zoli, P. (1979–1988). Dizionario etimologico della lingua italiana. Bologna: Zanichelli.
[Cov.]: Covarrubias Horozco, S. (2006 [1611]). Tesoro de la lengua castellana o española. Edición integral
e ilustrada de Ignacio Arellano y Rafael Zafra. Madrid: Iberoamericana.
Du Cange (1954). Glossarium mediae et infimae latinitatis, 1883–1887. Graz: Akademische Druck- und Verlagsanstalt.
Durán Guardeño, A. J. (2006). Vida de los números. Madrid: T Ediciones.
Ettinger, S. (1982). La variación lingüística en lexicografía. In G. Haensch et al, La lexicografía. De la lingüís
tica teórica a la lexicografía práctica (pp. 359–395). Madrid: Gredos.
Fajardo Aguirre, A. (1996–1997). Las marcas lexicográficas: conceptos y aplicación práctica en la Lexicografía española. Revista de Lexicografía, 3, 31–57.
Franci, R. (2010). The history of algebra in Italy in the 14th and 15th centuries. Some remarks on recent historiography. Actes d’Història de la Ciència i de la Tècnica, Nova Época, 3 (2), 175–194.
Franci, R., & Toti Rigatelli, L. (1988). The italian algebra of the fourtheen century. In C. Hay (Ed.). Mathema
tics from Manuscript to Print 1300–1600 (pp. 11–30). Oxford: Clarendon press.
Gaffiot, F. (2000 [1934]). Le Grand Gaffiot: dictionnaire latin-français. París: Hachette.
Glare, P. G. W. (1968–1982). Oxford Latin Dictionary. Oxford: Clarendon Press.
González Urbaneja, P. M. (2007). Pitágoras. El filósofo del número. Madrid: Nivola.
Gradenwitz, O. (1904). Laterculi Vocum latinarum. Leizpig: Verlag Von S. Hirzel.
Gutiérrez Cuadrado, J. (2008). Desafíos de la definición. In B. Elisenda, & J. DeCesaris (Eds.). Proceedings of
the 13th Euralex International Congress (pp. 525–533). Barcelona: Universitat Pompeu Fabra.
Gutiérrez Cuadrado, J. (2010). Niveles y procesos en la definición del diccionario. Quaderns de Filologia.
Estudis lingüístics, XV, 121–138.
Haensch, G. (1997). Los diccionarios del español en el umbral del siglo XXI. Salamanca: Ediciones Universidad
de Salamanca.
66
38 / 2017 / 2
DOSSIER thématique
Referencias bibliográficas
Itziar Molina Sangüesa
38 / 2017 / 2
Herrera, J. (1584). Institución de la Academia Real Mathemática. Madrid: Guillermo Droy.
[CNDHE]: Instituto de Investigación Rafael Lapesa - Real Academia Española. Corpus del Nuevo Diccionario
Histórico del español. <http:// //web.frl.es/CNDHE>
Instituto de Investigación Rafael Lapesa - Real Academia Española (2013). Mapa de diccionarios académicos.
<http://web.frl.es/ntllet/>
Lara, L. F. (1998–1999). Una hipótesis cognoscitiva sobre el orden de acepciones. Boletín de Filología, 37(1),
623–644.
[Lewis-Short]: Lewis, Ch. T. & Short, Ch. (1879). A Latin Dictionary: Founded on Andrew’s Edition of Freund’s
Latin Dictionary. Oxford: Clarendon Press.
[DICTER]: Mancho, Mª J. (dir.) (2014). Diccionario de la ciencia y de la técnica del Renacimiento español.
<http://DICTER.eusal.es/>
Mancho, Mª J. (dir.), & Quirós, M. (coord.) (2005). La ciencia y la técnica en la época de Cervantes: textos
e imágenes. Salamanca: Ediciones Universidad de Salamanca.
Meavilla Seguí, V. (2001). Aspectos históricos de las matemáticas elementales. Zaragoza: Prensas Universitarias
de Zaragoza.
Medina Guerra, A. Mª. (2003). La microestructura del diccionario: la definición. In A. Mª. Medina (coord.),
Lexicografía española (pp. 127–146) Barcelona: Ariel.
Molina Sangüesa, I. (2014a). El estudio de las matemáticas en el Renacimiento desde una perspectiva filológica: proyecto de una tesis doctoral. In C. Mata, A. J. Sáez, & A. Zúñiga (Eds.), «Sapere aude». Actas del III
Congreso Internacional Jóvenes Investigadores Siglo de Oro (JISO 2013). Pamplona: Servicio de publicaciones de la Universidad de Navarra, Colección BIADIG. <http://dadun.unav.edu/bitstream/10171/35941 /1/
JISO2013_20_Molina.pdf>
———— . (2014b). Sobre la variabilidad formal de los nombres de numerales cardinales en el léxico matemático del Renacimiento hispano. In C. Grande, L. Martín, & S. Salicio (coords.), Con una letra joven.
Avances en el estudio de la Historiografía e Historia de la Lengua Española (pp. 237–245). Salamanca:
Ediciones Universidad de Salamanca.
————. (2015). Las matemáticas en el Renacimiento hispano estudio léxico y glosario. Tesis Doctoral inédita.
Salamanca: Universidad de Salamanca.
Morales, C. (1998). Las relaciones del léxico en el diccionario. Bellaterra: Servei de Publicacions de la Universitat Autònoma de Barcelona.
Moreno Castillo, R. (2004). Fibonacci. El primer matemático medieval. Madrid: Nivola.
Niermeyer, J. F. & Van De Kieft, C. (2002). Mediae latinitatis lexicon minus. Boston: Brill Leiden.
Núñez, P. (1567). Libro de Álgebra en Arithmética y Geometría. Anvers: Herederos de Arnoldo Birckman.
Ortega, J. (1512). Conpusición de la arte de la Arismética y de Geometría. León de Francia, Maistro Nicolau
de Benedictis (por Joannes Trinxer).
Pascual Rodríguez, J. A., & García Pérez, R. (2007). Límites y horizontes en un diccionario histórico. Salamanca: Diputación de Salamanca.
Pena, J., & Campos, M. (2009). Propuesta metodológica para el establecimiento de familias léxicas en una
consideración histórica: el caso de hacer. Cuadernos del Instituto Historia de la Lengua, 2, 21–52.
Pérez de Moya, J. (1562). Arithmética práctica y speculativa. Salamanca: Mathías Gast.
———— . (1589). Manual de contadores. Madrid: Pedro Madrigal.
Picatoste y Rodríguez, F. (1862). Vocabulario matemático-etimológico. Madrid: Imprenta y Librería de D. E.
Aguado.
Porto Dapena, J. Á. (2002). Manual de técnica lexicográfica. Madrid: Arco/Libros.
Quirós, M. (2010). Conclusiones. In J. R. Carriazo & M. Gómez (Eds.), La marcación en lexicografía histórica
(pp. 317–335). San Millán de la Cogolla: Cilengua.
67
DOSSIER thématique
La confección del Glosario de aritmética y álgebra en el Renacimiento hispano como instrumento para el estudio ...
Itziar Molina Sangüesa
Real Academia Española (1933–36). Diccionario histórico de la lengua española. Madrid: Casa Editorial Hernando.
[Aut]: ———— . (1990 [1726–39]). Diccionario de Autoridades. Madrid: Gredos.
[NTLLE]: ———— . (2001). Nuevo tesoro lexicográfico de la lengua española (edición DVD). Madrid: Espasa
Calpe.
[DLE]: ———— . (201423). Diccionario de la lengua española. Madrid: Espasa Calpe.
[CORDE]: ———— . Corpus diacrónico del español. <http://www.rae.es>
Salavert Fabiani, V. L. (1994). Aritmética y sociedad en la España del siglo XVI. In S. Gama, D. Flament &
V. Navarro (Eds.), Contra los titanes de la rutina (pp. 51–70). Madrid: CSIC.
[Terr.]: Terreros y Pando, E. (1987 [1786–1793]). Diccionario castellano con las voces de las ciencias y las artes.
Madrid: Arco/Libros.
[TLFi]: Trésor de la Langue Française informatisé. <http://atilf.atilf.fr/tlf.html>
VV. AA. (200413). Diccionario de sinónimos y antónimos. Madrid: Espasa-Calpe.
Warusfel, A. (1972). Diccionario razonado de matemáticas. De las matemáticas clásicas a la matemática. Madrid: Tecnos.
Werner, R. (1982). La definición lexicográfica. In G. Haensch et al, La lexicografía. De la lingüística teórica
a la lexicografía práctica (pp. 259–328). Madrid: Gredos.
38 / 2017 / 2
DOSSIER thématique
La confección del Glosario de aritmética y álgebra en el Renacimiento hispano como instrumento para el estudio ...
68
|
|
https://openalex.org/W2051246024
|
https://journals.plos.org/plosone/article/file?id=10.1371/journal.pone.0111414&type=printable
|
English
| null |
Influence of Fuels, Weather and the Built Environment on the Exposure of Property to Wildfire
|
PloS one
| 2,014
|
cc-by
| 8,118
|
Introduction the intensity and rate of spread are lower and containment of the
fires through active suppression is more likely to be successful
[21,22,24]. Simulation studies have found individual fuel breaks
have the potential to reduce the size and intensity of wildfires
[25,26]. While studies examining the impact of fuel breaks on the
behaviour of individual fires are valuable, to quantify the extent to
which fuel breaks reduce risk to lives and property, we need to
examine the role of fuel breaks at the landscape scale [27]. Although simulation studies have shown management of fuels can
alter fire regimes in forested ecosystems, there is a need to quantify
how fuel breaks, in particular, can reduce the risk of exposure of
assets. Here we examine the performance of fuel breaks in
mitigating risk in Mediterranean shrubland (chaparral) landscapes. Wildfires can pose a significant risk to people and property [1–
4]. Large losses of property and life have been recorded from
individual fires or fire complexes in fire prone regions throughout
the globe [5–7]. Such events can impact individuals and
communities for many years [8–11]. As a result, fire management
agencies invest significant budgets in reducing the risk of loss from
wildfires, primarily through investment in fuel management and
active fire suppression [12,13]. Fuel management is a commonly used risk management tool in
fire prone landscapes [14,15]. The justification for the approach
derives from the fundamentals of fire behaviour with a reduction
in fuel loads expected to result in a subsequent lowering of the fire
intensity and rate of spread [16–20]. In turn, these changes to fire
behaviour are expected to increase the probability of successfully
containing the fire with active fire suppression [21,22]. Fuel breaks are the main fuel treatment carried out in the
chaparral shrublands of southern California, with a long history of
extensive deployment [23,24]. Thus case studies of their effective-
ness in mitigating risk can provide valuable insight into mitigation
of risk in a region with some of the highest exposure of fire-prone
urban and peri-urban developments in the world. Case studies in
this context may also be valuable for assessing fuel treatment
options in other fire-prone, temperate environments which share
similar elements of the problem [28]. Fuel breaks are a commonly applied type of fuel management
treatment in a variety of ecosystems. ceived June 11, 2014; Accepted September 12, 2014; Published October 31, 2014 Received June 11, 2014; Accepted September 12, 2014; Published October 31, 2014 This is an open-access article, free of all copyright, and may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for
any lawful purpose. The work is made available under the Creative Commons CC0 public domain dedication. ayesian Network files are available from the ABNMS data repository database (http://abnms.org/bnrepo/bn?bnId = 98). The data can also
http://figshare.com/articles/San_Diego_County_Fuel_Breaks_Bayesian_Network/1171225. ta Availability: All Bayesian Network files are available from the ABNMS data repository database (http://abnms.org/bnrepo/bn?bnId =
accessed at Figshare: http://figshare.com/articles/San_Diego_County_Fuel_Breaks_Bayesian_Network/1171225. Funding: Funding from the U.S. Geological Survey was provided to University of Wollongong to cover salary of LC for the study. The funders had no role in study
design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing Interests: The authors have declared that no competing interests exist. Competing Interests: The authors have declared that no competing interests exist. * Email: trent.penman@unimelb.edu.au ¤ Current address: Department of Forest and Ecosystem Science, University of Melbourne, Victoria, Australia. ¤ Current address: Department of Forest and Ecosystem Science, University of Melbourne, Victoria, Australia. nman1*¤, Luke Collins2, Alexandra D. Syphard3, Jon E. Keeley4,5, Ross A. Bradstock 1 Centre for Environmental Risk Management of Bushfires, Institute of Conservation Biology and Environmental Management, University of Wollongong, NSW, Australia,
2 Hawkesbury Institute for the Environment, University of Western Sydney, Penrith, NSW, Australia, 3 Conservation Biology Institute, La Mesa, California, United States of
America, 4 U.S. Geological Survey, Western Ecological Research Center, Sequoia-Kings Canyon Field Station, Three Rivers, California, United States of America,
5 Department of Ecology and Evolutionary Biology, University of California Los Angeles, Los Angeles, California, United States of America October 2014 | Volume 9 | Issue 10 | e111414 Abstract Wildfires can pose a significant risk to people and property. Billions of dollars are spent investing in fire management actions
in an attempt to reduce the risk of loss. One of the key areas where money is spent is through fuel treatment – either fuel
reduction (prescribed fire) or fuel removal (fuel breaks). Individual treatments can influence fire size and the maximum
distance travelled from the ignition and presumably risk, but few studies have examined the landscape level effectiveness of
these treatments. Here we use a Bayesian Network model to examine the relative influence of the built and natural
environment, weather, fuel and fuel treatments in determining the risk posed from wildfire to the wildland-urban interface. Fire size and distance travelled was influenced most strongly by weather, with exposure to fires most sensitive to changes in
the built environment and fire parameters. Natural environment variables and fuel load all had minor influences on fire size,
distance travelled and exposure of assets. These results suggest that management of fuels provided minimal reductions in
risk to assets and adequate planning of the changes in the built environment to cope with the expansion of human
populations is going to be vital for managing risk from fire under future climates. Citation: Penman TD, Collins L, Syphard AD, Keeley JE, Bradstock RA (2014) Influence of Fuels, Weather and the Built Environment on the Exposure of Property to
Wildfire. PLoS ONE 9(10): e111414. doi:10.1371/journal.pone.0111414
Edi
Ri
b
D
CSIRO A
li Editor: Ritaban Dutta, CSIRO, Australia What Influences the Exposure of Property to Wildfire? What Influences the Exposure of Property to Wildfire? Bayesian Networks (BN) provide a suitable methodology for the
analysis of risk management problems [29–31]. They are depicted
as directed acyclic graphs with nodes representing the variables
and arrows representing the directional relationships between
nodes [32]. There is a conditional probability table for each node
that contains the joint probability distributions for the variable
[33]. Root nodes occur at the top of the model and are not
influenced by other variables in the model. These nodes have a
conditional probability table containing a single probability for
each state in that node. Child nodes are variables that are
influenced by one or more variables (parent nodes). These nodes
have a conditional probability table that represents the probability
of a given state in the child node given the state(s) in the parent
node(s). Uncertainty
is
propagated
throughout
the
model,
providing probability distribution for all output nodes. Results
are
likelihoods
that
form
the
basis
for
risk
management
calculations [29]. Bayesian Networks have been found to be a
valuable method for examining fire risk management problems at
the landscape scale [31,34]. influence diagrams to depict the relationships of the conceptual
model and finally populate all the conditional probability tables
within the model. A basic conceptual model for the study was derived from
previous fire management research [31,34]. This model assumed
an ignition had occurred and predicted the subsequent spread and
potential impact of the wildfire upon property, which was
dependent upon environmental conditions and management
decisions. Influence diagrams encapsulated the conceptual framework and
the relevant environmental factors (Figure 2). These were devel-
oped for the study area by the authors through a series of
workshops held by the United States Geological Survey (Decem-
ber 2010, September 2011, May 2012) involving twelve research-
ers with expertise in fire management, Bayesian Network analysis
and landscape ecology. Iterations of the influence diagrams were
presented to the group, discussed and then further refined until a
consensus was reached. In the final set of influence diagrams, fire
size and distance travelled were assumed to be influenced by the
key variables considered in the simulation modelling – weather,
fuel moisture, landscape fuel load and the occurrence of fuel
breaks within the National Forest. Elements of the natural
environment at the ignition location, specifically fuel type, fuel
load and elevation, were considered to have an influence on fire
size and distance travelled. What Influences the Exposure of Property to Wildfire? The built environment also influences
fire spread and exposure of property. Exposure to fire was taken as
simple function considering the distance the fire could potentially
travel and the distance from the ignition point to property
(Figure 2). Here we develop a Bayesian Network model to examine the role
of fuel breaks in reducing the risk of assets being exposed to
wildfire using San Diego county as a case study area. San Diego
county has a history of major fire losses (circa 5000 houses
destroyed between 2000 and the present), which reflects extensive
and rapidly growing developments, exposure to regular episodes of
severe fire weather and terrain and vegetation conducive to the
rapid spread of intense fires [35]. Thus the county comprises
encapsulates key elements and a significant portion of the wildfire
risk problem in southern California. As a case study it therefore
provides the potential for key insights into fire management that
are regionally, nationally and globally significant. The BN model
combines data from a fire simulation model (FARSITE) and
environmental data. We specifically seek to determine the extent
to which risk posed by wildfires to properties at the wildland urban
interface is influenced by the environment (weather, fuel moisture,
natural environment), developmental patterns (built environment)
and fuel management (fuel load, fuel breaks). Data for the conditional probability tables in the analysis
(Table 1) were derived from either a simulation study or empirical
data for the study region. We undertook a comprehensive
simulation of fires in the area using the Fire Area Simulator
(FARSITE) using random ignition locations. FARSITE is a two
dimensional spatially explicit model that models fire spread using
Huygens’ principle [43]. The simulations examined all combina-
tions of fire weather (low, high and Santa Ana), live fuel moisture
(LMF 60% and 90%), fuel loading (low and high) and the presence
or absence of maintained fuel breaks. A total of 11,944 fires were
simulated a FARSITE framework. Complete details of the
approach are presented in Table S1 in Material S1. From the
simulation data, we derived values for the nodes fire size and
distance travelled. Weather, fuel moisture, landscape fuel load and
the occurrence of fuel breaks were implemented as decision nodes
to explore the relative influence of each factor. Environmental
variables (i.e. elevation, ignition fuel load, ignition fuel type and
distance to the interface) were derived from GIS data from the
region (www.landfire.gov Accessed October 2012). What Influences the Exposure of Property to Wildfire? All nodes and
methods of discretisation are described in more detail in Table 1. The Bayesian Network model is available from the ABNMS data
repository (www.abnms.org). Introduction These are mechanical
reductions or removal of fuel, typically as linear features along
ridgetops, to enable safe access for fire suppression crews to
manage fires [23]. Empirical analysis has found fuel breaks are
more effective when readily accessible and well-maintained and
when used for backfire operations [23,24]. Under these conditions October 2014 | Volume 9 | Issue 10 | e111414 1 PLOS ONE | www.plosone.org October 2014 | Volume 9 | Issue 10 | e111414 Methods The study area was San Diego County, California, USA, which
supports a population of approximately 3.2 million people in an
area of approximately 11 000 km2 (US Census, http://quickfacts. census.gov/qfd/states/06/06073.html, Accessed December 2013). In the county, there is a long and complex wildland urban
interface [36], along which thousands of homes have been
destroyed in major fire events in the last decade [37]. Native
vegetation of the area is dominated by chaparral, coastal sage
scrub, and oak woodland. The county experiences a Mediterra-
nean climate with hot dry summers and winter rainfall with
moderate temperatures. Fires in the county occur most frequently
in summer months, but most area burned occurs in the autumn
when annual fuel moisture is lowest and Santa Ana winds are most
frequent [38]. San Diego County was selected as it is dominated
by highly flammable shrubland vegetation and falls within known
Santa Ana wind corridors, making the region prone to recurrent
large fire events [39,40]. The simulation landscape was defined as
a 60 km660 km area east of San Diego (Figure 1). Two methods were used to examine the relative influence of
variables. Firstly, the relative influence of each of the modelled
factors was assessed using values of the terminal node – ‘‘exposure
to fire’’. This node reflects the risk of property being exposed to a
wildfire under the given conditions. We considered all 24
combinations of the key predictor variables – weather (3 levels),
fuel moisture (2 levels), landscape fuel load (2 levels) and fuel
breaks (2 levels) (Table 1). Secondly, the sensitivity of nodes was
assessed using the sensitivity to findings function in Netica (http://
www.norsys.com/netica.html, Accessed December 2013). This
function examines the extent to which changes in one variable
affects the variable of interest. We examined the sensitivity of A Bayesian Network model was used to examine the relative
influence of weather, the built and natural environment and fuel
breaks on the risk of exposure to wildfire. Here we broadly follow
the methods for developing Bayesian Networks recommended by
Marcot et al. [41] and Chen and Pollino [42]. The primary steps
used were to construct a conceptual model of the problem, develop October 2014 | Volume 9 | Issue 10 | e111414 PLOS ONE | www.plosone.org 2 What Influences the Exposure of Property to Wildfire? Figure 1. Location of the study area, San Diego County, USA. doi:10.1371/journal.pone.0111414.g001 Figure 1. Methods Location of the study area, San Diego County, USA. doi:10.1371/journal.pone.0111414.g001 findings for ‘‘exposure to fire’’ as the terminal node and ‘‘distance
travelled’’. findings for ‘‘exposure to fire’’ as the terminal node and ‘‘distance
travelled’’. distance travelled to the predictor variables were consistent. Fire
size and distance travelled increased with the severity of fire
weather and the landscape fuel load, and decreased with
increasing fuel moisture. The presence of fuel breaks had little
influence on either individual fire size or distance travelled
(Figure 3). Results The size of fires after a 12 hour simulation ranged from 0.1 ha
to 28,480 ha, with a mean of 2896 ha (640 S.E.) and a median
882 ha. The distance travelled ranged from 18 m to 47,300 m
with a mean of 6171 m (657 S.E.). Responses of fire size and Risk of exposure to fires was influenced most strongly by
weather (Figure 4). The risk of exposure was .99% for all fuel Figure 2. Influence diagrams for the Bayesian Network Model See Table 1 for node definitions and states. doi:10.1371/journal.pone.0111414.g002
. Figure 2. Influence diagrams for the Bayesian Network Model See Table 1 for node definitions and states. doi:10.1371/journal.pone.0111414.g002
. October 2014 | Volume 9 | Issue 10 | e111414 PLOS ONE | www.plosone.org PLOS ONE | www.plosone.org 3 What Influences the Exposure of Property to Wildfire? Table 1. Nodes, definitions and states used in the Bayesian Network model. Node
Description
Levels
Distance Travelled (O)
The maximum distance the fire travelled. 0 to 0.5 km; 0.5 to 1.5 km; 1.5 to 4 km; 4 to 10 km; 10 to 15 km; .15 km
Distance to the coast (GIS)
Distance from the ignition point to the coast. 0 to 10 km; 10 to 25 km; 25 to 42 km; 42 to 68 km; 68 to 80 km; . 80 km
Distance to road (GIS)
The distance from the ignition point
to the nearest mapped road. 0 to 50 m; 50 to 100 m; 100 to 500 m; 500 to 1000 m; 1000 to 2000 m;
.2000 m
Distance to structure (GIS)
The distance from the ignition point
to the nearest mapped house. Risk of exposure Weather had the strongest influence on fire size and distance
travelled (Figure 5b) and indirectly, the risk of exposure. It has
been well documented that fire weather strongly influences fire
size, the rate of spread, spotting distance, fire intensity and severity
[16,18,40,50–52]. As a result, wildfires burning under extreme
conditions account for the majority of area burnt for many regions
[44,53–55]. Furthermore, it is under extreme fire weather
conditions wildfires pose the greatest threat to people and property
[50,56–59]. Exposure to fires was most sensitive to changes in the built
environment, as well as fire parameters, i.e. fire size and distance
travelled (Figure 5a). Distance to structures from the ignition had
the strongest influence, followed by the distance travelled by a fire
and the fire size. This is expected as parent nodes are likely to have
the strongest influence on a node. Of the variables a greater
distance from the terminal node, variables depicting the built
environment (housing density and distance to road) were the next
most influential variables, followed by weather. Variables describ-
ing the natural environment had only a modest influence on
exposure to fires, with the fuel variables having no influence
(Figure 5a). Distance travelled by a fire was primarily influenced
by the weather on the day of the fire (Figure 5b). Natural
environment variables and fuel load all had minor influence (,
2.1%), with all built environment variables having low influence
(,1%) (Figure 5b). Exposure was also influenced by the built environment, namely
distance to road and housing density (Figure 5). Higher densities of
properties occur at lower elevations (less than 300 m) and these
have a higher risk of exposure compared with higher elevation
sites independent of weather (Figure 4). Fires starting close to the
interface are more likely to impact upon assets under any weather
conditions [60,61], whereas fires starting considerable distances
from property are only likely to impact on property under extreme
weather conditions conducive to fire spread [34,57]. These results
suggest that adequate planning of the changes in the built
environment to cope with the expansion of human populations is
going to be important for managing risk from fire [49,62,63]. We
do note that the model only considers exposure to fire and we have
not considered the size of the fire or the extent of the interface
exposed to the fire. Results 0 to 50 m; 50 to 100 m; 100 to 500 m; 500 to 1000 m; 1000 to 2000 m;
.2000 m
Elevation (GIS)
Elevation of the ignition point
0 to 300 m; 300 to 600 m; 6000 to 1000 m; 1000 to 4000 m; .4000 m
Exposure (C)
Are houses exposed by the fire
Yes; No
Fire size (O)
The final size of the simulated fire
0 to 20 ha: 20 to 150 ha: 150 to 1000 ha: 1000 to 5000 ha: 5000 to
10000 ha: .10000 ha
Fuel breaks (S)
Whether fuel breaks are present
Yes; No
Fuel load (S)
Landscape fuel load
High (2001); Low (2008)
Fuel moisture(S)
Live fuel moisture level at the start
of the simulation
60%; 90%
Fuel type at ignition (GIS)
Broad classification of the type of
vegetation at the point of ignition. Grass; Shrub; Tree
Housing density (GIS)
Number of houses per hectare
0 to 26; 26 to 33.5; 33.5 to 117; 117 to 205; 205 to 300; .300
Slope (GIS)
Slope under the ignition point
0 degrees; 0 to 5 degrees; 5 to 15 degrees; 15 to 30 degrees; .30
degrees
Weather (S)
Predominant conditions during
the simulated fire
Low; High; Santa Ana
(C) = calculated variable; (GIS) = GIS derived variable; (O) = output of the simulation model; (S) = variable set in the simulation model. doi:10.1371/journal.pone.0111414.t001 Table 1. Nodes, definitions and states used in the Bayesian Network model. = calculated variable; (GIS) = GIS derived variable; (O) = output of the simulation model; (S) = variable set in the simulation model. 10.1371/journal.pone.0111414.t001 scenarios considered under Santa Ana conditions. Under low and
high weather conditions, fuel load and fuel moisture had a strong
influence of risk of exposure, whereby risk was greater when
landscape fuels were high compared to low and when fuel
moisture was 60% compared to 90% (Figure 4). These effects were
lost under Santa Ana conditions, where fuel load and fuel moisture
had very little effect on the risk of exposure. Fuel breaks had very
little influence on the risk of exposure. Risk varied significantly
across elevation with fires starting in low elevation sites having
significantly higher risk than fires starting at higher elevation sites
(Figure 4). fire (size and distance travelled) and the nature of the built
environment [34,49]. Measured attributes of fuels had only a
minor influence on fire parameters and risk of exposure. Risk of exposure Fires that start away from populations will be
significantly larger when they do impact on the interface
compared with those that start nearby [60]. Larger fires would Results Fuel
breaks in the National Forest did not affect fire size, distance
travelled or the risk of exposure at the interface. October 2014 | Volume 9 | Issue 10 | e111414 Discussion Consistent with previous research, fire size and distance
travelled is most sensitive to changes in weather [18,44–48] and
the risk of exposure is most strongly influenced by attributes of the October 2014 | Volume 9 | Issue 10 | e111414 October 2014 | Volume 9 | Issue 10 | e111414 PLOS ONE | www.plosone.org 4 What Influences the Exposure of Property to Wildfire? Figure 3. Relationships from the FARSITE simulation data between weather and fire size with fuel moisture of a) 60% and b) 90%,
and distance travelled with fuel moisture of c) 60% and d) 90%. Open symbols are for simulations with no fuel breaks, closed symbols for
simulations with fuel breaks. Circles represent a high landscape fuel load scenario and triangles represent a low landscape fuel load scenario. NB 95%
confidence intervals were too small to depict in the graphics. doi:10.1371/journal.pone.0111414.g003 Figure 3. Relationships from the FARSITE simulation data between weather and fire size with fuel moisture of a) 60% and b) 90%,
and distance travelled with fuel moisture of c) 60% and d) 90%. Open symbols are for simulations with no fuel breaks, closed symbols for
simulations with fuel breaks. Circles represent a high landscape fuel load scenario and triangles represent a low landscape fuel load scenario. NB 95%
confidence intervals were too small to depict in the graphics. doi:10.1371/journal.pone.0111414.g003 be expected to impact on a greater number of assets than smaller
fires and these relationships require further investigation. levels (,60–80%) [65,66]. Early in the fire season live fuel
moisture is generally greater than 90% (Keeley et al. 2009), the
resulting area burnt in southern California typically is relatively
small [65,66]. Our selection of 60% and 90% may not have truly
captured the variable effect of live fuel moisture, particularly when
these values exceed 100% and fire activity is expected to be low. However, the greatest risk to assets comes during Santa Ana
weather conditions where there is no distinguishable effect of live
fuel moisture, providing further support for our existing results. Regardless, it is typically lowest at the end of summer drought
when Santa Ana winds and hence large fires are most likely [67]. Of the factors relating to fuels, landscape fuel load had the
strongest influence on fire size, distance travelled and risk of
exposure (Figure 3; Figure 4). October 2014 | Volume 9 | Issue 10 | e111414 Discussion Risk of exposure for the 24 scenarios modelled for a) fires igniting at elevations of 300 to 600 m; b) fires igniting a
elevations of 1000 to 4000 m; c) all locations across the landscape. Open symbols = fuel break scenarios; closed symbols = no fuel breaks
Grey symbols = fuel moisture of 60%; Black symbols = fuel moisture of 90%; Circles = high landscape fuel loads; Triangle = low landscape fuel loads
doi:10.1371/journal.pone.0111414.g004 Figure 4. Risk of exposure for the 24 scenarios modelled for a) fires igniting at elevations of 300 to 600 m; b) fires igniting at
elevations of 1000 to 4000 m; c) all locations across the landscape. Open symbols = fuel break scenarios; closed symbols = no fuel breaks;
Grey symbols = fuel moisture of 60%; Black symbols = fuel moisture of 90%; Circles = high landscape fuel loads; Triangle = low landscape fuel loads. doi:10.1371/journal.pone.0111414.g004 landscapes [60,68–70] and tend to occur close to roads and
development. This is important with regards to the result that fires
that start closer to homes are most likely to reach those homes. Simulations have revealed that fire size and burn probability are
sensitive to the use of random against non-random ignition
locations, though these biases are minimised under extreme
weather conditions [71] when the greatest risk of exposure occurs
(Figure 4). of suppression through direct attack of the fire front or indirect
attack from other breaks in the landscape, e.g. roads and rivers. Inclusion of the suppression at these points may have altered the
efficacy of fuel breaks when estimating risk. Similarly, we did not
consider the impact on fires of igniting backburns from fuel breaks. However, as the severity of the fire weather increases the
effectiveness of suppression actions are severely diminished
[22,48,72]. Therefore, we would only consider fuel breaks in
conjunction with suppression as having potential to further reduce
risk under low fire weather and not under moderate fire weather
or Santa Ana conditions [24,39]. Simulations have revealed that fire size and burn probability are
sensitive to the use of random against non-random ignition
locations, though these biases are minimised under extreme
weather conditions [71] when the greatest risk of exposure occurs
(Figure 4). The effectiveness of fuel breaks is contingent on suppression
resources and access [24]. In our study, the fuel breaks were
constructed in FARSITE in a manner that simulated suppression
along the fuel breaks. Discussion Fuel breaks tend to be constructed to allow
for control of the fire flanks and not the head fire, i.e. the point of
the greatest forward rate of spread. As a result, fuel breaks are
unlikely to affect the distance travelled by a fire and have negligible
impacts on total fire size. Our model did not model the interaction Discussion However, our model is more
sensitive to the effects of weather and the built environment
(Figure 5). Price et al. [64] found no effect of antecedent area
burnt on the annual area burnt by wildfire in southern California
coastal systems. The authors argue that the low effect of past fire is
related to the low level of wildfire in the system (,2% per annum)
and the rapid development of fuels (1–2 years). A large proportion
of the study area (,22%) burnt in wildfires during 2007 and was
consequently in the early stage of fuel development (i.e. 1 year old)
in the 2008 fuel layer, which may explain the strong influence of
fuel load under low and high fire weather. However, in San Diego
County, the majority of annual area burnt occurs under extreme
Santa Ana fire weather [55] where our model found no effect of
fuel load. These results support the finding of Price et al. [64] that
landscape fuel treatments in these systems are unlikely to reduce
the risk of fire to assets. (Figure 3; Figure 4). Scenarios with 90%
fuel moisture had significantly lower risk than those with 60%
under low and high conditions, but not under Santa Ana
(Figure 4). Live fuel moisture is related to fire activity in southern
California, with large fires generally being associated with low Fuel breaks were ineffective at altering risk of exposure of
property under any weather scenario in our study. Here we
modelled fires assuming that all mapped fuel breaks in San Diego
County were maintained (see Material S1; Figure 1), which
exceeds current practice. Fuel breaks have been found to affect
individual fire size and distance travelled [25,26]. The network of
fuel breaks in San Diego County is highly clustered (Figure 1)
presumably to protect particular assets from future wildfires,
although fuel breaks continue to be constructed. Clustering of the
fuel breaks will result in low encounter rates with wildfires that will
result in a low efficacy of this management technique [24] when
considering the landscape level risk. Here we assumed a random
ignition model, however ignitions do not occur randomly across October 2014 | Volume 9 | Issue 10 | e111414 October 2014 | Volume 9 | Issue 10 | e111414 PLOS ONE | www.plosone.org 5 What Influences the Exposure of Property to Wildfire? Figure 4. October 2014 | Volume 9 | Issue 10 | e111414 Fire management Management agencies seek to reduce risk to assets acknowledg-
ing that there are no practical means to remove the risk. Weather
is the primary determinate of risk to assets from fire [34,55,57]. October 2014 | Volume 9 | Issue 10 | e111414 PLOS ONE | www.plosone.org 6 Whil
i
b
ff
i
d
l i
l
b ildi
i
l
i k
id S
A
i d
id
[35
igure 5. Sensitivity to findings for nodes a) Exposure to fire and b) Distance travelled. White bars = fire variables, dark grey bars = buil
nvironment variables; light grey bars = natural environment variables; Black bars = simulation model variables. D2S = distance to structure
DistTrav = distance travelled by the fire; HouseDens = housing density; D2Rd = Distance to road; D2C = distance to the coast; IgFuelType = fuel type a
he point of ignition; IgFuels = fuel load at the point of ignition; FuelBreaks = presence of fuel breaks. oi:10.1371/journal.pone.0111414.g005
What Influences the Exposure of Property to Wildfire What Influences the Exposure of Property to Wildfire? ure 5. Sensitivity to findings for nodes a) Exposure to fire and b) Distance travelled. White bars = fire variables, dark grey bars = Figure 5. Sensitivity to findings for nodes a) Exposure to fire and b) Distance travelled. White bars = fire variables, dark grey bars = built
environment variables; light grey bars = natural environment variables; Black bars = simulation model variables. D2S = distance to structure;
DistTrav = distance travelled by the fire; HouseDens = housing density; D2Rd = Distance to road; D2C = distance to the coast; IgFuelType = fuel type at
the point of ignition; IgFuels = fuel load at the point of ignition; FuelBreaks = presence of fuel breaks. doi:10.1371/journal.pone.0111414.g005 building in low risk areas outside Santa Ana wind corridors [35]
and incorporating adequate offsets between vegetation and
structures [74,75]. In southern California, the best urban planning
practices would be to focus on infill-type development, as low to
intermediate housing density, and isolated clusters of development
are the strongest risk factors for a house being destroyed in a fire
[63]. Finally, reduce the vulnerability of residents and properties at
the urban interface. Residents can be educated to reduce the
vulnerability of their property through adequate preparation
[76,77]. Fire management Furthermore, properties can be built or retrofitted to
appropriate construction standards to be more resilient to the
impact of fire [78,79]. While each of these is likely to reduce risk,
only through an expanded analysis of these approaches across all
weather scenarios will it be possible to identify an optimal
management strategy. While management actions can be effective under relatively
benign fire weather, understanding the effectiveness of manage-
ment under extreme fire weather is fundamental to determining
the extent to which management can reduce risk to people and
property [48]. In our model, we considered the role of fuel
treatments both fuel breaks for suppression and fuel treatments
through examining the role of fuel loads. Neither of these was
effective under extreme fire weather despite our model considering
extreme levels of fuel treatment (.20% in 1 year old fuel) and fuel
breaks (all mapped breaks in the county). A range of other fire management approaches are available that
were not tested here. Three broad management areas have the
potential to significantly reduce risk to assets. Firstly, ignition
management to reduce the occurrence of ignitions and subsequent
fires will reduce the risk to assets [44]. Included in ignition
management would be rapid response or initial attack [47,73],
whereby resources are used to aggressively attempt to suppress
fires before they become established fires. Secondly, improved
urban planning policies to better develop the built environment to
reduce the extent of the exposure [49,62,74]. This would include References Sciences of the United States of America 106: 10706–10711 13. Stephens SL, Ruth LW (2005) Federal forest-fire policy in the United States. Ecological Applications 15: 532–542. 35. Syphard AD, Keeley JE, Massada AB, Brennan TJ, Radeloff VC (2012)
Housing Arrangement and Location Determine the Likelihood of Housing Loss
Due to Wildfire. PLoS ONE 7: e33954. 14. Fernandes PM, Botelho HS (2003) A review of prescribed burning effectiveness
in fire hazard reduction. International Journal of Wildland Fire 12: 117–128. 15. Penman TD, Christie FJ, Andersen AN, Bradstock RA, Cary GJ, et al. (2011)
Prescribed burning: How can it work to conserve the things we value? International Journal of Wildland Fire 20: 721–733. 36. Hammer RB, Radeloff VC, Fried JS, Stewart SI (2007) Wildland-urban
interface housing growth during the 1990 s in California, Oregon, and
Washington. International Journal of Wildland Fire 16: 255–265. J
16. Noble IR, Bary GAV, Gill AM (1980) Mcarthurs Fire Danger Meters Expressed
as Equations. Australian Journal of Ecology 5: 201–204. 37. Keeley JE, Syphard AD, Fotheringham CJ (2013) The 2003 and 2007 Wildfires
in Southern California. In: Boulter S, Palutikof J, Karoly D, Guitart D, editors. Natural Disasters and Adaptation to Climate Change. Oxford: Cambridge
University Press. pp. 42–52. 17. Price OF, Bradstock RA (2010) The effect of fuel age on the spread of fire in
sclerophyll forest in the Sydney region of Australia. International Journal of
Wildland Fire 19: 35–45. 38. Keeley JE, Fotheringham CJ (2003) Impact of past, present, and future fire
regimes on North American mediterranean shrublands. In: Veblen TT, Baker
WL, Montenegro G, Swetnam TW, editors. Fire and climatic change in
temperate ecosystems of the Western Americas. New York: Springer. pp. 218–
262. 18. Bradstock R, Hammill K, Collins L, Price O (2010) Effects of weather, fuel and
terrain on fire severity in topographically diverse landscapes of south-eastern
Australia. Landscape Ecology 25: 607–619. 19. Fule´ PZ, Crouse JE, Roccaforte JP, Kalies EL (2012) Do thinning and/or
burning treatments in western USA ponderosa or Jeffrey pine-dominated forests
help restore natural fire behavior? Forest Ecology and Management 269: 68–81. 39. Keeley JE, Safford H, Fotheringham CJ, Franklin J, Moritz M (2009) The 2007
Southern California Wildfires: Lessons in Complexity. Journal of Forestry 107:
287–296. 20. Collins L, Bradstock RA, Penman TD (2013) Can precipitation influence
landscape controls on wildfire severity? A case study within temperate eucalypt
forests of south-eastern Australia. References 1. Cohen JD (2000) Preventing Disaster: Home Ignitability in the Wildland-Urban
Interface. Journal of Forestry 98: 15–21. 21. Hirsch KG, Podur JJ, Janser RF, McAlpine RS, Martell DL (2004) Productivity
of Ontario initial-attack fire crews: results of an expert-judgement elicitation
study. Canadian Journal of Forest Research 34: 705–715. 1. Cohen JD (2000) Preventing Disaster: Home Ignitability in the Wildland-Urban
Interface. Journal of Forestry 98: 15–21. 2. Lampin-Maillet C, Jappiot M, Long M, Bouillon C, Morge D, et al. (2009)
Mapping wildland-urban interfaces at large scales integrating housing density
and vegetation aggregation for fire prevention in the South of France. Journal of
Environmental Management 91: 732–741. 2. Lampin-Maillet C, Jappiot M, Long M, Bouillon C, Morge D, et al. (2009)
Mapping wildland-urban interfaces at large scales integrating housing density
and vegetation aggregation for fire prevention in the South of France. Journal of
Environmental Management 91: 732–741. 22. Plucinski MP (2012) Factors Affecting Containment Area and Time of
Australian Forest Fires Featuring Aerial Suppression. Forest Science 58: 390–
398. 23. Syphard AD, Keeley JE, Brennan TJ (2011) Comparing the role of fuel breaks
across southern California national forests. Forest Ecology and Management
261: 2038–2048. 3. Gill AM, Stephens SL (2009) Scientific and social challenges for the
management of fire-prone wildland–urban interfaces. Environmental Research
Letters 4: 034014. 4. Gill AM, Stephens SL, Cary GJ (2013) The worldwide ‘‘wildfire’’ problem. Ecological Applications 23: 438–454. 24. Syphard AD, Keeley JE, Brennan TJ (2011) Factors affecting fuel break
effectiveness in the control of large fires on the Los Padres National Forest,
California. International Journal of Wildland Fire 20: 764–775. 5. McCaffrey S, Rhodes A (2009) Public Response to Wildfire: Is the
Australian’’Stay and Defend or Leave Early’’ Approach an Option for Wildfire
Management in the United States? Journal of Forestry 107: 9–15. 25. Agee JK, Bahro B, Finney MA, Omi PN, Sapsis DB, et al. (2000) The use of
shaded fuelbreaks in landscape fire management. Forest Ecology and
Management 127: 55–66. 6. Leonard J (2009) Report to the 2009 Victorian Bushfires Royal Commission
Building performacnce in Bushfires. Highett, Victoria, Australia: CSIRO
Sustainable Ecosystems. 74 p. 26. Finney MA (2001) Design of Regular Landscape Fuel Treatment Patterns for
Modifying Fire Growth and Behavior. Forest Science 47: 219–228. 27. Syphard AD, Scheller RM, Ward BC, Spencer WD, Strittholt JR (2011)
Simulating landscape-scale effects of fuels treatments in the Sierra Nevada,
California, USA. Supporting Information Material S1
Supplementary text outlines the modelling process
in farsite. Table S1, Fuel moisture conditions used in the
simulations. Dead fuel moisture values are from Scott and Burgin
(2005). See text for description of LFM categories. What Influences the Exposure of Property to Wildfire? What Influences the Exposure of Property to Wildfire? greatest, landscape fuel treatments are unlikely to have a
significant influence on risk. These results suggest that managing
the occurrence of fire and the spatial distribution of the built
environment across the landscape is likely to be the best way to
alter the risk profile. Further research is needed to examine the
cost trade-offs of each of these approaches. Acknowledgments Data for the study was provided by U.S. Geological Survey. The paper was
made possible due to the workshop participants - Teresa Brennan, Will
Forney, CJ Fotheringham, Rick Halsey, Bill Labiosa, Ben Landis, Owen
Price, Robert Taylor and Marti Winter. Stuart Brittain and Mark Finney
kindly provided a command line version of FARSITE and assistance with
its use. Robert Taylor provided advice on manipulating input layers for
FARSITE to simulate spotting in shrubland fuels. Author Contributions Conceived and designed the experiments: TP LC AS JK RB. Performed
the experiments: LC TP. Analyzed the data: TP LC AS. Contributed to
the writing of the manuscript: TP LC AS JK RB. Designed the specifics of
the computer simulation study: TP LC AS. Conceived and designed the experiments: TP LC AS JK RB. Performed
the experiments: LC TP. Analyzed the data: TP LC AS. Contributed to
the writing of the manuscript: TP LC AS JK RB. Designed the specifics of
the computer simulation study: TP LC AS. References International Journal of Wildland Fire 20: 364–383. 7. Boustras G, Boukas N, Katsaros E, Ziliaskopoulos A (2012) Wildland fire
preparedness is Greece and Cyprus: Lessons learned from the catastrophic fires
of 2007 and beyond. In: Paton D, Tedim F, editors. Wildfire and community:
Facilitating preparedness and resilience. Springfield, Illinois, USA: Charles C
Thomas Publisher Ltd. pp. 151–168. California, USA. International Journal of Wildland Fire 20: 364–383. 28. Keeley JE, Bond WJ, Bradstock RA, Pausas, editors (2012) Fire in
Mediterranean Ecosystems: Ecology, Evolution and Management. Cambridge:
Cambridge University Press. Thomas Publisher Ltd. pp. 151–168. 8. Ganewatta G (2008) The economics of bushfire management. In: Handmer J,
Haynes K, editors. Community Bushfire Safety. Collingwood: CSIRO
Publishing. pp. 151–159. 29. Marcot BG, Holthausen RS, Raphael MG, Rowland MM, Wisdom MJ (2001)
Using Bayesian belief networks to evaluate fish and wildlife population viability
under land management alternatives from an environmental impact statement. Forest Ecology and Management 153: 29–42. 9. McFarlane AC, Clayer JR, Bookless CL (1997) Psychiatric morbidity following a
natural disaster: An Australian bushfire. Social Psychiatry and Psychiatric
Epidemiology 32: 261–268. 30. Johnson S, Mengersen K, de Waal A, Marnewick K, Cilliers D, et al. (2010)
Modelling cheetah relocation success in southern Africa using an Iterative
Bayesian Network Development Cycle. Ecological Modelling 221: 641–651. 10. Langley A, Jones R (2005) Coping Efforts and Efficacy, Acculturation, and Post-
Traumatic Symptomatology in Adolescents Following Wildfire. Fire Technology
41: 125–143. 31. Penman TD, Price O, Bradstock RA (2011) Bayes Nets as a method for
analysing the influence of management actions in fire planning. International
Journal of Wildland Fire 20: 909–920. 11. Papadatou D, Giannopoulou I, Bitsakou P, Bellali T, Talias MA, et al. (2012)
Adolescents’ reactions after a wildfire disaster in Greece. Journal of Traumatic
Stress 25: 57–63. 32. Pearl J (1986) Fusion, propagation, and structuring in belief networks. Artificial
Intelligence 29: 241–288. 12. Schoennagel T, Nelson CR, Theobald DM, Carnwath GC, Chapman TB
(2009) Implementation of National Fire Plan treatments near the wildland-urban
interface in the western United States. Proceedings of the National Academy of
Sciences of the United States of America 106: 10706–10711. 33. Korb KB, Nicholson AE (2011) Bayesian Artificial Intelligence. Computer
Science and Data Analysis. Second Edition. Boca Raton: CRC/Chapman Hall. 34. Penman TD, Bradstock RA, Price OF (2014) Reducing wildfire risk to urban
developments: Simulation of cost-effective fuel treatment solutions in south
eastern Australia. Environmental Modelling & Software 52: 166–175. Conclusion Weather determines the risk of exposure for assets in the
landscape. Under extreme weather, where the risk of fire is October 2014 | Volume 9 | Issue 10 | e111414 PLOS ONE | www.plosone.org October 2014 | Volume 9 | Issue 10 | e111414 7 What Influences the Exposure of Property to Wildfire? What Influences the Exposure of Property to Wildfire? 41. Marcot BG, Steventon JD, Sutherland GD, McCann RK (2006) Guidelines for
developing and updating Bayesian belief networs applied to ecological modeling
and conservation. Canadian Journal of Forest Research 36: 3063–3074. 60. Syphard AD, Radeloff VC, Keuler NS, Taylor RS, Hawbaker TJ, et al. (2008)
Predicting spatial patterns of fire on a southern California landscape. International Journal of Wildland Fire 17: 602–613. 42. Chen SH, Pollino CA (2012) Good practice in Bayesian network modelling. Environmental Modelling & Software 37: 134–145. J
61. Price OF, Bradstock RA (2013) The spatial domain of wildfire risk and response
in the Wildland Urban Interface in Sydney, Australia. Natural Hazards and
Environmental Management 1: 4539–4564. Environmental Modelling & Software 37: 134– g
43. Finney MA (1998) FARSITE: Fire Area Simulator-Model Development and
Evaluation. United States Department of Agriculture Forest Service. Evaluation. United States Department of Agriculture Forest Serv 62. Syphard AD, Radeloff VC, Keeley JE, Hawbaker TJ, Clayton MK, et al. (2007)
Human influence on California fire regimes. Ecological Applications 17: 1388–
1402. 44. Cary GJ, Flannigan MD, Keane RE, Bradstock RA, Davies ID, et al. (2009)
Relative importance of fuel management, ignition management and weather for
area burned: Evidence from five landscape-fire-succession models. International
Journal of Wildland Fire 18: 147–156. 63. Syphard AD, Bar Massada A, Butsic V, Keeley JE (2013) Land Use Planning
and Wildfire: Development Policies Influence Future Probability of Housing
Loss. PLoS ONE 8: e71708. 45. Archibald S, Roy DP, Van Wilgen BW, Scholes RJ (2009) What limits fire? An
examination of drivers of burnt area in Southern Africa. Global Change Biology
15: 613–630. 64. Price OF, Bradstock RA, Keeley JE, Syphard AD (2012) The impact of
antecedent fire area on burned area in southern California coastal ecosystems. Journal of Environmental Management 113: 301–307. 46. Bessie WC, Johnson EA (1995) The Relative Importance of Fuels and Weather
on Fire Behavior in Subalpine Forests. Ecology 76: 747–762. 65. Dennison PE, Moritz MA (2009) Critical live fuel moisture in chaparral
ecosystems: a threshold for fire activity and its relationship to antecedent
precipitation. International Journal of Wildland Fire 18: 1021–1027. 47. Arienti MC, Cumming SG, Boutin S (2006) Empirical models of forest fire initial
attack success probabilities: the effects of fuels, anthropogenic linear features, fire
weather, and management. Canadian Journal of Forest Research 36: 3155–
3166. 66. What Influences the Exposure of Property to Wildfire? Dennison PE, Moritz MA, Taylor RS (2008) Evaluating predictive models of
critical live fuel mois-ture in the Santa Monica Mountains, California. International Journal of Wildland Fire 17: 18–27. 48. Penman TD, Collins L, Price OF, Bradstock RA, Metcalf S, et al. (2013)
Examining the relative effects of fire weather, suppression and fuel treatment on
fire behaviour – A simulation study. Journal of Environmental Management
131: 325–333. 67. Keeley JE, Zedler PH (2009) Large, high intensity fire events in southern
California shrublands: debunking the fine-grained age-patch model. Ecological
Applications 19: 69–94. 49. Syphard AD, Keeley JE, Bar Massada A, Brennan TJ, Radeloff VC (2012)
Housing Arrangement and Location Determine the Likelihood of Housing Loss
Due to Wildfire. PLoS ONE 7. 68. Penman TD, Bradstock RA, Price O (2013) Modelling the determinants of
ignition in the Sydney Basin, Australia: implications for future management. International Journal of Wildland Fire 22: 469–478. 50. Price OF, Bradstock RA (2012) The efficacy of fuel treatment in mitigating
property loss during wildfires: Insights from analysis of the severity of the
catastrophic fires in 2009 in Victoria, Australia. Journal of Environmental
Management 113: 146–157. 69. Hering A, Bell C, Genton M (2009) Modeling spatio-temporal wildfire ignition
point patterns. Environmental and Ecological Statistics 16: 225–250. 70. Diaz-Avalos C, Peterson DL, Alvarado E, Ferguson SA, Besag JE (2001) Space-
time modelling of lightning-caused ignitions in the Blue Mountains, Oregon. Canadian Journal of Forest Research-Revue Canadienne De Recherche
Forestiere 31: 1579–1593. 51. Collins BM, Kelly M, van Wagtendonk JW, Stephens SL (2007) Spatial patterns
of large natural fires in Sierra Nevada wilderness areas. Landscape Ecology 22:
545–557. 52. Thompson JR, Spies TA (2009) Vegetation and weather explain variation in
crown damage within a large mixed-severity wildfire. Forest Ecology and
Management 258: 1684–1694. 71. Bar Massada A, Syphard AD, Hawbaker TJ, Stewart SI, Radeloff VC (2011)
Effects of ignition location models on the burn patterns of simulated wildfires. Environmental Modelling & Software 26: 583–592. 53. Bradstock RA, Cohn JS, Gill AM, Bedward M, Lucas C (2009) Prediction of the
probability of large fires in the Sydney region of south-eastern Australia using
fire weather. International Journal of Wildland Fire 18: 932–943. g
72. Hirsch KG, Corey PN, Martell DL (1998) Using expert judgement to model
initial attack fire crew effectiveness. Forest Science 44: 539–549. 73. References International Journal of Wildland Fire
Accepted April 3rd 2013. 40. Moritz MA, Moody TJ, Krawchuk MA, Hughes M, Hall A (2010) Spatial
variation in extreme winds predicts large wildfire locations in chaparral
ecosystems. Geophysical Research Letters 37: L04801. PLOS ONE | www.plosone.org October 2014 | Volume 9 | Issue 10 | e111414 October 2014 | Volume 9 | Issue 10 | e111414 8 What Influences the Exposure of Property to Wildfire? Plucinski MP, McCarthy GJ, Hollis JJ, Gould JS (2012) The effect of aerial
suppression on the containment time of Australian wildfires estimated by fire
management personnel. International Journal of Wildland Fire 21: 219–229. 54. Minnich R, Chou Y (1997) Wildland Fire Patch Dynamics in the Chaparral of
Southern California and Northern Baja California. International Journal of
Wildland Fire 7: 221–248. 74. Gibbons P, van Bommel L, Gill AM, Cary GJ, Driscoll DA, et al. (2012) Land
Management Practices Associated with House Loss in Wildfires. PLoS ONE 7:
e29212. 55. Keeley JE, Fotheringham CJ (2001) Historic Fire Regime in Southern California
Shrublands. Conservation Biology 15: 1536–1548. 56. Blanchi R, Lucas C, Leonard J, Finkele K (2010) Meteorological conditions and
wildfire-related houseloss in Australia. International Journal of Wildland Fire 19:
914–926. 75. Cohen JD (2004) Relating flame radiation to home ignition using modeling and
experimental crown fires. Canadian Journal of Forest Research 34: 1616–1626. 76. Penman TD, Eriksen C, Blanchi R, Chladil M, Gill AM, et al. (2013) Defining
adequate means of residents to prepare property for protection from wildfire. International Journal of Disaster Risk Reduction 6: 67–77. 57. Bradstock RA, Gill AM, Kenny BJ, Scott J (1998) Bushfire risk at the urban
interface estimated from historical weather records: Consequences for the use of
prescribed fire in the Sydney region of south-eastern Australia. Journal of
Environmental Management 52: 259–271. International Journal of Disaster Risk Reduction 6: 67–77 77. McLennan J, Elliott G, Omodei M, Whittaker J (2013) Householders’ safety-
related decisions, plans, actions and outcomes during the 7 February 2009
Victorian (Australia) wildfires. Fire Safety Journal 61: 175–184. g
58. Keeley JE, Fotheringham CJ, Moritz MA (2004) Lessons from the October 2003
Wildfires in Southern California. Journal of Forestry Research (Harbin) 102: 26–
31. 78. Standards Australia (2009) A.S.3959–2009: Construction of buildings in bushfire
prone areas. Sydney, Australia: SAI Global Limited. pp. 129. prone areas. Sydney, Australia: SAI Global Limited. pp. 129. 59. Bar Massada A, Radeloff VC, Stewart SI, Hawbaker TJ (2009) Wildfire risk in
the wildland-urban interface: A simulation study in northwestern Wisconsin. Forest Ecology and Management 258: 1990–1999. 79. CBC (2007) California Build Code Materials and construction methods for
exterior wildfire exposure Title 24, part 2, volume 1 of 2. Sacramento, CA. October 2014 | Volume 9 | Issue 10 | e111414 PLOS ONE | www.plosone.org 9
|
https://openalex.org/W4394883535
|
https://www.ippsvfu.ru/jour/article/download/236/232
|
Russian
| null |
The administrative and legal status of a higher educational institution: features and problems
|
Vestnik Severo-Vostočnogo federalʹnogo universiteta imeni M.K. Ammosova. Seriâ "Istoriâ. Politologiâ. Pravo"
| 2,024
|
cc-by
| 1,608
|
Серия «Вестника СВФУ» № 4 (32) 2023 Серия «Вестника СВФУ» № 4 (32) 2023 DOI 10.25587/2587-5612-2023-4-70-72 УДК 342. 97 УДК 342. 97 Административно-правовой статус
высшего учебного заведения: особенности и проблемы Я. А. Петрова-Ядреева
СВФУ им. М. К. Аммосова, г. Якутск, Россия Аннотация. Административно-правовой статус вуза является важной составляющей его деятель-
ности. В данной статье рассматривается проблематика административно-правового статуса вуза
и возможные пути ее решения. Речь идет о необходимости обеспечения автономии и независимости
учебного заведения, контроля качества образовательного процесса, финансирования вузов и обеспе-
чения прав студентов и сотрудников. В статье также указывается на важность разработки и при-
менения соответствующего законодательства, которое бы гарантировало автономию университетов
и обеспечило их эффективное функционирование и развитие. Ключевые слова: администрация, высшее образование, статус вуза, автономия, университет, об-
разовательный процесс. ПЕТРОВА-ЯДРЕЕВА Яна Александровна – старший преподаватель кафедры «Арктического пра-
ва и права стран АТР», Юридический факультет, Северо-Восточный федеральный университет име-
ни М.К. Аммосова.
E-mail: yadreeva-jana@yandex.ru
PETROVA-YADREEVA Yana Alexandrovna – Senior Lecturer Department of Arctic law and the law of i E-mail: yadreeva-jana@yandex.ru ПЕТРОВА-ЯДРЕЕВА Яна Александровна – старший преподаватель кафедры «Арктического пра-
ва и права стран АТР», Юридический факультет, Северо-Восточный федеральный университет име-
ни М.К. Аммосова. PETROVA-YADREEVA Yana Alexandrovna – Senior Lecturer Department of Arctic law and the law of
the countries of the Asia-Pacific region, Faculty of Law, M.K. Ammosov North-Eastern Federal University. ПЕТРОВА ЯДРЕЕВА Яна Александровна старший преподаватель кафедры «Арктического пра
ва и права стран АТР», Юридический факультет, Северо-Восточный федеральный университет име-
ни М.К. Аммосова.
E-mail: yadreeva-jana@yandex.ru
PETROVA-YADREEVA Yana Alexandrovna – Senior Lecturer Department of Arctic law and the law of
the countries of the Asia-Pacific region, Faculty of Law, M.K. Ammosov North-Eastern Federal University. The administrative and legal status
of a higher educational institution: features and problems Ya. A. Petrova-Yadreeva
M. K. Ammosov North-Eastern Federal University, Yakutsk, Russia Abstract. The administrative and legal status of a university is an important component of its activities. This article discusses the problems of the administrative and legal status of a university and possible ways
to solve it. We are talking about the need to ensure the autonomy and independence of the educational
institution, control the quality of the educational process, finance universities and ensure the rights of
students and employees. The article also points out the importance of developing and applying appropriate
legislation that would guarantee the autonomy of universities and ensure their effective functioning and
development. Keywords: administration, higher education, university status, autonomy, university, educational
process. Административный статус вуза играет важную роль в его функционировании и раз-
витии. Он определяет основные права и обязанности учреждения высшего образования,
его отношения с государственными органами, а также взаимодействие с другими вузами
и обществом в целом. В данной статье мы рассмотрим особенности административного
статуса вуза, его влияние на учебный процесс и проблемы, связанные с этим вопросом. Административный статус вуза определяется законодательством страны, в которой
он функционирует. Обычно он включает в себя такие аспекты, как юридическое лицо,
финансовая автономия, самоуправление, участие в государственных программах и т. д.1 1 Федеральный закон от 29.12.2012 № 273-ФЗ «Об образовании в Российской Федерации» // Собрание
законодательства РФ, N 53, 31.12.2012, ст. 7598. ПЕТРОВА-ЯДРЕЕВА Яна Александровна – старший преподаватель кафедры «Арктического пра-
ва и права стран АТР», Юридический факультет, Северо-Восточный федеральный университет име-
ни М.К. Аммосова. y
j
@y
PETROVA-YADREEVA Yana Alexandrovna – Senior Lecturer Department of Arctic law and the law of
the countries of the Asia-Pacific region, Faculty of Law, M.K. Ammosov North-Eastern Federal University. 70 Серия «ИСТОРИЯ. ПОЛИТОЛОГИЯ. ПРАВО» В зависимости от конкретной страны и ее законодательства, у вуза может быть различ-
ный уровень автономии и гибкости в управлении. Административный статус вуза имеет
прямое влияние на учебный процесс. Например, автономные вузы могут самостоятельно
разрабатывать учебные планы, вводить новые специальности, привлекать иностранных
преподавателей и студентов. Таким образом, увеличивается конкурентоспособность вуза и
его привлекательность для студентов. Однако, административная автономия также требует
большей ответственности и профессионализма со стороны руководства вуза. Однако следует отметить, что современная практика управления в высших учебных
заведениях подтверждают важность реализации функции руководителя высшего учебно-
го заведения через наличие общепринятой формы управленческой структуры – ректорат
высшего учебного заведения. The administrative and legal status
of a higher educational institution: features and problems Конкретизация функций и полномочий ректората в системе
управления вузом нормативно нигде не определены, отсутствует его нормативно-правовое
понятие, как и регламентация его деятельности. Значит, следует легализовать основы пра-
вового статуса данного органа в образовательном законодательстве. В заключение следует
отметить, что административно-правовой статус вуза отражает его положение в системе
общественных отношений как коллективного субъекта, фиксирует систему управленче-
ских связей между вузами и органами исполнительной власти, служит основой стабильно-
сти и определенности этих связей. Он находится в тесной взаимосвязи с ролью, задачами
и перспективами развития высшего образования, так как обеспечивает надлежащее функ-
ционирование вуза в интересах решения стоящих перед ним задач. Успешное реформиро-
вание высшего образования в условиях рыночных отношений, реализация потребностей
личности в высшем профессиональном образовании, защита прав и интересов граждан
в этой сфере, развитие негосударственного образования – все это требует чёткого опреде-
ления статуса вуза, повышения эффективности и совершенствования управления высшим
образованием. Таким образом, под административно-правовым статусом высшего учебно-
го заведения следует понимать совокупность его прав, обязанностей и компетенции, опре-
делёнными целями, задачами деятельности, функциями высшего учебного заведения во
взаимоотношениях с органами государственного управления в зависимости от уровня его
автономии, организационно-правовой формы, наличия или отсутствия государственной
аккредитации [1]. Несмотря на преимущества, связанные с административной автономией, у вузов могут
возникать определенные проблемы. Во-первых, несовершенство законодательства может
приводить к неясности в вопросах, касающихся учебного процесса и управления вузом. Во-вторых, некоторые вузы могут злоупотреблять своими правами, игнорируя принципы
демократии и открытости. В-третьих, административная автономия может создавать не-
равенство между различными вузами, что приводит к ухудшению качества образования. Разработка и применение законодательства, гарантирующего автономию и независи-
мость университетов, является одной из ключевых задач в современной образовательной
сфере. В данной статье мы рассмотрим проблематику статуса вуза и возможные пути ее
решения через разработку и применение соответствующего законодательства. Один из основных аспектов проблематики административно-правового статуса вуза
связан с необходимостью обеспечения автономии университетов от вмешательства го-
сударственных органов. Автономия вуза предполагает возможность самостоятельного
определения образовательной программы, выбора факультетов и направлений, а также
проведение научных исследований и разработок. Для решения этой проблемы необходи-
мо разработать и принять законодательные акты, которые бы гарантировали и защищали
автономию вузов и предоставляли им необходимую независимость. 71 Серия «Вестника СВФУ» № 4 (32) 2023 Кроме того, одной из важных проблем является обеспечение качества образовательного
процесса и контроля за его соблюдением. Университеты должны иметь право назначать
квалифицированных и компетентных преподавателей, разрабатывать современные про-
граммы обучения и следить за их качеством. The administrative and legal status
of a higher educational institution: features and problems Предложение законодательных актов, кото-
рые бы устанавливали механизмы независимой оценки качества образования и обязывали
вузы проходить аудиты и мониторинг, может быть решением этой проблемы [2]. Однако финансирование вузов является одной из самых серьезных проблем, с кото-
рыми они сталкиваются. Качество образования и его доступность напрямую зависят от
финансового состояния университетов. Для решения этой проблемы необходимо разрабо-
тать и применять эффективные механизмы государственного и частного финансирования
вузов, а также использовать возможности привлечения инвестиций. Важной проблемой также является обеспечение прав студентов и сотрудников вуза. Не-
обходимо установить четкие правила и стандарты в отношении административных про-
цедур, приема и отчисления студентов, а также прав и обязанностей преподавателей и со-
трудников. Продвижение прозрачности и ответственности в административных процессах
может помочь решить данную проблему. В целом, разработка и применение законодательства, гарантирующего автономию и не-
зависимость университетов, является важным шагом в решении проблематики админи-
стративно-правового статуса вуза. Это поможет укрепить доверие к учебным заведениям,
повысить качество образования и способствовать их устойчивому развитию. При этом
необходимо учитывать особенности национального законодательства и стремиться к со-
гласованию национальных и международных стандартов в области высшего образования. Административный статус вуза играет ключевую роль в его развитии и эффективности. Он определяет уровень автономии и ответственности учреждения высшего образования,
его возможности в управлении и развитии. Однако, существует несколько проблем, связан-
ных с административной автономией, которые требуют внимания и решения со стороны
государственных органов и учреждений высшего образования. Литература 1. Никитина, А. В. Административно-правовой статус высших учебных заведений в Российской
Федерации / А. В. Никитина // Молодой ученый. – 2020. – № 51 (341). – С. 256-258 [Электронный
ресурс]. ‒ URL :https://moluch.ru/archive/341/76743/ 2. Певцова, Е. А. Образовательное право: актуальные вопросы современной теории и практики. Учебное пособие / Е. А. Певцова. – Москва : Международный юридический институт, 2012. – 253 c. 2. Певцова, Е. А. Образовательное право: актуальные вопросы современной теории и практики. Учебное пособие / Е. А. Певцова. – Москва : Международный юридический институт, 2012. – 253 c. 1. 1. Nikitina, A. V. Administrativno-pravovoj status vysshih uchebnyh zavedenij v Rossijskoj Federacii
/ A. V. Nikitina // Molodoj uchenyj. – 2020. – № 51 (341). – S. 256-258 [Jelektronnyj resurs]. ‒ URL :https://
moluch.ru/archive/341/76743/ 2. Pevcova, E. A. Obrazovatel’noe pravo: aktual’nye voprosy sovremennoj teorii i praktiki. Uchebnoe
posobie / E. A. Pevcova. – Moskva : Mezhdunarodnyj juridicheskij institut, 2012. – 253 c. References 1. 1. Nikitina, A. V. Administrativno-pravovoj status vysshih uchebnyh zavedenij v Rossijskoj Federacii
/ A. V. Nikitina // Molodoj uchenyj. – 2020. – № 51 (341). – S. 256-258 [Jelektronnyj resurs]. ‒ URL :https://
moluch.ru/archive/341/76743/ 1. 1. Nikitina, A. V. Administrativno-pravovoj status vysshih uchebnyh zavedenij v Rossijskoj Federacii
/ A. V. Nikitina // Molodoj uchenyj. – 2020. – № 51 (341). – S. 256-258 [Jelektronnyj resurs]. ‒ URL :https://
moluch.ru/archive/341/76743/ 2. Pevcova, E. A. Obrazovatel’noe pravo: aktual’nye voprosy sovremennoj teorii i praktiki. Uchebnoe
posobie / E. A. Pevcova. – Moskva : Mezhdunarodnyj juridicheskij institut, 2012. – 253 c. 2. Pevcova, E. A. Obrazovatel’noe pravo: aktual’nye voprosy sovremennoj teorii i praktiki. Uchebnoe
posobie / E. A. Pevcova. – Moskva : Mezhdunarodnyj juridicheskij institut, 2012. – 253 c. 2. Pevcova, E. A. Obrazovatel’noe pravo: aktual’nye voprosy sovremennoj teorii i praktiki. Uchebnoe
posobie / E. A. Pevcova. – Moskva : Mezhdunarodnyj juridicheskij institut, 2012. – 253 c. 72
|
https://openalex.org/W4385703465
|
https://pilotfeasibilitystudies.biomedcentral.com/counter/pdf/10.1186/s40814-023-01366-3
|
English
| null |
Hypnosis to reduce fear of falling in hospitalized older adults: a feasibility randomized controlled trial
|
Pilot and feasibility studies
| 2,023
|
cc-by
| 9,455
|
Pilot and Feasibility Studies Pilot and Feasibility Studies Cuvelier et al. Pilot and Feasibility Studies (2023) 9:139
https://doi.org/10.1186/s40814-023-01366-3 Open Access Hypnosis to reduce fear of falling
in hospitalized older adults: a feasibility
randomized controlled trial Clémence Cuvelier1, Mélany Hars1,2, Maria Pia Zamorani‑Bianchi1, François R. Herrmann1,3,
Catherine Ducharne Wieczorkiewicz1,3, Dina Zekry3, Gabriel Gold1 and Andrea Trombetti1,2* Abstract Background Fear of falling is associated with numerous negative health outcomes in older adults and can limit
the rehabilitation process. Hypnosis is now recognized as an effective treatment for a variety of conditions, especially
anxiety and pain, which can be integrated safely with conventional medicine. The objective of this study was to assess
the feasibility and acceptability of a hypnosis intervention in hospitalized older adults to reduce fear of falling. Methods In this feasibility randomized controlled trial, 32 older patients, hospitalized in geriatric rehabilitation wards,
were randomly allocated (1:1 ratio) to either an intervention group (hypnosis, 2 sessions, one per week, plus usual
rehabilitation program) or a control group (usual rehabilitation program only). Clinical assessors and statistician were
blinded to group allocation. Primary outcomes were recruitment rate, retention rate, and adherence to the inter‑
vention. Exploratory outcomes, analyzed according to the intention-to-treat principle, included impact of hypnosis
on fear of falling (assessed by a new scale perform-FES), functional status, in-hospital falls, and length of hospital stay. Results Recruitment rate was 1.3 patients per week. The recruitment of the population sample was achieved
in 5.5 months. The retention rate did not differ significantly between groups and a good adherence to the hypnosis
intervention was achieved (77% of patients received the full intervention). No adverse event related to the hypnosis
intervention was observed. Regarding exploratory clinical outcomes, no differences were found between groups
on any outcome. Conclusion Hypnosis is feasible and well accepted in a geriatric hospitalized population undergoing rehabilitation. Further pilot work should be conducted, with an increased number of hypnosis sessions, before conducting a full-
scale trial to conclude whether, or not, hypnosis is effective to reduce fear of falling. Trial registration NCT04726774. Keywords Fear of falling, Hypnosis, Rehabilitation, Feasibility study, Older adults *Correspondence:
Andrea Trombetti
andrea.trombetti@hcuge.ch; andrea.trombetti@unige.ch
Full list of author information is available at the end of the article *Correspondence:
Andrea Trombetti
andrea.trombetti@hcuge.ch; andrea.trombetti@unige.ch
Full list of author information is available at the end of the article © The Author(s) 2023. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which
permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the
original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or
other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line
to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory
regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this
licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativeco
mmons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. © The Author(s) 2023. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which
permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the
original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or
other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line
to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory
regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this
licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativeco
mmons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Page 2 of 11 Cuvelier et al. Pilot and Feasibility Studies (2023) 9:139 Cuvelier et al. Pilot and Feasibility Studies • What uncertainties existed regarding the feasibility? Whether hypnosis could be an effective interven-
tion strategy to reduce fear of falling in older adults
remains to be determined. We conducted a feasibility
randomized controlled trial to address if a hypnosis
intervention would be feasible and acceptable, and
explore effectiveness, to inform a potential future
large-scale study. Because this study focused on
hospitalized very old patients, including patients
with cognitive impairments, uncertainties existed
especially regarding recruitment, feasibility, and
adherence. Fear of falling can be evaluated using different tools,
mainly validated in the community and not in hospitalized
population. Whereas some tools measure the ability to
avoid a fall (i.e., Geriatric Fear of Falling Measure, GFFM)
[21], or the confidence in maintaining balance (i.e., Activ-
ities-specific Balance Confidence scale, ABC) [22], other
assess the concern about falling during activities (i.e., Fall
Efficacy Scale-International, FES-I) [23]. This variety and
heterogeneity of constructs can explain the difficulty to
measure the impact of interventions on fear of falling [14,
24]. Moreover, these tools are based on answers to a ques-
tionnaire and may not reflect a person’s feelings during
the actual functional performance. The answers may be
affected by some degree of cognitive deficit and not really
adapted to evaluate short term evolution, in which the
subject is not reexposed to the situations depicted in the
questionnaire. In that context and for the purpose of the
current study, we developed the Perform-FES scale spe-
cifically designed to measure the degree of concern about
falling in a hospital setting [25]. This scale demonstrated
excellent psychometric properties. • What are the key feasibility findings? i
Hypnosis was feasible to deliver and well
accepted by old patients hospitalized in a geriatric
hospital. No adverse event related to the hypnosis
intervention was observed. • What are the implications of the feasibility findings
for the design of the main study? The findings of this study provide important
information about recruitment, adherence, and
impact of a hypnosis intervention for a larger future
randomized controlled trial for fear of falling
reduction in hospitalized older adults. But further
pilot work should be conducted, with an increased
number of hypnosis sessions to demonstrate a
clinical effect. Key messages regarding feasibility length of stay [10–12]. Moreover, the high level of per-
ceived fall risk is likely to be associated with future falls,
independent of physiological risk [13], and may influence
physical capabilities, especially gait performances [8]. Considering this, strategies aimed at reducing fear of fall-
ing should be implemented in rehabilitation programs. In
a meta-analysis [14] exercise interventions were associ-
ated with a small or moderate reduction in fear of falling
in the short and long-term follow-up. Cognitive behav-
ioural therapy has also been shown to reduce fear of fall-
ing in older adults [15], especially in combination with
exercise [14, 16–18]. Studies suggested that multifactorial
assessment and combinations of interventions reduce the
risk of falls in older population [19, 20]. Background Falls are highly prevalent in older adults. They can result
in serious injuries and death [1]. They can also have psy-
chological consequences, such as fear or falling [2], which
may lead to restriction of activities resulting in further
worsening in functional performances, reduction of qual-
ity of life and loss of independence [3]. The relationship
between falls and fear of falling is close, and fear of fall-
ing and falls share risk factors [4]. Fear of falling further
increases the risk of falling [4]. Fear of falling can be the
consequence of a rational appraisal of reduced functional
abilities, or can be a construct reflecting the original pho-
bic condition and may be irrational [5]. Its prevalence is
estimated between 25 and 85% in older people according
to settings and type of measure [6, 7], and fear of falling
can also be present among those who never experienced
falls [8]. Main fear of falling related risk factors include
older age, female sex, functional impairment, or medi-
cations [9, 10]. In an inpatient geriatric rehabilitation
population, fear of falling has been associated with worse
functional recovery, more in-hospital falls, and increased Hypnosis is defined as “a state of consciousness
involving focused attention and reduced peripheral
awareness characterized by an enhanced capacity for
response to suggestion” [26]. During a hypnotic trance,
physiological, cognitive, and affective processes can be
modified. Several hypnosis techniques, such as medical
hypnosis, hypnotic communication, and hypnotherapy,
are recognized as safe and effective in different appli-
cations [27]. Some have suggested that older adults are
less receptive to hypnosis. However, studies involved
older adults with positive results for hypnosis in pain
[28], sleep disorders [29], and lumbar puncture related
distress [30]. The feasibility of hypnosis in older people
has been demonstrated in hospitalized [31] and home
care populations [32]. While few studies included older
persons with cognitive impairment [33], hypnotizabil-
ity does not seem to change with age. Hypnosis appears
to be particularly interesting in reducing pain, anxiety Page 3 of 11 Cuvelier et al. Pilot and Feasibility Studies (2023) 9:139 Cuvelier et al. Pilot and Feasibility Studies (2023) 9:139 Cuvelier et al. Pilot and Feasibility Studies and medications in this population with higher suscep-
tibility to adverse effects and drugs interactions. and medications in this population with higher suscep-
tibility to adverse effects and drugs interactions. Background Randomization and blinding procedure Randomization and blinding procedure After consent, patients were randomized to either the
intervention group or the control group. The rand-
omization sequence was computer-generated (ratio
1:1) and concealed until patient enrolment. The
nature of the intervention prevented us from blind-
ing patients and the hypnotherapist to allocation. All
clinical assessors, including physiotherapists and occu-
pational therapists, were blinded to group allocation. All statistical analyses were performed by a blinded
statistician (FRH). Specific aim and hypothesis Since there is no study, to our knowledge, assessing the
effect of hypnosis on fear of falling in a geriatric pop-
ulation, we conducted a feasibility trial to address if a
hypnosis intervention would be feasible and acceptable,
and explore effectiveness for fear of falling reduction, to
inform a potential future large-scale study.h The primary aims of this study were to determine the
feasibility and acceptability of a hypnosis intervention
in older fallers. The exploratory aims were to assess the
impact of hypnosis plus usual care (i.e., rehabilitation
program) on fear of falling, based on the Perform-FES,
other fear of falling scales, and other clinical outcomes
including functional status, medications use, in-hospi-
tal falls, and length of hospital stay, compared to usual
care alone. Interventions
Th The intervention consisted of two hypnosis session of
about 30 min, provided weekly by a physician certified
and trained in medical hypnosis (MPZB). MPZB is a
medical doctor practitioner with university-level train-
ing in medical hypnosis and holds a diploma delivered by
the Swiss Medical Society of Hypnosis. The intervention
was provided in addition to the usual rehabilitation pro-
gram. The participant was informed about medical hyp-
nosis in term of general concepts and objectives of the
study before inclusion. Hypnosis session were realized in
the patients’ room or in the physiotherapy room. Patient
and hypnotherapist were face to face. If possible, all the
hypnosis procedure was performed during walking,
depending on the agreement and the physical status of
the participant. During the consent interview, induction
channels were explored, to prepare the first hypnosis ses-
sion. The hypnosis procedure was divided into 4 phases:
induction, trans hypnotic with walk perception altera-
tion, post-hypnotic suggestion, and exit of hypnosis. Hypnosis was used as a communication tool adapted to
the patient’s situation. During the trans hypnotic period,
metaphors based on the patient’s personal history were
used. Thus, the patient can experience his symptoms
at another level of consciousness, non-analytical, non-
rational, but in relation to sensoriality. Fear of falling assessmenthfi – The Geriatric fear of falling measure [21]. It contains
15 items to assess older adult’s fear of falling through
psychosomatics symptoms, adopting a risk preven-
tion and modifying behaviour, and not only with
restriction activities [37]. Scores vary from 15 points
(never concerned) to 75 points (always concerned). – The Geriatric fear of falling measure [21]. It contains
15 items to assess older adult’s fear of falling through
psychosomatics symptoms, adopting a risk preven-
tion and modifying behaviour, and not only with
restriction activities [37]. Scores vary from 15 points
(never concerned) to 75 points (always concerned). The falls efficacy scale international (FES-I) was used to
assess participant’s perceived fall risk by asking about
concern about falling across a wide range of activities
of daily living and social activities [23, 34]. The FES-I
has excellent reliability (Cronbach’s α = 0.97, test–
retest = 0.94) [23], and good psychometrics properties
[35]. It comprises 16 items (housecleaning; dressing or
undressing; preparing simple meals; bathing or shower-
ing; shopping; sitting or rising from a chair; walking up or
down stairs; walking in the neighborhood; taking some-
thing above the level the head or from the ground; pick-
ing up the phone; walking on a slippery surface; visiting a
friend or relative; walking in crowded places; walking on
an uneven surface (with stones or holes); up or down a
slope; and attending a social event). The total score varied
between 16 (not worried) and 64 points (very worried).h Each scale was completed by each participant three
times (week 0 (at baseline): before intervention; week 1
(during intervention): between the two hypnosis sessions;
week 2 (at the end of the study): after all interventions). To avoid recall bias, scales were completed just after the
standardized scenario of the Perform-FES, which mimics
situations at risk of falling. If patients had difficulties in
completing the questionnaire, the occupational therapist
helped him. p
y
The Perform-FES, a new scale derived from FES-I short
version (7 items), but based on performance in real situ-
ations, was also used [25]. This scale was elaborated and
validated for the purpose of the study, to better assess the
fear of falling in older hospitalized patients who may have
difficulties to report their concern about falling in specific
daily-life tasks. Setting and participantsh g
This single-center, two-arm, feasibility randomized
controlled trial was conducted in a 296 bed acute care
and rehabilitation geriatric hospital of Geneva Univer-
sity Hospitals (Switzerland), from January 17, 2019 to
July 25, 2019 (NCT04726774). The study included hos-
pitalized patients aged 65 years and over, admitted to a
rehabilitation program specialized in falls and fracture
risk assessment and management (“CHutes Et Osteo-
PoroSe” program, CHEOPS) that has been described
elsewhere [20]. Exclusion criteria included patients
with behavioral disorders, or who did not speak French,
or lacked decision-making capacity. The participants
should have their capacity of discernment to subscribe
to the study, which mean they had the ability to under-
stand the objectives of the study, what it implied and
what were the risks and benefits, to finally decide if they
wanted to participate or not. Capacity of discernment
was assessed by the physician in charge of the patient. Behavioral disorders included opposition, aggressiv-
ity, or delirium, which would prevent the practice of
intervention (rehabilitation and/or hypnosis). All these
behavioral disorders were evaluated at inclusion by the
hypnotherapist (i.e., subjective to the hypnotherapist). The target sample size was 30 participants (15 partici-
pants in each group). The study was approved by the
Geneva Research Ethics Committee (2018–01550). All
patients provided written informed consent before any
study-related procedure. Both the intervention and control groups received
the usual rehabilitation program, a multifactorial fall-
and-fracture risk-based assessment and management
intervention, which has been shown to be effective in
improving physical parameters related to the risk of falls
among high-risk oldest old patients [20]. This program
includes an individually tailored intervention targeting
each patient’s individual risk factors and impairments,
including intensive physiotherapy for at least 2 weeks
(i.e., focused on gait, balance and muscular function, in
group and individual format and patient education on
falls prevention). In this program, there is 5 weekly group
sessions of 60-min duration and 3 to 5 weekly individual
sessions of 30–45-min duration. Page 4 of 11 Cuvelier et al. Pilot and Feasibility Studies (2023) 9:139 Cuvelier et al. Pilot and Feasibility Studies (2023) 9:139 Cuvelier et al. Pilot and Feasibility Studies Fear of falling assessmenthfi Each situation (i.e., dressing or undress-
ing: taking off one’s socks or dressing gown; bathing or
showering: step into the shower or the bath, turn on the
tap and get out; sitting and rising from a chair; walking
up and down stairs: four steps; taking something from the
ground; walking up and down a slope; and getting out:
walk on a stony path and sit on a bench) was reproduced
according to a standardized administration procedure,
under the supervision of an occupational therapist. The
total score varied between 7 and 28 points (1: not wor-
ried; 4: very worried). It had been shown previously that
the Perform-FES had a good internal consistency (Cron-
bach’s alpha coefficient = 0.78) and an excellent reliability
(intraclass correlation coefficient = 0.94) in a hospitalized
geriatric population [25]. This scale also revealed higher
performance than other fear of falling scales in discrimi-
nating patients with severe functional limitations. – The Activities-specific Balance Confidence scale [22,
36]. Patients are required to self-rate their degree of
confidence in their balance associated with the per-
formance of a series of daily living tasks. It was vali-
dated with high-functioning seniors. The simplified
version includes 15 items, with a 4-category response
format with descriptive anchors (i.e., 0: not at all con-
fident, 1: slightly confident, 2: moderately confident,
3: very confident). Minimal score is 0, and maximal
score is 45. Recruitment and participant flow l
Among 57 patients admitted in the CHEOPS unit
between January 17, 2019, and July 04, 2019 (168 days),
55 (96.5%) were assessed for eligibility. Of these, one
did not meet inclusion criteria (delirium), and 22 (40%)
declined to participate. The reasons for refusing to partic-
ipate to the study were due to the nature of intervention
in 6 cases and for other reasons in 16 cases (engage-
ment, questionnaires, no interest). In total, 32 (58.2%)
patients met eligibility criteria and provided informed
consent. The recruitment rate was 1.3 patients per week
(32 patients/24 weeks). After randomization 15 patients
were included in the hypnotherapy group, and 17 in the
control group. Seven patients were lost to follow-up (3
in hypnotherapy group, 4 in control group). The reasons
were an early discharge from hospital in all cases (change
of department (n = 1), transfer to a nursing home (n = 2)
and return home (n = 4)). Enrolment and follow-up of
patients are illustrated in Fig. 1. Analysis medications at the end of rehabilitation program, and
the presence or absence of anxiolitic and analgesic drugs,
were compared between groups using t tests. A two-
sided p value less than or equal to 0.05 was considered to
indicate statistical significance. Data were analyzed using
Stata version 16.1 (Stata Corp., College Station, TX) sta-
tistical software. Given the design of this study, we based the sample
size on the feasibility objective, rather than on a formal
power calculation to detect between-group difference
for patient reported outcomes. A sample of 30 patients
(15 per arm) was deemed large enough to provide use-
ful information about feasibility. The sample was based
on the same eligibility criteria that would be used in a
future full-scale randomized controlled trial. Descriptive
statistics were reported as mean ± standard deviation or
median [interquartile range (IQR)] for continuous vari-
ables or number (percent) for categorical variables. The
hypnosis and the control groups were compared at base-
line using t tests, Wilcoxon rank sum tests or Fisher’s
exact test as appropriate. All variables were tested for
normality and appropriate transformations were used to
transform non-normally distributed variables. Other assessments
D
h Demographics, comorbidities, cognitive, and nutritional
assessment were assessed at baseline. Socio-demo-
graphic data included age, sex, weight(kg) and body mass
index(kg/m2), place of living (home, residential home, or
nursing home). The Cumulative Illness Rating Scale-Ger-
iatric (CIRS-G) [38] estimated the number and the sever-
ity of comorbidities. Cognitive status was assessed by the
Mini Mental State Examination (MMSE) [39], and the
clock-drawing test [40]. Pain was assessed by the Visual
Analog Scale (VAS), and depression by the mini Geriat-
ric Depression Scale (mini-GDS) [41, 42]. The nutritional
status was measured by the Nutritional Risk Score
(NRS-2002) [43]. Number of drugs (total and specific
medications such as anxyolitic and/or hypnotic drugs),
physical performances (Short Physical Performance Bat-
tery (SPPB) [44]) and functional independence (Func-
tional Independence Measure (FIM) [45]), were evaluated
at baseline and at the final visit post intervention. SPPB is
a composite physical performance assessment tool includ-
ing a balance test, a gait speed test and a chair rise test. Poor performances on SPPB have been associated with
in-hospital falls, injurious falls, and fractures in our popu-
lation [46]. The FIM includes 18 items designed to assess
the degree of assistance required for a person with a dis-
ability to perform basic life activities safely and effectively. It is widely used to determine the progress that patients
make through programs of medical rehabilitation. The
number of falls within 6 months preceding admission was
collected by a nurse at admission based on self-report. In
hospital falls were prospectively recorded until discharge
using computer‐based incident report forms mandato-
rily completed after each fall by nurses and electronic
patients’case notes or medical reports (i.e., through the
integrated hospital information system). The other scales used to estimate the fear of falling
were: Cuvelier et al. Pilot and Feasibility Studies (2023) 9:139 Page 5 of 11 Cuvelier et al. Pilot and Feasibility Studies (2023) 9:139 Cuvelier et al. Pilot and Feasibility Studies Baseline characteristics
l
h Baseline characteristics of the study population are
shown in Table 1. The mean age of the patients was 85.6
(± 5.8 SD) years, and 53% were women. Most of them
(93.2%) lived at home before admission. The median
score of MMSE was 25 [19;26] and 14 (43.8%) patients
had a MMSE score ≤ 24. In terms of falls, only 3 (9%)
patients reported no falls, while 18 (56%) reported 2 or
more falls in the previous 6 months. Among the fallers,
22.7% had a fall related injury and 36.4% spent more than
30 min on the floor after a fall. At admission, the mean
MIF score was 86/126 (± 19.1 SD) and the mean SPPB
score was 5/12 (± 2.0 SD). Almost half of the included
patients were taking analgesics (52.3%) and/or anxiolitic-
hypnotic (43.2%) drugs at admission. The mean num-
ber of medications at admission was 8.8 (± 3.0 SD). The
mini-GDS and NRS-2002 score were significantly dif-
ferent between both groups at admission (p = 0.003 and
p = 0.025 respectively). Primary outcomes: feasibility and acceptability
of intervention Feasibility of recruitment and of the hypnosis inter-
vention was assessed by measuring recruitment rate,
retention rate, hypnosis adherence, and acceptance. Recruitment rate was analyzed by dividing the number of
included participants by the number of weeks it took to
include them. Retention rate was defined as the percent-
age of patients enrolled at baseline, who completed all
follow-up measurements. Hypnosis adherence was meas-
ured by the proportion of the total number of sessions
attended to the total number of sessions for which par-
ticipants were enrolled to attend and by the proportions
of participants who received all the sessions (i.e., 2 ses-
sions). Adverse events (including abreactions, falls during
hypnosis, headache and sleep disorders) were collected. Primary outcome: feasibility and acceptability of hypnosis Primary outcome: feasibility and acceptability of hypnosis
Recruitment of 32 patients (our recruitment target), took
5.5 months. Retention rates were similar between the
two groups: 10 patients (66.7%) in the hypnosis group
successfully completed the follow-up (i.e., 15 days reha-
bilitation program and completed all scores and ques-
tionnaires) compared to 11 (64.7%) in the control group. The length of hospital stay and the length of the reha-
bilitation program did not differ between the two groups
(p = 0.89 and p = 0.31 respectively) (Table 1). Exploratory outcomes Longitudinal data were analyzed according to the inten-
tion-to-treat (ITT) principle. Fear of falling, FIM and
SPPB scores were analyzed using linear mixed-effects
regression models, fitted using Stata “xtmixed” proce-
dure, taking into account random effects (participant)
and with the interaction group and visit entered in the
models. These models took the repeated measure design
of the study into account and allowed for a different
number of observations within subjects. The same anal-
ysis was conducted in the per-protocol (PP) population
(defined as participants who completed the interven-
tion period and assessment visit) and in the subgroup of
patients with a high concern of fear of falling (i.e., FES-I
score ≥ 28/64, Perform-FES score ≥ 9/28) [47]. The inci-
dence of inhospitals falls during hospital stay was ana-
lyzed using a negative binomial regression model. The
length of stay in hospital (in days), the number of total At baseline, the mean scores of fear of falling scales
were 32.3 ± 12.5 for the FES-I, 9.9 ± 3.1 for the Perform-
FES, 44.7 ± 12.5 for the GFFM, 23.4 ± 8.4 for the ABC-S,
without significant differences between the intervention
and the control groups (p > 0.05 for all). Half (50%) of Cuvelier et al. Pilot and Feasibility Studies (2023) 9:139 Page 6 of 11 Cuvelier et al. Pilot and Feasibility Studies Fig. 1 Flowchart of the feasibility randomized controlled trial Fig. 1 Flowchart of the feasibility randomized controlled trial per patient was 1.5 ± 0.8. Fourteen out of 23 sessions of
hypnosis were not performed while walking. The princi-
pal cause of non-walking hypnosis was patient’s choice. The most used induction channels were by memory and
kinesthesic. No adverse event related to the intervention
was reported. patients had a FES-I score ≥ 28/64, while 59% had a
Perform-FES score ≥ 9/28, which are known thresholds
above which fear of falling is a concern [47]. The two
groups were comparable regarding physiotherapy adher-
ence (9.5 ± 3.67 sessions in the intervention group vs
9.5 ± 4.14 sessions in the control group; p = 0.984). Exploratory outcomes
S
f
h Scores for the two groups at the different assessment
visits are summarized in Table 2. In the ITT analysis,
no significant randomization group × time interaction
effects were found at both intermediate and end visits
for any fear of falling outcomes (p > 0.05 for all). Also,
no significant randomization group × time interaction
effects were found at both intermediate and end visits for
SPPB and FIM outcomes (p > 0.05 for all). Similar to the
ITT analysis, neither the PP analysis nor the analysis in
the subgroup of patients with a high fear of falling found
any significant randomization group × time interaction
effects for any fear of falling and functional outcomes. No significant effect was observed for other secondary
outcomes, such as the number of drugs (p = 0.91), the
presence or absence of anxiolitic (p = 0.18) and analgesic
drugs (p = 0.70) at the end of hospitalization, or inhospi-
tals falls experienced during hospital stay (IRR = 0.875;
95%CI 0.46–1.68; p = 0.689). In total, 23 sessions of hypnosis were delivered, with
an adherence rate of 76.7% (95.8% after exclusion of 3
patients who did not receive any session). Four out of 15
patients in the hypnosis group did not have the full inter-
vention (i.e., two hypnosis sessions), with no session in 3
cases, and only one session in 1 case. The reasons were
medical (n = 2), logistic (n = 1), and refusal of the second
session (n = 1). The mean duration time of the first and
second sessions were 46.0 ± 13.0 min and 35.0 ± 8.2 min,
respectively. The mean number of hypnosis sessions Cuvelier et al. Pilot and Feasibility Studies (2023) 9:139 Cuvelier et al. Exploratory outcomes
S
f
h Pilot and Feasibility Studies Page 7 of 11 Table 1 Characteristics of patients at inclusion
Data presented as mean ± standard deviation, median [interquartile range] or number (percent)
* There was a significant difference in the Mini-Geriatric Depression Scale (p = 0.003) and in the Nutritional Risk Score (p = 0.025) between the hypnosis group and the
control group
Characteristic
Hypnosis group (n = 15)
Control
group
(n = 17)
Age, years
84.9 ± 5.9
86.1 ± 5.8
Sex, women
8 (53.3%)
9 (52.9%)
Place of living before admission
Home
13 (86.7%)
16 (94.1%)
Residential home
2 (13.3%)
0 (0.0%)
Nursing home
0 (0.0%)
1 (5.9%)
Weight (kg)
67.1 ± 17.6
76.1 ± 16.4
Body mass index (kg/m2)
24.4 ± 6.3
28.0 ± 6.5
Cumulative Illness Rating Scale-Geriatric [score range 0–56]
15.4 ± 5.4
15.2 ± 5.0
Mini Mental State Examination [score range 0–30]
25 [21;26]
23 [18;27]
Score ≤ 24
5 (33.3%)
9 (52.9%)
Clock-drawing test [score range 0–10]
9 [7;10]
8 [7;10]
Mini-Geriatric Depression Scale [score range 0–4]*
2 [1;2]
0 [0;1]
Nutritional Risk Score 2002 [score range 0–7]*
3.3 ± 1.4
2.4 ± 0.9
Score ≥ 3
8 (53.3%)
4 (23.5%)
Visual Analog Scale [score range 0–10]
0 [0;2]
0 [0;0]
Number of drugs
8.5 ± 3.4
9.1 ± 2.8
Use of analgesic drugs
5 (33.3%)
7 (41.2%)
Use of anxiolitic drugs
6 (40.0%)
6 (35.3%)
Functional Independance Measure [score range 18–126]
89 ± 20.4
83 ± 18.1
Score < 110
10 (66.7%)
14 (82.3%)
Short Physical Performance Battery [score range 0–12]
5 ± 2.1
5 ± 1.9
Score ≤ 6
13 (86.7%)
12 (70.6%)
Number of previous falls (past 6 months)
2 [1;3]
2 [1;3]
≥ 1 fall with serious injury
5 (33.3%)
2 (11.8%)
≥ 1 fall with time on floor > 30 min
5 (33.3%)
6 (35.3%)
Length of hospital stay (days)
30 [26;36]
29 [24;41]
Length of rehabilitation program (days)
21 [15;21]
20 [14;21] Data presented as mean ± standard deviation, median [interquartile range] or number (percent) * There was a significant difference in the Mini-Geriatric Depression Scale (p = 0.003) and in the Nutritional Risk Score (p = 0.025) between the hypnosis group and the
control group * There was a significant difference in the Mini-Geriatric Depression Scale (p = 0.003) and in the
control group In the study population as a whole, the SPPB score
improved with the rehabilitation program (time effect,
1.2; 95% CI 0.6–1.8; P < 0.001), as did the FIM score
(time effect, 11.5; 95% CI 7.1–16.0, p < 0.001). Exploratory outcomes
S
f
h without any adverse event reported. Exploratory out-
comes were not able to support the effectiveness of
hypnosis on fear of falling, self-efficacy and on func-
tional performance. A larger randomized control study
may be needed to assess the impact of hypnosis on fear
of falling. Discussion Pilot and Feasibility Studies Page 8 of 11 Table 2 Exploratory outcomes: scores at baseline, intermediate, and at the end of the study, for both groups
FES-I Falls Efficacy Scale-International, GFFM Geriatric Fear of Falling Measure, ABC-S Activities-Specific Balance Confidence Scale-Simplified, Perform-FES Performance-
Fall Efficacy Scale, SPPB Short Physical Performance Battery, FIM Functional Independence Measure
* Interaction effect between group and time at end visit
a A higher score indicates a greater concern about falling
b A higher score indicates a lower concern about falling
Scores
Group
Baseline visit
(week 0)
Intermediate visit
(week 1)
End visit
(week 2)
P* (interaction
group × time)
FES-I [score range 16–64]a
Hypnosis
33 ± 11
32 ± 11
27 ± 8
0.631
Control
31 ± 13
31 ± 14
26 ± 10
GFFM [score range 15–75]a
Hypnosis
45 ± 11
45 ± 9
45 ± 13
0.401
Control
44 ± 14
44 ± 9
44 ± 11
ABC-S [score range 0–45]b
Hypnosis
23 ± 8
26 ± 10
28 ± 10
0.611
Control
24 ± 9
25 ± 9
28 ± 9
Perform-FES [score range 7–28]a
Hypnosis
9.8 ± 2.7
9.7 ± 2.1
8.7 ± 2.7
0.117
Control
10.1 ± 3.4
9.0 ± 2.8
8.1 ± 1.4
SPPB [score range 0–12]
Hypnosis
4.8 ± 2.1
6.1 ± 2.9
6.4 ± 3.2
0.408
Control
4.8 ± 1.9
5.2 ± 1.7
5.6 ± 2.4
FIM [score range 18–126]
Hypnosis
88.9 ± 20.4
92.5 ± 17.0
97.1 ± 18.0
0.127
Control
82.9 ± 18.1
94.5 ± 19.1
98.0 ± 20.4 Table 2 Exploratory outcomes: scores at baseline, intermediate, and at the end of the study, for both groups short sessions lasting less than one hour, and induction
channels adapted to the sensory impairments of patients. No adverse event and especially no falls were reported
during hypnosis sessions. However, most of them (61%)
differed from the initial planned procedure because they
were not performed during walking. It was the patient’s
choice in most cases and it could be explained, accord-
ing to us, on the one hand by fear of falling and on the
other hand by the misconception of loss of consciousness
during hypnosis. The objective of hypnosis during walk-
ing was to strengthen the anchors when walking. To be
useful, the association of positive emotions, good feelings
and good sensations with walking needs to be repeated. Discussion Extending the use of hypnotic communication to the
entire healthcare team would permit a continuation of
the effect of hypnosis sessions. per week for 57 patients admitted in the rehabilitation
program during this period. Adherence to the hypnosis
intervention was relatively good, which suggests that the
intervention was suitable for this particular sample of
older patients. Retention rate was about 65%, most drop
out cases were due to a request for an early return home
(only 4 persons did not complete the entire rehabilitation
program due to medical reasons), without a difference
in the number of patients lost to follow-up between the
two groups. These requests were related with a functional
improvement allowing a return home and an early suc-
cess of the rehabilitation program. The study population was characterised by low physi-
cal performance (mean SPPB below 6) with high risk of
falling and high fear of falling (mean ABC score < 50 [48]
and mean FES-I > 28 [47]). Almost half of them (46.8%)
had a Perform-FES score > 9/28. It has been previously
shown that there was a correlation between SPPB and
Perform-FES with a good ability for the Perform-FES
to discriminate patients with several functional limita-
tions. Moreover, functional limitations, as assessed by
the SPPB, has been shown as an independent predictor of
inhospital falls and fractures in our setting [49]. f
The number of hypnosis sessions could be questioned. We chose two hypnosis sessions per patient for feasibil-
ity with only one hypnotherapist, and to be realistic in
a two weeks rehabilitation program. It was possible that
the hypnotic effect has a limited time effect [31, 56, 57]. In the review of Jensen and Pattersen about hypnotic
treatment for chronic pain, almost all studies included a
minimum of four sessions of hypnosis, and most of them
encourage the additional practice of self-hypnosis [56]. Self-hypnosis could allow a longer effect of hypnosis. As already reported in a study focusing on hypnosis in
older persons, hypnosis sessions could be shorter due to
a decrease in attentional capacities, especially in case of
cognitive impairment [32]. In studies evaluating hypnosis in rehabilitation, authors
conclude that hypnosis could facilitate psychological and
physical changes by reducing pain, fear or anxiety and
increase patient’s motivation in rehabilitation [50–55]. To our knowledge, none of them concerned hypnosis and
fear of falling or a specific geriatric population. Discussion Concerning recruitment, there was no reluctance for
hypnosis in our inpatient geriatric population. Only 11%
(6/55) refused to participate in the study due to hypnosis
intervention. One identified barrier to participation was
multiple and self-administered questionnaires. We used
several scales because, as discussed before, they not all
similarly measure the fear of falling. The recruitment was
acceptable (56%) with a rate of recruitment of 1.3 patients Our results support the feasibility and acceptability of
medical hypnosis in an inpatient geriatric rehabilita-
tion population. We demonstrated a recruitment rate
of 1.3 patients per week with a period of inclusion of
5.5 months for 32 patients and a good hypnosis adher-
ence rate (95.8%). No adverse event of hypnosis was
reported. Only one person refused the second session, Cuvelier et al. Pilot and Feasibility Studies (2023) 9:139 Cuvelier et al. Discussion Modali-
ties of hypnosis were adapted to this inpatient geriatric
population and could be useful for further studies, as Page 9 of 11 Page 9 of 11 Cuvelier et al. Pilot and Feasibility Studies (2023) 9:139 Cuvelier et al. Pilot and Feasibility Studies (2023) 9:139 Cuvelier et al. Pilot and Feasibility Studies patient’s characteristics, although it was a randomized
study, there was difference at baseline for MMSE and
NRS-2002 scores between the two groups. Both of these
characteristics could influence falls and risk of falling
but there was no difference in in-hospital falls and func-
tional score between both groups during all the study. It
worth to say that although cognitive disorders were not
an exclusion criterion in our study, the mean MMSE
score at baseline was relatively high (> 20/30). It could be
explained by the inclusion criteria of the rehabilitation
program and the need to consent to the study. For further
larger studies, the prevalence of cognitive impairment
should be considered because it can impact both adher-
ence to the protocol and the response to the self-ques-
tionnaires. The Perform-FES scale seems to be a good
alternative in this population as this scale has been devel-
oped in a hospitalized older population with approxi-
mately 40% of patients with cognitive disorders. The
sensitivity to change of this scale is under evaluation by
our group. Finally, as walking during hypnosis was almost
not realized, we cannot conclude about this practice. Although our study population reflects a geriatric pop-
ulation with low physical performance scores, high risk
of falls and high fear of falling, we were not able to show
the effect of hypnosis on fear of falling in this feasibility
study. We did however demonstrate feasibility support-
ing the performance of a future more largely powered
study. Scales of fear of falling had good validity and relia-
bility, but their sensitivity to change is less studied. There
is no longitudinal study yet concerning the new scale Per-
form-FES. Moreover, we could not show a reduction in
medication use with hypnosis, which would have been an
interesting benefit in this population with a high risk of
drug-induced iatrogenic disease. In the whole study population, patients improved their
SPPB and FIM scores with the rehabilitation program. Conclusion Activities-specific Balance Confidence scale
“ Chutes et OsteoPoroSe”
Fall Efficacy Scale-International
Functional Independence Measure
Geriatric Depression Scale
Geriatric Fear of Falling Measure
Intention-to-treat
Mini Mental State Examination
Nutritional Risk Score
Per protocol
Short Physical Performance Battery
Visual Analogue Scale Discussion These results confirm the beneficial effects of a multifac-
torial fall-and-fracture risk assessment and management
program, applied in a dedicated geriatric hospital unit, to
improve fall-related physical performances and the level
of independence in activities of daily living in high-risk
patients [20]. Abbreviations
ABC
CHEOPS program
FES-I
FIM
GDS
GFFM
ITT
MMSE
NRS-2002
PP
SPPB
VAS Conclusion The strengths of this study are the randomized con-
trolled design and the standardized rehabilitation pro-
gram, which may facilitate the design for a future larger
randomized control trial. The feasibility and safety of
hypnosis in this inpatient geriatric population with few
refusals could encourage other studies with hypnosis in
this specific old and frail population. The current study
has limitations that need to be addressed. Some of these
limitations are the small sample size and the number of
lost to follow-up. However, the good recruitment rate,
suggests the feasibility of a larger study. The number of
patients lost to follow-up should be taken into account
for further studies. In our experience, 20% (3/15) of the
patients in hypnosis group did not receive hypnosis ses-
sions at all. This could mask the potential effect of hyp-
nosis on fear of falling. Other limitations are the number
of secondary outcomes and the impossibility to conclude
about the effectiveness of hypnosis on fear of falling
due to a lack of power. The primary aim of this feasibil-
ity study was to evaluate feasibility rather than effective-
ness. Regarding fear of falling, the number of outcome
measures was extensive, especially given the absence of
a gold standard measure in the hospitalized population. The FES-I scale is widely used in community-dwelling
populations, since it has been validated in frail commu-
nity-dwelling older adult with and without cognitive
impairment [58] and reconciles the best sensitivity to
change and the least missing data [58, 59]. Further works
are still required to identify the best outcome measure
to be used in our hospitalized population. Regarding In conclusion, this study confirmed that hypnosis is fea-
sible and well accepted in an inpatient geriatric rehabili-
tation population. Based on these results, further pilot
work should be performed with an increased number
of hypnosis sessions and the use of hypnotic communi-
cation by the healthcare team in order to reinforce the
hypnotic effect, before conducting a full-scale trial to
conclude whether or not hypnosis is effective to reduce
fear of falling. To identify the best outcome measures to
be used in this full-scale trial to address fear of falling is
also of utmost importance. Activities-specific Balance Confidence scale
“ Chutes et OsteoPoroSe”
Fall Efficacy Scale-International
Functional Independence Measure
Geriatric Depression Scale
Geriatric Fear of Falling Measure
Intention-to-treat
Mini Mental State Examination
Nutritional Risk Score
Per protocol
Short Physical Performance Battery
Visual Analogue Scale Acknowledgements g
The authors thank all members of the CHEOPS staff for their dedication and
professionalism. They are especially indebted to the occupational thera‑
pists’ team for their hard work on the project: Catherine Ducharne (head
occupational therapist), Raoua Ben Aziza, Florence Chiren Cellier, and Noémie
Duc Mba. They also thank Elizabeth Bolomey-Koreneff for her strong support
throughout the project. Page 10 of 11 Cuvelier et al. Pilot and Feasibility Studies (2023) 9:139 Cuvelier et al. Pilot and Feasibility Studies (2023) 9:139 Cuvelier et al. Pilot and Feasibility Studies 9. Makino K, Makizako H, Doi T, Tsutsumimoto K, Hotta R, Nakakubo S,
Suzuki T, Shimada H. Impact of fear of falling and fall history on dis‑
ability incidence among older adults: Prospective cohort study. Int
J Geriatr Psychiatry. 2018;33(4):658–62. https://doi.org/10.1002/gps.
4837. Availability of data and materials The datasets used and analyzed during the current study are available from
the corresponding author on reasonable request. 14. Kendrick D, Kumar A, Carpenter H, Zijlstra GA, Skelton DA, Cook JR, et al. Exercise for reducing fear of falling in older people living in the commu‑
nity. Cochrane Database Syst Rev. 2014;2014(11):CD009848. https://doi.
org/10.1002/14651858.CD009848.pub2. Declarations 15. Parry SW, Bamford C, Deary V, Finch TL, Gray J, MacDonald C, et al. Cognitive-behavioural therapy-based intervention to reduce fear of
falling in older people: therapy development and randomised controlled
trial - the Strategies for Increasing Independence, Confidence and Energy
(STRIDE) study. Health Technol Assess. 2016;20(56):1–206. Consent for publication
Not applicable. Consent for publication
Not applicable. Received: 15 November 2022 Accepted: 18 July 2023 21. Huang TT. Geriatric fear of falling measure: development and psychomet‑
ric testing. Int J Nurs Stud mars. 2006;43(3):357–65. Competing interests 17. Brouwer BJ, Walker C, Rydahl SJ, Culham EG. Reducing fear of falling in
seniors through education and activity programs: a randomized trial. J
Am Geriatr Soc. 2003;51(6):829–34. The authors declare that they have no competing interests. Funding
O Open access funding provided by University of Geneva. This research was
funded by Geneva University Hospitals (Research and development project
9–2018-I). The funding source had no role in the design of the study and col‑
lection, analysis, and interpretation of the data, and in writing the manuscript. 12. Deshpande N, Metter EJ, Lauretani F, Bandinelli S, Guralnik J, Ferrucci L. Activity restriction induced by fear of falling and objective and subjective
measures of physical function: a prospective cohort study. J Am Geriatr
Soc avr. 2008;56(4):615–20. 13. Delbaere K, Close JCT, Brodaty H, Sachdev P, Lord SR. Determinants of
disparities between perceived and physiological risk of falling among
elderly people: cohort study. BMJ. 2010;341:c4165. Authors’ contributions AT had full access to the data in the study and takes responsibility for the
integrity of the data and the accuracy of the data analysis. Study design and
research question: CC, MH, MPZB, FR, CD, DZ, GG, AT. Data collection and
assembly: CC, MH, MPZB, AT. Statistical analysis: MH, FR, AT. Data interpreta‑
tion: CC, MH, MPZB, AT. Drafting manuscript: CC, MH, AT. Revising manu‑
script content: CC, MH, MPZB, FR, CD, DZ, GG, AT. Approving final version of
manuscript: CC, MH, MPZB, FR, CD, DZ, GG, AT. Administrative, technical, and
material support: MPZB, FR, CD, DZ, GG, AT. All authors read and approved the
final manuscript. 10. Denkinger MD, Lukas A, Nikolaus T, Hauer K. Factors associated with fear
of falling and associated activity restriction in community-dwelling older
adults: a systematic review. Am J Geriatr Psychiatry. 2015;23(1):72‑86. 11. Denkinger MD, Igl W, Lukas A, Bader A, Bailer S, Franke S, et al. Relation‑
ship between fear of falling and outcomes of an inpatient geriatric
rehabilitation population–fear of the fear of falling. J Am Geriatr Soc avr. 2010;58(4):664–73. References 22. Powell LE, Myers AM. The Activities-specific Balance Confidence (ABC)
Scale. J Gerontol A Biol Sci Med Sci. 1995;50A(1):M28–34. 1. Stevens JA, Rudd RA. Circumstances and contributing causes of fall
deaths among persons aged 65 and older: United States, 2010. J Am
Geriatr Soc mars. 2014;62(3):470–5. 1. Stevens JA, Rudd RA. Circumstances and contributing causes of fall
deaths among persons aged 65 and older: United States, 2010. J Am
Geriatr Soc mars. 2014;62(3):470–5. 23. Yardley L, Beyer N, Hauer K, Kempen G, Piot-Ziegler C, Todd C. Develop‑
ment and initial validation of the Falls Efficacy Scale-International (FES-I). Age Ageing. 2005;34(6):614–9. 2. Legters K. Fear of falling. Phys Ther mars. 2002;82(3):264–72. 3. Lord SR. Predictors of nursing home placement and mortality of resi‑
dents in intermediate care. Age Ageing. 1994;23(6):499–504. 24. Jørstad EC, Hauer K, Becker C, Lamb SE, ProFaNE Group. Measuring the
psychological outcomes of falling: a systematic review. J Am Geriatr Soc. 2005;53(3):501‑10. 4. Friedman SM, Munoz B, West SK, Rubin GS, Fried LP. Falls and fear of
falling: which comes first? A longitudinal prediction model suggests
strategies for primary and secondary prevention. J Am Geriatr Soc
août. 2002;50(8):1329–35. 25. Ferrer Soler C, Cuvelier C, Hars M, Herrmann FR, Charpiot A, Ducharne
Wieczorkiewicz C, et al. Validation of the Perform-FES: a new fear of falling
scale for hospitalized geriatric patients. Aging Clin Exp Res. 2021;33(1):67–
76. https://doi.org/10.1007/s40520-020-01726-6. 5. Bhala RP, O’Donnell J, Thoppil E. Ptophobia. Phobic fear of falling and
its clinical management. Phys Ther. 1982;62(2):187‑90. 26. Elkins GR, Barabasz AF, Council JR, Spiegel D. Advancing research and
practice: the revised APA Division 30 definition of hypnosis. Int J Clin Exp
Hypn. 2015;63(1):1–9. 6. Esbrí-Víctor M, Huedo-Rodenas I, López-Utiel M, Navarro-López JL,
Martínez-Reig M, Serra-Rexach JA, et al. Frailty and Fear of Falling: The
FISTAC Study. J Frailty Aging. 2017;6(3):136–40. 27. Häuser W, Hagl M, Schmierer A, Hansen E. The Efficacy, Safety and Appli‑
cations of Medical Hypnosis. Dtsch Arzteblatt Int. 2016;113(17):289‑96. y
y
7. Scheffer AC, Schuurmans MJ, van Dijk N, van der Hooft T, de Rooij SE. Fear of falling: measurement strategy, prevalence, risk factors and con‑
sequences among older persons. Age Ageing janv. 2008;37(1):19–24. 28. Norelli LJ, Harju SK. Behavioral approaches to
elderly. Clin Geriatr Med. 2008;24(2):335‑44. 28. Norelli LJ, Harju SK. Behavioral approaches to pain management i
elderly. Clin Geriatr Med. 2008;24(2):335‑44. elderly. Clin Geriatr Med. 2008;24(2):335‑44. 8. Author details 18. Chua CHM, Jiang Y, Lim DS, Wu VX, Wang W. Effectiveness of cognitive
behaviour therapy-based multicomponent interventions on fear of fall‑
ing among community-dwelling older adults: A systematic review and
meta-analysis. J Adv Nurs. 2019;75(12):3299–315. 1 Division of Geriatrics, Department of Rehabilitation and Geriatrics, Geneva
University Hospitals and Faculty of Medicine, Thônex, Switzerland. 2 Division
of Bone Diseases, Department of Medicine, Geneva University Hospitals
and Faculty of Medicine, Geneva, Switzerland. 3 Division of Internal Medicine
for the Aged, Department of Rehabilitation and Geriatrics, Geneva University
Hospitals and Faculty of Medicine, Geneva, Switzerland. y
19. Tricco AC, Thomas SM, Veroniki AA, Hamid JS, Cogo E, Strifler L, et al. Comparisons of Interventions for Preventing Falls in Older Adults: A
Systematic Review and Meta-analysis. JAMA. 2017;318(17):1687‑99. 20. Trombetti A, Hars M, Herrmann F, Rizzoli R, Ferrari S. Effect of a multifacto‑
rial fall-and-fracture risk assessment and management program on gait
and balance performances and disability in hospitalized older adults: a
controlled study. Osteoporos Int. 2013;24(3):867‑76. Received: 15 November 2022 Accepted: 18 July 2023 Received: 15 November 2022 Accepted: 18 July 2023 Ethics approval and consent to participate The study was approved by the Geneva Research Ethics Committee (protocol
2018–01550). All procedures performed were in accordance with the ethical
standards of the institutional research committee and with the 1964 Helsinki
Declaration and its later amendments or comparable ethical standards. 16. Zijlstra GAR, van Haastregt JCM, Ambergen T, van Rossum E, van Eijk JTM,
Tennstedt SL, et al. Effects of a multicomponent cognitive behavioral
group intervention on fear of falling and activity avoidance in commu‑
nity-dwelling older adults: results of a randomized controlled trial. J Am
Geriatr Soc. 2009;57(11):2020–8. References Makino K, Makizako H, Doi T, Tsutsumimoto K, Hotta R, Nakakubo S,
et al. Fear of falling and gait parameters in older adults with and with‑
out fall history. Geriatr Gerontol Int. 2017;17(12):2455‑9. 29. Cordi MJ, Hirsiger S, Mérillat S, Rasch B. Improving sleep and cognition by
hypnotic suggestion in the elderly. Neuropsychologia. 2015;69:176–82. Cuvelier et al. Pilot and Feasibility Studies (2023) 9:139 Page 11 of 11 Cuvelier et al. Pilot and Feasibility Studies 30. Simon EP, Canonico MM. Use of hypnosis in controlling lumbar puncture
distress in an adult needle-phobic dementia patient. Int J Clin Exp Hypn. 2001;49(1):56–67. 54. Deeley Q. Hypnosis as therapy for functional neurologic disorders. Handb
Clin Neurol. 2016;139:585–95. 55. Formica C, Micchia K, Cartella E, De Salvo S, Bonanno L, Corallo F, et al. Analgesic hypnotic treatment in a post-stroke patient. Am J Clin Hypn. 2021;63(3):192‑201. 31. Ardigo S, Herrmann FR, Moret V, Déramé L, Giannelli S, Gold G, et al. Hypnosis can reduce pain in hospitalized older patients: a randomized
controlled study. BMC Geriatr. 2016;16:14. 56. Jensen M, Patterson DR. Hypnotic treatment of chronic pain. J Behav
Med. 2006;29(1):95–124. 32. Billot M, Jaglin P, Rainville P, Rigoard P, Langlois P, Cardinaud N, et al. Hypnosis program effectiveness in a 12-week home care interven‑
tion to manage chronic pain in elderly women: a pilot trial. Clin Ther. 2020;42(1):221–9. 57. Tennstedt S, Howland J, Lachman M, Peterson E, Kasten L, Jette A. A ran‑
domized, controlled trial of a group intervention to reduce fear of falling
and associated activity restriction in older adults. J Gerontol B Psychol Sci
Soc Sci. 1998;53(6):P384-392. 33. Hartman BJ. Hypnotic susceptibility assessed in a group of mentally ill
geriatric patients. J Am Geriatr Soc. 1965;13:460–1. 58. Hauer KA, Kempen GIJM, Schwenk M, Yardley L, Beyer N, Todd C, et al. Validity and sensitivity to change of the falls efficacy scales international
to assess fear of falling in older adults with and without cognitive impair‑
ment. Gerontology. 2011;57(5):462–72. 34. Tinetti ME, Baker DI, McAvay G, Claus EB, Garrett P, Gottschalk M, et al. A multifactorial intervention to reduce the risk of falling among elderly
people living in the community. N Engl J Med. 1994;331(13):821‑7. 35. Kempen GIJM, Zijlstra G a. R, van Haastregt JCM. [The assessment of fear
of falling with the Falls Efficacy Scale-International (FES-I). Development
and psychometric properties in Dutch elderly]. Tijdschr Gerontol Geriatr. 2007;38(4):204‑12. 59. References Denkinger MD, Igl W, Coll-Planas L, Nikolaus T, Bailer S, Bader A, et al. Practicality, validity and sensitivity to change of fear of falling self-report
in hospitalised elderly—a comparison of four instruments. Age Ageing. 2009;38(1):108‑12. 36. Filiatrault J, Gauvin L, Fournier M, Parisien M, Robitaille Y, Laforest S, et al. Evidence of the psychometric qualities of a simplified version of the
Activities-specific Balance Confidence scale for community-dwelling
seniors. Arch Phys Med Rehabil. 2007;88(5):664–72. •
fast, convenient online submission
•
thorough peer review by experienced researchers in your field
•
rapid publication on acceptance
•
support for research data, including large and complex data types
•
gold Open Access which fosters wider collaboration and increased citations
maximum visibility for your research: over 100M website views per year
•
At BMC, research is always in progress.
Learn more biomedcentral.com/submissions
Ready to submit your research
Ready to submit your research ? Choose BMC and benefit from:
? Choose BMC and benefit from: Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in pub‑
lished maps and institutional affiliations. Springer Nature remains neutral with regard to jurisdictional claims in pub‑
lished maps and institutional affiliations. Springer Nature remains neutral with regard to jurisdictional claims in pub‑
lished maps and institutional affiliations. 37. Franzoni S, Rozzini R, Boffelli S, Frisoni GB, Trabucchi M. Fear of falling in
nursing home patients. Gerontology. 1994;40(1):38–44. 38. Miller MD, Towers A. A manual of guidelines for scoring the Cumulative
Illness Rating Scale for Geriatrics (CIRS-G). Pittsburgh, PA: University of
Pittsburgh; 1991. p. 2–31. 39. Folstein MF, Folstein SE, McHugh PR. « Mini-mental state ». A practical
method for grading the cognitive state of patients for the clinician. J Psychiatr Res. 1975;12(3):189‑98. 40. Ploenes C, Sharp S, Martin M. The Clock Test: drawing a clock for
detection of cognitive disorders in geriatric patients. Z Gerontol août. 1994;27(4):246–52. 41. Yesavage JA. Geriatric Depression Scale. Psychopharmacol Bull. 1988;24(4):709–11. 42. Clément JP, Nassif RF, Léger JM, Marchan F. Development and contribu‑
tion to the validation of a brief French version of the Yesavage Geriatric
Depression Scale. Encephale. 1997;23(2):91–9. 43. Kondrup J, Rasmussen HH, Hamberg O, Stanga Z, Ad Hoc ESPEN Working
Group. Nutritional risk screening (NRS 2002): a new method based on an
analysis of controlled clinical trials. Clin Nutr. 2003;22(3):321‑36. 44. Guralnik JM, Simonsick EM, Ferrucci L, Glynn RJ, Berkman LF, Blazer DG,
et al. A short physical performance battery assessing lower extrem‑
ity function: association with self-reported disability and prediction of
mortality and nursing home admission. J Gerontol. 1994;49(2):M85-94. 45. Kidd D, Stewart G, Baldry J, Johnson J, Rossiter D, Petruckevitch A, et al. The Functional Independence Measure: a comparative validity and reli‑
ability study. Disabil Rehabil. 1995;17(1):10–4. 46. Cruz-Jentoft AJ, Baeyens JP, Bauer JM, Boirie Y, Cederholm T, Landi F, et al. Sarcopenia: European consensus on definition and diagnosis. Age Age‑
ing. 2010;39(4):412–23. 47. Delbaere K, Close JCT, Mikolaizak AS, Sachdev PS, Brodaty H, Lord SR. The
Falls Efficacy Scale International (FES-I). A comprehensive longitudinal
validation study. Age Ageing. 2010;39(2):210‑6. 48. Myers AM, Fletcher PC, Myers AH, Sherk W. Discriminative and evaluative
properties of the activities-specific balance confidence (ABC) scale. J Gerontol A Biol Sci Med Sci. 1998;53(4):M287–294. Publisher’s Note •
fast, convenient online submission
•
thorough peer review by experienced researchers in your field
•
rapid publication on acceptance
•
support for research data, including large and complex data types
•
gold Open Access which fosters wider collaboration and increased citations
maximum visibility for your research: over 100M website views per year
•
At BMC, research is always in progress. Learn more biomedcentral.com/submissions
Ready to submit your research
Ready to submit your research ? Choose BMC and benefit from:
? Choose BMC and benefit from: 49. Hars M, Audet MC, Herrmann F, De Chassey J, Rizzoli R, Reny JL, et al. Functional performances on admission predict in-hospital falls, injurious
falls, and fractures in older patients: a prospective study. J Bone Miner Res
Off J Am Soc Bone Miner Res. 2018;33(5):852‑9. f
50. Appel PR. The use of hypnosis in physical medicine and rehabilitation. Psychiatr Med. 1992;10(1):133–48. f
50. Appel PR. The use of hypnosis in physical medicine and rehabilitation. Psychiatr Med. 1992;10(1):133–48. y
51. Wright ME. Hypnosis and rehabilitation. Rehabil Lit. 1960;21:2–12. 51. Wright ME. Hypnosis and rehabilitation. Rehabil Lit. 1960;21:2–12. 52. Becker F. Medical hypnosis in physical medicine and rehabilitation. J Med
Assoc Ga. 1960;49:233–5. 52. Becker F. Medical hypnosis in physical medicine and rehabilitation. J Med
Assoc Ga. 1960;49:233–5. 53. Crasilneck HB, Hall JA. The use of hypnosis in the rehabilitation of com‑
plicate vascular and post-traumatic neurological patients. Int J Clin Exp
Hypn. 1970;18(3):145–59. 53. Crasilneck HB, Hall JA. The use of hypnosis in the rehabilitation of com‑
plicate vascular and post-traumatic neurological patients. Int J Clin Exp
Hypn. 1970;18(3):145–59.
|
https://openalex.org/W2510970602
|
https://www.nature.com/articles/srep31524.pdf
|
English
| null |
The ontogeny of Butyrophilin-like (Btnl) 1 and Btnl6 in murine small intestine
|
Scientific reports
| 2,016
|
cc-by
| 6,014
|
Cristina Lebrero-Fernández & Anna Bas-Forsberg Cristina Lebrero-Fernández & Anna Bas-Forsberg Murine Butyrophilin-like (Btnl) 1 and Btnl6 are primarily restricted to intestinal epithelium where
they regulate the function of intraepithelial T lymphocytes. We recently demonstrated that Btnl1 and
Btnl6 can form an intra-family heterocomplex and that the Btnl1-Btnl6 complex selectively expands
Vγ7Vδ4 TCR IELs. To define the regulation of Btnl expression in the small intestine during ontogeny
we examined the presence of Btnl1 and Btnl6 in the small bowel of newborn to 4-week-old mice. Although RNA expression of Btnl1 and Btnl6 was detected in the small intestine at day 0, Btnl1 and
Btnl6 protein expression was substantially delayed and was not detectable in the intestinal epithelium
until the mice reached 2–3 weeks of age. The markedly elevated Btnl protein level at week 3 coincided
with a significant increase of γδ TCR IELs, particularly those bearing the Vγ7Vδ4 receptor. This was
not dependent on gut microbial colonization as mice housed in germ-free conditions had normal Btnl
protein levels. Taken together, our data show that the expression of Btnl1 and Btnl6 is delayed in the
murine neonatal gut and that the appearance of the Btnl1 and Btnl6 proteins in the intestinal mucosa
associates with the expansion of Vγ7Vδ4 TCR IELs. Members of the Butyrophilin (Btn) and Btn-like (Btnl) family are recognized as novel immune regulators. The
BTN and BTNL proteins are characterized by their structural relatedness to the B7 co-stimulatory molecules,
and as with B7, several of human and murine BTN and BTNL family members are reported to dampen or aug-
ment αβ and γδ T lymphocyte responses1–10. Moreover, BTN3A1 reportedly mediates the activation of Vγ9Vδ2
T cells by phosphoantigens11–13, suggesting that BTN and BTNL proteins not only have the capacity to regulate
T cell-mediated immune responses but are additionally involved in phosphoantigen sensing. In addition, some
of the BTN and BTNL molecules have been associated with inflammatory disorders and cancer. Thus: polymor-
phisms in the human BTNL2 gene have been linked to pulmonary sarcoidosis, ulcerative colitis, rheumatoid
arthritis, myositis and prostate cancer14–18; human BTN3 has been associated with ovarian cancer9,19,20; Btn2a2
deficiency was recently described to exacerbate experimental autoimmune encephalomyelitis and to potentiate
anti-tumor responses21; and we recently presented altered expression of human BTN and BTNL genes in intestinal
inflammation and colon cancer22. Cristina Lebrero-Fernández & Anna Bas-Forsberg These accumulating data provide growing evidence that members of the Btn
and Btnl family play a diverse and essential role in orchestrating the immune system. We previously demonstrated
that Btnl1 can regulate intraepithelial lymphocyte (IEL)–epithelial cell interactions in the murine small intestinal
mucosa by attenuating the epithelial response to activated IELs5. Furthermore, we recently reported that Btnl1,
in a heteromeric complex with Btnl6, selectively enhances the expansion of γδ IELs bearing the Vγ7Vδ4 TCR in
the absence of exogenous activation6. To define the expression of Btnl1 and Btnl6 in the intestine during ontog-
eny we investigated the presence of Btnl mRNA and protein in the small intestine of day 0–4-week-old mice. Additionally, we examined the expression of γδ TCR IELs in the neonate gut and characterized the usage of Vγ7
and Vδ4 chains during the first weeks of neonatal development. www.nature.com/scientificreports www.nature.com/scientificreports www.nature.com/scientificreports The ontogeny of Butyrophilin-like
(Btnl) 1 and Btnl6 in murine small
intestine received: 09 June 2016
accepted: 21 July 2016
Published: 16 August 2016 received: 09 June 2016
accepted: 21 July 2016
Published: 16 August 2016 Materials and Methods Tissue sections were viewed using confocal microscopy (Zeiss LSM700
Inverted) and analyzed with ZEN lite 2011 microscope software (Carl Zeiss, Oberkochen, Germany). Confocal microscopy. Methanol-Carnoy-fixed and paraffin embedded sections were stained with
anti-Btnl1 rabbit polyclonal antiserum or pre-immune serum. Sections were incubated with TRITC-conjugated
AffiniPure F(ab’)2 fragment donkey anti-rabbit IgG (H+L) (Jackson ImmunoResearch, West Grove, PA). Sections
were mounted in ProLong® Gold antifade reagent containing 4′,6-diamidino-2-phenylindole (DAPI) (Molecular
Probes®, Life Technologies, Eugene, OR). Tissue sections were viewed using confocal microscopy (Zeiss LSM700
Inverted) and analyzed with ZEN lite 2011 microscope software (Carl Zeiss, Oberkochen, Germany). Western blotting. Small intestine, harvested from GF and CV C57BL/6 mice, was homogenized in cell
lysis buffer (50 mM Tris-HCl, pH 8, 150 mM NaCl, 1% Triton X-100) containing complete protease inhibi-
tors cocktail tablets (Roche Diagnostics, Mannheim, Germany). Small intestinal epithelial cells isolated from
CV C57BL/6 mice and Btnl6-transfected MODE-K cells were lysed in cell lysis buffer. Cell and tissue lysates
were clarified by centrifugation, and protein concentration was measured with BCA Protein Assay Kit (Pierce,
Rockford, IL). Twenty micrograms of protein were denatured in reducing sample buffer (NuPAGE LDS 4×;
Novex®, Life Technologies, Carlsbad, CA) containing 1M DTT (Sigma-Aldrich, St. Louis, MO) and loaded onto
a NuPage 4–12% Bis-Tris Gel (Novex®, Life Technologies, Carlsbad, CA). Separated proteins were transferred
onto nitrocellulose transfer membranes (Merck Millipore, Darmstadt, Germany) that were immunoblotted using
anti-FLAG antibody (Sigma-Aldrich, St. Louis, MO), anti-Btnl6 rabbit polyclonal antiserum (Moravian-Biotech,
Brno, Czech Republic), rabbit pre-immune serum, or anti-β-actin antibody (Sigma-Aldrich, St. Louis, MO), and
detected with HRP-conjugated goat anti-mouse antibody or HRP-conjugated goat anti-rabbit antibody (Jackson
ImmunoResearch, West Grove, PA). Flow cytometric analysis. Cell-surface antigen expression was analyzed using the following antibodies:
anti-Btnl1 rabbit polyclonal antiserum, pre-immune serum, anti-CD45-Alexa Fluor 700 (30-F11; eBioscience, San
Diego, CA), anti-CD3ε-FITC (145–2C11; BD PharmigenTM, San Diego, CA), anti-pan TCRγδ-eFluor450 (eBioGL3;
eBioscience, San Diego, CA), anti-TCRβ-APC (H57–597; eBioscience, San Diego, CA), anti-TCR Vγ1.1/Cr4-PE
(2.11; BioLegend, San Diego, CA), anti-TCR Vδ4-eFluor660 (GL2; eBioscience, San Diego, CA), anti-TCR Vγ7-biotin
(kindly provided by Dr Pablo Pereira, Institut Pasteur, Paris, France), 7-aminoactinomycin D (7AAD; Sigma-Aldrich,
St. Louis, MO), and LIVE/DEAD® Fixable Red Dead Cell Stain (Molecular Probes®, Life Technologies, Eugene, OR). Detection of anti-Btnl1 and anti-TCR Vγ7 was achieved with APC-conjugated AffiniPure F(ab’)2 fragment donkey
anti-rabbit IgG (H+L) (Jackson ImmunoResearch, West Grove, PA) and streptavidin-APC-CyTM7 (BD Biosciences,
San Diego, CA), respectively. Materials and Methods Mice. Germ-free (GF) 8–12-week-old C57BL/6 mice and conventional (CV) C57BL/6 mice were housed in
the Laboratory of Experimental Biomedicine (EBM), University of Gothenburg (Gothenburg, Sweden). GF mice
were maintained in flexible plastic film isolators under a 12-h light cycle and fed autoclaved chow diet (Labdiet)
ad libitum. GF status was verified regularly by anaerobic culturing in addition to PCR for bacterial 16S rDNA. CV C57BL/6 mice were bred at EBM and were used in the experiments at 0–8 weeks of age. Pregnant mice were Department of Microbiology and Immunology, Institute of Biomedicine, University of Gothenburg, Gothenburg
weden. Correspondence and requests for materials should be addressed to A.B.-F. (email: anna.forsberg1@gu.se) Scientific Reports | 6:31524 | DOI: 10.1038/srep31524 www.nature.com/scientificreports/ identified and monitored daily until delivery. The day of birth was identified as day 0 of life. The pups were kept
in the parental cage until day 21 of age and were thereafter weaned from their mother. Protocols were approved
by the Gothenburg animal ethics committee (Göteborgs djurförsöksetiska nämnd; permit no. 335–2012), and
institutional animal use and care guidelines were followed. identified and monitored daily until delivery. The day of birth was identified as day 0 of life. The pups were kept
in the parental cage until day 21 of age and were thereafter weaned from their mother. Protocols were approved
by the Gothenburg animal ethics committee (Göteborgs djurförsöksetiska nämnd; permit no. 335–2012), and
institutional animal use and care guidelines were followed. RNA extraction and cDNA preparation. Murine small intestinal tissue was lysed and homogenized
(TissuelyserII, Qiagen, Valencia, CA) and total RNA was isolated using RNeasy® mini kit (Qiagen, Valencia, CA),
including DNAse I digestion. RNA concentration was determined spectrophotometrically (NanoDrop ND-1000,
Wilmington, DE). SuperScriptTM III Reverse Transcriptase kit (InvitrogenTM, Life Technologies, Carlsbad, CA)
was used for cDNA synthesis, using 1000 ng RNA as template in total reaction volume of 20 μl. RNA extraction and cDNA preparation. Murine small intestinal tissue was lysed and homogenized
(TissuelyserII, Qiagen, Valencia, CA) and total RNA was isolated using RNeasy® mini kit (Qiagen, Valencia, CA),
including DNAse I digestion. RNA concentration was determined spectrophotometrically (NanoDrop ND-1000,
Wilmington, DE). SuperScriptTM III Reverse Transcriptase kit (InvitrogenTM, Life Technologies, Carlsbad, CA)
was used for cDNA synthesis, using 1000 ng RNA as template in total reaction volume of 20 μl. Quantitative real-time PCR. Materials and Methods Expression of Btnl1 and Btnl6 complementary DNA (cDNA) was measured
by quantitative real-time PCR using GoTaq® qPCR Master Mix, according to the manufacturer’s instructions
(Promega, Madison, WI). The qPCR was performed on a LightCycler480 thermal cycler (Roche Diagnostics,
Mannheim, Germany). The PCR primers, spanning exon-exon borders to avoid amplification of genomic DNA,
were: 5′-tgaccaggagaaatcgaagg-3′ and 5′-caccgagcaggaccaatagt-3′ for Btnl1, and 5′-atccttggagatccacagtgaa-3′
and 5′-gggagagaccttgggaaaga-3′ for Btnl6. RNA expression was normalized to the expression of β-actin
(5′-cttctttgcagctccttcgtt-3′ and 5′-aggagtccttctgacccatgc-3′). Each qPCR analysis was duplicated. Quantitative real-time PCR. Expression of Btnl1 and Btnl6 complementary DNA (cDNA) was measured
by quantitative real-time PCR using GoTaq® qPCR Master Mix, according to the manufacturer’s instructions
(Promega, Madison, WI). The qPCR was performed on a LightCycler480 thermal cycler (Roche Diagnostics,
Mannheim, Germany). The PCR primers, spanning exon-exon borders to avoid amplification of genomic DNA,
were: 5′-tgaccaggagaaatcgaagg-3′ and 5′-caccgagcaggaccaatagt-3′ for Btnl1, and 5′-atccttggagatccacagtgaa-3′
and 5′-gggagagaccttgggaaaga-3′ for Btnl6. RNA expression was normalized to the expression of β-actin
(5′-cttctttgcagctccttcgtt-3′ and 5′-aggagtccttctgacccatgc-3′). Each qPCR analysis was duplicated. Generation of stably transfected N-terminal FLAG-tagged Btnl6-pMX-IRES-GFP-MODE-K
cells. Stably transfected murine intestinal epithelial MODE-K cells were generated, as previously described5. Cells
were maintained at 37 °C, 5% CO2 in Dulbecco’s modified essential medium (DMEM; Gibco®, Life Technologies,
Carlsbad, CA) plus 10% FCS (PAA Laboratories, Linz, Austria), 100 Ug/ml penicillin, 100 μg/ml streptomycin,
0.292 mg/ml glutamine, and 1 × non-essential amino acids (Gibco®, Life Technologies, Carlsbad, CA). Isolation of murine small intestinal epithelial cells and intraepithelial lymphocytes. Intestinal
epithelial cells (iECs) and intraepithelial lymphocytes from murine small bowel were isolated according to previ-
ously described procedures23. IELs were recovered at the interface between 80% and 40% Percoll (GE Healthcare
Bio-sciences AB, Uppsala, Sweden), and iECs were recovered at the interface between 40% and 20% Percoll. Isolation of murine small intestinal epithelial cells and intraepithelial lymphocytes. Intestinal
epithelial cells (iECs) and intraepithelial lymphocytes from murine small bowel were isolated according to previ-
ously described procedures23. IELs were recovered at the interface between 80% and 40% Percoll (GE Healthcare
Bio-sciences AB, Uppsala, Sweden), and iECs were recovered at the interface between 40% and 20% Percoll. Confocal microscopy. Methanol-Carnoy-fixed and paraffin embedded sections were stained with
anti-Btnl1 rabbit polyclonal antiserum or pre-immune serum. Sections were incubated with TRITC-conjugated
AffiniPure F(ab’)2 fragment donkey anti-rabbit IgG (H+L) (Jackson ImmunoResearch, West Grove, PA). Sections
were mounted in ProLong® Gold antifade reagent containing 4′,6-diamidino-2-phenylindole (DAPI) (Molecular
Probes®, Life Technologies, Eugene, OR). Results 2A, Btnl1 and Btnl6 RNA was readily expressed in GF animals and the level of
RNA expression, although slightly lower, was comparable to RNA levels found in animals housed in CV condi-
tions (Fig. 2A). To determine protein levels, small intestinal tissue derived from GF mice was immunostained
with anti-Btnl1 antibody for Btnl1 detection, or used for western blot analysis for detection of Btnl6. Btnl1+
epithelial cells were detected in the intestine of GF animals with a frequency comparable to CV mice (Fig. 2B). Likewise, Btnl6 was present in lysates of small intestinal tissue of GF mice (Fig. 2C). Hence, these data show that
Btnl1 and Btnl6 are expressed in the absence of intestinal microbiota and indicate that the delay in Btnl protein
expression in the intestine of CV mice is not due to gut colonization. Expansion of Vγ7Vδ4 TCR IELs in the neonate gut associates with the appearance of the Btnl1
and Btnl6 proteins. Our recent data show that Btnl1 promotes IEL proliferation and that Btnl1-Btnl6 heter-
omers increase the frequency of Vγ7Vδ4 TCR IELs in particular6. In view of the current results showing a delay
in Btnl1 and Btnl6 protein expression in neonatal intestinal mucosa the relative frequencies of γδ and αβ TCR
IELs and the relative frequencies of Vγ7 and Vγ1 chains, the principal chains utilized in C57BL/6 mice24, were
determined. This analysis identified γδ TCR T cells as the predominant intraepithelial T cell subset comprising
80% (78% ± 9) of the total IELs at birth (Fig. 3A). The ratio between γδ:αβ IELs then dramatically inverted and
at day 7, and until 2 weeks of age, αβ TCR T cells represented the dominant IEL subset (63% ± 8 αβ TCR IELs
vs. 32% ± 7 γδ TCR IELs) (Fig. 3A). At the age of 3 weeks the frequency of γδ TCR T cells increased substantially
(58% ± 5 γδ TCR IELs vs. 36% ± 4 αβ TCR IELs) and at this age the IEL composition was similar to the composi-
tion of IEL population in adult mice (Fig. 3A). The significant enrichment of γδ TCR T cells at week 3 coincided
with the substantial increase in Btnl1 levels and the appearance of Btnl6 in the intestinal mucosa. To determine
whether the appearance of the Btnl proteins coincided with increased Vγ7Vδ4 TCR levels, we examined the rel-
ative frequencies of Vγ7, Vδ4 and Vγ1 IELs. Results Appearance of Btnl1 and Btnl6 in the small intestinal epithelium is delayed during ontog-
eny. At birth and at weaning, mammals are exposed to multiple novel antigens from diet and microbial
colonization. To gain insight into how these events regulate the expression of Btnl1 and Btnl6, small intestinal
epithelial cells isolated from newborn mice and from mice pre- and post-weaning were analyzed for Btnl1 and
Btnl6 expression. Although Btnl1 and Btnl6 transcripts were detected in the small intestine of newborn animals
(Fig. 1A), the mRNA expression was not reflected at the protein level by flow cytometry (Fig. 1B), or immuno-
histochemistry (Fig. 1C), for Btnl1, or by western blot for Btnl6 (Fig. 1D). Directly ex vivo, viable 7AAD− CD45−
small intestinal cells with side and forward scatter typical of epithelial cells showed increasing levels of surface
Btnl1 during intestinal ontogeny; no protein was detected in newborn and 1-week-old pups, the protein appeared
when the mice had reached the age of 2 weeks (2.4% ± 0.3) and then increased to reach levels comparable to
those detected in adult animals (19% ± 5) by 4 weeks of age (24% ± 7) (Fig. 1B). Whereas flow cytometry analysis
detected Btnl1 cell surface expression, the immunostaining of intestinal tissue visualized total protein levels; in
4 μm sections of small intestinal tissue Btnl1 was first detected at 2 weeks of age and, consistent with flow cytom-
etry data, protein expression was found to increase with increasing age (Fig. 1C). As for Btnl1, the expression of
Btnl6 proteins was substantially delayed and was not detectable in the intestinal epithelium until 3 weeks of age
(Fig. 1D). Thus, the appearance of Btnl proteins occurred in pre-weaning pups kept mainly on a milk diet and was
not associated with weaning. The delay in Btnl1 and Btnl6 protein appearance in the neonate gut is not associated with gut
colonization. In order to determine if the expression of Btnl1 and Btnl6 in the intestinal mucosa in the early
postnatal period is dependent on the presence of intestinal microbiota, crude intestinal samples from adult mice
maintained under strict GF conditions were analyzed with real-time PCR for the presence of Btnl1 and Btnl6
transcripts. As is evident in Fig. Results Our data show that within the γδ TCR IEL subset, cells expressing
Vγ7 and cells expressing Vγ7Vδ4 TCR increased markedly between postnatal weeks 2 and 3 (from 50% ± 5 to
78% ± 5 and from 9% ± 1 to 16% ± 4, respectively) (Fig. 3B,C), and that the expansion of these cells correlated
with a substantial increase in Btnl1 protein levels (Fig. 3B,C). In contrast, the ratio of Vγ1-bearing IELs was not
specifically affected by the appearance of the Btnl proteins (Fig. 3D). Materials and Methods Cells were gated on 7AAD or LIVE/DEAD® Fixable Red negative cells to exclude
non-viable cells, and positive staining for Btnl1 was determined by comparison with pre-immune serum. Cell sam-
ples were acquired on LSR II flow cytometer using the DIVA software (BD Biosciences, San Diego, CA), and analysis
of data was perfomed using the FlowJo Software version 7.6.5 (Ashland, OR). Statistical analysis. Data were generated using GraphPad Prism version 6.04 (San Diego, CA). The unpaired
two-tailed t test was used for comparison between two independent groups, while One-Way ANOVA followed
by Holm-Sidak´s multiple comparisons test was applied to evaluate differences between three or more groups. Pearson correlation test was performed to determine the correlation between parameters. Differences were con-
sidered as statistically significant when p < 0.05 (*P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001 and ****P ≤ 0.0001). Scientific Reports | 6:31524 | DOI: 10.1038/srep31524 2 www.nature.com/scientificreports/ Scientific Reports | 6:31524 | DOI: 10.1038/srep31524 Discussion 4–8 mice/time-point were analyzed and One-Way
ANOVA followed by Holm-Sidak´s multiple comparisons test was used for statistical analysis (*P ≤ 0.05,
**P ≤ 0.01, ***P ≤ 0.001 and ****P ≤ 0.0001) (A,B). (C) Small intestinal sections were immunostained with
anti-Btnl1 rabbit polyclonal antiserum (red) and counterstained with DAPI (blue) to visualize nuclei. No
staining was detected using pre-immune serum. Original magnification 20x. Four mice were stained for each
time-point and representative stainings are shown. (D) Isolated iECs from small intestinal tissue of C57BL/6
mice (20 μg) were analyzed for Btnl6 protein expression. Lysates from MODE-K cells transfected with FLAG-
tagged Btnl6 cDNA pMX-IRES-GFP served as a positive control. The predicted protein migrating under
reducing conditions at the theoretical molecular weight of ~59 kDa for FLAG-tagged Btnl6 and ~58 kDa for
non tagged Btnl6 was detected with anti FLAG antibody or Btnl6 specific polyclonal antibody No bands were Figure 1. Btnl1 and Btnl6 expression in murine intestine during ontogeny. Expression of Btnl1 and Btnl6
genes (A), and Btnl1 (B,C) and Btnl6 proteins (D) was examined in small intestinal tissue of newborn (day 0),
1, 2, 3, and 4-week-old and adult C57BL/6 mice. (A) Btnl1 and Btnl6 gene expression was assessed by qPCR, run
in duplicates, and normalized against β-actin. (B) Murine intestinal epithelial cells, gated on CD45− epithelial
cells and 7AAD negative cells to exclude non-viable cells, were stained with anti-Btnl1 rabbit polyclonal
antiserum (solid black line) or pre-immune serum (shaded histogram) which served as a negative control. Representative histogram for each time-point is shown. 4–8 mice/time-point were analyzed and One-Way
ANOVA followed by Holm-Sidak´s multiple comparisons test was used for statistical analysis (*P ≤ 0.05,
**P ≤ 0.01, ***P ≤ 0.001 and ****P ≤ 0.0001) (A,B). (C) Small intestinal sections were immunostained with
anti-Btnl1 rabbit polyclonal antiserum (red) and counterstained with DAPI (blue) to visualize nuclei. No
staining was detected using pre-immune serum. Original magnification 20x. Four mice were stained for each
time-point and representative stainings are shown. (D) Isolated iECs from small intestinal tissue of C57BL/6
mice (20 μg) were analyzed for Btnl6 protein expression. Lysates from MODE-K cells transfected with FLAG-
tagged Btnl6 cDNA pMX-IRES-GFP served as a positive control. The predicted protein migrating under
reducing conditions at the theoretical molecular weight of ~59 kDa for FLAG-tagged Btnl6 and ~58 kDa for
non-tagged Btnl6 was detected with anti-FLAG antibody or Btnl6-specific polyclonal antibody. Discussion We have analyzed the ontogeny of Btnl1 and Btnl6 in the small intestine of neonatal mice. Our data identify a
substantial delay in Btnl1 and Btnl6 protein expression in the neonatal intestinal epithelium and demonstrate
that these proteins are absent in the small intestinal mucosa at birth and during the first 2–3 weeks of life. The
delay in Btnl1 and Btnl6 protein expression was not reflected at the RNA level suggesting post-transcriptional
mechanisms that regulate Btnl protein expression in the postnatal gut. Such mechanisms may involve microR-
NAs that function in RNA silencing and have the capacity to repress the translation of transcribed mRNAs25, or
post-translational turnover of the protein. During early neonatal life, important changes occur in the intestine. The developing gut immune system is challenged by milk and microbial flora and later the diet of mice changes
from milk to pelleted food leading to changes in microbial contents. This period is essential for a complete devel-
opment of the mucosal immune system and we therefore assessed the impact of gut microbiota on Btnl expres-
sion by investigating the presence of Btnl1 and Btnl6 proteins in GF animals. We found that Btnl expression does
not correlate with microbial exposure as mice housed in GF conditions had Btnl1 and Btnl6 protein levels which
were not different from conventional mice. Although the presence of Btnl1 in 14-day-old pre-weaning pups kept
mainly on a milk diet suggests that expression of Btnl proteins is not dependent on change of diet, we cannot
determine whether Btnl expression is regulated by increasing exposure to dietary antigen before weaning, or if
the expression is regulated by an unknown developmental factor. Scientific Reports | 6:31524 | DOI: 10.1038/srep31524 3 www.nature.com/scientificreports/ Figure 1. Btnl1 and Btnl6 expression in murine intestine during ontogeny. Expression of Btnl1 and Btnl6
genes (A), and Btnl1 (B,C) and Btnl6 proteins (D) was examined in small intestinal tissue of newborn (day 0),
1, 2, 3, and 4-week-old and adult C57BL/6 mice. (A) Btnl1 and Btnl6 gene expression was assessed by qPCR, run
in duplicates, and normalized against β-actin. (B) Murine intestinal epithelial cells, gated on CD45− epithelial
cells and 7AAD negative cells to exclude non-viable cells, were stained with anti-Btnl1 rabbit polyclonal
antiserum (solid black line) or pre-immune serum (shaded histogram) which served as a negative control. Representative histogram for each time-point is shown. Discussion No bands were
detected on gels immunoblotted with pre-immune serum. The β-actin immunoblot acts as a loading control. Data are representative of four experiments. Scientific Reports | 6:31524 | DOI: 10.1038/srep31524 4 www.nature.com/scientificreports/ ntificreports/
Figure 2. Btnl1 and Btnl6 expression in germ-free mice. Expression of Btnl1 and Btnl6 transcripts (A), and
Btnl1 (B) and Btnl6 proteins (C) was analyzed in small intestinal tissue of GF and CV adult C57BL/6 mice. (A) Btnl1 and Btnl6 gene expression was assessed by qPCR, run in duplicates, and normalized against β-actin. 4–8 mice/group were included and unpaired two-tailed t test was used for statistical analysis (*P ≤ 0.05,
**P ≤ 0.01, ***P ≤ 0.001 and ****P ≤ 0.0001). (B) Small intestinal sections were immunostained with anti-
Btnl1 rabbit polyclonal antiserum (red) and counterstained with DAPI (blue) to visualize nuclei. No staining
was detected using pre-immune serum. Original magnification 20x. Four mice were stained for each group and
representative stainings are shown. (C) Lysates from small intestinal tissue of GF and CV adult C57BL/6 mice
(20 μg) were analyzed for Btnl6 protein expression. Lysates from MODE-K cells transfected with FLAG-tagged
Btnl6 cDNA pMX-IRES-GFP served as a positive control. The predicted protein migrating under reducing
conditions at the theoretical molecular weight of ~59 kDa for FLAG-tagged Btnl6 and ~58 kDa for non-tagged
Btnl6 was detected with anti-FLAG antibody or Btnl6-specific polyclonal antibody. No bands were detected
on gels immunoblotted with pre-immune serum. The β-actin immunoblot acts as a loading control. Data are
representative of two experiments. GF: germ-free; CV: conventional. Figure 2. Btnl1 and Btnl6 expression in germ-free mice. Expression of Btnl1 and Btnl6 transcripts (A), and
Btnl1 (B) and Btnl6 proteins (C) was analyzed in small intestinal tissue of GF and CV adult C57BL/6 mice. (A) Btnl1 and Btnl6 gene expression was assessed by qPCR, run in duplicates, and normalized against β-actin. 4–8 mice/group were included and unpaired two-tailed t test was used for statistical analysis (*P ≤ 0.05,
**P ≤ 0.01, ***P ≤ 0.001 and ****P ≤ 0.0001). (B) Small intestinal sections were immunostained with anti-
Btnl1 rabbit polyclonal antiserum (red) and counterstained with DAPI (blue) to visualize nuclei. No staining
was detected using pre-immune serum. Original magnification 20x. Four mice were stained for each group and
representative stainings are shown. (C) Lysates from small intestinal tissue of GF and CV adult C57BL/6 mice
(20 μg) were analyzed for Btnl6 protein expression. Discussion Lysates from MODE-K cells transfected with FLAG-tagged
Btnl6 cDNA pMX-IRES-GFP served as a positive control. The predicted protein migrating under reducing
conditions at the theoretical molecular weight of ~59 kDa for FLAG-tagged Btnl6 and ~58 kDa for non-tagged
Btnl6 was detected with anti-FLAG antibody or Btnl6-specific polyclonal antibody. No bands were detected
on gels immunoblotted with pre-immune serum. The β-actin immunoblot acts as a loading control. Data are
representative of two experiments. GF: germ-free; CV: conventional. We previously reported that Btnl1 can promote the expansion of small intestinal IELs, and that Btnl1 in a heter-
omeric protein complex with Btnl6, augments the expansion of γδ T cells bearing the Vγ7Vδ4 TCR in particular6. To examine if the appearance of the Btnl1 and Btnl6 proteins in the intestinal epithelium was associated with the
expansion of γδ TCR cells we examined the percentage of Vγ7 and Vγ7Vδ4 bearing IELs in the neonatal mice. In agreement with previous reports investigating neonatal mouse small intestine26, we found that between 2 and
3 weeks of age γδ TCR IELs increased substantially, whereas the percentage of αβ TCR IELs decreased. Within
the expanding γδ TCR IEL compartment Vγ7 and Vγ7Vδ4 TCR bearing T cells showed the greatest increase that
correlated with the appearance of the Btnl proteins in the intestine. The expansion of γδ ΤCR IELs is indicative of
a major influx of γδ TCR lymphocytes and/or a local expansion of resident γδ TCR IELs in the neonatal gut epi-
thelium. Observations from rat IEL ontogeny in which relative IEL numbers were similar in normal and athymic Scientific Reports | 6:31524 | DOI: 10.1038/srep31524 5 www.nature.com/scientificreports/ 6
6:31524 | DOI: 10 1038/srep31524
Figure 3. Btnl1 expression correlates with the presence of IELs bearing the Vγ7Vδ4 TCR. Small intestinal
IELs from newborn (day 0), 1, 2, 3, and 4-week-old and adult C57BL/6 mice were analyzed for the expression
of αβ and γδ TCR (A), and the γδ TCR IELs for the expression of Vγ7, Vγ1, and Vδ4 chains (B–D). 3–11
mice/group (A) and 7–16 mice/group (B–D) were analyzed and One-Way ANOVA followed by Holm-
Sidak´s multiple comparisons test was used for statistical analysis. Correlation between Btnl1 expression and
the percentage of Vγ7 TCR IELs (B) or Vγ7Vδ4 TCR IELs (C) during the mouse ontogeny was determined
using the Pearson correlation test (*P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001 and ****P ≤ 0.0001). References P. MicroRNAs: genomics, biogenesis, mechanism, and function. Cell 116, 281–297 (2004). l
d
l
ff
l
f T
ll
b
h l
d
l 25. Bartel, D. P. MicroRNAs: genomics, biogenesis, mechanism, an ,
R
g
,
g
,
,
,
(
)
26. Kuo, S., El Guindy, A., Panwala, C. M., Hagan, P. M. & Camerini, V. Differential appearance of T cell subsets in the large and
intestine of neonatal mice. Pediatric research 49, 543–551, doi: 10.1203/00006450-200104000-00017 (2001). l
l
d
d
l
d
l d
l
f
h l l
d l
d l h 27. Helgeland, L., Brandtzaeg, P., Rolstad, B. & Vaage, J. T. Sequential development of intraepithelial gamma delta and alpha beta
T lymphocytes expressing CD8 alpha beta in neonatal rat intestine: requirement for the thymus. Immunology 92, 447–456 (1997). 28. Rocha, B., Vassalli, P. & Guy-Grand, D. Thymic and extrathymic origins of gut intraepithelial lymphocyte populations in mice. The
Journal of experimental medicine 180, 681–686 (1994). T lymphocytes expressing CD8 alpha beta in neonatal rat intestine: requirement for the thymus. Immunology 92, 447 456 (1997). 28. Rocha, B., Vassalli, P. & Guy-Grand, D. Thymic and extrathymic origins of gut intraepithelial lymphocyte populations in mice. The
Journal of experimental medicine 180, 681–686 (1994). f
p
29. Lefrancois, L., LeCorre, R., Mayo, J., Bluestone, J. A. & Goodman, T. Extrathymic selection of TCR gamma delta +T cells by cl
major histocompatibility complex molecules. Cell 63, 333–340, doi: 0092-8674(90)90166-C (1990). Discussion Each dot in the
correlation analysis in Fig. 3B,C represents % of Btnl1+ epithelial cells vs % of Vγ7 (Vδ4) TCR IELs of one
mouse at a particular time-point. Newborn −8-week-old mice were used in this correlation analysis. Figure 3. Btnl1 expression correlates with the presence of IELs bearing the Vγ7Vδ4 TCR. Small intestinal
IELs from newborn (day 0), 1, 2, 3, and 4-week-old and adult C57BL/6 mice were analyzed for the expression
of αβ and γδ TCR (A), and the γδ TCR IELs for the expression of Vγ7, Vγ1, and Vδ4 chains (B–D). 3–11
mice/group (A) and 7–16 mice/group (B–D) were analyzed and One-Way ANOVA followed by Holm-
Sidak´s multiple comparisons test was used for statistical analysis. Correlation between Btnl1 expression and
the percentage of Vγ7 TCR IELs (B) or Vγ7Vδ4 TCR IELs (C) during the mouse ontogeny was determined
using the Pearson correlation test (*P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001 and ****P ≤ 0.0001). Each dot in the
correlation analysis in Fig. 3B,C represents % of Btnl1+ epithelial cells vs % of Vγ7 (Vδ4) TCR IELs of one
mouse at a particular time-point. Newborn −8-week-old mice were used in this correlation analysis. Scientific Reports | 6:31524 | DOI: 10.1038/srep31524 6 www.nature.com/scientificreports/ neonates27, and from mouse studies in which γδ TCR IELs were regenerated in the small intestine in the absence
of the thymus28, suggest that the IEL colonization process may be regulated by gut micro-environmental factors
rather than by immigration of thymus-derived T cells. Intriguingly, representation of Vδ4-expressing IEL subsets
is influenced by genes linked to the MHC Class II region24,29, which also contains the Btnl1 and Btnl6 genes. Although further experiments, for example using Btnl-/- approaches, will be necessary to confirm the association
between Btnl1 and Btnl6 and the expansion of γδ TCR IELs, particularly those bearing the Vγ7Vδ4 TCR, these
data support our recent in vitro results6 and further add strength to the evidence of a link between Btnl genes and
the γδ expressing IEL repertoire in the intestinal intraepithelial compartment. References 1. Nguyen, T., Liu, X. K., Zhang, Y. & Dong, C. BTNL2, a butyrophilin-like molecule that functions to inhibit T cell activation
J Immunol. 176, 7354–7360, doi: 176/12/7354 (2006). 2. Yamazaki, T. et al. A butyrophilin family member critically inhibits T cell activation. J Immunol. 185, 5907–5914, doi: 10.4049/
jimmunol.1000835 (2010). 2. Yamazaki, T. et al. A butyrophilin family member critically inhibits T cell activation. J Immunol. 185, 5907–5914, doi: 10.4049/
jimmunol.1000835 (2010). j
(
)
3. Arnett, H. A. et al. BTNL2, a butyrophilin/B7-like molecule, is a negative costimulatory molecule modulated in intestinal
inflammation. J Immunol. 178, 1523–1533, doi: 178/3/1523 (2007). l
4. Chapoval, A. I. et al. BTNL8, a butyrophilin-like molecule that costimulates the primary immune response. Mol Immunol 56
819–828, doi: 10.1016/j.molimm.2013.08.003 (2013). l
4. Chapoval, A. I. et al. BTNL8, a butyrophilin-like mo y
819–828, doi: 10.1016/j.molimm.2013.08.003 (2013). j
5. Bas, A. et al. Butyrophilin-like 1 encodes an enterocyte protein that selectively regulates functional interactions with T lymphocytes. Proc Natl Acad Sci USA 108, 4376–4381, doi: 10.1073/pnas.1010647108 (2011). p
6. Lebrero-Fernandez, C., Bergstrom, J. H., Pelaseyed, T. & Bas-Forsberg, A. Murine Butyrophilin-Like 1 and Btnl6 Form Heteromeric
Complexes in Small Intestinal Epithelial Cells and Promote Proliferation of Local T Lymphocytes. Frontiers in immunology 7, 1,
doi: 10.3389/fimmu.2016.00001 (2016). i
7. Yamashiro, H., Yoshizaki, S., Tadaki, T., Egawa, K. & Seo, N. Stimulation of human butyrophilin 3 molecules results in negative
regulation of cellular immunity. Journal of leukocyte biology 88, 757–767, doi: 10.1189/jlb.0309156 (2010). g
y
f
y
gy
j
8. Smith, I. A. et al. BTN1A1, the mammary gland butyrophilin, and BTN2A2 are both inhibitors of T cell activation. J Immunol. 184
3514–3525, doi: 10.4049/jimmunol.0900416 (2010). j
9. Cubillos-Ruiz, J. R. et al. CD277 is a negative co-stimulatory molecule universally expressed by ovarian cancer microenvironmenta
cells. Oncotarget 1, 329–338 (2010).f g
10. Messal, N. et al. Differential role for CD277 as a co-regulator of the immune signal in T and NK cells. Eur J Immunol. 41, 3443–3454,
doi: 10.1002/eji.201141404 (2011). j
1. Vavassori, S. et al. Butyrophilin 3A1 binds phosphorylated antigens and stimulates human gammadelta T cells. Nat Immunol. 14
908–916, doi: 10.1038/ni.2665 (2013).h 12. Sandstrom, A. et al. The intracellular B30.2 domain of butyrophilin 3A1 binds phosphoantigens to mediate activation of human
Vgamma9Vdelta2 T cells. Immunity 40, 490–500, doi: 10.1016/j.immuni.2014.03.003 (2014). g
y
j
13. Sebestyen, Z. et al. References RhoB Mediates Phosphoantigen Recognition by Vgamma9Vdelta2 T Cell Receptor. Cell reports 15, 1973–1985,
doi: 10.1016/j.celrep.2016.04.081 (2016). 4. Valentonyte, R. et al. Sarcoidosis is associated with a truncating splice site mutation in BTNL2. Nat Genet 37, 357–364, doi: 10.1038
ng1519 (2005).i g
5. Pathan, S. et al. Confirmation of the novel association at the BTNL2 locus with ulcerative colitis. Tissue Antigens 74, 322–329
doi: 10.1111/j.1399-0039.2009.01314.x (2009).i j
6. Mitsunaga, S. et al. Exome sequencing identifies novel rheumatoid arthritis-susceptible variants in the BTNL2. Journal of human
genetics 58, 210–215, doi: 10.1038/jhg.2013.2 (2013).l g
j g
17. Price, P. et al. Two major histocompatibility complex haplotypes influence susceptibility to sporadic inclusion body myositis: c
evaluation of an association with HLA-DR3. Tissue Antigens 64, 575–580, doi: 10.1111/j.1399-0039.2004.00310.x (2004). g
j
(
)
18. Fitzgerald, L. M. et al. Germline missense variants in the BTNL2 gene are associated with prostate cancer susceptibility. Cancer
epidemiology, biomarkers & prevention: a publication of the American Association for Cancer Research, cosponsored by the American
Society of Preventive Oncology 22, 1520–1528, doi: 10.1158/1055-9965.EPI-13-0345 (2013). y f
gy
19. Peedicayil, A. et al. Risk of ovarian cancer and inherited variants in relapse-associated genes. Plos one 5, e8884, doi: 10.1371/journal. pone.0008884 (2010).i p
0. Le Page, C. et al. BTN3A2 expression in epithelial ovarian cancer is associated with higher tumor infiltrating T cells and a bette
prognosis. Plos one 7, e38541, doi: 10.1371/journal.pone.0038541 (2012).h 21. Sarter, K. et al. Btn2a2, a T cell immunomodulatory molecule coregulated with MHC class II genes. The Journal of experimental
medicine 213, 177–187, doi: 10.1084/jem.20150435 (2016).l medicine 213, 177–187, doi: 10.1084/jem.20150435 (2016). j
(
)
22. Lebrero-Fernández, C. et al. Altered expression of Butyrophilin (BTN) and BTN-like (BTNL) genes in intestinal inflammation and
colon cancer. Immunity, Inflammation and Disease 4, 191–200, doi: 10.1002/iid3.105 (2016). 22. Lebrero-Fernández, C. et al. Altered expression of Butyroph colon cancer. Immunity, Inflammation and Disease 4, 191 200, doi: 10.1002/iid3.105 (2016). 23. Lundqvist, C., Hammarstrom, M. L., Athlin, L. & Hammarstrom, S. Isolation of functionally active intraepithelial lymphocyte
enterocytes from human small and large intestine. J Immunol Methods 152, 253–263, doi: 0022-1759(92)90147-L (1992). 24. Pereira, P., Lafaille, J. J., Gerber, D. & Tonegawa, S. The T cell receptor repertoire of intestinal intraepithelial gammadelta
T lymphocytes is influenced by genes linked to the major histocompatibility complex and to the T cell receptor loci. Proc Natl Acad
Sci USA 94, 5761–5766 (1997). 25. Bartel, D. Author Contributions C.L.-F. designed and performed experiments, analyzed data, prepared figures and revised the manuscript; A.B.-F. supervised the project, designed experiments, interpreted the data and wrote the manuscript. Acknowledgements g
We wish to thank Fredrik Bäckhed (Gothenburg University) for providing germ-free mice, Dr. Pablo Pereira (Institut
Pasteur) for providing antibodies, and Paul Bland (Gothenburg University) for critical reading of the manuscript. We also thank the Centre of Cellular Imaging at the Sahlgrenska Academy for providing microscope training and
equipment. This work was supported by the Swedish Research Council (ABF, grant no. 621-2011-4917). Scientific Reports | 6:31524 | DOI: 10.1038/srep31524 7 www.nature.com/scientificreports/ Scientific Reports | 6:31524 | DOI: 10.1038/srep31524 Additional Informationi Competing financial interests: The authors declare no competing financial interests. Competing financial interests: The authors declare no competing financial interests. How to cite this article: Lebrero-Fernández, C. and Bas-Forsberg, A. The ontogeny of Butyrophilin-like (Btnl)
1 and Btnl6 in murine small intestine. Sci. Rep. 6, 31524; doi: 10.1038/srep31524 (2016). How to cite this article: Lebrero-Fernández, C. and Bas-Forsberg, A. The ontogeny of Butyrophilin-like (Btnl)
1 and Btnl6 in murine small intestine. Sci. Rep. 6, 31524; doi: 10.1038/srep31524 (2016). How to cite this article: Lebrero-Fernández, C. and Bas-Forsberg, A. The ontogeny of Butyrophilin-like (Btnl)
1 and Btnl6 in murine small intestine. Sci. Rep. 6, 31524; doi: 10.1038/srep31524 (2016). This work is licensed under a Creative Commons Attribution 4.0 International License. The images
or other third party material in this article are included in the article’s Creative Commons license,
unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license,
users will need to obtain permission from the license holder to reproduce the material. To view a copy of this
license, visit http://creativecommons.org/licenses/by/4.0/ Scientific Reports | 6:31524 | DOI: 10.1038/srep31524 8
|
https://openalex.org/W4211183069
|
https://diposit.ub.edu/dspace/bitstream/2445/120386/1/657131.pdf
|
English
| null |
Hepatocellular carcinoma
|
Nature reviews. Disease primers
| 2,016
|
public-domain
| 21,703
|
Hepatocellular carcinoma 10Instituto de Investigación Sanitaria de Navarra (IDISNA) and Centro de Investigación
Biomédica en Red de Enfermedades Hepáticas y Digestivas (Ciberehd), Pamplona, Spain. 11Department of Gastroenterology, University Health Network, University of Toronto,
Canada. 12Mayo Clinic, Mayo College of Medicine, Rochester, Minnesota, USA. Correspondence to JML
email: Josep.Llovet@mssm.edu Hepatocellular carcinoma Josep M. Llovet1,2,3, Jessica Zucman-Rossi4,5,6,7, Eli Pikarsky8, Bruno Sangro9,10, Myron
Schwartz1, Morris Sherman11 and Gregory Gores12 1Liver Cancer Program, Division of Liver Diseases, Tisch Cancer Institute, Department of
Medicine, Icahn School of Medicine at Mount Sinai, Madison Ave 1425, 11F-70, Box 1123
New York, NY10029, New York, USA. 1Liver Cancer Program, Division of Liver Diseases, Tisch Cancer Institute, Department of
Medicine, Icahn School of Medicine at Mount Sinai, Madison Ave 1425, 11F-70, Box 1123
New York, NY10029, New York, USA. 2Liver Cancer Translational Research Laboratory, Barcelona Clínic Liver Cancer Group (BCLC),
Liver Unit, IDIBAPS- Hospital Clínic, CIBEREHD, University of Barcelona, Catalonia, Spain. 3 2Liver Cancer Translational Research Laboratory, Barcelona Clínic Liver Cancer Group (BCLC),
Liver Unit, IDIBAPS- Hospital Clínic, CIBEREHD, University of Barcelona, Catalonia, Spain. 3I
tit
ió C t l
d R
i E t di A
t (ICREA) B
l
C t l
i
S
i ció Catalana de Recerca i Estudis Avançats (ICREA), Barcelona, Catalonia, Spain. 3Institució Catalana de Recerca i Estudis Avançats (ICREA), Barcelona, Catalonia, Spain. 4Inserm, UMR-1162, Génomique fonctionnelle des Tumeurs solides, Equipe Labellisée Ligue
Contre le Cancer, Institut Universitaire d’Hematologie, Paris, France. 4Inserm, UMR-1162, Génomique fonctionnelle des Tumeurs solides, Equipe Labellisée Ligue
Contre le Cancer, Institut Universitaire d’Hematologie, Paris, France. 4Inserm, UMR-1162, Génomique fonctionnelle des Tumeurs solides, Equipe Labellisée Ligue
Contre le Cancer, Institut Universitaire d’Hematologie, Paris, France. 5Université Paris Descartes, Labex Immuno-Oncology, Sorbonne Paris Cité, Faculté de
Médecine, Paris, France. 5Université Paris Descartes, Labex Immuno-Oncology, Sorbonne Paris Cité, Faculté de
Médecine, Paris, France. 6Université Paris 13, Sorbonne Paris Cité, Unité de Formation et de Recherche Santé,
Médecine, Biologie humaine, Bobigny, France. 6Université Paris 13, Sorbonne Paris Cité, Unité de Formation et de Recherche Santé,
Médecine, Biologie humaine, Bobigny, France. 7Université Paris Diderot, Paris, France. 7Université Paris Diderot, Paris, France. 8Lautenberg Center for Immunology and Cancer Research and Department of Pathology,
Hebrew University Hadassah- Medical School, Jerusalem, Israel. 8Lautenberg Center for Immunology and Cancer Research and Department of Pathology,
Hebrew University Hadassah- Medical School, Jerusalem, Israel. 9Liver Unit, Clínica Universidad de Navarra, Pamplona, Spain. 10 9Liver Unit, Clínica Universidad de Navarra, Pamplona, Spain. 9Liver Unit, Clínica Universidad de Navarra, Pamplona, Spain. 10Instituto de Investigación Sanitaria de Navarra (IDISNA) and Centro de Investigación
Biomédica en Red de Enfermedades Hepáticas y Digestivas (Ciberehd), Pamplona, Spain. 11Department of Gastroenterology, University Health Network, University of Toronto,
Canada. Abstract Hepatocellular carcinoma (HCC) represents approximately 90% of all cases of primary liver
cancer, which is the second leading cause of cancer related deaths globally and has an
incidence of 850,000 new cases per year. The main risk factors for developing HCC are well-
known and include infection with hepatitis B and C viruses, alcohol intake and ingestion of
the fungal metabolite aflatoxin B1. Nonetheless, knowledge is emerging regarding additional
risk factors such as non-alcoholic steatohepatitis. Advances in the understanding of the
molecular pathogenesis of HCC led to identification of critical driver mutations, however the
most prevalent of these are not yet druggable targets. The molecular classification of HCC is
not established, and the Barcelona-Clinic-Liver Cancer Classification is the main clinical
algorithm for the stratification of patients according to prognosis and treatment allocation. Surveillance programmes enable detection of early-stage tumours that are amenable to
curative therapies — resection, liver transplantation or local ablation. At more-developed
stages, only chemoembolization (for intermediate HCC) and sorafenib (for advanced HCC)
have shown survival benefits. There are major unmet needs in HCC management that might
be addressed through discovery of new therapies and their combinations for use in the
adjuvant setting and for intermediate and advanced stage disease, biomarkers for therapy
stratification, patient-tailored strategies targeting driver mutations and/or activating
signalling cascades and validated measurements of quality of life. Recent failures in testing
systemic drugs for intermediate and advanced stages have pointed towards a refinement in
trial design and defining novel approaches. Author contributions: Introduction (J.M.L.); Epidemiology (G.G.); Mechanisms/pathophysiology (J.Z.R. and E.P.);
Diagnosis, screening and prevention (M.Sh. and G.G.); Management (M.Sc, B.S. and J.M.L.,);
Quality of life (M.Sh.); Outlook (J.M.L., G.G. and EP); overview of Primer (J.M.L.). Acknowledgments: J.M.L. has grants from the European Commission Horizon 2020 (HEP-
CAR, proposal number 667273-2), the Samuel Waxman Cancer Research Foundation, the 1 Grant I+D Program (SAF2013-41027) and the Asociación Española Contra el Cáncer (AECC). J.Z-R has received funding from INSERM, the French National Cancer Institute (INCa) and The
Ligue Contre le Cancer (équipe Labellisée). E.P receives funding from the Dr. Miriam and
Sheldon G. Adelson foundation, the European Research Council and the Israel Science
Foundation. We would like to thank Robert Montal, MD (Liver Cancer Translational Lab,
BCLC Group, IDIBAPS – Hospital Clínic) for his support in designing figures and tables for this
manuscript. Competing interests: J.M.L. receives research support and grants from Bayer
Pharmaceuticals, Bristol Myers Squibb, Boehringer-Ingelheim and Blueprint Medicines, and
is a consultant for Bayer Pharmaceuticals, Bristol Myers Squibb, Blueprint Medicines, Eli Lilly,
Celsion, Biocompatibles, Boehringer-Ingelheim, Novartis and GlaxoSmithKline. M.Sh. is
consultant for Bayer Pharmaceuticals, Celsion, ArQule, H3 Biomedicine and Merck. J.Z-R. is a
consultant for IntegraGen. B.S. has received lecturing and consulting fees from Bayer
Healthcare and Sirtex Medical. G.G. is in the Data Safety and Monitoring committee for a
Bayer Pharmaceuticals Trial in hepatocellular carcinoma. M.Sc. and E.P. have nothing to
disclose. 2 2 [H1] Introduction Classification of HCC is not based on the TNM system, as opposed to the majority of
neoplasms, but on the Barcelona-Clinic-Liver Cancer (BCLC) Classification, which is endorsed
by European and American clinical practice guidelines3,15,16. This staging system defines five
prognostic subclasses and allocates specific treatments for each stage. Five treatments can
extend the life expectancy of patients with HCC: surgical resection, liver transplantation,
radiofrequency ablation, chemoembolization and the multikinase inhibitor sorafenib3. Around 40% of patients (early stages 0 and A) are eligible for potentially curative therapies
— resection, transplantation or local ablation — which can provide median survival rates of
60 months and beyond, in contrast with an historical natural history survival of 36 months. For patients with more-advanced disease, only two treatments have demonstrated survival
advantages in the setting of randomized-controlled trials (RCTs). Patients at intermediate
stage (stage B) with preserved liver function benefit from chemoembolization17,18 and have
an estimated median survival of 26 months. Patients at advanced stage (stage C) benefit
from systemic sorafenib, which extends survival by approximately 3 months (from 8 to 11
months) and represents the standard of care in this setting19. During the past years, several
studies have tested therapies in the adjuvant setting, in combination or alternatives to
chemoembolization, and alternative systemic first-line and second-line treatments. This Primer provides an overview and up-dated summary of the current knowledge on the
epidemiology, pathogenesis and treatment of HCC. We describe current prevention,
evidence-based standards of care and novel therapies emerging in light of the understanding
of the pathogenesis of HCC, such as proof-of-concept studies and immunotherapy. [H1] Introduction Liver cancer is a major health problem, with more than 850,000 cases annually worldwide1. This neoplasm is currently the second leading cause of cancer-related death globally, a figure
that is on the rise2. Among all primary liver cancers, hepatocellular carcinoma (HCC) is the
most common neoplasm, accounting for 90% of cases1,3–12. Various risk factors for HCC
development are well-defined, such cirrhosis (chronic liver damage caused by fibrosis),
hepatitis B virus (HBV) infection, hepatitis C virus (HCV) infection, alcohol abuse and
metabolic syndrome3. Other co-factors, such as tobacco inhalation and aflatoxin B1 (a fungal
carcinogen present in food supplies associated with mutations in the tumour suppressor
gene TP53) intake are well-characterized contributors to HCC (Figure 1)1,3–10. Recent
discoveries have pointed to infection with adeno-associated virus 2 (AAV2) as a novel cause
of the disease, particularly in individuals without cirrhosis13. Primary prevention of HCC
through HBV vaccination has been demonstrated. Similarly, in patients with chronic
infection, effective anti-viral therapies against HBV and HCV that produce sustained
virological responses are associated with a profound decrease in HCC incidence. Guidelines
recommend surveillance by using ultrasonography every 6 months in at risk-populations. 3 Over the past decade, there has been an improvement in the understanding of the
molecular pathogenesis of the disease14. Genomic analysis has provided a clear picture of
the main drivers responsible for tumour initiation and progression. Each HCC has an average
of 40 genomic aberrations among which few are considered drivers. Common mutations
affect telomere maintenance (mutations in telomere reverse transcriptase (TERT)), Wnt
pathway activation (mutations in catenin beta 1 (CTNNB1)), inactivation of cellular tumour
antigen p53 (p53, encoded by TP53), chromatin remodelling (mutations in AT-rich
interaction domain 1A (ARID1A)), Ras signalling, mammalian target of rapamycin (mTOR)
signalling and oxidative stress pathway activation. Only a handful of these drivers are
currently druggable targets, such as amplification of fibroblast growth factor 19 (FGF19). Classification of HCC is not based on the TNM system, as opposed to the majority of
neoplasms, but on the Barcelona-Clinic-Liver Cancer (BCLC) Classification, which is endorsed
by European and American clinical practice guidelines3,15,16. This staging system defines five
prognostic subclasses and allocates specific treatments for each stage. Five treatments can
extend the life expectancy of patients with HCC: surgical resection, liver transplantation,
radiofrequency ablation, chemoembolization and the multikinase inhibitor sorafenib3. [H1] Epidemiology Worldwide, liver cancer is the sixth most common cancer (approximately 850,000 new cases
each year), and the second leading cause of cancer-related deaths (approximately 800,000
per year) (Figure 1)1,3–10. 85-90% of all primary liver cancers are HCC. Unlike other human 4 4 malignancies, the risk factors for HCC are well established (Table 1). Indeed, HCC is common
in patients with advanced hepatic fibrosis or cirrhosis due to chronic liver disease, and in
particular with liver damage caused by HBV and HCV infection and unhealthy alcohol use. The worldwide incidence of HCC and chronic viral hepatitis parallel each other. HBV is a DNA
virus that can cause insertional mutagenesis. In the context of chronic HBV infection, HCC
usually occurs on a background of cirrhosis (up to 85% in selected studies)20. Moreover, in
patients with chronic HCV infection, which is caused by an RNA virus, HCC rarely occurs in
the absence of advanced hepatic fibrosis or cirrhosis21. The highest incidence rates of HCC
are in Asia and sub-Saharan Africa owing to the high prevalence of HBV infection (Figure 2)1. In Africa, aflatoxin B1 appears to be synergistic with HBV in causing HCC22. This synergy is a
likely explanation for the earlier onset of HCC in this continent compared with the rest of the
world. In North America, Japan and Europe, HCV is the leading cause of HCC. Other causes of
cirrhosis associated with HCC include unhealthy alcohol use, nonalcoholic steatohepatitis
(NASH), alpha-1 antitrypsin deficiency and hemochromatosis23. By contrast, HCC is less
common in cirrhosis that is caused by autoimmune hepatitis, Wilson’s disease and
cholestatic liver disorders. Along with viral hepatitis and cirrhosis, other factors can contribute to disease risk. HCC has
a strong gender predilection, being threefold more common in men than women24 (Figure
2)1. Most patients with HCC are over 45 years of age, except in sub-Saharan Africa, given the
latency between the onset of virus-mediated liver inflammation and the development of
cirrhosis. Unidentified host, viral and environmental interactions are likely responsible for
the lower age of onset in Africa. The association between tobacco use and HCC, even in the
presence of HBV or HCV, has been inconsistent. An emerging cause of HCC is the metabolic
syndrome due to diabetes and obesity, and the associated liver disease nonalcoholic fatty
liver disease (NAFLD) and NASH25. [H1] Epidemiology NAFLD or NASH might be — along with chronic HBV
infection — the exception to the rule that HCC is always associated with advanced hepatic
fibrosis or cirrhosis. A recent study that needs careful verification suggests that
approximately 40% of patients with HCC and NAFLD or NASH might not have cirrhosis25. As discussed in greater detail below, HCC is one of the cancers for which prevention is
possible. HBV vaccination has been shown to reduce the incidence of HCC in populations
with a high prevalence of HBV26. Antiviral therapy for HBV with nucleotide and nucleoside
analogues reduces, but does not eliminate, the risk of HCC in treated cohorts of patients27. Likewise, anti-viral therapy with interferon might reduce the risk of HCC in patients with HCV
infection that is associated chronic liver disease28; however, this effect will need to be re-
assessed with non-interferon based direct acting antiviral drugs that are now approved for
practice. Finally, statin use and coffee consumption are associated with a reduced risk of 5 HCC in population studies29. In the future, potent liver directed antifibrotic therapies might
also reduce the risk of developing HCC. HCC in population studies29. In the future, potent liver directed antifibrotic therapies might
also reduce the risk of developing HCC. [H3] Early molecular alterations during hepatocarcinogenesis. HCC development is a complex multistep process that usually occurs in the context of liver
cirrhosis and is related to the diversity of aetiologies of the underlying liver disease. The
natural history of HCC in cirrhosis follows a sequence of events starting with the successive
development of pre-cancerous cirrhotic nodules with low grade dysplasia, called low-grade
dysplastic nodules (LGDN). This is followed by high-grade dysplastic nodules (HGDN) that can
transform into early-stage HCC (stages 0 and A) and progress into more advanced HCC
(stages B and C). Malignant transformation into HCC can originate from various cell types
including mature hepatocyte and stem or progenitor cells30. Similar to other epithelial solid tumours, HCC is the result of the accumulation of somatic
genomic alterations in passenger and driver cancer genes. As previously defined31, a cancer
driver would be a cell-autonomous or non–cell-autonomous alteration that contributes to
tumour evolution at any stage — including initiation, progression, metastasis and resistance
to therapy — by promoting a variety of functions including proliferation, survival, invasion or
immune evasion. In each HCC nodule, a mean number of 40 functional somatic alterations
are accumulated in coding regions and consequently each tumour is the result of a unique
combination of genetic alterations mixed together with epigenetic modifications32,33. This
general observation underlines the complexity of hepatocarcinogenesis and the huge
diversity of HCC (Table 2; Figure 3)33–38. However, the fact that genomic alterations are not
accumulated randomly suggests that several pathways can cooperate to promote
oncogenesis and that some of them can be related to specific risk factors33. Molecular markers identified to discriminate early-stage HCC from the pre-cancerous
nodules (LGDN and HGDN) have provided some clues about the early steps of carcinogenesis
that occurs in the context of cirrhosis39–43. On the basis of these markers, the mechanism of
hepatocyte malignant transformation has been shown to include Wnt/β-catenin pathway
activation, re-expression of fetal genes, deregulation of protein folding machinery and the
response to oxidative stress. Moreover, several lines of evidence have shown that telomere
maintenance and the telomerase complex that controls the nucleotide TTAGGG, a repeated
sequence at the end of chromosomes, play a major part in initiation and promotion of HCC
in cirrhosis44,45. First, mice deficient of telomerase RNA component (Terc), which codes for 6 the catalytic unit of the telomerase complex, showed short telomeres and developed
cirrhosis followed by HCC46. [H3] Early molecular alterations during hepatocarcinogenesis. Second, constitutive inactivating mutations in TERT, which
encodes the telomerase reverse transcriptase, are associated with an increased risk of
cirrhosis in humans47–49. However, progression to HCC involves a second step, with
telomerase reactivation required to promote liver carcinogenesis and to allow uncontrolled
hepatocyte proliferation (the ‘telomerase switch’)46,50. In humans, TERT is not expressed in
normal hepatocytes, but becomes re-expressed early during hepatocarcinogenesis in LGDN
and particularly HGDN42,51. In these lesions, telomerase re-expression is related to the
occurrence of point mutations at two hotspots in the TERT promoter. These alterations are
the most frequent recurrent somatic mutations identified in LGDN (6%), HGDN (20 %) and
HCC (60%)(Figure 3)33–38. Together, these results suggest that TERT promoter mutations are
oncogenic and, in most cases of HCC (>90%), telomerase activation is selected during
malignant transformation and tumour progression in cirrhotic and non-cirrhotic livers34,37. Within this paradigm, TERT promoter activation is required at an early step of
transformation in order to bypass the replicative senescence of cirrhotic hepatocytes. By
contrast, acquisition of genomic diversity appears to be a late event in liver carcinogenesis14. Viral infections by HCV and HBV are also associated with early molecular alterations involved
in malignant transformation through the induction of chronic inflammation, the expression
of viral proteins and the viral life cycle52,53 (Figure 3)33–38. Specific molecular alterations are
frequently related to HBV infection that can induce mutagenesis by insertion of viral DNA
into major driver genes of hepatocarcinogenesis54. The most frequent insertions of HBV DNA
in hepatocytes occur within the TERT promoter and activate telomerase and other
oncogenes including lysine (K)-specific methyltransferase 2B (KMT2B, also called MLL4),
cyclin E1 (CCNE1) and SUMO1/sentrin specific peptidase 5 (SENP5)55. In cooperation with
HBV infection, exposure to aflatoxin B1 , which is common in subtropical regions, induces
DNA adducts and the occurrence of frequent mutations, in particular within TP53, that have
a specific nucleotide signature22,33. Recently, deep sequencing analysis revealed another
DNA virus — AAV2 — that is related to a frequent harmless infection in the general
population and causes insertional mutagenesis in rare cases of HCC13. AAV2 insertions were
mainly identified in HCC that had developed in normal liver tissues, without other classical
risk factors. Viral insertions are found within normal and cancerous cells, and in the tumour
genome normally occur within typical oncogenes such as TERT, cyclin A2 (CCNA2), CCNE1,
tumour necrosis factor superfamily member 10 (TNFSF10) and KMT2B. [H3] Cancer drivers in progressed HCC. 7 7 Several pathways and processes have been implicated in HCC progression (Table 2). First,
telomere maintenance contributes to the evasion of cellular senescence. As previously
mentioned, telomerase is overexpressed in 90% of HCC and this is overexpression is related
to TERT promoter mutations in 60% of cases and to gene amplification in 5% of the
cases33,34,38. The two hotspots of mutations are located at nucleotide positions -124 and -146
upstream of the ATG. Mutations at both sites can create a new binding site recognized by a
transcription factor inducing TERT mRNA expression. Second, the Wnt/β-catenin pathway is frequently activated in HCC through CTNNB1
mutations that activate β-catenin (11-37% HCC cases), particularly in patients without HBV
infection and well-differentiated tumours56,57. Inactivating mutations or deletions are also
frequently identified in axin 1 (AXIN1, 10% of HCC cases) or more rarely in adenomatous
polyposis coli (APC, 1-2% of HCCs) and zinc and ring finger 3 (ZNRF3, 3% of HCCs). All of these
mutations result in activation of the Wnt/β-catenin pathway33,38. Third, inactivation of p53 and alterations of cell cycle are major defects in HCC, particularly in
cases related to HBV infection. In this context, TP53 mutations are the most frequent
alterations with a specific hotspot of mutation (R249S) in patients with aflatoxin B1
exposure22,33,38,58. Inactivation of the retinoblastoma pathway is also frequently observed Third, inactivation of p53 and alterations of cell cycle are major defects in HCC, particularly in
cases related to HBV infection. In this context, TP53 mutations are the most frequent
alterations with a specific hotspot of mutation (R249S) in patients with aflatoxin B1
exposure22,33,38,58. Inactivation of the retinoblastoma pathway is also frequently observed
through retinoblastoma 1 (RB1) mutations (3-8% of HCCs) or cyclin-dependent kinase
inhibitor 2A deletions (CDKN2A, 2-12% of HCCs). Most of these molecular defects are
associated with poor prognosis and could contribute to a more-aggressive phenotype33,35,59. Fourth, chromatin remodelling complexes and epigenetic regulators are frequently altered in
HCC. These alterations include mutations in the BRG1- or HRBM-associated factors (BAF) and
polybromo-associated BAF (PBAF) chromatin complex (ARID1A mutations in 4-17% of cases
and ARID2 mutation in 3-18% of cases) or in the histone methylation writer family (KMT2 –
MLL genes mutated in 2-4% of cases), which can also be modified by HBV insertions in
KMT2B (MLL4) (10% of cases)33,38,55. Recently, H3K9 modifier histone-lysine N-
methyltransferase SETDB1 was identified as over-expressed in HCC60,61. [H3] Cancer drivers in progressed HCC. Interestingly, these observations
suggest that NFE2L2 can protect from HCC occurrence during the development of chronic
liver disease but its constitutive activation can also contribute to tumour progression at a
late stage of tumour progression32,64. DNA amplifications are also associated with HCC. The most common high-level
amplifications in HCC are in chromosome regions 11q13 and 6p21 (5–10% of cases)32,35,65,66. Cyclin D1 (CCND1) and FGF19 are bona fide oncogenes in 11q13 and represent potential
therapeutic targets66. Similarly, high-level gains of 6p21 that contain more than four copies
of vascular endothelial growth factor A (VEGFA) have been identified in 4–8% of HCC cases65. VEGFA amplification induces both neoangiogenesis and tumour proliferation resulting from
the induction of macrophage-mediated hepatocyte growth factor secretion67. Whether
these amplifications represent targetable oncogenic addiction loops in HCC remains to be
elucidated by the testing of selective molecules in clinical trials. DNA amplifications are also associated with HCC. The most common high-level
amplifications in HCC are in chromosome regions 11q13 and 6p21 (5–10% of cases)32,35,65,66. Cyclin D1 (CCND1) and FGF19 are bona fide oncogenes in 11q13 and represent potential
therapeutic targets66. Similarly, high-level gains of 6p21 that contain more than four copies
of vascular endothelial growth factor A (VEGFA) have been identified in 4–8% of HCC cases65. VEGFA amplification induces both neoangiogenesis and tumour proliferation resulting from
the induction of macrophage-mediated hepatocyte growth factor secretion67. Whether
these amplifications represent targetable oncogenic addiction loops in HCC remains to be
elucidated by the testing of selective molecules in clinical trials. Modelling of HCC development in mice is a very useful approach for achieving better
understanding of the mechanisms of hepatocarcinogenesis and for testing for new
therapies. For drug screening, HCC cell lines or primary tumours xenografted in immuno-
deficient mice are easy to test for drug response. In more sophisticated models, candidate
oncogenes or tumour suppressor genes can be genetically modified in classical transgenic or
knockout animal models. More recently, siRNA68 and CRISPR/Cas969 techniques have
emerged as powerful methods to generate tumours in mice and to subsequently test for
drug response. [H3] Cancer drivers in progressed HCC. SETDB1 over-
expression promotes cancer cell growth via p53 methylation and is associated with tumour
aggressiveness and poor prognosis. Interestingly, DNA methylation is globally altered in HCC
and aberrant modifications are associated with prognosis62 or HBV infection63. Fourth, chromatin remodelling complexes and epigenetic regulators are frequently altered in
HCC. These alterations include mutations in the BRG1- or HRBM-associated factors (BAF) and
polybromo-associated BAF (PBAF) chromatin complex (ARID1A mutations in 4-17% of cases
and ARID2 mutation in 3-18% of cases) or in the histone methylation writer family (KMT2 –
MLL genes mutated in 2-4% of cases), which can also be modified by HBV insertions in
KMT2B (MLL4) (10% of cases)33,38,55. Recently, H3K9 modifier histone-lysine N-
methyltransferase SETDB1 was identified as over-expressed in HCC60,61. SETDB1 over-
expression promotes cancer cell growth via p53 methylation and is associated with tumour
aggressiveness and poor prognosis. Interestingly, DNA methylation is globally altered in HCC
and aberrant modifications are associated with prognosis62 or HBV infection63. Fifth, the Ras/Raf/MAP and the PI3K/AKT-mTOR pathways are frequently activated in HCC. These changes are caused by amplification a region that includes fibroblast growth factor 3
(FGF3), FGF4 and FGF19 in approximately 5% of tumours, and can also be related to
inactivating mutations in tuberous sclerosis 1 (TSC1) or TSC2 (3-8% of cases), or in
phosphatase and tensin homolog (PTEN) (1-3% of cases). Ribosomal protein S6 kinase,
90kDa, polypeptide 3 (RPS6KA3) mutations that cause inactivation of ribosomal protein S6 Fifth, the Ras/Raf/MAP and the PI3K/AKT-mTOR pathways are frequently activated in HCC. These changes are caused by amplification a region that includes fibroblast growth factor 3
(FGF3), FGF4 and FGF19 in approximately 5% of tumours, and can also be related to
inactivating mutations in tuberous sclerosis 1 (TSC1) or TSC2 (3-8% of cases), or in
phosphatase and tensin homolog (PTEN) (1-3% of cases). Ribosomal protein S6 kinase,
90kDa, polypeptide 3 (RPS6KA3) mutations that cause inactivation of ribosomal protein S6 8 kinase alpha-3 (also called RSK2) (5-9% cases) lead to an activation of ras/MAPK
signalling32,33, whereas mutations that activate Ras proteins themselves (KRAS, HRAS, NRAS
of BRAF) are rarely identified (<1% of cases). However, additional mechanisms of pathway
activation remain to be identified. Finally, the oxidative stress pathway is constitutively activated in HCC owing to mutations
that activate nuclear factor erythroid 2-related factor 2 (NFE2L2) or that inactivate kelch-like
ECH-associated protein 1 (KEAP1) in 5-15% of HCC cases. [H2] Molecular classes Genomic studies have revealed molecular subclasses of HCC (reviewed in14)65,70–73. Two main
molecular classes, each representing approximately 50% of patients, have been identified:
proliferative and non-proliferative HCC65,71,73. The proliferative subclass is enriched by
activation of Ras, mTOR, and insulin-like growth factor (IGF) signalling and FGF19
amplification, and is associated with HBV-related aetiologies and poor outcomes65,71,72,74. Some authors have proposed that there are two subtypes of the proliferative class: the
Wnt/transforming growth factor β (TGF-β) group and the progenitor-cell group. The 9 9 progenitor cell group is enriched in progenitor cell markers, such as epithelial cell adhesion
molecule (EpCAM), and the overexpression of α-fetoprotein65,71,72,74,75. By contrast, the non-
proliferative subclass is more heterogeneous, but there is still a clear subtype characterized
by CTNNB1 mutations that are associated with alcohol-related and HCV-related HCCs75. Direct translation of molecular HCC subclasses into clinical management is yet to be
achieved. [H2] Role of the microenvironment
[H3] Chronic inflammation. An altered microenvironment is now perceived to be a key enabling characteristic of cancer
and is known to participate in all stages of malignant progression, from the initial
transformation phases, through invasion and all the way to metastasis76,77. Pathologists have
long recognized that some tumours are densely infiltrated by cells of both the innate and
adaptive arms of the immune system and thereby mirror inflammatory conditions arising in
non-neoplastic tissues76,78. HCC is a prototypical inflammation associated cancer, with
approximately 90% of HCC burden associated with prolonged hepatitis due to viral hepatitis,
excessive alcohol intake or NAFLD-NASH. This indicates that the immune microenvironment
plays pivotal parts in the pathogenesis of this disease79. Interestingly, in fully developed
HCCs, the presence of immune infiltrates is associated with a better prognosis, likely owing
to more effective anti- tumour immunity80. Remarkably, an unbiased screen of gene
expression patterns in HCC revealed that activation of the key innate inflammatory signalling
mediators nuclear factor of κ light polypeptide gene enhancer in B-cells (NF-κB), epidermal
growth factor (EGF) and interleukin-6 (IL6) in the liver parenchyma, but not in tumour cells,
is associated with poor prognosis81. Furthermore, as noted above, genomic screens of HCC
also did not reveal mutations that activate inflammatory signalling pathways. Thus, it seems
that the inflamed liver promotes the seminal HCC cells at early phases of development82. Multiple cell types interact with hepatocytes in the chronically inflamed liver. These include
among others macrophages, stellate cells, endothelial cells and lymphocytes; all of which
were shown in certain mouse models to favour tumour growth. Importantly, although the
liver is envisaged as a metabolic organ, it maintains a uniquely tolerant immune system,
which his necessary to prevent the induction of immunity against multiple antigens and
immunostimulatory molecules, such as gut-derived nutrients and microbiota derived signals
which constantly flood the liver via the portal system. Understanding this unique hepatic
immune system is likely important in the context of the complex interaction between
malignant hepatocytes and the liver immune system83,84. 10 The mechanisms through which immune cells promote growth of early-stage HCC are
beginning to be elucidated. Multiple experimental models have substantiated that secretion
of various cytokines by immune cells can change function of the interacting hepatocyte,
rendering it less sensitive to intracellular tumour suppressor pathways (and possibly also to
extracellular ones such as anti-tumour adaptive immune responses). [H2] Role of the microenvironment
[H3] Chronic inflammation. An example is secretion
of TNF by macrophages in the chronic inflammatory hepatic infiltrate, which activates the
NF-κB pathway in hepatocytes, rendering the latter less sensitive to apoptosis and thus
promoting carcinogenesis85.Two important inflammatory signalling pathways that are
activated by inflammatory cytokines in chronically inflamed livers and promote HCC are the
NF-κB and JAK-STAT pathways86. Similarly, multiple molecules that are secreted by
microenvironmental constituents have been shown to promote hepatocyte growth; these
include TNF, lymphotoxin-α, lymphotoxin-β86, IL-687 and hepatocyte growth factor (HGF)88. The ability of inflammatory cells to produce potentially mutagenic reactive oxygen species
(ROS) and reactive nitrogen species (RNS) is considered by many to underlie some of their
pro- tumourigenic activity89; yet this must still be substantiated by rigorous testing in animal
models. However, it is clear that the increased hepatocyte proliferation that occurs in
chronic hepatitis can potentiate DNA damage induced mutations90. [H3] Fibrosis. Hepatic stellate cells normally reside in the liver sinusoids and have multiple roles in hepatic
homeostasis including retinoid storage, immunomodulation, liver regeneration and
vasoregulation. Importantly, upon liver injury they are the primary cellular mediators of
hepatic fibrosis and cirrhosis, which are strongly associated with HCC79. The presence of a
stellate cell gene expression signature is a poor prognosis indicator in human HCC91 and
overexpression of platelet-derived growth factor C (PDGFC) in the mouse liver induces
stellate cell activation and hepatic fibrosis followed by HCC92. Taken together, these data
indicate that hepatic fibrosis and cirrhosis have a causative role in HCC. The mechanisms
through which activated stellate cells could drive HCC progression include cytokine
secretion, angiogenesis promotion, protumorigenic extracellular matrix components,
increased tissue stiffness and immunosuppression; yet the human relevance of each of these
mechanisms and its relative contribution remain to be elucidated93. In summary, chronic
liver inflammation generates a pathologic microenvironment, infiltrated by adaptive and
innate immune cells and stellate cells, together producing a pathological milieu composed of
collagen, multiple extracellular matrix proteins, growth factors and cytokines which can form
a protumorigenic stroma. [H3] Vaccination. The most effective approach for preventing HCC is prevention of the underlying liver disease,
the best example of which is hepatitis B vaccination. Universal vaccination against hepatitis B
would be expected to reduce HCC incidence, and this has indeed been demonstrated26,95. The first evidence of the ability of the HBV vaccine to reduce the incidence of HCC came
from Taiwan after the introduction of universal neonatal vaccination in 1984. This
programme was associated with a reduction in the incidence of childhood HCC from 0.7 per
100,000 individuals to 0.36 per 100,000 individuals between 1981 and 1994 (P<0.01)95. As
the vaccinated cohort aged, the decrease in HCC incidence was carried through to
adolescents and is now being detected in young adults26. However, there are currently approximately 400 million adults who are infected with
hepatitis B. Not all of those with hepatitis B are at equal risk of developing HCC. Various risk
factors are well known, including hepatitis B viral load, male sex, older age and active liver
disease96–103. For those who have active disease and who receive treatment for hepatitis B,
the risk of HCC is reduced, although not eliminated104. This has been demonstrated in a
single RCT using the antiviral medication lamivudine105, a study that unfortunately did not
have HCC incidence as an endpoint. However, similar results have now been obtained in
cohort studies using various methods to match treated and untreated populations, including
historical controls106. The most effective approach for preventing HCC is prevention of the underlying liver disease,
the best example of which is hepatitis B vaccination. Universal vaccination against hepatitis B
would be expected to reduce HCC incidence, and this has indeed been demonstrated26,95. The first evidence of the ability of the HBV vaccine to reduce the incidence of HCC came
from Taiwan after the introduction of universal neonatal vaccination in 1984. This
programme was associated with a reduction in the incidence of childhood HCC from 0.7 per
100,000 individuals to 0.36 per 100,000 individuals between 1981 and 1994 (P<0.01)95. As
the vaccinated cohort aged, the decrease in HCC incidence was carried through to
adolescents and is now being detected in young adults26. However, there are currently approximately 400 million adults who are infected with
hepatitis B. Not all of those with hepatitis B are at equal risk of developing HCC. [H3] Targeting the microenvironment in HCC. 11 11 Because the tumour microenvironment plays a pivotal part in the natural history of HCC,
there is a strong rationale for modulating the dynamic cross talk between hepatocytes and
the stroma as treatment for this disease79, in particular in HCC prevention. Reversing liver
fibrosis is already feasible in patients with chronic liver disease using antifibrotic therapies94. Conceivably, targeting various immune cells is also achievable, as is the targeting of specific
molecules which are present in the pro tumourigenic hepatic microenvironment. Despite
these possibilities, improved knowledge of the relative contribution of the different cells and
molecules composing the chronically inflamed hepatic network to HCC pathogenesis, and
elucidation of the network's hierarchy, is needed for planning potent preventive or
therapeutic schemes. Such developments will entail better understanding of the complex
natures of the interactions that take place in chronically inflamed human livers, in concert
with meticulous testing of their functional relevance in animal models recapitulating the
different aetiologies of hepatic inflammation (viral, metabolic and other). The ideal
preventive treatment should convert a pro-tumourigenic microenvironment into an anti-
tumourigenic one. [H3] Lifestyle modification. Whether abstinence from alcohol in those with alcoholic liver disease decreases the
incidence of HCC is not yet known. Of course, primary prevention by counselling against
unhealthy use would reduce the incidence of alcohol associated cirrhosis and the attendant
risk of HCC. Similarly, for those with NAFLD who successfully lose weight or otherwise
control their disease, there is no information as to whether HCC incidence is reduced113. Thus, in summary, prevention of HCC in large part depends on prevention or treatment of
the underlying liver disease. [H3] Vaccination. Various risk
factors are well known, including hepatitis B viral load, male sex, older age and active liver
disease96–103. For those who have active disease and who receive treatment for hepatitis B,
the risk of HCC is reduced, although not eliminated104. This has been demonstrated in a
single RCT using the antiviral medication lamivudine105, a study that unfortunately did not
have HCC incidence as an endpoint. However, similar results have now been obtained in
cohort studies using various methods to match treated and untreated populations, including
historical controls106. 12 Similarly, treatment and eradication of hepatitis C results in a decreased HCC incidence21. This has only been documented in cohort studies. There are no RCTs comparing HCC
incidence in treated patients compared with untreated patients. Such a study would not be
ethical. However, cohort studies have documented that the HCC incidence in those
successfully treated for hepatitis C is lower than in historical untreated controls107–110. Until
recently, standard treatment of hepatitis C was based on interferon, and sustained virologic
responses in patients with hepatitis C cirrhosis have been associated with a substantial
reduced risk in HCC development (hazard ratio (HR) 0.23, confidence interval (CI) 0.16-0.35)111. It is too early to ascertain the effect of new direct acting antiviral (DAA) drugs for HCV – that
achieve sustained virological responses in 90% of cases- on the occurrence of HCC112. If one
extrapolates from the interferon era to the current DAA therapy setting, as a greater
proportion of individuals who are infected with HCV are treated, the overall hepatitis
burden, and the absolute number of patients with hepatitis who become at risk for HCC, will
decrease. However, the overall effect of DAA therapy will be modest because most
individuals with HCV remain undiagnosed, and treatment penetrance is limited owing to
socioeconomic factors. Assuming that one-third of patients with HCV come to medical
attention and achieve 90% sustained virologic responses with DAA agents, the burden of
HCC in the United Stated might be reduced by 10-15% over the next decade113,114. [H3] Chemoprevention. There are several putative chemopreventative agents that have been proposed to reduce
HCC incidence in at-risk populations, including statins and metformin115. The data on
metformin are becoming more robust, as demonstrated by a meta-analysis that suggested
metformin decreases the risk of HCC in people with diabetes116. Moreover, a study has
shown that metformin may be safely continued in people with diabetes who also have
cirrhosis117. Nonetheless, the evidence for the ability of metformin to reduce HCC risk is not
strong enough to recommend using it in at-risk patients, and prospective studies testing this
agent are required. [H2] Surveillance for HCC 13 When HCC causes symptoms, the disease is most often in an advanced stage, and not
amenable to potentially curative treatment. Death usually ensues within a few months. However, HCC has a prolonged sub-clinical course that provides the opportunity for early
detection. Early-stage HCC lesions are small and frequently curable, often by minimally
invasive methods. These considerations have led to the development of protocols for
surveillance for HCC in patients at risk for this cancer (Table 1). Surveillance, however,
remains controversial since the only RCT that demonstrated decreased mortality was
probably statistically incorrectly analyzed118. Nonetheless, there is a wealth of evidence of
lesser quality, including cohort studies with comparisons between screened and unscreened
populations, and demonstration that cure is more likely with HCC that is detected by
surveillance compared with cases that are diagnosed once symptoms have become
apparent119–124. Several cost-efficacy analyses have also demonstrated benefits of
surveillance125–137. Furthermore, in considering the potential benefits from surveillance
compared with the potential harms it is clear that benefits outweigh harms. Benefits include
early detection and potential cure of HCC. Harms include some additional investigations and
potentially some unnecessary interventions. However, as discussed below, a diagnostic
algorithm has been developed that was designed to minimize false-positives, and thereby
reduce the likelihood of harm. More recently, there have been a number of risk scores
developed to improve identification of the at-risk population. None have yet received
widespread application. The techniques for surveillance are also controversial. There is no question that ultrasound
should be part of the algorithm, but the use of biomarkers remains controversial. There is
some suggestion that biomarkers improve early detection138–144, but there is as yet no
evidence that this leads to improved cure rates compared with ultrasound alone. Most
importantly, the usual serum biomarkers that are used — α-fetoprotein (AFP), des-γ
carboxyprothrombin (DCP) and the L3 fraction of AFP (AFP-L3) — are all more frequently
associated with advanced-stage disease than early-stage disease and would therefore be
theoretically unsuitable for detection of early disease138–144. Among these, AFP is the most
widely used and this still remains an area of controversy. The ideal surveillance interval for at-risk patients (Table 1) according to guidelines is 6
months3,23,145–148. Studies have demonstrated that a 3-month interval does not enhance
outcomes146, and survival is lower with 12 month than with 6 month intervals145. [H2] Diagnosis Patients enrolled in a surveillance programme are diagnosed by identification of a new liver
nodule on abdominal ultrasound, and diagnostic confirmation using either non-invasive 14 criteria or biopsy. These patients are generally asymptomatic and have early stage HCC. Conversely, patients diagnosed outside surveillance usually present at advanced stages with
large symptomatic tumours and/or portal vein invasion. Symptoms include malaise, weight
loss, anorexia, abdominal discomfort or signs related to advanced liver dysfunction. Diagnosis can be made by non-invasive (radiological) or invasive (biopsy) approaches. Radiological diagnosis is achieved with a high degree of confidence if the lesion is found in a
patient with cirrhosis. Using contrast imaging, the lesion shows the radiological hallmarks of
HCC, which are hypervascularity in the arterial phase of a contrast study (CT or MRI) and a
decreased signal compared with the rest of the liver in the venous and/or delayed phases of
the study149. When these typical features are observed the diagnosis is confirmed and a
biopsy is not necessary3,15,150,151. Latest generation CT and/or MRI following reported
protocols are recommended152. MRI with liver specific contrast agents might help in the
diagnosis of HCC, but the specificity of these agents is still suboptimal. The caveat of the non-invasive radiological criteria is that this algorithm only applies to those
who have an elevated risk of HCC. A biopsy is required for patients who do not have any
special risks for HCC, for the most part patients without cirrhosis. The recommended
algorithm for investigation of lesions in at-risk patients is as follows3,15: For nodules < 1 cm in
size, ultrasound follow-up at 3 months is recommended. For lesions > 1 cm, the radiological
hallmarks of HCC define diagnosis. If the radiology is not typical in at least one of two
imaging techniques (CT and MRI), a liver biopsy is recommended3,15. Of note, this accepted
practice puts the assessment of HCC at a disadvantage compared with most cancers since
tumour tissue for molecular studies would not be routinely obtained in clinical practice. Recent guidelines recommend obtaining tissue samples in the setting of all research studies
in HCC3. Diagnostic difficulty occurs mainly with early-stage HCC lesions, in which the radiological
appearance might be atypical, necessitating a biopsy. However, biopsy is not an ideal gold
standard, because of variation introduced by sampling and complications. The risk of
complications, such as tumour seeding and bleeding, after liver biopsy is less than 3%153. [H2] Overview of evidence-based management and staging systems [H2] Overview of evidence-based management and staging systems The BCLC staging system provides an easy-to-use algorithm that links tumour stages with
treatment allocation policies based on evidence (Figure 4)3,158,159. Treatments are classified
as radical therapies with potential to cure HCC — such as surgical resection, liver
transplantation or percutaneous ablation — or palliative therapies which are aimed at
improving survival —chemoembolization and sorafenib3. Treatment allocation for standard
of care follows the levels of evidence defined by the National Cancer Institute, which rely on
strengths of study design and end points (Table 3). Controversy regarding HCC staging
systems remains. In fact, alternative staging or scoring systems have been proposed, such as
the Hong Kong classification160, the Cancer of the Liver Italian Program (CLIP) score161, the
TNM system162 and the Japan Integrated Staging (JIS) score163. None of these systems has
acquired global consensus. Scoring systems, such as the CLIP, are not widely use since they
do not incorporate treatment allocation to distinct prognostic stages, whereas others are
specifically applied in Asia (the Hong Kong system and the JIS). [H2] Diagnosis Although the sensitivity of liver biopsies ranges between 70-90% for all tumour sizes3, small
lesions might be missed, giving a false-negative result. In addition, in early-stage HCC,
morphological changes may be minimal compared with dysplastic hepatocytes, making the
diagnosis uncertain154. In this setting, the use of special stains may help to resolve diagnostic
uncertainties. For example, HCC cells expresses glypican 3, glutamine synthetase, heat shock
protein 70 and clathrin heavy chain155–157. Positive staining for two of these four markers is
highly specific for HCC. In addition, comparison of the tumour biopsy to biopsy of the
surrounding non-tumourous liver may be helpful in highlighting the early features of
malignancy. Patients with suspicious lesions in whom a diagnosis cannot be confirmed on a
biopsy should be closely monitored, and might even require a second biopsy. 15 [H3] Intermediate-stage disease. Patients with intermediate-stage HCC (BCLC B) are characterized by multinodular disease,
preserved liver function and the absence of tumour-related symptoms, vascular invasion and
extrahepatic spread. Chemoembolization in the form of transcatheter arterial
chemoembolization (TACE) — a minimally invasive technique that combines local delivery of
beads to restrict tumour blood supply with local administration of chemotherapy — is
recommended based on results from RCTs and one meta-analysis of pooled data17,18,176. Improvements in the selection of candidates, supra-selective embolization – selective
blockade of the artery feeding the tumour that minimizes collateral hepatic toxicity — and
improvements in embolization devices (drug-eluting beads) have led to current median
survival times of 26 months177, and beyond 40 months in referral centres178. Alternative
embolization strategies, such as radioembolization using beads coated with yttrium-90 (Y-
90) — an isotope that emits short-range β radiation — are effective and have a favourable
safety profile179, but only well-designed, properly powered RCTs will determine the
therapeutic niche of this intervention. [H3] Early-stage disease. Surgical resection is the first-line option for patients with early-stage HCC (BCLC 0 or A) with
solitary tumours, and confers 5-year survival rates of 70%3,15,164,165. Introduction of
restrictive selection of candidates for resection — patients with single nodules, absence of
portal hypertension and well-preserved liver function — along with anatomic resection — in
which tissue removal is performed in-line with the location of functional segments of the
liver — has minimized complications and reduced recurrence3,15,164,166. Metastases and de
novo HCC account for 70% 5-year recurrence after resection 167, and no adjuvant therapies
are able to prevent this complication168. Adoptive immunotherapy169 and the vitamin A
derivatives acyclic retinoids170 were reported as effective treatments in this setting;
however, these trials were conducted in small study populations and the results have not
been reproduced. Liver transplantation is the best first-line option for patients with tumours within Milan
criteria (defined as a single tumour <5 cm in size or three nodules <3 cm in size without
vascular invasion) (BCLC A) that are unsuitable for resection3,171. These criteria have been
independently validated internationally and have been adopted by guidelines3,15,165 and Liver
Transplant Units. Nonetheless, some studies have pointed to the fact that moderate
expansion of Milan criteria might lead to acceptable and competitive long-term
outcomes172–174. Local ablation using radiofrequency has been proposed as an optimal Liver transplantation is the best first-line option for patients with tumours within Milan
criteria (defined as a single tumour <5 cm in size or three nodules <3 cm in size without
vascular invasion) (BCLC A) that are unsuitable for resection3,171. These criteria have been
independently validated internationally and have been adopted by guidelines3,15,165 and Liver
Transplant Units. Nonetheless, some studies have pointed to the fact that moderate
expansion of Milan criteria might lead to acceptable and competitive long-term
outcomes172–174. Local ablation using radiofrequency has been proposed as an optimal 16 alternative to resection in patients with a single tumour of <2 cm in size who are unsuitable
for transplantation16,175, but no RCTs specifically addressing this issue have been reported. On the other hand, in patients with early tumours (BCLC 0 or A) who are not suitable for
surgery, local ablation is the standard of care with 5-year survival rates of 50–70%3. [H3] Advanced and end-stage disease. The Sorafenib HCC Assessment Randomized Protocol (SHARP) trial demonstrated that
sorafenib, a multiple tyrosine kinase inhibitor, was able to substantially increase survival in
patients with advanced-stage HCC (BCLC C) from 7.9 months to 10.7 months (HR 0.69)19. The
beneficial effects of sorafenib occur regardless of HCC aetiology, as was validated in Asian
patients who were infected with HBV180. On the basis of these data, sorafenib became the
standard of care in this setting. Eight RCTs testing alternative systemic molecular treatments
reported since the SHARP trial have not shown survival benefits in HCC. These findings can
be partially explained by the specific toxicity of some agents owing to the underlying liver
cirrhosis, the high molecular heterogeneity of HCC and a relative resistance to conventional
chemotherapy181. Patients with end-stage disease (BCLC D) should be considered for
nutritional and psychological support and proper management of pain, but are not
candidates for entering clinical trials. [H2] Surgical therapies 17 17 Surgery is the treatment of choice for most patients with early-stage HCC, with 5-year
survival in appropriately selected cases exceeding 70%; its role in more-advanced HCC is
more controversial3,165. The large majority of patients with HCC have underlying liver
disease, and both the location and extent of the tumour and the status of the non-malignant
tumour liver tissue must be considered in the choice of surgical procedure. The prognosis of
chronic liver disease is commonly assessed using the Child-Pugh score, which uses five
clinical measures — total bilirubin, serum albumin, prothrombin time, ascites severity and
hepatic encephalopathy grade — to classify patients into one of three groups (A-C) of
predicted survival rates. In brief, Child-Pugh’s A reflects well-preserved liver function, Child’s
B moderate liver dysfunction with a median life expectancy of approximately 3 years and
Child C severe liver dysfunction with life expectancy of approximately 1 year182. Surgery is the treatment of choice for most patients with early-stage HCC, with 5-year
survival in appropriately selected cases exceeding 70%; its role in more-advanced HCC is
more controversial3,165. The large majority of patients with HCC have underlying liver
disease, and both the location and extent of the tumour and the status of the non-malignant
tumour liver tissue must be considered in the choice of surgical procedure. The prognosis of
chronic liver disease is commonly assessed using the Child-Pugh score, which uses five
clinical measures — total bilirubin, serum albumin, prothrombin time, ascites severity and
hepatic encephalopathy grade — to classify patients into one of three groups (A-C) of
predicted survival rates. In brief, Child-Pugh’s A reflects well-preserved liver function, Child’s
B moderate liver dysfunction with a median life expectancy of approximately 3 years and
Child C severe liver dysfunction with life expectancy of approximately 1 year182. [H3] Transplantation. Liver transplantation is indicated for patients with early-stage HCC who fall within Milan
criteria and who are not candidates for partial hepatectomy3,15,171. For these patients,
transplantation yields a 5-year survival rate of 70% with a recurrence rate of approximately
10%, and 10-year survival rates > 50%192(Table 3; Figure 4)3,158,159. Transplantation offers the
appeal of removing unrecognized intrahepatic metastases and essentially eliminating the
risk of de novo tumour development, but this benefit is offset by higher perioperative and
late non- tumour-related mortality such that post-transplant survival, at least to 5 years, is
not substantially better than after resection188. Donor organ scarcity varies geographically. To the extent that patients must wait for organs, drop-out from the waiting list owing to
tumour progression reduces transplant survival on an intention-to-treat basis166, and loco-
regional treatment to prevent progression while waiting is an integral part of the process193 . Living donor transplantation is a way for patients with suitable donors to avoid the risk of
waiting list drop-out; results are comparable to those achieved with deceased donors194. The
mTOR inhibitors sirolimus or everolimus are often added to the immunosuppressive
regimen195. However, their effectiveness as anti-tumour therapies to prevent recurrence-
free-survival after transplantation failed in a recent RCT196,197. The Milan criteria are widely used as the basis for transplant eligibility, and adherence to
them yields good post-transplant survival. Increasingly, however, the benefit provided by
transplant compared with the alternatives (for example, resection and loco-regional
therapies), as opposed to the absolute survival rate, is being considered in discussions about
donor organ allocation; basing allocation on benefit of transplant rather than on absolute
survival could lead to different choices of transplant candidates. For example, the benefit of
transplant for a patient with single HCC within Milan criteria as compared with resection
may be considerably less than the benefit of transplant for multinodular HCC beyond Milan
criteria as compared to chemoembolization, even though the absolute survival after
transplantation in the second scenario is lower198. Several cohorts have explored the
implication of extending Milan criteria for transplantation in HCC199. Extended criteria such
as the University of California San Francisco (UCSF)200 and Up-to-7172 proposals have been
reported to yield survival rates similar to the Milan criteria, as has down-staging of HCC
beyond the Milan criteria by loco-regional treatment201. Selection criteria for extended
indications based on genomic information have reported good outcomes175. [H3] Resection. Patients with a single technically resectable HCC without macrovascular invasion, with
preserved liver function (Child-Pugh class A with bilirubin < 1 mg per dl) and no portal
hypertension are optimal candidates for partial hepatectomy, and experience low (1-2%)
perioperative mortality15,164–166(Table 3; Figure 4)3,158,159. Limited, yet anatomical, resection
is preferred when possible in order to spare uninvolved liver parenchyma and to remove
satellite tumours that result from local vascular invasion183; when major (≥ 3 segment)
resection is required, preliminary portal embolization184 or lobar radioembolization185 is
performed in some centres to induce growth of the future liver remnant. Over a period of 4-
6 weeks following portal embolization, the volume of the contralateral liver lobe typically
increases by 20-25%, thereby decreasing the risk of postoperative hepatic insufficiency. These practical approaches have not yet been adopted by guidelines. With improving
technology and experience, laparoscopic resection is increasingly employed with improved
early outcomes186, and percutaneous thermal ablation has become an acceptable alternative
for accessible tumours < 2cm175. Resection is often applied outside of guidelines for patients
with multifocal HCC or portal hypertension, particularly in Asia where availability of
transplantation is limited, albeit with decreased 5-year survival (50-60% compared with >
70% in optimal candidates)187. Patients with a single technically resectable HCC without macrovascular invasion, with
preserved liver function (Child-Pugh class A with bilirubin < 1 mg per dl) and no portal
hypertension are optimal candidates for partial hepatectomy, and experience low (1-2%)
perioperative mortality15,164–166(Table 3; Figure 4)3,158,159. Limited, yet anatomical, resection Recurrence of HCC in the liver is common after resection (up to 70% at 5 years)165,188,
because the remaining liver is both the most common site of metastasis of the primary HCC,
and is at risk for developing de novo HCC. With careful follow-up, recurrence can often be
treated effectively by repeat resection189, thermal ablation or liver transplantation190 with
resultant long-term survival. There is no proven adjuvant therapy for HCC resection; small
trials have produced positive results for retinoids170, immunotherapy169 and I-131 lipiodol191, 18 but these have not been confirmed in larger studies, and a large RCT including more than
1000 patients of sorafenib in this setting showed no benefit168. but these have not been confirmed in larger studies, and a large RCT including more than
1000 patients of sorafenib in this setting showed no benefit168. [H3] Percutaneous ablation. Percutaneous and intraarterial therapies are usually performed by interventional radiologists
and are the mainstay of the treatment of patients with early-stage and intermediate-stage
HCC (BCLC 0-B) who are not candidates for surgery (Table 3). Percutaneous tumour ablation
involves the insertion of a needle through the skin to access the inside of a tumour. This
approach can be used to inject agents that induce tumour cell killing (usually absolute
ethanol) or to insert a probe that delivers energy that induces a deleterious increase in
temperature (radiofrequency, microwave or laser)203. Tumour size (≤ 3 cm), number (≤ 3
tumours) and location (accessible with ultrasound guidance) limit the applicability of
percutaneous ablation204. Several RCTs have demonstrated a significant benefit of radiofrequency ablation compared
with percutaneous ethanol injection in terms of complete response rate (absence of contrast
uptake within the treated lesion in the arterial phase of CT or MRI) and time to
recurrence205,206. As a result, radiofrequency is the standard ablative therapy at early stages
of the disease since it provides better results than ethanol205 (Figure 4)3,158,159. Ethanol
injection is nevertheless a valuable option for tumours located near the large hepatic vessels
and bile ducts, or in centers with limited access to technology. Five-year survival rates after
radiofrequency ablation averages 60%207. Although tumour progression or relapse is higher
after percutaneous ablation than after liver resection, the long-term outcomes are similar
and ablation has been proposed as first-line therapy for tumours that are < 2 cm in size16. [H3] Transplantation. The
reproducibility of these proposals on a large scale awaits confirmation. 19 Postsurgical outcomes are dependent on the nature of the underlying liver disease. Historically, results have been better in HBV-related HCC compared with HCV-related HCC188
because viral control has been possible for HBV, leading some authors to propose more
liberal surgical guidelines for patients with HBV-related HCC160.There is good reason to
expect that with the recent development of effective HCV treatment202, improved results of
both resection and transplant might lead to a broadening of the accepted indications for
surgical treatment in patients with hepatitis C. Postsurgical outcomes are dependent on the nature of the underlying liver disease. Historically, results have been better in HBV-related HCC compared with HCV-related HCC188
because viral control has been possible for HBV, leading some authors to propose more
liberal surgical guidelines for patients with HBV-related HCC160.There is good reason to
expect that with the recent development of effective HCV treatment202, improved results of
both resection and transplant might lead to a broadening of the accepted indications for
surgical treatment in patients with hepatitis C. [H3] Intraarterial therapy. The most commonly used intraarterial therapy is TACE, which involves the sequential
injection into one or more branches of the hepatic artery of chemotherapeutic drugs
(doxorubicin, mitomycin C, cisplatin or combinations) loaded to the particles or emulsified in
Lipiodol (an oily contrast agent that is selectively retained in the HCC nodules) and
embolizing particles that interrupt blood flow208,209. The result is the induction of acute
ischaemic necrosis and eventually a prolonged exposure of tumour cells to the drugs. TACE is 20 nowadays usually performed ‘on demand’, with patients being evaluated every 6-8 weeks
with contrast-enhanced CT or MRI and additional selective TACE sessions performed only if
active tumour areas are found210. The use of the more costly drug-eluting particles results in
a simpler and more standardized procedure that increases tumour response rates but does
not improve survival211. Strong scientific evidence makes TACE the standard of care for patients with large or
multiple tumours (BCLC B) or those small tumours that cannot be resected or
percutaneously ablated (BCLC A unsuitable for surgery or local ablation), if the patient has a
preserved liver function, no cancer-related symptoms and no vascular invasion or
extrahepatic spread3. Two RCTs 17,176 and one meta-analysis of pooled data established a
significant benefit of TACE versus supportive care or suboptimal therapies (tamoxifen and
oral 5--fluorouracil (5-FU) in this patient population18 (Figure 4)3,158,159. Median survival after
TACE ranges from 16-45 months in the early stage (BCLC 0-A), 15.6-26.3 months in
intermediate stage (BCLC B) and 6.8-13.6 in the advanced stage (BCLC C)177,212. The largest
RCT ever reported for TACE describes a median survival duration of 26 months for patients
with intermediate-stage HCC177. Combination strategies with TACE and systemic therapies
(brivanib or sorafenib) have not resulted in clinical benefit177,213. A different meta-analysis
that included the combination of different loco-regional therapies (TACE or radiofrequency
ablation in the active arms) has questioned the benefits of TACE214. The use of TACE alone or
in combination with sorafenib for patients with advanced-stage HCC is not supported by
scientific evidence or recommended by guidelines3,23. Alternatives to TACE include Y90-radioembolization, which uses much smaller embolizing
particles that are injected into the hepatic arteries to provide selective internal irradiation of
the tumours179,215. Evidence of survival benefit for radioembolization has not yet been
proven in the setting of RCTs compared with the standard of care, which is TACE in
intermediate-stage HCC and sorfenib in advanced-stage HCC. [H3] Intraarterial therapy. This technique is usually
applied to those patients that are not good candidates for TACE because of large tumour
burden, vascular invasion or progression to prior TACE216. A second alternative to TACE is
bland transarterial embolization (TAE), which involves occluding the tumour blood supply
using microbeads without simultaneous administration of chemotherapeutic agents. TAE has
provided lower response rates compared with TACE in RCTs217, and a meta-analysis showed
that it provides suboptimal survival compared with TACE18. Finally, hepatic arterial infusion
chemotherapy — a catheter-based procedure for the long-term administration of agents
directly into the liver — is frequently used in Japan to treat patients who are poor candidates
for TACE, however there is no scientific evidence supporting this approach. [H2] Systemic therapies and future treatment approaches [H2] Systemic therapies and future treatment approaches 21 21 More than 100 RCTs have been reported testing chemotherapy or other types of systemic
therapies in HCC, but only one drug, sorafenib, has proven survival advantages19,180,218. Treatment with systemic chemotherapy and anti-oestrogen therapies has been shown to be
ineffective in HCC18. Systemic chemotherapy with doxorubicin, PIAF regime (platinum,
interferon, doxorubicin and 5-FU)181 and FOLFOX regime219 lacked survival advantages and
was accompanied in some instances with important toxicity. Treatment with sorafenib is
associated with manageable adverse events (diarrhoea and skin reactions on the hands and
feet) and an absolute increase in median survival of 3 months19. Subsequent studies
revealed a stable benefit of this drug in all regions of the world and in all HCC aetiologies220,
an in recent trials sorafenib consistently showed a median survival of approximately 10
months. Unfortunately, no predictive biomarkers of responsiveness to sorafenib have been
identified221. Although sorafenib is in a unique position of primacy in the management of
HCC, it has some restrictions in the target population (not indicated in patients with poor
liver function or in the adjuvant setting) and in the understanding of the mechanism of
action. The efficacy of sorafenib probably results from a balance between targeting cancer
cells and the microenvironment by blocking multiple kinases — up to 40, including vascular
endothelial growth factor receptor (VEGFR), platelet-derived growth factor receptor
(PDGFR), mast/stem cell growth factor receptor (c-Kit) and/or serine/threonine-protein
kinase B-raf (BRAF)222. Interestingly, this wide blockade does not induce liver damage or
other life-threatening complications. The main characteristics of the SHARP trial have been adopted by guidelines of trial design223
and replicated by almost all subsequent studies testing molecular therapies in HCC218. This
seminal study enrolled patients with well-preserved liver function (as indicated by Child-
Pugh A class), with advanced disease (BCLC C) or those with intermediate stage disease
(BCLC B) that progressed following TACE, and defined overall survival as the primary end
point (Table 5). Nonetheless, none of the therapies tested after the SHARP trial, including
brivanib (a fibroblast growth factor receptor (FGFR) and VEGFR inhibitor)224, sunitinib (a c-
Kit, VEGFR and PDGFR inhibitor)225, linifanib (a VEGFR and PDGFR inhibitor)226, erlotinib (an
epidermal growth factor receptor (EGFR) inhibitor)227 in the first line-setting and brivanib228,
everolimus (an mTOR inhibitor)229 and ramucirumab (a VEGFR2 inhibitor)230 in the second-
line setting have demonstrated survival benefits in patients with HCC. [H2] Systemic therapies and future treatment approaches The reasons for the
disappointing Phase III clinical trial results are reviewed elsewhere231 and include a marginal
anti-tumour potency, liver toxicity, flaws in trial design and lack of biomarker-based
enrichment. Randomized Phase II studies are recommended prior to conducting pivotal
Phase III trials. Phase II studies are essential for identifying signals of efficacy, futility or
toxicity, and might prevent the devotion of resources to Phase III trials that test therapies
with marginal chances of success. Overall survival is the mainstay end-point for Phase III
studies. Although time-to-progression remains relevant in Phase II studies223, recent results
pointing to a lack of correlation between TTP and OS225,226,228 suggest assessing 22 simultaneously other end-points such as objective response using the modified Response
Evaluation Criteria In Solid Tumours (mRECIST)231,232. Regarding biomarkers, none of the
Phase III studies were enriched by biomarker analysis, which could have indicated
effectiveness of the drug in selected subpopulations. Most of the drugs currently being tested in Phase III trials are antiangiogenic agents, cell
cycle inhibitors, receptor tyrosine kinase inhibitors and checkpoint inhibitors (Table 4; Figure
5). These broad spectrum compounds are tested in HCC in ‘all comers’, such as levantinib (a
VEGFR2 and VEGFR3 inhibitor) for first-line treatment and regorafenib (multikinase inhibitor,
including a VEGFR2 and angiopoietin-1 receptor (TIE2) inhibitor) or cabozantinib (a VEGFR
and hepatocyte growth factor receptor (MET) inhibitor). However, the fact that many agents
in Phase III trials failed with non-biomarker enriched populations is currently being
counteracted by a more precise approach. Primarily, Phase II proof-of-concept studies that
test drugs blocking potential oncogenic addiction loops and Phase II and III studies using
biomarker-based trial enrichment strategies to define activation of signalling pathways in
HCC subgroups are currently on-going. For instance, genomic studies defined a human HCC
subclass characterized by TGF-β signalling activation233 that is associated with an aggressive
phenotype74. The small molecule galunisertib blocks TGF-β signalling and has been tested in
Phase II as a single agent or in combination with sorafenib234. Similarly, overexpression
and/or amplification of FGF19 is characteristic of 10-20% of cases of HCC65,66. New
generation FGFR-specific kinase inhibitors have been tested in pre-clinical models
demonstrating that FGF19 amplification is a predictive biomarker of response235. More
recently, BLU9931, a highly specific FGFR-4 inhibitor, demonstrated pre-clinical activity in
patient derived xenografts, providing the rationale for exploring this compound in phase I or
II clinical trials236. [H2] Systemic therapies and future treatment approaches Finally, universal activation of the RAS/MAPK axis is common in patients
with advanced HCC237. First attempts to block this pathway with mitogen-activated protein
kinase kinase (MEK) inhibitors (selumetinib) have failed in detecting significant objective
radiological tumour responses238. The Assessing BAY86-9766 Plus Sorafenib for the
Treatment of Liver Cancer (BASIL) trial tested refametinib (a MEK1-2 inhibitor) in advanced
HCC, and results indicated that RAS mutations could be a potential biomarker of treatment
response 239. Alternatively, recent data has emerged pointing towards biomarker-driven selection of
candidates in testing drugs in Phase III trials. This has been the case for ramucirumab, which
is currently in Phase III for second-line treatment of patients with advanced HCC and AFP >
400 ng per mL, based on subgroups analysis data229. Similarly, tivantinib, a tubulin and MET
inhibitor240, is currently being tested in Phase III only in MET-positive patients in a second-
line setting. Activation of MET is estimated to occur in approximately 50% of these patients. 23 A similar drug, cabozantinib , is being tested as a second-line therapy in an all-comers after
preliminary positive effect in clinical studies241. Immune checkpoint inhibitors have been approved by regulatory agencies owing to their
considerable activity in patients with advanced-stage melanoma and lung cancer242. The
rationale for an immunological approach to treat HCC has been proposed for years. Indirect
evidence includes the pivotal role that the immune system has in the development of
chronic liver disease and HCC, dendritic cell-based approaches showing certain anti- tumour
activity243,244 and occasional reports of cases of spontaneous remission. Although pilot
studies with the cytotoxic T-lymphocyte protein 4 (CTLA-4) blocking antibody tremelimumab
did not produce signals of efficay245, a recent phase I-II trial with the programmed cell death
protein 1 (PD-1) inhibitor nivolumab showed a manageable adverse event profile and
produced durable responses in patients with HCC patients who had tumours that were
resistant to sorafenib246. Phase III trials have been designed to test this drug. Finally, two
alternative types of molecular-based therapies are currently being explored in early clinical
research studies: epigenetic modifying therapies247,248 and microRNAs249. Molecular targeted therapies are associated with acquired drug resistance. [H2] Systemic therapies and future treatment approaches The most
common traits of acquired resistance mechanisms are the persistent activation of the
oncogenic target itself owing to secondary mutations — for example, mutation of EGFR
(T790M) in gefitinib/erlotinib-resistant patients with non-small-cell lung cancer250 — or
acquired mutations in alternative drivers of the pathway — such as mutations in genes
encoding MAPKs in vemurafenib-resistant BRAF melanoma251. Acquired resistance to
sorafenib in HCC has mostly been explored in experimental models. Several mechanisms
have been implicated, including activation of MAPK14 signalling 68, enrichment of tumour-
initiating cells and re-activation of IGF/FGF signalling252. [H1] Quality of life As expected, QOL was much
better after radiofrequency ablation than after resection, and remained superior up to 36
months post-treatment. In addition, QOL following radioembolization has been compared
with TACE256. In this study, there was no overall difference in QOL between the two groups,
but the sample size was small. Despite the lack of statistically significant differences, in the
TACE group QOL was decreased at 2 and 4 weeks, whereas in the radioembolization group
some aspects of QOL actually improved. However, in this study the patients who underwent
radioembolization had more advanced disease than those who underwent TACE, so the
results are not directly comparable. Finally, the SHARP trial demonstrating survival benefits
of sorafenib compared to placebo also tested time-to-symptomatic progression — as
measured by the Functional Assessment of Cancer Therapy–Hepatobiliary Symptom Index 8
(FHSI8) — as co-primary end-point. The negative results of this end-point contrasted with
the survival benefits obtained by sorafenib, thus challenging the accuracy the tool used19. [H1] Quality of life Unsurprisingly, as liver disease and HCC tumour progresses quality of life (QOL) suffers. The
purpose of measuring QOL should be to compare outcomes between treatment arms, even
if one is a placebo. There is little agreement as to the best method of measuring QOL in HCC
research. Although general instruments are in use, there are two that are specific to HCC. These are the European Organisation for Research and Treatment of Cancer Quality of Live
Questionnaire (EORTC QLQ) 18253,254 and the Functional Assessment of Cancer Therapy-
Hepatobiliary (FACT-Hep)255 questionnaire. Both of these tools are derivatives of more
general QOL questionnaires for cancer and have been validated externally, but there is little
indication that they are measuring the same thing, or that they are comparable. 24 There is very little literature on QOL in HCC. Most studies report changes in QOL in single
arm studies, and very few studies have designed primarily to compare two different
treatments in patients with similar stage disease. For example, one study compared QOL
after resection with QOL following radiofrequency ablation255. As expected, QOL was much
better after radiofrequency ablation than after resection, and remained superior up to 36
months post-treatment. In addition, QOL following radioembolization has been compared
with TACE256. In this study, there was no overall difference in QOL between the two groups,
but the sample size was small. Despite the lack of statistically significant differences, in the
TACE group QOL was decreased at 2 and 4 weeks, whereas in the radioembolization group
some aspects of QOL actually improved. However, in this study the patients who underwent
radioembolization had more advanced disease than those who underwent TACE, so the
results are not directly comparable. Finally, the SHARP trial demonstrating survival benefits
of sorafenib compared to placebo also tested time-to-symptomatic progression — as
measured by the Functional Assessment of Cancer Therapy–Hepatobiliary Symptom Index 8
(FHSI8) — as co-primary end-point. The negative results of this end-point contrasted with
the survival benefits obtained by sorafenib, thus challenging the accuracy the tool used19. There is very little literature on QOL in HCC. Most studies report changes in QOL in single
arm studies, and very few studies have designed primarily to compare two different
treatments in patients with similar stage disease. For example, one study compared QOL
after resection with QOL following radiofrequency ablation255. [H3] Standard therapy. Only five treatments are recommended by the European Association for the Study of the
Liver (EASL) and the American Association for the Study of Liver Diseases (AASLD) guidelines
for HCC management (Table 3; Figure 4)3,158,159. Several other treatments have been tested
in RCTs since the last successful therapy (sorafenib) was approved, and all of these studies
produced negative or inconclusive results. In early HCC, no adjuvant therapy has shown
efficacy168 and this represents an important unmet medical need. In addition, since the
advent of radiofrequency as the mainstream of local ablation, no other proposed approach,
such as microwave, has led to a major control of the disease. Chembomebolization remains
the sole proven effective therapy in intermediate HCC. None of the combination therapies
with chemoembolization have shown additive outcome advantages. Phase III trials testing
alternative therapies, such as radioemebolization, are awaited. Indications beyond
guidelines for resection259 and TACE are widely applied in clinical practice, particularly in
Asia260, and this represents an important challenge that needs to be addressed with robust
and well-designed studies. [H1] Outlook
[H2] Global disease burden The global burden of HCC is increasing and considerable challenges are ahead for improving
the understanding and treatment of this complex disease. The main unmet medical needs
for HCC are summarized in Box 1. Considering that HBV and HCV infection are the main risk
factors for HCC development, it can be presumed that the implementation of new more
effective anti-viral therapies might decrease the incidence of HCC on a global scale in the
following decade. Antiviral therapies approved for HCV infection achieve sustained viral
clearance in more than 90% of cases202,257, and well-established anti-HBV therapies lead to
undetectable viral titres (circulating HBV-DNA) in most patients. Nonetheless, the fact that
the risk factors can be eliminated (efficacy), does not always translate into global
improvements (effectiveness) owing to suboptimal implementation of treatments in
underdeveloped areas and other complex reasons. Similarly, despite that fact that
surveillance is cost-effective in HCC, the global implementation of such programmes is still
suboptimal and is estimated to engage <50% of the target population in the West. Public
health policies encouraging the implementation of such programmes in well-defined
populations should lead to an increase in early tumour detection, and hence survival
benefits. Finally, in parallel with advances in the treatment of viral hepatitis, other
aetiologies of HCC are emerging, particularly NASH-related HCC which is associated with
obesity and diabetes. The effect of these unfolding risk factors on HCC burden remains to be
elucidated, but might counterbalance the decreases expected with HCV control. 25 [H2] Drug and biomarker discovery High-throughput genomic studies reporting gene sequencing of large cohorts have already
established the main oncogenic drivers of HCC. However, most of these drivers, such as the
TERT promoter, TP53 and CTNNB1, have not proven to be druggable and as such
understanding of their role in HCC has not translated into improving the management of the
disease. Drug discovery targeting these complex proteins and regulatory mechanisms should
represent a major breakthrough in HCC research258. On the other hand, the identification of
driver mutations or amplifications in relevant genes — such as FGF19, CCND1 and VEGF —
has not yet translated into proof-of-concept early clinical trials based on biomarkers. Understanding the targets of the microenvironment in tumour progression and response to
therapies has been an area clearly underexplored79,231, considering the clinical relevance of
this compartment in the risk of tumour development, prognosis and immunomodulation76–
79. In addition, despite results reporting preclinical testing of drugs in genetically modified
models or patient-derived xenografts, clinically relevant models recapitulating the spectrum
of human disease are still suboptimal. High-throughput genomic studies reporting gene sequencing of large cohorts have already
established the main oncogenic drivers of HCC. However, most of these drivers, such as the
TERT promoter, TP53 and CTNNB1, have not proven to be druggable and as such
understanding of their role in HCC has not translated into improving the management of the
disease. Drug discovery targeting these complex proteins and regulatory mechanisms should
represent a major breakthrough in HCC research258. On the other hand, the identification of
driver mutations or amplifications in relevant genes — such as FGF19, CCND1 and VEGF —
has not yet translated into proof-of-concept early clinical trials based on biomarkers. Understanding the targets of the microenvironment in tumour progression and response to
therapies has been an area clearly underexplored79,231, considering the clinical relevance of
this compartment in the risk of tumour development, prognosis and immunomodulation76–
79. In addition, despite results reporting preclinical testing of drugs in genetically modified
models or patient-derived xenografts, clinically relevant models recapitulating the spectrum
of human disease are still suboptimal. [H3] Progress in treating advanced disease. Management of advanced HCC has attracted major attention during the past decade. Trials
testing systemic chemotherapy have been almost abandoned and the few that were
conducted showed disappointing results219. Sorafenib approval represented a breakthrough, 26 challenged the concept of advanced HCC as a cancer that is not druggable19 and paved the
way for testing novel drugs. What has been unexpected is the failure of all molecular agents
tested afterward sorafenib224–230. As a consequence, this major cause of cancer-related
death remains as one of the few solid tumours with only one systemic therapy available. New avenues are currently being explored to overcome this situation. Among the drugs
currently being tested in Phase II and Phase III trials (Table 4; Figure 5), some are exploring
novel areas that led to major advancements in cancer management, such as immune
checkpoint inhibition261. Others are targeting specific subpopulations of HCC patients based
on biomarkers identified in phase I/II studies, such as tivantinib for MET-positive patients240,
or ramucirumab for those patients progressing on sorafenib with an AFP of >400 ng per
mL230. Finally, a few studies are exploring targeting specific tumour driver genes, which are
those responsible for ‘oncogenic addiction’. Oncogene addiction is defined as the state of
dependency of tumour cells on a given driver. Since solid tumours are expected to share 4-8
drivers per tumour, the addiction to one specific oncogene confers status of primacy to this
target262. Thus, clinically relevant anti-tumour responses are achieved when targeting those
drivers, as is the case for inhibiting the effects of antiplastic lymphoma receptor tyrosine
kinase (ALK) fusion rearrangement with crizotinib in patients with non-small-cell lung
cancer263. Trial design in HCC should follow specific recommendations that are provided in
Table 5. [H3] The future of precision medicine for HCC. HCC is one of the few cancers where non-invasive radiological criteria suffice for the
diagnosis of the disease in patients with an underlying risk factor (cirrhosis)3,15. This clinical
approach conflicts with the concept of precision medicine, which is based on the
administration of selective therapies targeting molecular alterations relevant to tumour
progression in a given individual. In order to implement precision medicine in HCC, a tumour
biopsy is required and, as such, routine biopsy has now been adopted by guidelines for
clinical trials in HCC research3. The reliability of this strategy is based on the assumption that
one tumour biopsy suffices for recapitulating the molecular information found in the whole
neoplasm. This concept is currently debatable in oncology owing to evidence of inter-
tumoural and intra-tumoural heterogeneity in all malignancies, including in HCC31,264. This
observation prompts an important question: can we rely upon a single biopsy for decision-
making or should we obtain multiple biopsies, even though this more thorough approach
appears clinically impractical? To explore heterogeneity we need to understand the concept of trunk, branch and
passenger mutations31,265. Trunk mutations occur at the onset of the disease and are potent
transforming drivers present in all cells of a given tumour at early stages31. Conversely, 27 27 branch mutations develop late in the natural history of the tumours or as result of acquired
resistance under the pressure of therapies and thus are only present in a subgroup of
tumour cells. Finally passenger mutations, the most common type of mutation, are of
marginal relevance in terms of cell transformation, progression or dissemination, but they
might be helpful in defining the clonality of tumours or their immunogenicity. Therefore, as
reported in solid cancers such as non-small-cell cancer266, breast cancer and melanoma,
molecular therapies that achieve good tumour response and survival benefits can do so by
targeting trunk mutations that have been identified with a single biopsy. Conversely,
heterogeneity might be cumbersome when branch mutations have a more dominant role,
for instance at very advanced stages266 or in the setting of acquired resistance265. In these
instances, liquid biopsy —checking either tumour DNA and mRNA in cell-free plasma or
circulating tumour cells — emerges as the most promising alternative for the molecular
monitoring of tumour progression and relapse267,268. Recent reports point to the benefits of
liquid biopsy in recapitulating tissue trunk driver mutations269 and in capturing unique
branch subclonal mutations acquired under treatment pressure270. [H3] The future of precision medicine for HCC. Finally, heterogeneity has
more important therapeutic implications for molecular agents than for immune check-point
inhibitors. 28 Figure 1. The global burden of hepatocellular carcinoma. Incidence of hepatocellular
carcinoma (HCC) according to data from Globocan 20121,3–9. ASR, Age-standardised rate per
100,000 (Modified from10). Figure 1. The global burden of hepatocellular carcinoma. Incidence of hepatocellular
carcinoma (HCC) according to data from Globocan 20121,3–9. ASR, Age-standardised rate per
100,000 (Modified from10). Figure 2. Liver cancer incidence according to region and sex. Figures reflect age-
standardized rate per 100,000 inhabitants. (Adapted from1.) Figure 2. Liver cancer incidence according to region and sex. Figures reflect age-
standardized rate per 100,000 inhabitants. (Adapted from1.) Figure 2. Liver cancer incidence according to region and sex. Figures reflect age-
standardized rate per 100,000 inhabitants. (Adapted from1.) Figure 3. Cancer progression and driver genes. Major recurrent molecular defects observed
early in liver carcinogenesis. TERT promoter mutations are common early events identified in
most of HCC that develops in a cirrhotic liver. Other mechanisms are specifically related to
risk factors: HBV and AAV2 viral infections induce insertional mutagenesis that recurrently
targets oncogenes. Also, hepatocellular adenoma, a rare benign liver tumour occurring most
frequently in women who take oral contraception, can transform into HCC with sequential
accumulation of CTNNB1 and TERT promoter mutations with or without STAT3 activation33–
38. AAV2, adeno-associated virus 2; CCN, cyclin; CTNNB1, catenin β 1; FRK, fyn-related Src
family tyrosine kinase; GNAS, GNAS complex locus; HBV, hepatitis B virus; HCA,
hepatocellular adenoma; HCC, hepatocellular carcinoma; HCV, hepatitis C virus; HNF1A,
HNF1 homeobox A; JAK1, Janus kinase 1; KMT2B (also called MLL4), lysine (K)-specific
methyltransferase 4; IL6ST, interleukin 6 signal transducer; STAT3, signal transducer and
activator of transcription 3 (acute-phase response factor); TERT, telomere reverse
transcriptase. Figure 4: BCLC staging system and therapeutic strategy. Classification comprises five stages
that select the best candidates for the best therapies currently available. Patients with
asymptomatic early tumours (stages 0–A) are candidates for radical therapies (resection,
transplantation or local ablation). Asymptomatic patients with multinodular HCC (stage B)
are suitable for transcatheter arterial chemoembolization (TACE), whereas patients with
advanced symptomatic tumours and/or an invasive tumoural pattern (stage C) are
candidates to receive sorafenib. End-stage disease (stage D) includes patients with poor
prognosis that should be treated by best supportive care. *Patients in Child-Pugh class C
should be first considered for liver transplantation. **Treatment stage migration: Consider
the next efficacious treatment in the algorithm when previous therapies fail. BCLC,
Barcelona Clinic Liver Cancer; DDLT, deceased donor liver transplantation; EASL, European
Association for the Study of Liver Disease; EORTC, European Organisation for Research and
Treatment of Cancer; GRADE, grading of recommendations assessment, development and
evaluation; HCC, hepatocellular carcinoma; LDLT, living donor liver transplantation; PEI,
percutaneous ethanol injection; RF, radiofrequency ablation; TACE, transcatheter arterial
chemoembolization; OS, overall survival; (Modified from3,158,159). 29 29 Figure 5. Molecular targeted therapies for HCC and their target signalling pathways. Summary of treatments tested in Phase II-III clinical trials. Orange receptors have tyrosine
kinase activity and red receptors have serine/threonine kinase activity. Figure 2. Liver cancer incidence according to region and sex. Figures reflect age-
standardized rate per 100,000 inhabitants. (Adapted from1.) Green boxes contain
drugs with positive Phase III studies, red boxes contain drugs with negative results from
Phase III trials and drugs in grey boxes have been tested in phase II studies. AA3R, adenosine
A(3) receptor; AKT, protein kinase B; c-Kit, mast/stem cell growth factor receptor; AR,
androgen receptor; CDK, cyclin dependent kinase; CTLA-4, cytotoxic T-lymphocyte protein 4;
SDF1, stromal cell-derived factor 1; EGF, epidermal growth factor; EGFR, epidermal growth
factor receptor; ERK, mitogen activated protein kinase kinase kinase; HDAC, histone
deacetylase; HGF, hepatocyte growth factor;MEK, mitogen-activated protein kinase kinase;
MET, hepatocyte growth factor receptor; mTOR, mammalian target of rapamycin; MYC, myc
proto-oncogene protein; PD-1, programmed cell death protein 1; PD-L1, programmed cell
death protein 1 ligand 1; PDGF, platelet-derived growth factor; PDGFR, platelet-derived
growth factor receptor; PI3K, phosphatidylinositol-4,5-bisphosphate 3-kinase; PLK1,
serine/threonine-protein kinase PLK1; RAF, Raf family protein; RAS, Ras family protein;
STAT3, signal transducer and activator of transcription 3; TAF, tumour-associated fibroblast;
TAM: tumour-associated macrophage; TGFβ; transforming growth factor beta; TGFβR1,
transforming growth factor beta receptor type-1; TIE2, angiopoietin-1 receptor; VEGF,
vascular endothelial growth factor. 7.
El-Serag, H. B. Hepatocellular carcinoma. N. Engl. J. Med. 365, 1118–27 (2011). References 2. Global, regional, and national age–sex specific all-cause and cause-specific mortality
for 240 causes of death, 1990–2013: a systematic analysis for the Global Burden of
Disease Study 2013. Lancet 385, 117–71 (2014). 2. Global, regional, and national age–sex specific all-cause and cause-specific mortality
for 240 causes of death, 1990–2013: a systematic analysis for the Global Burden of
Disease Study 2013. Lancet 385, 117–71 (2014). 3. EASL-EORTC clinical practice guidelines: management of hepatocellular carcinoma. J. Hepatol. 56, 908–43 (2012). 3. EASL-EORTC clinical practice guidelines: management of hepatocellular carcinoma. J. Hepatol. 56, 908–43 (2012). 4. Liu, J. & Fan, D. Hepatitis B in China. Lancet 369, 1582–1583 (2007). 4. Liu, J. & Fan, D. Hepatitis B in China. Lancet 369, 1582–1583 (2007). 5. Mohd Hanafiah, K., Groeger, J., Flaxman, A. D. & Wiersma, S. T. Global epidemiology
of hepatitis C virus infection: new estimates of age-specific antibody to HCV
seroprevalence. Hepatology 57, 1333–42 (2013). 6. Mohamoud, Y. A., Mumtaz, G. R., Riome, S., Miller, D. & Abu-Raddad, L. J. The
epidemiology of hepatitis C virus in Egypt: a systematic review and data synthesis. BMC Infect. Dis. 13, 288 (2013). 7. El-Serag, H. B. Hepatocellular carcinoma. N. Engl. J. Med. 365, 1118–27 (2011). 30 8. Hyattsville, M. Health, United States, 2014: With Special Feature on Adults Aged 55–
64. Natl. Cent. Heal. Stat. (2015). 9. Omer, R. E. et al. Population-attributable risk of dietary aflatoxins and hepatitis B virus
infection with respect to hepatocellular carcinoma. Nutr. Cancer 48, 15–21 (2004). 10. Laursen, L. A preventable cancer. Nature 516, S2–3 (2014). 11. Sartorius, K., Sartorius, B., Aldous, C., Govender, P. S. & Madiba, T. E. Global and
country underestimation of hepatocellular carcinoma (HCC) in 2012 and its
implications. Cancer Epidemiol. 39, 284–290 (2015). 12. Ferlay, J. et al. Cancer incidence and mortality worldwide: sources, methods and
major patterns in GLOBOCAN 2012. Int. J. Cancer 136, E359–86 (2015). 13. Nault, J.-C. et al. Recurrent AAV2-related insertional mutagenesis in human
hepatocellular carcinomas. Nat. Genet. 10, 1187–93 (2015). 14. Zucman-Rossi, J., Villanueva, A., Nault, J.-C. & Llovet, J. M. The genetic landscape and
biomarkers of hepatocellular carcinoma. Gastroenterology 5, 1226–1239 (2015). 15. Bruix, J. & Sherman, M. Management of hepatocellular carcinoma: An update. Hepatology 53, 1020–1022 (2011). 16. Forner, A., Llovet, J. M. & Bruix, J. Hepatocellular carcinoma. Lancet 379, 1245–1255
(2012). 17. Llovet, J. M. et al. References Arterial embolisation or chemoembolisation versus symptomatic
treatment in patients with unresectable hepatocellular carcinoma: a randomised
controlled trial. Lancet 359, 1734–9 (2002). 18. Llovet, J. M. & Bruix, J. Systematic review of randomized trials for unresectable
hepatocellular carcinoma: Chemoembolization improves survival. Hepatology 37,
429–442 (2003). 19. Llovet, J. M. et al. Sorafenib in advanced hepatocellular carcinoma. N. Engl. J. Med. 359, 378–90 (2008). 20. Yang, J. D. et al. Cirrhosis is present in most patients with hepatitis B and
hepatocellular carcinoma. Clin. Gastroenterol. Hepatol. 9, 64–70 (2011). 21. Lok, A. S. et al. Incidence of hepatocellular carcinoma and associated risk factors in
hepatitis C-related advanced liver disease. Gastroenterology 136, 138–48 (2009). 22. Hsu, I. C. et al. Mutational hotspot in the p53 gene in human hepatocellular
carcinomas. Nature 350, 427–428 (1991). 31 23. Bruix, J. & Sherman, M. Management of hepatocellular carcinoma. Hepatology 42,
1208–1236 (2005). 24. Yang, D. et al. Impact of sex on the survival of patients with hepatocellular carcinoma:
a Surveillance, Epidemiology, and End Results analysis. Cancer 120, 3707–3716 (2014). 25. Mittal, S. et al. Temporal trends of nonalcoholic fatty liver disease-related
hepatocellular carcinoma in the veteran affairs population. Clin. Gastroenterol. Hepatol. 13, 594–601 e1 (2015). 26. Chang, M. H. et al. Decreased incidence of hepatocellular carcinoma in hepatitis B
vaccinees: a 20-year follow-up study. J. Natl. Cancer Inst. 101, 1348–1355 (2009). 27. Wong, G. L. et al. Entecavir treatment reduces hepatic events and deaths in chronic
hepatitis B patients with liver cirrhosis. Hepatology 58, 1537–1547 (2013). 28. Singal, A. G., Volk, M. L., Jensen, D., Di Bisceglie, A. M. & Schoenfeld, P. S. A sustained
viral response is associated with reduced liver-related morbidity and mortality in
patients with hepatitis C virus. Clin. Gastroenterol. Hepatol. 8, 280–8, 288 e1 (2010). 29. Singh, S., Singh, P. P., Singh, A. G., Murad, M. H. & Sanchez, W. Statins are associated
with a reduced risk of hepatocellular cancer: a systematic review and meta-analysis. Gastroenterology 144, 323–332 (2013). 30. Marquardt, J. U., Andersen, J. B. & Thorgeirsson, S. S. Functional and genetic
deconstruction of the cellular origin in liver cancer. Nat. Rev. Cancer 15, 653–67
(2015). 31. Alizadeh, A. A. et al. Toward understanding and exploiting tumor heterogeneity. Nat. Med. 21, 846–853 (2015). 32. Guichard, C. et al. Integrated analysis of somatic mutations and focal copy-number
changes identifies key genes and pathways in hepatocellular carcinoma. Nat. Genet. 44, 694–8 (2012). 33. References Schulze, K. et al. Exome sequencing of hepatocellular carcinomas identifies new
mutational signatures and potential therapeutic targets. Nat. Genet. 47, (2015). 34. Nault, J. C. et al. High frequency of telomerase reverse-transcriptase promoter
somatic mutations in hepatocellular carcinoma and preneoplastic lesions. Nat. Commun. 4, 2218 (2013). 35. Ahn, S.-M. et al. Genomic portrait of resectable hepatocellular carcinomas:
implications of RB1 and FGF19 aberrations for patient stratification. Hepatology 60,
1972–82 (2014). 32 36. Nault, J. C. et al. Telomerase reverse transcriptase promoter mutation is an early
somatic genetic alteration in the transformation of premalignant nodules in
hepatocellular carcinoma on cirrhosis. Hepatology 60, 1983–92 (2014). 37. Pilati, C. et al. Genomic profiling of hepatocellular adenomas reveals recurrent FRK-
activating mutations and the mechanisms of malignant transformation. Cancer Cell
25, 428–41 (2014). 38. Totoki, Y. et al. Trans-ancestry mutational landscape of hepatocellular carcinoma
genomes. Nat. Genet. 46, 1267–73 (2014). 39. Di Tommaso, L. et al. Diagnostic value of HSP70, glypican 3, and glutamine synthetase
in hepatocellular nodules in cirrhosis. Hepatology 45, 725–34 (2007). 40. Di Tommaso, L. et al. The application of markers (HSP70 GPC3 and GS) in liver biopsies
is useful for detection of hepatocellular carcinoma. J. Hepatol. 50, 746–54 (2009). 41. Llovet, J. M. et al. A molecular signature to discriminate dysplastic nodules from early
hepatocellular carcinoma in HCV cirrhosis. Gastroenterology 131, 1758–67 (2006). 42. Wurmbach, E. et al. Genome-wide molecular profiles of HCV-induced dysplasia and
hepatocellular carcinoma. Hepatology 45, 938–47 (2007). 43. Paradis, V. et al. Molecular profiling of hepatocellular carcinomas (HCC) using a large-
scale real-time RT-PCR approach: determination of a molecular diagnostic index. Am. J. Pathol. 163, 733–41 (2003). 44. Günes, C. & Rudolph, K. L. The role of telomeres in stem cells and cancer. Cell 152,
390–3 (2013). 45. Satyanarayana, A. et al. Mitogen stimulation cooperates with telomere shortening to
activate DNA damage responses and senescence signaling. Mol. Cell. Biol. 24, 5459–74
(2004). 46. Farazi, P. A. et al. Differential impact of telomere dysfunction on initiation and
progression of hepatocellular carcinoma. Cancer Res. 63, 5021–7 (2003). 47. Hartmann, D. et al. Telomerase gene mutations are associated with cirrhosis
formation. Hepatology 53, 1608–17 (2011). 48. Calado, R. T. et al. Constitutional telomerase mutations are genetic risk factors for
cirrhosis. Hepatology 53, 1600–7 (2011). 49. Rudolph, K. L., Chang, S., Millard, M., Schreiber-Agus, N. & DePinho, R. A. Inhibition of
experimental liver cirrhosis in mice by telomerase gene delivery. References Science 287, 1253–8
(2000). 33 50. Lechel, A. et al. Telomerase deletion limits progression of p53-mutant hepatocellular
carcinoma with short telomeres in chronic liver disease. Gastroenterology 132, 1465–
75 (2007). 51. Kotoula, V. et al. Expression of human telomerase reverse transcriptase in
regenerative and precancerous lesions of cirrhotic livers. Liver 22, 57–69 (2002). 52. Bartosch, B., Thimme, R., Blum, H. E. & Zoulim, F. Hepatitis C virus-induced
hepatocarcinogenesis. J. Hepatol. 51, 810–20 (2009). 53. Neuveut, C., Wei, Y. & Buendia, M. A. Mechanisms of HBV-related
hepatocarcinogenesis. J. Hepatol. 52, 594–604 (2010). 54. Wang, J., Chenivesse, X., Henglein, B. & Bréchot, C. Hepatitis B virus integration in a
cyclin A gene in a hepatocellular carcinoma. Nature 343, 555–7 (1990). 55. Sung, W.-K. et al. Genome-wide survey of recurrent HBV integration in hepatocellular
carcinoma. Nat. Genet. 44, 765–9 (2012). 56. De La Coste, A. et al. Somatic mutations of the beta-catenin gene are frequent in
mouse and human hepatocellular carcinomas. Proc. Natl. Acad. Sci. U. S. A. 95, 8847–
51 (1998). 57. Audard, V. et al. Cholestasis is a marker for hepatocellular carcinomas displaying beta-
catenin mutations. J. Pathol. 212, 345–52 (2007). 58. Bressac, B., Kew, M., Wands, J. & Ozturk, M. Selective G to T mutations of p53 gene in
hepatocellular carcinoma from southern Africa. Nature 350, 429–31 (1991). 59. Amaddeo, G. et al. Integration of tumour and viral genomic characterizations in HBV-
related hepatocellular carcinomas. Gut 64, 820–9 (2015). 60. Fei, Q. et al. Histone methyltransferase SETDB1 regulates liver cancer cell growth
through methylation of p53. Nat. Commun. 6, 8651 (2015). 61. Wong, C.-M. et al. Up-regulation of histone methyltransferase SETDB1 by multiple
mechanisms in hepatocellular carcinoma promotes cancer metastasis. Hepatology
(2015). 62. Villanueva, A. et al. DNA methylation-based prognosis and epidrivers in hepatocellular
carcinoma. Hepatology 61, 1945–56 (2015). 63. Herceg, Z. & Paliwal, A. Epigenetic mechanisms in hepatocellular carcinoma: how
environmental factors influence the epigenome. Mutat. Res. 727, 55–61 (2011). 64. Sporn, M. B. & Liby, K. T. NRF2 and cancer: the good, the bad and the importance of
context. Nat. Rev. Cancer 12, 564–71 (2012). 34 65. Chiang, D. Y. et al. Focal gains of VEGFA and molecular classification of hepatocellular
carcinoma. Cancer Res. 68, 6779–6788 (2008). 66. Sawey, E. T. et al. Identification of a therapeutic strategy targeting amplified FGF19 in
liver cancer by Oncogenomic screening. Cancer Cell 19, 347–58 (2011). 67. Horwitz, E. et al. References Human and mouse VEGFA-amplified hepatocellular carcinomas are
highly sensitive to Sorafenib treatment. Cancer Discov. 4, 730–743 (2014). 68. Rudalska, R. et al. In vivo RNAi screening identifies a mechanism of sorafenib
resistance in liver cancer. Nat. Med. 20, 1138–46 (2014). 69. Weber, J. et al. CRISPR/Cas9 somatic multiplex-mutagenesis for high-throughput
functional cancer genomics in mice. Proc. Natl. Acad. Sci. U. S. A. 112, 13982–7 (2015). 70. Hoshida, Y. et al. Molecular classification and novel targets in hepatocellular
carcinoma: recent advancements. Semin. Liver Dis. 30, 35–51 (2010). 71. Boyault, S. et al. Transcriptome classification of HCC is related to gene alterations and
to new therapeutic targets. Hepatology 45, 42–52 (2007). 72. Lee, J.-S. et al. Classification and prediction of survival in hepatocellular carcinoma by
gene expression profiling. Hepatology 40, 667–76 (2004). 73. Lee, J.-S. et al. A novel prognostic subtype of human hepatocellular carcinoma derived
from hepatic progenitor cells. Nat. Med. 12, 410–6 (2006). 74. Hoshida, Y. et al. Integrative transcriptome analysis reveals common molecular
subclasses of human hepatocellular carcinoma. Cancer Res. 69, 7385–92 (2009). 75. Lachenmayer, A. et al. Wnt-pathway activation in two molecular classes of
hepatocellular carcinoma and experimental modulation by sorafenib. Clin. Cancer Res. 18, 4997–5007 (2012). 76. Hanahan, D. & Weinberg, R. A. Hallmarks of cancer: the next generation. Cell 144,
646–674 (2011). 77. Grivennikov, S. I., Greten, F. R. & Karin, M. Immunity, inflammation, and cancer. Cell
140, 883–899 (2010). 78. Dvorak, H. F. Tumors: wounds that do not heal. Similarities between tumor stroma
generation and wound healing. N Engl J Med 315, 1650–1659 (1986). 79. Hernandez-Gea, V., Toffanin, S., Friedman, S. L. & Llovet, J. M. Role of the
microenvironment in the pathogenesis and treatment of hepatocellular carcinoma. Gastroenterology 144, 512–527 (2013). 35 35 80. Wada, Y., Nakashima, O., Kutami, R., Yamamoto, O. & Kojiro, M. Clinicopathological
study on hepatocellular carcinoma with lymphocytic infiltration. Hepatology 27, 407–
414 (1998). 81. Hoshida, Y. et al. Gene expression in fixed tissues and outcome in hepatocellular
carcinoma. N Engl J Med 359, 1995–2004 (2008). 82. Finkin, S. et al. Ectopic lymphoid structures function as microniches for tumor
progenitor cells in hepatocellular carcinoma. Nat. Immunol. 16, 1235–44 (2015). 3. Crispe, I. N. The liver as a lymphoid organ. Annu Rev Immunol 27, 147–163 (2009). 84. Thomson, A. W. & Knolle, P. A. Antigen-presenting cell function in the tolerogenic liver
environment. Nat Rev Immunol 10, 753–766 (2010). 85. Pikarsky, E. 94.
Lee, Y. A., Wallace, M. C. & Friedman, S. L. Pathobiology of liver fibrosis: a
translational success story. Gut 64, 830–41 (2015). References et al. NF-kappaB functions as a tumour promoter in inflammation-
associated cancer. Nature 431, 461–6 (2004). 86. Bauer, J. et al. Lymphotoxin, NF-kB, and cancer: the dark side of cytokines. Dig Dis 30,
453–468 (2012). 87. Taniguchi, K. & Karin, M. IL-6 and related cytokines as the critical lynchpins between
inflammation and cancer. Semin Immunol 26, 54–74 (2014). 88. LeCouter, J. et al. Angiogenesis-independent endothelial protection of liver: role of
VEGFR-1. Science (80-. ). 299, 890–893 (2003). 89. Hussain, S. P., Hofseth, L. J. & Harris, C. C. Radical causes of cancer. Nat Rev Cancer 3,
276–285 (2003). 90. Kiraly, O., Gong, G., Olipitz, W., Muthupalani, S. & Engelward, B. P. Inflammation-
induced cell proliferation potentiates DNA damage-induced mutations in vivo. PLoS
Genet 11, e1004901 (2015). 91. Zhang, D. Y. et al. A hepatic stellate cell gene expression signature associated with
outcomes in hepatitis C cirrhosis and hepatocellular carcinoma after curative
resection. Gut (2015). 92. Campbell, J. S. et al. Platelet-derived growth factor C induces liver fibrosis, steatosis,
and hepatocellular carcinoma. Proc. Natl. Acad. Sci. U. S. A. 102, 3389–94 (2005). 93. Dapito, D. H. & Schwabe, R. F. Hepatic Stellate Cells and Liver Cancer. Stellate Cells in
Health and Disease (Elsevier, 2015). 94. Lee, Y. A., Wallace, M. C. & Friedman, S. L. Pathobiology of liver fibrosis: a
translational success story. Gut 64, 830–41 (2015). 36 36 95. Chang, M. H. et al. Universal hepatitis B vaccination in Taiwan and the incidence of
hepatocellular carcinoma in children. Taiwan Childhood Hepatoma Study Group. N. Engl. J. Med. 336, 1855–9 (1997). 96. Yuen, M.-F. et al. Independent risk factors and predictive score for the development
of hepatocellular carcinoma in chronic hepatitis B. J. Hepatol. 50, 80–8 (2009). 97. Wong, V. W.-S. et al. Clinical scoring system to predict hepatocellular carcinoma in
chronic hepatitis B carriers. J. Clin. Oncol. 28, 1660–5 (2010). 98. Wong, G. L.-H. et al. Liver stiffness-based optimization of hepatocellular carcinoma
risk score in patients with chronic hepatitis B. J. Hepatol. 60, 339–45 (2014). 99. Chen, C.-J. Risk of Hepatocellular Carcinoma Across a Biological Gradient of Serum
Hepatitis B Virus DNA Level. JAMA 295, 65 (2006). 100. Yang, H.-I. et al. Nomograms for risk of hepatocellular carcinoma in patients with
chronic hepatitis B virus infection. J. Clin. Oncol. 28, 2437–44 (2010). 101. Yang, H.-I. et al. References Risk estimation for hepatocellular carcinoma in chronic hepatitis B
(REACH-B): development and validation of a predictive score. Lancet. Oncol. 12, 568–
74 (2011). 102. Lee, M.-H. et al. Prediction models of long-term cirrhosis and hepatocellular
carcinoma risk in chronic hepatitis B patients: risk scores integrating host and virus
profiles. Hepatology 58, 546–54 (2013). 103. Wong, G. L.-H. et al. Accuracy of risk scores for patients with chronic hepatitis B
receiving entecavir treatment. Gastroenterology 144, 933–44 (2013). 104. Arends, P. et al. Entecavir treatment does not eliminate the risk of hepatocellular
carcinoma in chronic hepatitis B: limited role for risk scores in Caucasians. Gut 64,
1289–95 (2015). 105. Liaw, Y.-F. et al. Lamivudine for patients with chronic hepatitis B and advanced liver
disease. N. Engl. J. Med. 351, 1521–31 (2004). 106. Papatheodoridis, G. V et al. Incidence and predictors of hepatocellular carcinoma in
Caucasian chronic hepatitis B patients receiving entecavir or tenofovir. J. Hepatol. 62,
363–70 (2015). 107. Ogawa, E. et al. Efficacy of pegylated interferon alpha-2b and ribavirin treatment on
the risk of hepatocellular carcinoma in patients with chronic hepatitis C: a prospective,
multicenter study. J. Hepatol. 58, 495–501 (2013). 37 37 108. Harada, N. et al. Risk factors for hepatocellular carcinoma in hepatitis C patients with
normal alanine aminotransferase treated with pegylated interferon and ribavirin. J. Viral Hepat. 21, 357–65 (2014). 109. Dohmen, K. et al. The incidence and risk factors for the development of hepatocellular
carcinoma after peginterferon plus ribavirin therapy for chronic hepatitis C. Hepatogastroenterology. 60, 2034–8 (2013). 110. Van der Meer, A. J. et al. Association between sustained virological response and all-
cause mortality among patients with chronic hepatitis C and advanced hepatic
fibrosis. JAMA 308, 2584–93 (2012). 111. Morgan, R. L. et al. Eradication of hepatitis C virus infection and the development of
hepatocellular carcinoma: a meta-analysis of observational studies. Ann. Intern. Med. 158, 329–37 (2013). 112. Foster, G. R. et al. Sofosbuvir and Velpatasvir for HCV Genotype 2 and 3 Infection. N. Engl. J. Med. 373, 2608–17 (2015). 113. Singal, A. G. & El-Serag, H. B. Hepatocellular Carcinoma from Epidemiology to
Prevention: Translating Knowledge into Practice. Clin. Gastroenterol. Hepatol. 13,
2140–51 (2015). 114. Younossi, Z. M. et al. The impact of hepatitis C burden: an evidence-based approach. Aliment. Pharmacol. Ther. 39, 518–31 (2014). 115. Bosch, J. & Forns, X. Therapy. Statins and liver disease: from concern to ‘wonder’
drugs? Nat. Rev. Gastroenterol. Hepatol. References 12, 320–1 (2015). 116. Zhang, H., Gao, C., Fang, L., Zhao, H.-C. & Yao, S.-K. Metformin and reduced risk of
hepatocellular carcinoma in diabetic patients: a meta-analysis. Scand. J. Gastroenterol. 48, 78–87 (2013). 117. Zhang, X. et al. Continuation of metformin use after a diagnosis of cirrhosis
significantly improves survival of patients with diabetes. Hepatology 60, 2008–16
(2014). 118. Zhang, B.-H., Yang, B.-H. & Tang, Z.-Y. Randomized controlled trial of screening for
hepatocellular carcinoma. J. Cancer Res. Clin. Oncol. 130, 417–22 (2004). 119. Yeh, Y.-P. et al. Evaluation of abdominal ultrasonography mass screening for
hepatocellular carcinoma in Taiwan. Hepatology 59, 1840–9 (2014). 120. Tong, M. J., Sun, H.-E., Hsien, C. & Lu, D. S. K. Surveillance for hepatocellular
carcinoma improves survival in Asian-American patients with hepatitis B: results from
a community-based clinic. Dig. Dis. Sci. 55, 826–35 (2010). 38 121. Tanaka, H. et al. Surveillance of hepatocellular carcinoma in patients with hepatitis C
virus infection may improve patient survival. Liver Int. 26, 543–51 (2006). 122. Wong, G. L.-H. et al. Surveillance programme for hepatocellular carcinoma improves
the survival of patients with chronic viral hepatitis. Liver Int. 28, 79–87 (2008). 123. Taura, N. et al. Clinical benefits of hepatocellular carcinoma surveillance: a single-
center, hospital-based study. Oncol. Rep. 14, 999–1003 (2005). 124. Chan, A. C. Y. et al. Changing paradigm in the management of hepatocellular
carcinoma improves the survival benefit of early detection by screening. Ann. Surg. 247, 666–73 (2008). 125. Nouso, K. et al. Cost-effectiveness of the surveillance program of hepatocellular
carcinoma depends on the medical circumstances. J. Gastroenterol. Hepatol. 23, 437–
44 (2008). 126. Lin, O. S., Keeffe, E. B., Sanders, G. D. & Owens, D. K. Cost-effectiveness of screening
for hepatocellular carcinoma in patients with cirrhosis due to chronic hepatitis C. Aliment. Pharmacol. Ther. 19, 1159–72 (2004). 127. Arguedas, M. R., Chen, V. K., Eloubeidi, M. A. & Fallon, M. B. Screening for
hepatocellular carcinoma in patients with hepatitis C cirrhosis: a cost-utility analysis. Am. J. Gastroenterol. 98, 679–90 (2003). 128. Andersson, K. L., Salomon, J. A., Goldie, S. J. & Chung, R. T. Cost effectiveness of
alternative surveillance strategies for hepatocellular carcinoma in patients with
cirrhosis. Clin. Gastroenterol. Hepatol. 6, 1418–24 (2008). 129. Cucchetti, A. et al. Cost-effectiveness of semi-annual surveillance for hepatocellular
carcinoma in cirrhotic patients of the Italian Liver Cancer population. J. Hepatol. 56,
1089–96 (2012). 130. Sarasin, F. P., Giostra, E. & Hadengue, A. References Cost-effectiveness of screening for detection
of small hepatocellular carcinoma in western patients with Child-Pugh class A
cirrhosis. Am. J. Med. 101, 422–34 (1996). 131. Thompson Coon, J. et al. Surveillance of cirrhosis for hepatocellular carcinoma:
systematic review and economic analysis. Health Technol. Assess. 11, 1–206 (2007). 132. Saab, S. et al. Hepatocellular carcinoma screening in patients waiting for liver
transplantation: a decision analytic model. Liver Transpl. 9, 672–81 (2003). 133. Naugler, W. E. & Sonnenberg, A. Survival and cost-effectiveness analysis of competing
strategies in the management of small hepatocellular carcinoma. Liver Transpl. 16,
1186–94 (2010). 39 134. Thompson Coon, J. et al. Surveillance of cirrhosis for hepatocellular carcinoma: a cost-
utility analysis. Br. J. Cancer 98, 1166–75 (2008). 135. Patel, D., Terrault, N. A., Yao, F. Y., Bass, N. M. & Ladabaum, U. Cost-effectiveness of
hepatocellular carcinoma surveillance in patients with hepatitis C virus-related
cirrhosis. Clin. Gastroenterol. Hepatol. 3, 75–84 (2005). 136. Kang, J. Y., Lee, T. P., Yap, I. & Lun, K. C. Analysis of cost-effectiveness of different
strategies for hepatocellular carcinoma screening in hepatitis B virus carriers. J. Gastroenterol. Hepatol. 7, 463–8 (1992). 137. Shih, S. T.-F., Crowley, S. & Sheu, J.-C. Cost-effectiveness analysis of a two-stage
screening intervention for hepatocellular carcinoma in Taiwan. J. Formos. Med. Assoc. 109, 39–55 (2010). 138. Marrero, J. A. et al. Alpha-fetoprotein, des-gamma carboxyprothrombin, and lectin-
bound alpha-fetoprotein in early hepatocellular carcinoma. Gastroenterology 137,
110–8 (2009). 139. Asaoka, Y. et al. Frequency of and predictive factors for vascular invasion after
radiofrequency ablation for hepatocellular carcinoma. PLoS One 9, e111662 (2014). 140. Kudo, A. et al. Does the preoperative alpha-fetoprotein predict the recurrence and
mortality after hepatectomy for hepatocellular carcinoma without macrovascular
invasion in patients with normal liver function? Hepatol. Res. 44, E437–46 (2014). 141. Shirabe, K. et al. New scoring system for prediction of microvascular invasion in
patients with hepatocellular carcinoma. Liver Int. 34, 937–41 (2014). 142. Park, H. et al. Clinical usefulness of double biomarkers AFP and PIVKA-II for
subdividing prognostic groups in locally advanced hepatocellular carcinoma. Liver Int. 34, 313–21 (2014). 143. Nakazawa, T. et al. Early increase in α-fetoprotein for predicting unfavorable clinical
outcomes in patients with advanced hepatocellular carcinoma treated with sorafenib. Eur. J. Gastroenterol. Hepatol. 25, 683–9 (2013). 144. Park, W.-H. et al. Clinical utility of des-γ-carboxyprothrombin kinetics as a
complement to radiologic response in patients with hepatocellular carcinoma
undergoing transarterial chemoembolization. J. Vasc. Interv. Radiol. References 23, 927–36
(2012). 145. Han, K.-H. et al. Survival of hepatocellular carcinoma patients may be improved in
surveillance interval not more than 6 months compared with more than 6 months: a
15-year prospective study. J. Clin. Gastroenterol. 47, 538–44 (2013). 40 40 146. Trinchet, J.-C. et al. Ultrasonographic surveillance of hepatocellular carcinoma in
cirrhosis: a randomized trial comparing 3- and 6-month periodicities. Hepatology 54,
1987–97 (2011). 147. Santi, V. et al. Semiannual surveillance is superior to annual surveillance for the
detection of early hepatocellular carcinoma and patient survival. J. Hepatol. 53, 291–7
(2010). 148. Wang, J.-H. et al. Hepatocellular carcinoma surveillance at 4- vs. 12-month intervals
for patients with chronic viral hepatitis: a randomized study in community. Am. J. Gastroenterol. 108, 416–24 (2013). 149. Mitchell, D. G., Bruix, J., Sherman, M. & Sirlin, C. B. LI-RADS (Liver Imaging Reporting
and Data System): summary, discussion, and consensus of the LI-RADS Management
Working Group and future directions. Hepatology 61, 1056–65 (2015). 150. Khalili, K. et al. Optimization of imaging diagnosis of 1-2 cm hepatocellular carcinoma:
an analysis of diagnostic performance and resource utilization. J. Hepatol. 54, 723–8
(2011). 151. Sangiovanni, A. et al. The diagnostic and economic impact of contrast imaging
techniques in the diagnosis of small hepatocellular carcinoma in cirrhosis. Gut 59,
638–44 (2010). 152. Lencioni, R., Cioni, D., Della Pina, C., Crocetti, L. & Bartolozzi, C. Imaging diagnosis. Semin. Liver Dis. 25, 162–70 (2005). 153. Silva, M. A. et al. Needle track seeding following biopsy of liver lesions in the diagnosis
of hepatocellular cancer: a systematic review and meta-analysis. Gut 57, 1592–6
(2008). 154. Pathologic diagnosis of early hepatocellular carcinoma: a report of the international
consensus group for hepatocellular neoplasia. Hepatology 49, 658–64 (2009). 155. Libbrecht, L. et al. Glypican-3 expression distinguishes small hepatocellular
carcinomas from cirrhosis, dysplastic nodules, and focal nodular hyperplasia-like
nodules. Am. J. Surg. Pathol. 30, 1405–11 (2006). 156. Tremosini, S. et al. Prospective validation of an immunohistochemical panel (glypican
3, heat shock protein 70 and glutamine synthetase) in liver biopsies for diagnosis of
very early hepatocellular carcinoma. Gut 61, 1481–7 (2012). 157. Di Tommaso, L. et al. Diagnostic accuracy of clathrin heavy chain staining in a marker
panel for the diagnosis of small hepatocellular carcinoma. Hepatology 53, 1549–1557
(2011). 41 41 158. Bruix, J., Han, K., Gores, G., Llovet, J. M. & Mazzaferro, V. Liver cancer: Approaching a
personalized care. J. Hepatol. 62, 144–156 (2015). 159. References Llovet, J. M., Brú, C. & Bruix, J. Prognosis of hepatocellular carcinoma: the BCLC
staging classification. Semin. Liver Dis. 19, 329–38 (1999). 160. Yau, T. et al. Development of Hong Kong Liver Cancer staging system with treatment
stratification for patients with hepatocellular carcinoma. Gastroenterology 146, 1691–
700.e3 (2014). 161. A new prognostic system for hepatocellular carcinoma: a retrospective study of 435
patients: the Cancer of the Liver Italian Program (CLIP) investigators. Hepatology 28,
751–5 (1998). 162. Sobin, L. H. & Compton, C. C. TNM seventh edition: what’s new, what's changed:
communication from the International Union Against Cancer and the American Joint
Committee on Cancer. Cancer 116, 5336–9 (2010). 163. Kudo, M., Chung, H. & Osaki, Y. Prognostic staging system for hepatocellular
carcinoma (CLIP score): its value and limitations, and a proposal for a new staging
system, the Japan Integrated Staging Score (JIS score). J. Gastroenterol. 38, 207–15
(2003). 164. Bruix, J. et al. Surgical resection of hepatocellular carcinoma in cirrhotic patients:
prognostic value of preoperative portal pressure. Gastroenterology 111, 1018–22
(1996). 165. Llovet, J. M., Schwartz, M. & Mazzaferro, V. Resection and Liver Transplantation for
Hepatocellular Carcinoma. Semin Liver Dis 25, 181–200 (2005). 166. Llovet, J. M., Fuster, J. & Bruix, J. Intention-to-treat analysis of surgical treatment for
early hepatocellular carcinoma: resection versus transplantation. Hepatology 30,
1434–1440 (1999). 167. Imamura, H. et al. Risk factors contributing to early and late phase intrahepatic
recurrence of hepatocellular carcinoma after hepatectomy. J. Hepatol. 38, 200–7
(2003). 168. Bruix, J. et al. Adjuvant sorafenib for hepatocellular carcinoma after resection or
ablation (STORM): a phase 3, randomised, double-blind, placebo-controlled trial. Lancet. Oncol. 16, 1344–54 (2015). 169. Takayama, T. et al. Adoptive immunotherapy to lower postsurgical recurrence rates of
hepatocellular carcinoma: a randomised trial. Lancet 356, 802–7 (2000). 42 170. Muto, Y. et al. Prevention of second primary tumors by an acyclic retinoid,
polyprenoic acid, in patients with hepatocellular carcinoma. Hepatoma Prevention
Study Group. N. Engl. J. Med. 334, 1561–1567 (1996). 171. Mazzaferro, V. et al. Liver transplantation for the treatment of small hepatocellular
carcinomas in patients with cirrhosis. N. Engl. J. Med. 334, 693–9 (1996). 172. Mazzaferro, V. et al. Predicting survival after liver transplantation in patients with
hepatocellular carcinoma beyond the Milan criteria: a retrospective, exploratory
analysis. Lancet. Oncol. 10, 35–43 (2009). 173. Toso, C. et al. Total tumor volume and alpha-fetoprotein for selection of transplant
candidates with hepatocellular carcinoma: A prospective validation. References Hepatology 62,
158–65 (2015). 174. Miltiadous, O. et al. Progenitor cell markers predict outcome of patients with
Hepatocellular Carcinoma beyond Milan criteria undergoing liver transplantation. J. Hepatol. (2015). 175. Livraghi, T. et al. Sustained complete response and complications rates after
radiofrequency ablation of very early hepatocellular carcinoma in cirrhosis: Is
resection still the treatment of choice? Hepatology 47, 82–89 (2008). 176. Lo, C.-M. et al. Randomized controlled trial of transarterial lipiodol
chemoembolization for unresectable hepatocellular carcinoma. Hepatology 35, 1164–
71 (2002). 177. Kudo, M. et al. Brivanib as adjuvant therapy to transarterial chemoembolization in
patients with hepatocellular carcinoma: A randomized phase III trial. Hepatology 60,
1697–707 (2014). 178. Burrel, M. et al. Survival of patients with hepatocellular carcinoma treated by
transarterial chemoembolisation (TACE) using Drug Eluting Beads. Implications for
clinical practice and trial design. J. Hepatol. 56, 1330–5 (2012). 179. Salem, R. et al. Radioembolization results in longer time-to-progression and reduced
toxicity compared with chemoembolization in patients with hepatocellular carcinoma. Gastroenterology 140, 497–507.e2 (2011). 180. Cheng, A.-L. et al. Efficacy and safety of sorafenib in patients in the Asia-Pacific region
with advanced hepatocellular carcinoma: a phase III randomised, double-blind,
placebo-controlled trial. Lancet Oncol. 10, 25–34 (2009). 181. Yeo, W. et al. A randomized phase III study of doxorubicin versus cisplatin/interferon
alpha-2b/doxorubicin/fluorouracil (PIAF) combination chemotherapy for unresectable
hepatocellular carcinoma. J. Natl. Cancer Inst. 97, 1532–8 (2005). 43 182. D’Amico, G., Garcia-Tsao, G. & Pagliaro, L. Natural history and prognostic indicators of
survival in cirrhosis: A systematic review of 118 studies. J. Hepatol. 44, 217–231
(2006). 183. Eguchi, S. et al. Comparison of the outcomes between an anatomical
subsegmentectomy and a non-anatomical minor hepatectomy for single
hepatocellular carcinomas based on a Japanese nationwide survey. Surgery 143, 469–
75 (2008). 184. Palavecino, M. et al. Major hepatic resection for hepatocellular carcinoma with or
without portal vein embolization: Perioperative outcome and survival. Surgery 145,
399–405 (2009). 185. Vouche, M. et al. Radiation lobectomy: time-dependent analysis of future liver
remnant volume in unresectable liver cancer as a bridge to resection. J. Hepatol. 59,
1029–36 (2013). 186. Cherqui, D. Laparoscopic liver resection: A new paradigm in the management of
hepatocellular carcinoma? J. Hepatol. 63, 540–2 (2015). 187. Ishizawa, T. et al. Neither Multiple Tumors Nor Portal Hypertension Are Surgical
Contraindications for Hepatocellular Carcinoma. Gastroenterology 134, 1908–1916
(2008). 188. Franssen, B. et al. References Differences in surgical outcomes between hepatitis B- and hepatitis
C-related hepatocellular carcinoma: a retrospective analysis of a single North
American center. Ann. Surg. 260, 650–6; discussion 656–8 (2014). 189. Roayaie, S., Bassi, D., Tarchi, P., Labow, D. & Schwartz, M. Second hepatic resection
for recurrent hepatocellular cancer: a Western experience. J. Hepatol. 55, 346–50
(2011). 190. Fuks, D. et al. Benefit of initial resection of hepatocellular carcinoma followed by
transplantation in case of recurrence: an intention-to-treat analysis. Hepatology 55,
132–40 (2012). 191. Lau, W. Y. et al. Adjuvant intra-arterial iodine-131-labelled lipiodol for resectable
hepatocellular carcinoma: a prospective randomised trial. Lancet (London, England)
353, 797–801 (1999). 192. Clavien, P.-A. et al. Recommendations for liver transplantation for hepatocellular
carcinoma: an international consensus conference report. Lancet. Oncol. 13, e11–22
(2012). 193. Lesurtel, M., Müllhaupt, B., Pestalozzi, B. C., Pfammatter, T. & Clavien, P.-A. Transarterial chemoembolization as a bridge to liver transplantation for 44 hepatocellular carcinoma: an evidence-based analysis. Am. J. Transplant 6, 2644–50
(2006). hepatocellular carcinoma: an evidence-based analysis. Am. J. Transplant 6, 2644–50
(2006). 194. Kulik, L. M. et al. Outcomes of living and deceased donor liver transplant recipients
with hepatocellular carcinoma: results of the A2ALL cohort. Am. J. Transplant 12,
2997–3007 (2012). 195. Toso, C., Merani, S., Bigam, D. L., Shapiro, A. M. J. & Kneteman, N. M. Sirolimus-based
immunosuppression is associated with increased survival after liver transplantation
for hepatocellular carcinoma. Hepatology 51, 1237–43 (2010). 196. Schnitzbauer, A. A. et al. A prospective randomised, open-labeled, trial comparing
sirolimus-containing versus mTOR-inhibitor-free immunosuppression in patients
undergoing liver transplantation for hepatocellular carcinoma. BMC Cancer 10, 190
(2010). 197. Geissler, E. K. et al. Sirolimus Use in Liver Transplant Recipients With Hepatocellular
Carcinoma: A Randomized, Multicenter, Open-Label Phase 3 Trial. Transplantation
(2015). 198. Vitale, A. et al. Barcelona Clinic Liver Cancer staging and transplant survival benefit for
patients with hepatocellular carcinoma: a multicentre, cohort study. Lancet. Oncol. 12, 654–62 (2011). 199. Menon, K. V, Hakeem, A. R. & Heaton, N. D. Review article: liver transplantation for
hepatocellular carcinoma - a critical appraisal of the current worldwide listing criteria. Aliment. Pharmacol. Ther. 40, 893–902 (2014). 200. Yao, F. Y. et al. Liver transplantation for hepatocellular carcinoma: expansion of the
tumor size limits does not adversely impact survival. Hepatology 33, 1394–403 (2001). 201. Yao, F. Y. et al. Excellent outcome following down-staging of hepatocellular carcinoma
prior to liver transplantation: an intention-to-treat analysis. Hepatology 48, 819–27
(2008). 202. References Kowdley, K. V et al. Ledipasvir and sofosbuvir for 8 or 12 weeks for chronic HCV
without cirrhosis. N. Engl. J. Med. 370, 1879–88 (2014). 203. Ahmed, M. et al. Image-guided tumor ablation: standardization of terminology and
reporting criteria--a 10-year update. Radiology 273, 241–60 (2014). 204. Lencioni, R. & Crocetti, L. Image-guided ablation for hepatocellular carcinoma. Recent
Results Cancer Res 190, 181–94 (2013). 205. Cho YK, Kim JK, Kim MY, Rhim H, H. J. Systematic review of randomized trials for
hepatocellular carcinoma treated with percutaneous ablation therapies. Hepatology
49, 453–9 (2009). 45 206. Orlando, A., Leandro, G., Olivo, M., Andriulli, A. & Cottone, M. Radiofrequency
Thermal Ablation vs. Percutaneous Ethanol Injection for Small Hepatocellular
Carcinoma in Cirrhosis: Meta-Analysis of Randomized Controlled Trials. Am. J. Gastroenterol. 104, 514–524 (2009). 207. Shiina, S. et al. Radiofrequency ablation for hepatocellular carcinoma: 10-year
outcome and prognostic factors. Am. J. Gastroenterol. 107, 569–77; quiz 578 (2012). 208. Salem, R. et al. Research reporting standards for radioembolization of hepatic
malignancies. J. Vasc. Interv. Radiol. 22, 265–78 (2011). 209. Bruix, J., Sala, M. & Llovet, J. M. Chemoembolization for hepatocellular carcinoma. Gastroenterology 127, S179–88 (2004). 210. Raoul, J.-L. et al. Evolving strategies for the management of intermediate-stage
hepatocellular carcinoma: available evidence and expert opinion on the use of
transarterial chemoembolization. Cancer Treat. Rev. 37, 212–20 (2011). 211. Golfieri, R. et al. Randomised controlled trial of doxorubicin-eluting beads vs
conventional chemoembolisation for hepatocellular carcinoma. Br. J. Cancer 111,
255–64 (2014). 212. Sangro, B. & Salem, R. Transarterial chemoembolization and radioembolization. Semin. Liver Dis. 34, 435–43 (2014). 213. Lencioni, R. et al. Sorafenib or Placebo plus TACE with Doxorubicin-Eluting Beads for
Intermediate-Stage HCC: Phase II, Randomized, Doubled-Blind SPACE Trial. J Hepatol
In press, (2015). 214. Oliveri, R. S., Wetterslev, J. & Gluud, C. Transarterial (chemo)embolisation for
unresectable hepatocellular carcinoma. Cochrane database Syst. Rev. CD004787
(2011). 215. Sangro, B., Iñarrairaegui, M. & Bilbao, J. I. Radioembolization for hepatocellular
carcinoma. J. Hepatol. 56, 464–73 (2012). 216. Sangro, B. et al. Survival after yttrium-90 resin microsphere radioembolization of
hepatocellular carcinoma across Barcelona clinic liver cancer stages: a European
evaluation. Hepatology 54, 868–78 (2011). 217. Meyer, T. et al. A randomised phase II/III trial of 3-weekly cisplatin-based sequential
transarterial chemoembolisation vs embolisation alone for hepatocellular carcinoma. Br. J. Cancer 108, 1252–9 (2013). 218. Llovet, J. M., Villanueva, A., Lachenmayer, A. & Finn, R. S. References Advances in targeted
therapies for hepatocellular carcinoma in the genomic era. Nat. Rev. Clin. Oncol. (2015). 46 219. Qin, S. et al. Randomized, multicenter, open-label study of oxaliplatin plus
fluorouracil/leucovorin versus doxorubicin as palliative chemotherapy in patients with
advanced hepatocellular carcinoma from Asia. J. Clin. Oncol. 31, 3501–8 (2013). 220. Bruix, J. et al. Efficacy and safety of sorafenib in patients with advanced hepatocellular
carcinoma: subanalyses of a phase III trial. J. Hepatol. 57, 821–9 (2012). 221. Llovet, J. M. et al. Plasma biomarkers as predictors of outcome in patients with
advanced hepatocellular carcinoma. Clin. Cancer Res. 18, 2290–300 (2012). 222. Wilhelm, S. M. et al. Preclinical overview of sorafenib, a multikinase inhibitor that
targets both Raf and VEGF and PDGF receptor tyrosine kinase signaling. Mol. Cancer
Ther. 7, 3129–3140 (2008). 223. Llovet, J. M. et al. Design and endpoints of clinical trials in hepatocellular carcinoma. J. Natl. Cancer Inst. 100, 698–711 (2008). 224. Johnson, P. J. et al. Brivanib versus sorafenib as first-line therapy in patients with
unresectable, advanced hepatocellular carcinoma: results from the randomized phase
III BRISK-FL study. J. Clin. Oncol. 31, 3517–24 (2013). 225. Cheng, A.-L. et al. Sunitinib versus sorafenib in advanced hepatocellular cancer: results
of a randomized phase III trial. J. Clin. Oncol. 31, 4067–75 (2013). 226. Cainap, C. et al. Linifanib versus Sorafenib in patients with advanced hepatocellular
carcinoma: results of a randomized phase III trial. J. Clin. Oncol. 33, 172–9 (2015). 227. Zhu, A. X. et al. SEARCH: a phase III, randomized, double-blind, placebo-controlled trial
of sorafenib plus erlotinib in patients with advanced hepatocellular carcinoma. J. Clin. Oncol. 33, 559–66 (2015). 228. Llovet, J. M. et al. Brivanib in patients with advanced hepatocellular carcinoma who
were intolerant to sorafenib or for whom sorafenib failed: results from the
randomized phase III BRISK-PS study. J. Clin. Oncol. 31, 3509–3516 (2013). 229. Zhu, A. X. et al. Effect of everolimus on survival in advanced hepatocellular carcinoma
after failure of sorafenib: the EVOLVE-1 randomized clinical trial. JAMA 312, 57–67
(2014). 230. Zhu, A. X. et al. Ramucirumab versus placebo as second-line treatment in patients
with advanced hepatocellular carcinoma following first-line therapy with sorafenib
(REACH): a randomised, double-blind, multicentre, phase 3 trial. Lancet Oncol. 859–
870 (2015). 231. Llovet, J. M. & Hernandez-Gea, V. Hepatocellular carcinoma: reasons for phase III
failure and novel perspectives on trial design. Clin. Cancer Res. 20, 2072–9 (2014). 47 232. Lencioni, R. References & Llovet, J. Modified RECIST (mRECIST) Assessment for Hepatocellular
Carcinoma. Semin. Liver Dis. 30, 052–060 (2010). 233. Coulouarn, C., Factor, V. M. & Thorgeirsson, S. S. Transforming growth factor-beta
gene expression signature in mouse hepatocytes predicts clinical outcome in human
cancer. Hepatology 47, 2059–67 (2008). 234. Faivre, S. J., Santoro, A. & Kelley, R. K. A phase 2 study of a novel transforming growth
factor-beta (TGF-β 1) receptor I kinase inhibitor, LY2157299 monohydrate (LY), in
patients with advanced hepatocellular carcinoma. J. Clin. Oncol 32, LBA173 (2014). 235. Finn, R. S. Abstract 3858: Gains in FGF19 are predictive of response to the fibroblast
growth factor receptor (FGFR) small molecule tyrosine kinase inhibitor BGJ 398 in
vitro. Cancer Res. 72, 3858 (2012). 236. Hagel, M. et al. First Selective Small Molecule Inhibitor of FGFR4 for the Treatment of
Hepatocellular Carcinomas with an Activated FGFR4 Signaling Pathway. Cancer Discov. 5, 424–37 (2015). 237. Calvisi, D. F. et al. Ubiquitous Activation of Ras and Jak/Stat Pathways in Human HCC. Gastroenterology 130, 1117–1128 (2006). 238. O’Neil, B. H. et al. Phase II study of the mitogen-activated protein kinase 1/2 inhibitor
selumetinib in patients with advanced hepatocellular carcinoma. J. Clin. Oncol. 29,
2350–6 (2011). 239. Lim, H. Y. et al. A phase II study of the efficacy and safety of the combination therapy
of the MEK inhibitor refametinib (BAY 86-9766) plus sorafenib for Asian patients with
unresectable hepatocellular carcinoma. Clin. Cancer Res. 20, 5976–85 (2014). 240. Goyal, L., Muzumdar, M. D. & Zhu, A. X. Targeting the HGF/c-MET pathway in
hepatocellular carcinoma. Clin. Cancer Res. 19, 2310–8 (2013). 241. Xiang, Q. et al. Cabozantinib suppresses tumor growth and metastasis in
hepatocellular carcinoma by a dual blockade of VEGFR2 and MET. Clin. Cancer Res. 20,
2959–2970 (2014). 242. Robert, C. et al. Nivolumab in Previously Untreated Melanoma without BRAF
Mutation. N. Engl. J. Med. 372, 141116004513004 (2014). 243. Butterfield, L. H. et al. A phase I/II trial testing immunization of hepatocellular
carcinoma patients with dendritic cells pulsed with four alpha-fetoprotein peptides. Clin. Cancer Res. 12, 2817–25 (2006). 244. Palmer, D. H. et al. A phase II study of adoptive immunotherapy using dendritic cells
pulsed with tumor lysate in patients with hepatocellular carcinoma. Hepatology 49,
124–32 (2009). 48 245. Sangro, B. et al. A clinical trial of CTLA-4 blockade with tremelimumab in patients with
hepatocellular carcinoma and chronic hepatitis C. J. Hepatol. 59, 81–8 (2013). 246. El-Khoueiry, A. References et al. Phase I/II safety and antitumor activity of nivolumab in patients
with advanced hepatocellular carcinoma (HCC): CA209-040. J Clin Oncol 33, (suppl;
abstr LBA101) (2015). 247. Yeo, W. et al. Epigenetic therapy using belinostat for patients with unresectable
hepatocellular carcinoma: a multicenter phase I/II study with biomarker and
pharmacokinetic analysis of tumors from patients in the Mayo Phase II Consortium
and the Cancer Therapeutics Res. J. Clin. Oncol. 30, 3361–7 (2012). 248. Bitzer, M. et al. Efficacy, safety, tolerability, and PK of the HDAC inhibitor resminostat
in sorafenib-refractory hepatocellular carcinoma (HCC): Phase II SHELTER study. J Clin
Oncol 30, (Suppl;abst 4115) (2012). 249. https://clinicaltrials.gov/ct2/show/NCT01507168. A Study of RO5137382 (GC33) in
Patients With Advanced or Metastatic Hepatocellular Carcinoma. US Natl. Libr. Med. (2015). 250. Pao, W. et al. Acquired resistance of lung adenocarcinomas to gefitinib or erlotinib is
associated with a second mutation in the EGFR kinase domain. PLoS Med. 2, e73
(2005). 251. Van Allen, E. M. et al. The genetic landscape of clinical resistance to RAF inhibition in
metastatic melanoma. Cancer Discov. 4, 94–109 (2014). 252. Tovar, V. et al. Tumor initiating cells and IGF/FGF signaling contribute to sorafenib
resistance in hepatocellular carcinoma. Gut In press, (2015). 253. Chie, W.-C. et al. International cross-cultural field validation of an European
Organization for Research and Treatment of Cancer questionnaire module for patients
with primary liver cancer, the European Organization for Research and Treatment of
Cancer quality-of-life ques. Hepatology 55, 1122–9 (2012). 254. Blazeby, J. M. et al. Development of a questionnaire module to supplement the
EORTC QLQ-C30 to assess quality of life in patients with hepatocellular carcinoma, the
EORTC QLQ-HCC18. Eur. J. Cancer 40, 2439–44 (2004). 255. Huang, G. et al. Quality of life after surgical resection compared with radiofrequency
ablation for small hepatocellular carcinomas. Br. J. Surg. 101, 1006–15 (2014). 256. Salem, R. et al. Increased quality of life among hepatocellular carcinoma patients
treated with radioembolization, compared with chemoembolization. Clin. Gastroenterol. Hepatol. 11, 1358–1365.e1 (2013). 49 257. Afdhal, N. et al. Ledipasvir and sofosbuvir for untreated HCV genotype 1 infection. N. Engl. J. Med. 370, 1889–98 (2014). 258. Kahn, M. Can we safely target the WNT pathway? Nat. Rev. Drug Discov. 13, 513–32
(2014). 259. Roayaie, S. et al. The role of hepatic resection in the treatment of hepatocellular
cancer. Hepatology 62, 440–51 (2015). 260. Park, J.-W. et al. 270. Thress, K. S. et al. Acquired EGFR C797S mutation mediates resistance to AZD9291 in
non-small cell lung cancer harboring EGFR T790M. Nat. Med. 21, 560–2 (2015). Annotated References EASL-EORTC clinical practice guidelines: management of hepatocellular carcinoma. J. Hepatol. 56, 908–43 (2012). This article outlines the european consensus guidelines for the management for HCC,
including definition of treatment allocation criteria according to evidence Hepatol. 56, 908–43 (2012). This article outlines the european consensus guidelines for the management for HCC,
including definition of treatment allocation criteria according to evidence
Llovet, J. M. et al. Arterial embolisation or chemoembolisation versus symptomatic
treatment in patients with unresectable hepatocellular carcinoma: a randomised
controlled trial. Lancet 359, 1734–9 (2002). This paper reports a positive RCT supporting the use of TACE in intermediate HCC
Llovet, J. M. et al. Sorafenib in advanced hepatocellular carcinoma. N. Engl. J. Med. 359,
378–90 (2008). This paper details a positive RCT supporting the use of sorafenib in advanced HCC . It
is the first RCT demosntarting survival benefits for a systemic drug in HCC. The study
provided the rationale for approval of this drug in the management of HCC
Schulze, K. et al. Exome sequencing of hepatocellular carcinomas identifies new mutational
signatures and potential therapeutic targets. Nat. Genet. 47, (2015). This is the largest study reporting whole exome sequencing in a cohort of HCC
patients from the West
Totoki, Y. et al. Trans-ancestry mutational landscape of hepatocellular carcinoma genomes. Nat. Genet. 46, 1267–73 (2014). This is the largest study reporting whole exome sequencing in a cohort of HCC
patients from the Asia
Hoshida, Y. et al. Gene expression in fixed tissues and outcome in hepatocellular carcinoma. N Engl J Med 359, 1995–2004 (2008). This is the first manuscript to define the importance of the ‘cancer field effect’ in the
prognosis of HCC patients after resection
Chang, M. H. et al. Universal hepatitis B vaccination in Taiwan and the incidence of
hepatocellular carcinoma in children. Taiwan Childhood Hepatoma Study Group. N. Engl. J. Med. 336, 1855–9 (1997). This study demostrates the effect of universal HBV vaccination in decreasing the
incidence of HCC in Taiwan Llovet, J. M. et al. Arterial embolisation or chemoembolisation versus symptomatic
treatment in patients with unresectable hepatocellular carcinoma: a randomised
controlled trial. Lancet 359, 1734–9 (2002). This paper reports a positive RCT supporting the use of TACE in intermediate HCC Llovet, J. M. et al. Sorafenib in advanced hepatocellular carcinoma. N. Engl. J. Med. 359,
378–90 (2008). This paper details a positive RCT supporting the use of sorafenib in advanced HCC . References Global patterns of hepatocellular carcinoma management from
diagnosis to death: the BRIDGE Study. Liver Int. 35, 2155–2166 (2015). 261. Robert, C. et al. Pembrolizumab versus Ipilimumab in Advanced Melanoma. N. Engl. J. Med. 372, 2521–32 (2015). 262. Weinstein, I. B. Cancer. Addiction to oncogenes--the Achilles heal of cancer. Science
297, 63–4 (2002). 263. Shaw, A. T. et al. Crizotinib in ROS1 -Rearranged Non–Small-Cell Lung Cancer. N. Engl. J. Med. 371, 1963–71 (2014). 264. Nault, J.-C. & Villanueva, A. Intratumor molecular and phenotypic diversity in
hepatocellular carcinoma. Clin. Cancer Res. 21, 1786–8 (2015). 265. McGranahan, N. & Swanton, C. Biological and Therapeutic Impact of Intratumor
Heterogeneity in Cancer Evolution. Cancer Cell 27, 15–26 (2015). 266. Gridelli, C. et al. Non-small-cell lung cancer. Nat. Rev. Dis. Prim. 1, 1–16 (2015). 267. Tabernero, J. et al. Analysis of circulating DNA and protein biomarkers to predict the
clinical activity of regorafenib and assess prognosis in patients with metastatic
colorectal cancer: a retrospective, exploratory analysis of the CORRECT trial. Lancet
Oncol. 16, 937–48 (2015). 268. Crowley, E., Di Nicolantonio, F., Loupakis, F. & Bardelli, A. Liquid biopsy: monitoring
cancer-genetics in the blood. Nat. Rev. Clin. Oncol. 10, 472–84 (2013). 269. Thierry, A. R. et al. Clinical validation of the detection of KRAS and BRAF mutations
from circulating tumor DNA. Nat. Med. 20, 430–5 (2014). 270. Thress, K. S. et al. Acquired EGFR C797S mutation mediates resistance to AZD9291 in
non-small cell lung cancer harboring EGFR T790M. Nat. Med. 21, 560–2 (2015). 50 50 Annotated References Annotated References It
is the first RCT demosntarting survival benefits for a systemic drug in HCC. The study
provided the rationale for approval of this drug in the management of HCC Chang, M. H. et al. Universal hepatitis B vaccination in Taiwan and the incidence of
hepatocellular carcinoma in children. Taiwan Childhood Hepatoma Study Group. N. Engl. J. Med. 336, 1855–9 (1997). Thi
t d
d
t
t
th
ff
t
f
i
l HBV
i
ti
i
d
i
th Chang, M. H. et al. Universal hepatitis B vaccination in Taiwan and the incidence of
hepatocellular carcinoma in children. Taiwan Childhood Hepatoma Study Group. N. Engl. J. Med. 336, 1855–9 (1997). This study demostrates the effect of universal HBV vaccination in decreasing the
incidence of HCC in Taiwan This study demostrates the effect of universal HBV vaccination in decreasing the
incidence of HCC in Taiwan 51 Mazzaferro, V. et al. Liver transplantation for the treatment of small hepatocellular
carcinomas in patients with cirrhosis. N. Engl. J. Med. 334, 693–9 (1996). This paper defines the ‘Milan criteria’ for selection of candidates for liver
transplantation. These criteria are currrently adopted by most transplant units
globally Lo, C.-M. et al. Randomized controlled trial of transarterial lipiodol chemoembolization for
unresectable hepatocellular carcinoma. Hepatology 35, 1164–71 (2002). This RCT demosntrated the efficacy of TACE in patients with intermediate HCC This RCT demosntrated the efficacy of TACE in patients with intermediate HCC Clavien, P.-A. et al. Recommendations for liver transplantation for hepatocellular carcinoma:
an international consensus conference report. Lancet. Oncol. 13, e11–22 (2012). This publication outlines consensus guidelines for the management of HCC with liver
transplantation Llovet, J. M. et al. Design and endpoints of clinical trials in hepatocellular carcinoma. J. Natl. Cancer Inst. 100, 698–711 (2008). This paper describes consensus guidelines of the design of clinical trials in HCC Llovet, J. M. et al. Design and endpoints of clinical trials in hepatocellular carcinoma. J. Natl. Cancer Inst. 100, 698–711 (2008). This paper describes consensus guidelines of the design of clinical trials in HCC 52
|
https://openalex.org/W4225143669
|
https://zenodo.org/records/8001294/files/A69490511122.pdf
|
English
| null |
Comparative Analysis of Routing Algorithms to Enhance the Quality of Service in MANET
|
International journal of recent technology and engineering
| 2,022
|
cc-by
| 6,841
|
Comparative Analysis of Routing Algorithms to
Enhance the Quality of Service in MANET K. Vimala, D. Maruthanayagam Abstract: Mobile Ad hoc Network is a self-organizing,
infrastructure-free, distributed wireless networks made up of
various mobile devices. Quality-of-Service routing is most
difficult task in MANET due to inherent characteristics–for
example frequent dynamic topology, node mobility, resource
scarcity, absence of centralized control, etc as well. The QoS
variables of any MANET routing algorithms determine its
performance. QoS routing is process of routing packets from
source (S) to destination (D) based on QoS resource constraints
such as bandwidth, delay, packet loss rate, cost, security, link
stability, and so on. Swarm intelligence, which mimics the
collective behaviour of biological organisms to handle routing
problems and improve QoS in the network, has been one of most
popular studies for network routing in recent years. Particle
Swarm Optimization algorithm (PSO), Genetic Algorithm (GA)
and Ant Colony Optimization algorithm (ACO) have all been
shown to be effective for developing routing algorithms by
improving QoS metrics in ad hoc networks using Swarm
Intelligence (SI). The primary objective of this comparative study
paper is to improve QoS parameters by applying swarm
intelligence to MANET routing algorithms. Swarm intelligence-
based routing algorithms will be more promising for the specific
nature
of
adhoc
networks,
outperforming
in
real
scenarios/constraints/environmental conditions and will be tuned
and simulated to obtain an efficient and effective MANET
routing protocol. This paper investigates four potential pre-
existing approaches proposed for MANET routing problems. These
routing
algorithms
are
evaluated
using
various
performance metrics such as packet delivery ratio, routing
overhead, link failure prevention, energy consumption, accuracy,
and throughput, among others. MANET is quick and simple to organizing, making them
ideal for real time applications such as environmental
monitoring, military surveillance and etc.,[3]. In MANETs,
routing for QoS is critical. In further to finding a best route
from a source (S) to destination (D), QoS routing must
ensure end user quality, usually respecting bandwidth or
delay [4].To develop a secure and more efficient routing
algorithms, that algorithm can also improve overall quality
of network services throughput the routing procedure is a
more challenge task for MANETs, because MANET mobile
nodes communicate with other nodes only they are inside
the communication range. International Journal of Recent Technology and Engineering (IJRTE)
ISSN: 2277-3878 (Online), Volume-11 Issue-1, May 2022 International Journal of Recent Technology and Engineering (IJRTE)
ISSN: 2277-3878 (Online), Volume-11 Issue-1, May 2022 International Journal of Recent Technology and Engineering (IJRTE)
ISSN: 2277-3878 (Online), Volume-11 Issue-1, May 2022 International Journal of Recent Technology and Engineering (IJRTE)
ISSN: 2277-3878 (Online), Volume-11 Issue-1, May 2022 Comparative Analysis of Routing Algorithms to
Enhance the Quality of Service in MANET When the receiver mobile node is
outlying from the transmitter range, the dynamic behaviour
of MANET makes ensuring QoS difficult because the node
to node connection and link quality change dynamically,
resulting in frequent link failures then affecting nodes to
connect with other nodes [5]. Another critical important
issue in MANET is security, because malicious nodes can
purpose fully misbehave, altering packet information and
disrupting packet routing to preferred destinations, dropping
packet delivery percentage and reliability. Many MANET
routing algorithms, for example (Dynamic Source Routing
(DSR), Ad hoc On Demand Distance Vector (AODV), than
Cluster Based Routing (CBRP) already been designed to
address link failure and security issues in MANETs. Most
solutions, however, are principal effective for a specific type
of attack or provide better security at the expense of
sacrificing QoS. Until now, no research effort has been able
to provide the optimum routing efficiency to ensure high
network performance. Because of their self-organizing
nature, SI-inspired routing systems have become a research
focus in recent years [6], which is very appropriate to
routing problems and QoS-related concerns in Mobile Ad
hoc Networks (MANETs). Inspired by natural biological
swarms, swarm intelligence possesses a number of potent
properties for distributed problem solving in more complex
real world applications like optimization and control. Natural systems with swarm intelligence assets include ants,
bees, and birds, in which unsophisticated agents interact
locally with their atmosphere to explore collective problem
solving without centralised control. This paper describes and
analyses the performance of some existing routing
algorithms in order to apply swarm intelligence to routing
algorithms used in MANETs to improve QoS. Keywords: MANET, Routing, Link failure, Security, Swarm
Intelligence and Quality of Services. I. INTRODUCTION MANET has received a lot of attention in latest years,
owing to the general use of cheaper and more powerful
wireless networks [1]. It’s composed a collection of wireless
mobile nodes, mobile nodes are communicated with each
other through one or more connections it does not have any
centralized authority [2]. Each mobile node in MANET can
acts as both a terminal node and router, which means each
mobile node can make its own traffic though also receiving
packets from other mobile nodes and sending them to
neighbouring nodes. Manuscript received on 18 April 2022. Revised Manuscript received on 28 April 2022. Manuscript published on 30 May 2022. * Correspondence Author
K Vi
l * R
h S h l
S i Vij
Vid
l
C ll
f A
& Manuscript received on 18 April 2022. * Correspondence Author The
QoS is improved by routing algorithms using to find an
alternative path or repair local route failure when the link
failure time is predicted in advance. The literature discusses
various routing techniques for estimating link failure
estimated time that are based on statistical, probabilistic, and
trigonometric approaches. Furthermore, security is an
important QoS factor; however, the network security
requirements for QoS provisioning in MANET networks
have received little attention, which is a major issue. A
secure QoS aware scheme can be used in a variety of real
world applications, for example defending any critical
network against DoS than fabrication attacks. Data integrity,
timely information deliveries are the primary requirements
of patient monitoring systems. Recent advancements in
MANET research have resulted in methods for mitigating a
wide variety of malicious attacks [8]. Earlier research
focused on using a variety of security schemes to ensure
secure routing performance for Ad hoc. To prevent
malicious nodes from entering the network, vast majority of
routing
approaches
employ
key
management
or
cryptography techniques [9]. However, these security
techniques suffer from a general drawback on key
verification process than key exchange, which results in
high network traffic. This leads to battery and computational
capacity limitations, resulting in bandwidth constraints for
MANET nodes and expensive schemes. A bio-inspired
model is a novel field influenced by social behaviour. It
primarily solves network problems due to its unique natural
features and bottom to up approach. Because MANET is a
distributed network system, researchers are investigating
network best route analysis using bio-inspired methods. Bio- distributed and autonomous manner in dynamically
changing environments; for that reason, they can maximise
routing performance, reduce control overhead, and quickly
recover from a path failure due to modifications in network
topology [10]. Bio-inspired routing techniques have the
following advantages over traditional MANET routing
protocols: First, because a swarm of ants or bees finds the
optimal route to food in a distributed manner in dynamic
environments, a routing protocol inspired by swarm
behaviour is suitable for autonomously finding the shortest
path to the destination in MANETs. Second, because swarm
intelligence maintains the optimal path flexibly in a resource
limited environment, the bio-inspired routing protocol can
efficiently maintain the shortest path while reducing the
occurrence of overhead. Third, because the swarm
effectively solves various unpredictable problems in the real
world, the bio-inspired routing protocol can quickly repair a
link or node failure [6]. * Correspondence Author K. Vimala*, Research Scholar, Sri Vijay Vidyalaya College of Arts &
Science, Dharmapuri (Tamil Nadu), India. QoS in MANETs, according to RFC 2386, is an observance
requirement that must be met by the MANET network while
transporting a data packet flow from the source (S) to its
destination (D) [7]. p
Dr. D. Maruthanayagam, Professor & Head, PG and Research,
Department of Computer Science, Sri Vijay Vidyalaya College of Arts &
Science, Dharmapuri (Tamil Nadu), India. 108
Published By:
Blue Eyes Intelligence Engineering
and Sciences Publication (BEIESP)
© Copyright: All rights reserved. destination (D) [7]. © The Authors. Published by Blue Eyes Intelligence Engineering and
Sciences Publication (BEIESP). This is an open access article under the
CC-BY-NC-ND license http://creativecommons.org/licenses/by-nc-nd/4.0/ 108
Published By:
Blue Eyes Intelligence Engineering
and Sciences Publication (BEIESP)
© Copyright: All rights reserved. Retrieval Number: 100.1/ijrte.A69490511122
DOI: 10.35940/ijrte.A6949.0511122
Journal Website: www.ijrte.org Retrieval Number: 100.1/ijrte.A69490511122
DOI: 10.35940/ijrte.A6949.0511122
Journal Website: www.ijrte.org 108 omparative Analysis of Routing Algorithms to Enhance the Quality of Service in MANET Regards of end-to-end performance, the ad hoc network is
expected to provide users with significant pre specified
service attributes like bandwidth requirement, data packet
damage probability, latency or jitter variation, route
acquisition delay, communication overhead, and so on. Finding a feasible route that must satisfies the QoS
parameters of a given MANET application is one of the key
issues in QoS routing [7]. However, due to node mobility,
limited resource constraints, multi-hop communications,
channel access contention, and a lack of central
coordination, QoS routing in MANETs is extremely
difficult. The main QoS issues are link failures, which
reduce
overall
network
performance;
mobile
node
movement in network changes the topology, resulting in
frequent connection failures. This causes a slew of
problems, including data packet loss than packet delay, both
are degrade wireless network performance. Link failures
between nodes occur when the network structure changes by
a no of reasons, including channel interference. One
important criterion that will determine service quality is link
failure recovery (QoS). Path breaks between neighbouring
nodes may occur as a result of the wireless environment's
instability. The main causes of unintentional node failure in
MANETs are channel quality, link quality, and node energy
level. If any established link fails during data transmission,
many researchers solve the problems by repaired the route
locally or discover a new route, this procedure degrade the
performance of network in existing routing algorithms. * Correspondence Author Several approaches proposed
various routing algorithms with QoS support for MANETs
over last decade. However, none of them are sufficient to
incorporate all of the available QoS factors for efficient
packet routing. The provisioning of QoS is well studied in
ad hoc, with the main goal of the research being to improve
the QoS factors by recovering link failure than detecting
malicious nodes. The following section looks at some of the
most recent existing MANET routing algorithms. A. CBMNDM
(Clusters
Based
Malicious
Nodes
Detection Methodology in MANET) A. CBMNDM
(Clusters
Based
Malicious
Nodes
Detection Methodology in MANET) For
detecting
and
removing
malicious
nodes,
S. Gopalakrishnan et al [11] proposed a CBMNDM. A cluster
key is assigned by cluster head for each node, than the
assigned key is used for packet data transmission between
the node and cluster head. The CH (cluster head) examines
each data transaction from a node to see if it matches the
key in their cluster table (CT). If the node is identity is valid
the node is recognise as a member of this cluster; otherwise
the node determined as a malicious node. In this paper, the
author also discussed link failure occurs by the malicious
node in MANET and gain of each link. The performance of
the proposed method was assessed using the PDR, network
life time, and energy consumption. Retrieval Number: 100.1/ijrte.A69490511122
DOI: 10.35940/ijrte.A6949.0511122
Journal Website: www.ijrte.org C.
Proficient Trusted Node ID Based Resource
Reservation Protocol (PT-NIDBRRP) The PT-NIDBRRP was proposed by G. Jegan, a researcher
[13]. Using a weighted end delay based approach, shortest
path was found here. Proposed algorithm will find shortest
path from starting point to target, potentially increasing
detection rate. Sequence numeral and hop address are added
to algorithm when route detection process is started. After
determining shortest path, failed link was discovered. Link
failure localization structure of employed trusted algorithm
is better at predicting and resolving link failure issues. Posterior probability assessment was then used to determine
the type of attack that caused the link failure. Figure 2: Link Failure Detection Fi
2 Li k F il
D
i 2. Detection of Node Link Faults Author used a link cost algorithm to detect link failure. For
example, the author used three primary source (S) nodes: s1,
s2, and s3. All three source nodes (S) send this packet data to
the sink node. The Node s1 connects to the sink node via
links l2 and l1. Node s2 connects to the sink via links l3 and
l1. Node s3 connects to the sink node via links l4 and l1. Node
s3 may also send data to sink node via link l5. There is a
single unit link cost for each network link. Assume that links
l1 and l5 are both lossy. The link cost algorithm, which is
described below, can be used to find these lossy links: C. Proficient Trusted Node ID Based Resource
Reservation Protocol (PT-NIDBRRP) B.
Improved Failure Aware Third Party Auditor
(IFTPA) Based Homomorphism Linear Authenticator
(HLA) Mechanism (IFHM) B. Improved Failure Aware Third Party Auditor
(IFTPA) Based Homomorphism Linear Authenticator
(HLA) Mechanism (IFHM) Researchers K. Vanitha., et al [12] proposed an (IFTPA)
based HLA technique IFHM with SAODV to overcome the
problems and have a well-organized malicious mobile node
detection process than provide better performances. The
proposed method used the trust verification method for data
packet damage information and detects malicious node this
process are done by using dropping routing, packet
identification. In SAODV data packet dropping occurred by
link failure or malicious nodes. 1. To determine shortest path between nodes, the weighed end
delay was used. In this case, author relied on the expert
trusted Node ID-based (RRP) resource reservation protocol. Proposed weighed end delay method calculated shortest path
by determining weightage length of path from the source to
each node. Weighed path is shortest path with no link
breaks. Following the calculation of the path, information
data packets can be sent directly to the destination mobile
node. It is a time and energy saving method. 1. Clusters Based Malicious Nodes Detection Clusters of nodes are formed by dividing the mobile nodes
into smaller regions. Each cluster is led by a (CH) cluster
head, which is in charge of controlling all nodes within their
capacity. MANET keeps several CHs that are all connected
to sink. The packet can be sent directly to sink or through
other Cluster Heads by a single Cluster Head. Every CH
keeps a CT (cluster table), and every mobile node keeps a
NT (neighbour table). The CT includes information about all
mobile nodes, distance between the CH and each node
within the cluster, and the CK (cluster key). The CH assigns
a cluster key to each mobile node in the cluster, which is
used for packet data transactions between the CH and the
node. The CH examines each data transaction from a node
to see if it matches a key in their cluster table. Published By:
Blue Eyes Intelligence Engineering
and Sciences Publication (BEIESP)
© Copyright: All rights reserved. Published By:
Blue Eyes Intelligence Engineering
and Sciences Publication (BEIESP)
© Copyright: All rights reserved. Retrieval Number: 100.1/ijrte.A69490511122
DOI: 10.35940/ijrte.A6949.0511122
Journal Website: www.ijrte.org Retrieval Number: 100.1/ijrte.A69490511122
DOI: 10.35940/ijrte.A6949.0511122
Journal Website: www.ijrte.org 109 International Journal of Recent Technology and Engineering (IJRTE)
ISSN: 2277-3878 (Online), Volume-11 Issue-1, May 2022 International Journal of Recent Technology and Engineering (IJRTE)
ISSN: 2277-3878 (Online), Volume-11 Issue-1, May 2022 International Journal of Recent Technology and Engineering (IJRTE)
ISSN: 2277-3878 (Online), Volume-11 Issue-1, May 2022 enables data encryption by utilising the secret key
generating process, which is an arbitrary producing
procedure. The proposed algorithm is based on FTPA, and
HLA is a metadata and data integrity verification process
(FTPA). As a result, this method is referred to as FHM, than
security is referred to as IFHM. Detection process identifies
malicious nodes by correlating lost packets. Following
transmission, loss occurrences are identified, and the reason
for dropping is specified. As a result of this process, the
effects will not be visible in the next communication. Figure 3: Shortest Path Prediction After calculating distance between each node, the nodes
must be searched for a link breakage. Because a broken link
allows data packets to return to their origin and the process
to resume. As a result, the protocol used here was simply
modified to find the link breakage, and the author developed
a capable trusted Node ID-based RRP for easy prediction of
link breakage. The author's primary goal was to locate the malicious node
and the network failure link. For verification and linear
process homomorphism, the HLA mechanism was used, and
for authentication, the FTPA was used. By combining both
techniques, the security process was defined as an FTPA
based HLA technique IFHM. The FHM method was
developed than clearly explained during previous phase of
the author's work. Thereby, the proposed research improves
the security and the pseudo random function (PRF), these
are included in FHM. The PRF process was used to
implement the privacy-preserving-public-auditing scheme
and to integrate the method as a dynamic and linear process
of generating the method in wireless network. In this case, a
random process's PRF generator is deterministic than
efficient, returning the sequence based on received input. It 1.
Shortest path selection by weighed end delay
technique 1. Shortest path selection by weighed end delay
technique Retrieval Number: 100.1/ijrte.A69490511122
DOI: 10.35940/ijrte.A6949.0511122
Journal Website: www.ijrte.org Path selection, Route update Density of pheromones along the route and probability of
periodinterval between the paths determine the path chosen
from sender (x) to receiver (y). The time interval indirectly
indicates distance of path. Thereby, one of major factors
determines best network path. Forward ANTs typically
move from sources in all directions around the node. The
deposit of forward ANTs is pheromone, than it updates
pheromone level in each mobile node. Multiple paths from
sender to receiver are generated by the ant colony concept. To determine the best path to take, possibility of shortest
paths and pheromone deviation on path are used. Assume that Baye's theorem applied to calculate variable's
probability distribution. It is computed by multiplying
probability distribution values by normalising constant. E.
Trusted path Selection The three factors such as average deviation of battery level,
shortest route path, and lowest hop count between mobile
nodes are used in this proposed method. Best path equation
probability is regularly used to determine the shortest path
(1).When compared to other direction route and battery
deviation the trust path is selected by smaller no of hops. Both hop and battery deviation conditions provide a secure
communication path while reducing packet drops and selfish
attacks caused by vulnerable network nodes. Trust path's
conditions are outlined below. The posterior probability can be directly proportional to
likelihood, which is a multiple of previous probability. Where P(t)xy represents best path from (x) to(y), p(t)xy
represents probability of shortest paths from (x) to(y), than
qxy represents pheromone deviation. Uncertainty malicious nodes are dropped the data packets,
the ant agent forwards the data packets that were data
dropped by the vulnerable mobile node, and the reduced
pheromone density is set to zero, causing a time delay
increase. Nonetheless, they only forward a percentage of
data packets to next hop, and the pheromone density
between two adjacent nodes is reduced compare to a level
less than another path density level. Thereby, malicious
nodes within network can be easily identified. The likelihood task was essentially used to map out
function's
likelihood. After
determining
previous
probability, attack was determined. D.
Trust Path Ant Colony Optimization (TPACO) The path of Trust S. Sugumaran's (TPACO) algorithm [14]
is a swarm intelligent scheme for selecting a trust path from
multiple routes between source (S) and destination (D). Proposed algorithm controls the Data Packet dropping
occurrence and consent in network by serving as best route. TPACO algorithm, trust mobile node from sender (S) to
destination (D) employed with ant colony algorithm, it
consists of three steps: (1) discovered potential path, (2)
path selection and updating, than (3) trust path selection. In
this case, Trust value is calculated by choosing a probable
path between the source (S) and destination (D), counting
the no of hops, and calculating node's battery deviation in
percentage. 2.
Attack Estimation The attack will be determined using the posterior possibility
estimation technique. Published By:
Blue Eyes Intelligence Engineering
and Sciences Publication (BEIESP)
© Copyright: All rights reserved. Published By:
Blue Eyes Intelligence Engineering
and Sciences Publication (BEIESP)
© Copyright: All rights reserved. Retrieval Number: 100.1/ijrte.A69490511122
DOI: 10.35940/ijrte.A6949.0511122
Journal Website: www.ijrte.org Retrieval Number: 100.1/ijrte.A69490511122
DOI: 10.35940/ijrte.A6949.0511122
Journal Website: www.ijrte.org 110 omparative Analysis of Routing Algorithms to Enhance the Quality of Service in MANET This technique a probability distribution method for
determining the unidentified prediction of a value, thanthe
process is dependent on arbitrary variable selection. Main
goal is for it to look for a related experience and arbitrarily
observe the relevant data. To achieve goal, it is essential to
provide a consistent intervals. Basically put, it can observe
the member's possibility and reflect uncertainty value. level in every node in network. If retraced path fails to reach
sources on time due to a lengthy distance, a broken link, or a
malicious node, pheromone density is fixed to zero after a
time interval, than updated details are stored in the node
register. Retrieval Number: 100.1/ijrte.A69490511122
DOI: 10.35940/ijrte.A6949.0511122
Journal Website: www.ijrte.org 3.1 Simulation Environment The delivery ratios of the CBMNDM, IFHM, PT-
NIDBRRP, and TPACO methods are shown in Figure 4. When compared to other methods, the TPACO method had
a higher packet delivery ratio and higher packet
transmission efficiency. TPACO routing could detect
congestion quickly and easily find the shortest route. The
TPACO protocol was able to detect link breaks easily during
the route recovery process prior to data transfer, resulting in
gradual data loss reduction and significantly improved
packet delivery rate. UC Berkeley is working on a discrete even driven simulator
called NS2 (Network Simulator). It is a component of the
VINT project. NS2 simulation software objective is to
support networking investigation and learning. NS2 is an
excellent tool for designing different routing algorithms,
comparing different protocol algorithms, and evaluating
traffic. It has a reputation for being a collaborative
environment. It is freely distributed and open source. NS2 is
used, maintained, and developed by a large number of
organisations and individuals involved in development and
research. The simulation environment consists of 50
wireless mobile nodes that form an evenly distributed
mobile ad hoc network that travels over a 1500 x 1500 metre
area in 500 seconds. The network's mobile nodes are all set
up to use pre-existing techniques like CBMNDM, IFHM,
PTNIDBRRP, and TPACO. Figure 4: No of Nodes vs. PDR Table 1: Simulation Environment Table 1: Simulation Environment Table 1: Simulation Environment Figure 4: No of Nodes vs. PDR 1. and the battery deviation of each wireless mobile node is
lower than fifty present, secure communications is expected. Trusted mobile nodes provides efficient than possible data
packet transmission paths. TPACO optimization algorithm
is using to implement route protection techniques. The
deposit pheromone of choose path was assigned the greatest
route for data packet transmission. Where the source
chooses the shortest route and other routes are saved for
future use. MANET enables a larger number of nodes to
enter and exit the network. A transmission delay in this case
collapses the wireless network route and depletes node
energy. Above this problem is avoided by employing the
frequent path selection probability method and choosing the
route from wireless network periodicals. ▪
Packet Delivery Ratio (PDR): The percentage of data
packets received by the destination (D) versus data
packets sent by the source (S). ▪
Packet Delivery Ratio (PDR): The percentage of data
packets received by the destination (D) versus data
packets sent by the source (S). p
y
▪
End-to-End Delay (ETED): End Delay is the quantity
of time it takes to send packets from a source (S) to a
destination (D). ▪
End-to-End Delay (ETED): End Delay is the quantity
of time it takes to send packets from a source (S) to a
destination (D). ▪
Routing Overhead (RO): Overload in the network is
proportional to latency and occurs as a result of
frequent node movement, network congestion, or high
traffic. ▪
Routing Overhead (RO): Overload in the network is
proportional to latency and occurs as a result of
frequent node movement, network congestion, or high
traffic. ▪
Link break prediction: It identifies breakpoints in the
overall route. ▪
Link break prediction: It identifies breakpoints in the
overall route. ▪
Residual Energy (RE): The remaining energy of the
sensor node can be calculated by adding up the energy
consumed while the node is in each state. ▪
Residual Energy (RE): The remaining energy of the
sensor node can be calculated by adding up the energy
consumed while the node is in each state. 2. Routing Overhead Figure 5 compares the performance of the CBMNDM,
IFHM, PT-NIDBRRP, and TPACO methods based on their
overhead. Despite a slight increase in TPACO, packet
receiving efficiency has improved. 3.2 Performance Parameters 1. Because sources (S) do not have a route to destination (D) at
first, the ant's agent initiates the source requested message
and distributes it to all the neighbouring nodes in the
network. When the Forward ANT reaches a neighbour node
that is not a destination node, it proceeds to the next hop
with updated information about the neighbour node. Otherwise, Destination destroys Forward ANT requests and
provides Backward ANT replay to sources. The Ant
pheromone density is being deposited on paths at this time. If the pheromone value is less than the threshold value, the
path is unsuitable for data transmission than should be
terminated. The density of pheromones between the source
and destination, as well as each adjacent node, determines
path connectivity. If the Ants arrive at their destination, they
must retrace their path to sources and update pheromone In this case, percentage of node battery of their deviations is
shown by below equation, Hop count, Pheromone Deviation, and node battery
deviation of each mobile node in wireless network is
described
using
secure
communication. When
the
pheromone deviation is lower than forty present the hop
count from the total mobile node in the ad hoc network is
lower than the threshold, Published By:
Blue Eyes Intelligence Engineering
and Sciences Publication (BEIESP)
© Copyright: All rights reserved. , Published By:
Blue Eyes Intelligence Engineering
and Sciences Publication (BEIESP)
© Copyright: All rights reserved. Retrieval Number: 100.1/ijrte.A69490511122
DOI: 10.35940/ijrte.A6949.0511122
Journal Website: www.ijrte.org Retrieval Number: 100.1/ijrte.A69490511122
DOI: 10.35940/ijrte.A6949.0511122
Journal Website: www.ijrte.org 111 International Journal of Recent Technology and Engineering (IJRTE)
ISSN: 2277-3878 (Online), Volume-11 Issue-1, May 2022 and the battery deviation of each wireless mobile node is
lower than fifty present, secure communications is expected. Trusted mobile nodes provides efficient than possible data
packet transmission paths. TPACO optimization algorithm
is using to implement route protection techniques. The
deposit pheromone of choose path was assigned the greatest
route for data packet transmission. Where the source
chooses the shortest route and other routes are saved for
future use. MANET enables a larger number of nodes to
enter and exit the network. A transmission delay in this case
collapses the wireless network route and depletes node
energy. Above this problem is avoided by employing the
frequent path selection probability method and choosing the
route from wireless network periodicals. 5. Residual Energy Figure 8 depicts the variation of energy as the amount of
nodes increases. The remaining energy drops as the number
of nodes increases to 10, 20, 30, and 50. The results show
that the number of nodes has a negative effect on the
remaining energy for all protocols. The TPACO technique
is shown to conserve (save) energy more than other
protocols because it establishes an ideal route from the
source (S) to the destination (D) node. As a result, this path
has the highest energy level; taking this approach reduces
the likelihood of exhausting nodes. 113
Published By:
Blue Eyes Intelligence Engineering
and Sciences Publication (BEIESP)
© Copyright: All rights reserved. Figure 8: Residual Energy Figure 6: Node Mobility vs. End-to-End delay Fi
8 R
id
l E 3.2 Performance Parameters The TPACO technique
is shown to conserve (save) energy more than other
protocols because it establishes an ideal route from the
source (S) to the destination (D) node. As a result, this path
has the highest energy level; taking this approach reduces
the likelihood of exhausting nodes. omparative Analysis of Routing Algorithms to Enhance the Quality of Service in MANET Figure 5: No of Nodes vs. Routing Overhead (a) Figure 5: No of Nodes vs. Routing Overhead (b)
Figure 7: (a) Effect of mobility on throughput, (b) Effect
of pause time on throughput
5. Residual Energy 3.2 Performance Parameters Based on the performance metrics listed below, evaluate the
performance of existing routing algorithms. The routing balancing of the network's load is causing the
overhead promenade. According to the graph, the routing
overhead of TPACO is less than that of CBMNDM, IFHM,
and PT-NIDBRRP as the network size increases. ▪
Throughput: Defined as the amount of data packets
transported over the course of the simulation. Routing
algorithm proficiency is measured by throughput when
receiving packets per destination (D). The following
formula is used to calculate throughput: Throughput =
number of delivered packets multiplied by packet size
multiplied by 8 bits divided by total simulation time. ▪
Throughput: Defined as the amount of data packets
transported over the course of the simulation. Routing
algorithm proficiency is measured by throughput when
receiving packets per destination (D). The following
formula is used to calculate throughput: Throughput =
number of delivered packets multiplied by packet size
multiplied by 8 bits divided by total simulation time. 112
Published By:
Blue Eyes Intelligence Engineering
and Sciences Publication (BEIESP)
© Copyright: All rights reserved. g
=
e 112
Published By:
Blue Eyes Intelligence Engineering
and Sciences Publication (BEIESP)
© Copyright: All rights reserved. Retrieval Number: 100.1/ijrte.A69490511122
DOI: 10.35940/ijrte.A6949.0511122
Journal Website: www.ijrte.org Retrieval Number: 100.1/ijrte.A69490511122
DOI: 10.35940/ijrte.A6949.0511122
Journal Website: www.ijrte.org Comparative Analysis of Routing Algorithms to Enhance the Quality of Service in MANET
Figure 5: No of Nodes vs. Routing Overhead
3. End to End Delay
The outcome of network mobility of the CBMNDM, IFHM,
PT-NIDBRRP, and TPACO algorithms are depicted by
Figure 6.The end delay increases as node mobility rises
from 10 to 100 metres per second. Higher amount of
mobility causes more links halts and frequent redirecting,
resulting in longer end delays. The broken links may require another route recovery and
route discovery progression. In every mobility condition,
TPACO outperforms the competition. Figure 6: Node Mobility vs. End-to-End delay
4. Throughput
Figures 7(a) and 7(b) show how mobility and pause time
(a)
(b)
Figure 7: (a) Effect of mobility on throughput, (b) Effect
of pause time on throughput
5. Residual Energy
Figure 8 depicts the variation of energy as the amount of
nodes increases. The remaining energy drops as the number
of nodes increases to 10, 20, 30, and 50. The results show
that the number of nodes has a negative effect on the
remaining energy for all protocols. 3. End to End Delay The outcome of network mobility of the CBMNDM, IFHM,
PT-NIDBRRP, and TPACO algorithms are depicted by
Figure 6.The end delay increases as node mobility rises
from 10 to 100 metres per second. Higher amount of
mobility causes more links halts and frequent redirecting,
resulting in longer end delays. The broken links may require another route recovery and
route discovery progression. In every mobility condition,
TPACO outperforms the competition. Figure 6: Node Mobility vs. End-to-End delay (b)
Figure 7: (a) Effect of mobility on throughput, (b) Effect
of pause time on throughput (b)
Figure 7: (a) Effect of mobility on throughput, (b) Effect
of pause time on throughput
5 R
id
l E (b) Retrieval Number: 100.1/ijrte.A69490511122
DOI: 10.35940/ijrte.A6949.0511122
Journal Website: www.ijrte.org 7. Accuracy Prediction The precision of the CBMNDM, IFHM, PT-NIDBRRP, and
TPACO methods is depicted in Figure 10. Because the type
of attack can be accurately predicted, the TPACO can
acquire (obtain) greater accuracy than other existing
methods. The attack in this case was determined to be a
passive DoS attack, which was difficult to predict but was
made possible by the implementation of the probability
distribution method. In this experiment the accuracy
prediction was determined by passive DoS attack, which
attack was more difficult to predict. TPACO algorithms
predict DoS attack efficient than other algorithms and also
predict link breakage, its provide better results secure data
communication. IV. CONCLUSION In this paper various routing algorithms for mobile ad hoc
networks are compared, including one inspired by bio-
inspired. The simulation results concluded the swarm
intelligence-based algorithm performs better than others in
different types of scenarios involving, for example, data
loads, network mobility and network density. In terms of
packet delivery ratio, end-to-end delay, throughput, and
other quality of service metrics, TPACO algorithms
outperform CBMNDM, IFHM, and PT-NIDBRRP routing
protocols. TPACO, which is based on swarm intelligence, is
found to outperform the competition. REFERENCES 1. DevarajanJinilErsis, T.Paul Robert, “Review of ad-hoc on-demand
distance vector protocol and its swarm intelligent variants for Mobile
Ad-hoc NETwork”, The Institution of Engineering and Technology
2017. 1. DevarajanJinilErsis, T.Paul Robert, “Review of ad-hoc on-demand
distance vector protocol and its swarm intelligent variants for Mobile
Ad-hoc NETwork”, The Institution of Engineering and Technology
2017. Figure 10: No of Nodes vs. Accuracy 2. P. Vijayalakshmi, S. A. J. Francis, and J. A. Dinakaran, ``A robust
energy efficient ant colony optimization routing algorithm for multi-
hop ad hoc networks in MANETs,'' Wireless Netw., vol. 22, no. 6, pp. 1_20, 2015. 3. Malathi, M., Jayashri, S. “Robust against route failure using power
proficient reliable routing in MANET”. AEJ J. 2016. [CrossRef] 4. Chavhan, S.; Venkataram, P. “Emergent Intelligence Based QoS
Routing in MANET”. ProcediaComput. Sci. 2015, 52, 659–664. [CrossRef] 5. S. Kalaivanan, “Quality of service (QoS) and priority aware models for
energy efficient and demand routing procedure in mobile ad hoc
networks,” Journal of Ambient Intelligence and Humanized
Computing, vol. 12, no. 3, pp. 4019–4026, 2021. p
g
pp
6. R. T. Merlin and R. Ravi, “Novel trust based energy aware routing
mechanism for mitigation of black hole attacks in MANET,” Wireless
Personal Communications, vol. 104, no. 4, pp. 1599–1636, 2019. 7. Calarany, C., &Manoharan, R. (2019). “An Efficient Evaluation of Bi-
Objective Optimization of Path Stability Model for Mobile Ad Hoc
Networks Using EN2RP”. Journal of Computational and Theoretical
Nanoscience, 16(4), 1454-1464. Figure 10: No of Nodes vs. Accuracy 4. Throughput Figures 7(a) and 7(b) show how mobility and pause time
affect throughput for the protocols CBMNDM, IFHM, PT-
NIDBRRP, and TPACO. 7(a)-During high mobility,
methods such as CBMNDM, IFHM, and PT-NIDBRRP
exhibit a small degradation in throughput due to high-link
breakage, whereas. TPACO exhibits a 3% improvement in throughput over
other algorithms. Figure 7 (b) shows that there is no
difference between algorithms; however, as the pause time
increases, the difference between those throughputs grows
significantly. Higher throughput is attributed to increased
network longevity as a result of proper transmission power
adjustment during the link selection phase and energy
savings during the sleep scheduling phase. Figure 8: Residual Energy Figure 8: Residual Energy 113
Published By:
Blue Eyes Intelligence Engineering
and Sciences Publication (BEIESP)
© Copyright: All rights reserved. reased
power
nergy
Figure 8: Residual Energy 113
Published By:
Blue Eyes Intelligence Engineering
and Sciences Publication (BEIESP)
© Copyright: All rights reserved. d
r
y
g
gy 113
Published By:
Blue Eyes Intelligence Engineering
and Sciences Publication (BEIESP)
© Copyright: All rights reserved. 113
Published By:
Blue Eyes Intelligence Engineering
and Sciences Publication (BEIESP)
© Copyright: All rights reserved. Retrieval Number: 100.1/ijrte.A69490511122
DOI: 10.35940/ijrte.A6949.0511122
Journal Website: www.ijrte.org 113 International Journal of Recent Technology and Engineering (IJRTE)
ISSN: 2277-3878 (Online), Volume-11 Issue-1, May 2022 International Journal of Recent Technology and Engineering (IJRTE)
ISSN: 2277-3878 (Online), Volume-11 Issue-1, May 2022 International Journal of Recent Technology and Engineering (IJRTE)
ISSN: 2277-3878 (Online), Volume-11 Issue-1, May 2022 6. Link Breakage Prediction a decrease in the total number of normal nodes. Figure 11
concluded that swarm intelligence-based routing algorithms
have a higher detection rate than other routing algorithms
because they solve many optimization problems and provide
the best solution. Figure 9 show that the TPACO algorithm has fewer link
breaks than all other existing protocols. Before the data
communication process begins, the link breakage can be
predicted (calculated). The link breaks as the node moves
more. So, that was overcome in this case. Figure 11: Detection Rate Figure 11: Detection Rate Figure 9: Mobility Speed vs. Link Breakage Figure 11: Detection Rate Figure 9: Mobility Speed vs. Link Breakage Published By:
Blue Eyes Intelligence Engineering
and Sciences Publication (BEIESP)
© Copyright: All rights reserved. Comparative Analysis of Routing Algorithms to Enhance the Quality of Service in MANE 8. Khanna, G., Chaturvedi, S. K., &Soh, S. (2019). “Reliability
evaluation of mobile ad hoc networks by considering link expiration
time and border time”. International Journal of System Assurance
Engineering and Management, 10(3), 399-415. Engineering and Management, 10(3), 399-415. 9. Kumar, J., &Kathirvel, A. (2019). “Analysis and Ideas for Improved
Routing in MANET”. 10. Kumar, K. P., &Babu, B. P. (2019, April). “A Simple and Cost-
Effective Anomaly Detection Paradigm on the Basis of Computational
Intelligence for Mobile Ad-Hoc Networks from a Security Viewpoint”. In Computer Science On-line Conference (pp. 78-86). Springer, Cham. 11. S. Gopalakrishnan, P. Mohan Kumar, “Performance Analysis of
Malicious Node Detection and Elimination Using Clustering Approach
on MANET”, Circuits and Systems, 2016, 7, 748-758 Published
Online May 2016. 12. Vanitha Kumar, A.M.J.Md.ZubairRahman, “Preventing malicious
packet dropping nodes in MANET using IFHM based SAODV routing
protocol”, The journal of Networks, software tools and applications,
ISSN-1386-7857,2019. 13. G. Jegan, D. Kamalakkannan, P. Samundiswary, “A Trusted Method
for Early Data Link Failure Prediction”, International Journal of Recent
Technology and Engineering (IJRTE) ISSN: 2277-3878, Volume-8
Issue-5, January 2020. y
14. S. Sugumaran, P. Venkatesan, “Optimized Trust Path for Control the
Packet Dropping and Collusion Attack using Ant Colony in MANET”,
International Journal of Engineering and Advanced Technology
(IJEAT) ISSN: 2249–8958, Volume-8, Issue-6, August 2019. AUTHOR PROFILE K.Vimala,
received
her
M.Phil
Degree
from
Bharathidasan University, Tiruchirappalli 2004. She
received her M.Sc., Degree from Nehru Memorial
College
affiliated
to
Bharathidasan
University,
Tiruchirapalli in 2001. She is pursuing her Ph.D
Degree (Part-Time) in Sri Vijay Vidyalaya College of
Arts &Science, Dharmapuri, Tamilnadu, India. She is
working as HOD Cum Assistant Professor in
Department of Computer Science at Pavai Arts and Science College for
Women, Anaipalayam, Rasipuram, Namakkal. Her current research of
interests includes Wireless Sensor Network and Computer Networks. Dr. D. Maruthanayagam, received his Ph.D Degree
from Manonmaniam Sundaranar University, Tirunelveli
in the year 2014. He received his M.Phil Degree from
Bharathidasan University, Trichy in the year 2005. He
received his M.C.A Degree from Madras University,
Chennai in the year 2000. He is working as HOD Cum
Professor, PG and Research Department of Computer
Science, Sri Vijay Vidyalaya College of Arts &
Science, Dharmapuri, Tamilnadu, India. He has above 21 years of
experience in academic field. He has published 6 books, more than 45
papers in International Journals and 30 papers in National & International
Conferences so far. His areas of interest include Computer Networks, Grid
Computing, Cloud Computing and Mobile Computing. Dr. D. Maruthanayagam, received his Ph.D Degree
from Manonmaniam Sundaranar University, Tirunelveli
in the year 2014. He received his M.Phil Degree from
Bharathidasan University, Trichy in the year 2005. He
received his M.C.A Degree from Madras University,
Chennai in the year 2000. He is working as HOD Cum
Professor, PG and Research Department of Computer
Science, Sri Vijay Vidyalaya College of Arts &
Science, Dharmapuri, Tamilnadu, India. He has above 21 years of
experience in academic field. He has published 6 books, more than 45
papers in International Journals and 30 papers in National & International
Conferences so far. His areas of interest include Computer Networks, Grid
Computing, Cloud Computing and Mobile Computing. 115
Published By:
Blue Eyes Intelligence Engineering
and Sciences Publication (BEIESP)
© Copyright: All rights reserved. 115
Published By:
Blue Eyes Intelligence Engineering
and Sciences Publication (BEIESP)
© Copyright: All rights reserved. 115
Published By:
Blue Eyes Intelligence Engineering
and Sciences Publication (BEIESP)
© Copyright: All rights reserved. 8. Detection Rate Published By:
Blue Eyes Intelligence Engineering
and Sciences Publication (BEIESP)
© Copyright: All rights reserved. Figure 11 depicts the ability of the TPACO algorithm to
detect malicious activity. The detection rate is compared to
various scenarios in Fig. 11. Three scenarios were
considered with 50 nodes in total. The detection rate
decreases as the number of malicious nodes increases due to Published By:
Blue Eyes Intelligence Engineering
and Sciences Publication (BEIESP)
© Copyright: All rights reserved. 1
Retrieval Number: 100.1/ijrte.A69490511122
DOI: 10.35940/ijrte.A6949.0511122
Journal Website: www.ijrte.org Retrieval Number: 100.1/ijrte.A69490511122
DOI: 10.35940/ijrte.A6949.0511122
Journal Website: www.ijrte.org Retrieval Number: 100.1/ijrte.A69490511122
DOI: 10.35940/ijrte.A6949.0511122
Journal Website: www.ijrte.org 114 Comparative Analysis of Routing Algorithms to Enhance the Quality of Service in MANET
G
Ch t
edi
S
K
&S h
S
(2019)
“Reli bilit Retrieval Number: 100.1/ijrte.A69490511122
DOI: 10.35940/ijrte.A6949.0511122
Journal Website: www.ijrte.org Retrieval Number: 100.1/ijrte.A69490511122
DOI: 10.35940/ijrte.A6949.0511122
Journal Website: www.ijrte.org
|
https://openalex.org/W2938553549
|
http://vestnik.vsmu.by/downloads/2019/2/2019_18_2_37-43.pdf
|
Russian
| null |
Acute effect of single introduction of taurine: specific or nonspecific?
|
Vestnik Vitebskogo gosudarstvennogo medicinskogo universiteta
| 2,019
|
cc-by
| 3,844
|
Резюме. Целью исследования явилось раскрытие динамических изменений аминокислотного пула плазмы после одно-
кратного введения фармакологической дозы таурина. Материал и методы. Эксперимент проводили на 29 крысах-самках массой 120-140 г при свободном доступе жи-
вотных к пище и воде. Животным внутрижелудочно вводили таурин в дозе 500 мг/кг массы. Декапитацию жи-
вотных осуществляли через 15, 30 и 90 мин после введения таурина. Для анализа использовали плазму крови. Определение свободных аминокислот производили методом обращеннофазной ВЭЖХ. Результаты и обсуждение. Однократное внутрижелудочное введение таурина (500 мг/кг) приводит к увеличению
концентрации этой аминокислоты в плазме крови, максимальный уровень которой регистрировали через 30 мин
(837,5±45,89 мкмоль/л, тогда как в контроле 142,0±18,95 мкмоль/л), однако наиболее выраженные изменения
аминокислотного пула плазмы крови наблюдали через 90 мин после его введения. Однократное введение таурина
снижало общее количество аминокислот и их азотсодержащих метаболитов в плазме крови. Заключение. Таким образом, вызванное введением таурина снижение общего количества аминокислот и их азот-
содержащих производных в плазме крови крыс, учитывая многочисленные функции таурина в организме, вероят-
но, свидетельствует о неспецифической стимуляции синтеза белка. Очевидно, что осморегуляторные, антиокси-
дантные и гормональные эффекты таурина в наибольшей степени будут обусловлены концентрация-зависимыми
изменениями аминокислот в плазме крови и внеклеточной жидкости, тогда как длительное его введение в не-
больших (близких к физиологическим) дозах в большей степени предполагает более тонкое воздействие на сиг-
нальные/регуляторные механизмы. р
у
р
Ключевые слова: таурин, свободные аминокислоты, плазма крови, крысы, азотсодержащие метаболиты ами-
нокислот. Ключевые слова: таурин, свободные аминокислоты, плазма крови, крысы, азотсодержащие мет
нокислот. SHEIBAK V.M., PAULIUKAVETS A.Y., DOROSHENKO E.M., OLEKHNOVICH E.A. Grodno State Medical University, Grodno, Republic of Belarus ВЕСТ
БИОХИМИЯ ВЕСТ
БИОХИМИЯ DOI: https://doi.org/10.22263/2312-4156.2019.2.37 Abstract. Objectives. To reveal the dynamic changes in the amino acid pool of the plasma after a single injection of a pharmacological
dose of taurine. Material and methods. The experiment was performed on 29 female rats weighing 120-140 g, with free access of animals
to food and water. Taurine was administered to rats intragastrically at a dose of 500 mg / kg of body weight. Animals were
decapitated in 15, 30, and 90 min after taurine administration. The blood plasma was used for analysis. Determination of
free amino acids was carried out by the method of reversed-phase HPLC. y
p
Results. A single intragastric administration of taurine (500 mg/kg) leads to an increase in the concentration of this amino
acid in the blood plasma, the maximum level of which was recorded after 30 min (837.5±45.89 μmol / l, whereas in the
control it was 142.0±18.95 μmol/l), however, the most pronounced changes in the amino acid pool of the blood plasma
were observed in 90 minutes after taurine administration. A single administration of taurine reduced the total amount of
amino acids and their nitrogen-containing metabolites in the blood plasma. 37 ESTNIK VITEBSKOGO GOSUDARSTVENNOGO MEDITSINSKOGO UNIVERSITETA, 2019, VOL. 18, N2 Conclusions. Thus, a decrease in the total amount of amino acids and their nitrogen-containing derivatives in rats’ blood
plasma, caused by the administration of taurine, is likely to indicate a nonspecific stimulation of protein synthesis. Obviously, the osmoregulatory, antioxidant and hormonal effects of taurine will be affected to the greatest extent by the
concentration-dependent changes of amino acids in the blood plasma and extracellular fluid, while long-term administration
of it in small (close to physiological) doses largely implies a more subtle effect on the signal / regulatory mechanisms. Key words: taurine, free amino acids, blood plasma, rats, nitrogen-containing metabolites of amino acids. Показано, что таурин выполняет множе-
ство функций в клетках млекопитающих, одна-
ко его цитопротекторные свойства привлекают
наибольшее внимание [1, 2]. Положительные
свойства таурина обусловлены, в частности, его
участием в неферментативной антиоксидантной
системе, энергетическом метаболизме митохон-
дрий, изменением гомеостаза Ca2+, осморегуля-
торными эффектами и центральными регуля-
торными механизмами, детерминированными
воздействием на ГАМКА- и глициновые рецепто-
ры [3, 4]. концентрации в клетках ЦНС, он может частич-
но заменять ГАМК и глицин, выполняя функцию
тормозного нейротрансмиттера. При патологиче-
ских состояниях, сопровождающихся гипоксией/
ишемией мозга или гипоосмотическим стрессом,
таурин высвобождается из различных клеток
ЦНС и функционирует как нейропротектор [11]. Abstract. Другим важным механизмом цитопротек-
торного действия таурина является ослабление
эндоплазматического (ER) стресса. ER-стресс
является важным механизмом регуляции, пред-
назначенным для восстановления функции эн-
доплазматической сети и баланса между дегра-
дацией, биосинтезом и фолдингом белков [12]. Когда клетка испытывает чрезмерный ER-стресс,
стимулируются катаболические пути. Вероятно,
таурин может понижать уровень стресса путем
оптимизации фолдинга белка либо стабилизируя
внутриклеточную рН [13]. Цитопротекция обусловлена прямой анти-
оксидантной активностью таурина [5]. Проти-
вовоспалительное действие таурина связано с
нейтрализацией в нейтрофилах хлорноватистой
кислоты, продуцируемой с участием миелопе-
роксидазы. Конъюгирование таурина с уридином
tRNALeu (UUR) в митохондриях оптимизирует
биосинтез белковых комплексов ЦТД, генерацию
АТФ и уменьшает образование супероксид-ани-
она [6]. Кроме того, таурин, изменяя проницае-
мость транзиторных потенциал-зависимых Са2+-
каналов митохондриальной мембраны, тормозит
апоптоз [7]. Таурин оказывает эффект и на энергетиче-
ский обмен в клетках, активируя чувствительные
к NADH ферменты путем снижения соотношения
NADH/NAD+ [14]. Показано, что при длительном
введении таурина мышам, получающим стан-
дартный рацион, снижается уровень глюкозы в
плазме и одновременно увеличивается размер
островков Лангерганса в поджелудочной железе
(вероятно, стимуляция наработки инсулина) [15]. Таурин, наряду с мочевиной и маннито-
лом, является органическим осмолитом, уровни
которых изменяются в ответ на повышение ос-
мотической нагрузки и снижаются при гипоос-
мотическом стрессе. Это важные механизмы не-
специфической защиты клетки от чрезмерного
набухания или сжатия в ответ на осмотический
вне- и внутриклеточный дисбаланс катионов [8]. Одним из важнейших сигнальных эффек-
тов таурина является его воздействие на перерас-
пределение вне- и внутриклеточных потоков ио-
нов кальция (Ca2+) [9]. Таурин ингибирует приток
Ca2+ в клетки, стимулируемый глутаматом, через
Ca2+-каналы L-, P / Q- и N-типа, а также канал
N-метил-D-аспартатного (NMDA) рецептора в
нейронах и клетках глии [10]. Таурин, наряду с мочевиной и маннито-
лом, является органическим осмолитом, уровни
которых изменяются в ответ на повышение ос-
мотической нагрузки и снижаются при гипоос-
мотическом стрессе. Это важные механизмы не-
специфической защиты клетки от чрезмерного
набухания или сжатия в ответ на осмотический
вне- и внутриклеточный дисбаланс катионов [8]. Очевидно, что для более полного понима-
ния механизмов действия таурина и выявления
его специфических и неспецифических (фарма-
кологических) эффектов важным является опре-
деление динамических изменений содержания
таурина в тканях после однократного введения
этого соединения в количествах, превышающих
физиологические. Выявление характера измене-
ний пула свободных аминокислот плазмы крови
позволяет не только обнаруживать изменения
концентрации самого таурина, но и показать
колебания уровней аминосодержащих метабо-
литов, генерируемых в процессе метаболизма в
организме. телей может рассматриваться как элемент мета-
боломики. персий. Характеристика изучаемых показателей
проводилась с помощью параметрической стати-
стики (t-критерий Стьюдента для независимых
выборок). Правомерность использования t-теста
Стьюдента проверялась с помощью критериев
применения параметрической статистики с ис-
пользованием теста Холмогорова-Смирнова с
поправкой Лиллиефорса для выборок с заранее
неизвестными дисперсиями. В таблицах пред-
ставлены средняя арифметическая и стандартная
ошибка средней арифметической. Целью исследования явилось раскрытие
динамических изменений аминокислотного пула
плазмы после однократного введения фармако-
логической дозы таурина. Материал и методы Эксперимент проводили на 29 крысах-
самках массой 120-140 г при свободном доступе
животных к пище и воде. Животные были раз-
делены на 2 группы: 1-ой контрольной группе
(n=8) – внутрижелудочно вводили физраствор
(0,9% раствор натрия хлорида), 2-ой группе жи-
вотных (n=21) внутрижелудочно вводили тау-
рин 500 мг/кг массы. Декапитацию животных
осуществляли через 15, 30 и 90 мин после вве-
дения таурина. Для анализа использовали плаз-
му крови. Определение свободных аминокислот
производили методом обращеннофазной ВЭЖХ. Все определения осуществляли с помощью хро-
матографической системы Agilent 1100, прием
и обработка данных – с помощью программы
Agilent ChemStation A10.01. Все полученные
данные подвергнуты статистической обработ-
ке. Анализ данных выполнен с использованием
пакета программ Statistica 6.0.и Microsoft Excel
2002. Для этого в полученных данных проверяли
нормальность распределения и равенство дис- Abstract. Одновременное описание совокуп-
ности метаболически ассоциированных показа- Одним из важнейших сигнальных эффек-
тов таурина является его воздействие на перерас-
пределение вне- и внутриклеточных потоков ио-
нов кальция (Ca2+) [9]. Таурин ингибирует приток
Ca2+ в клетки, стимулируемый глутаматом, через
Ca2+-каналы L-, P / Q- и N-типа, а также канал
N-метил-D-аспартатного (NMDA) рецептора в
нейронах и клетках глии [10]. Таурин является слабым агонистом ГАМКА-
и NMDA рецептора. Однако, благодаря высокой 38 ТНИК ВИТЕБСКОГО ГОСУДАРСТВЕННОГО МЕДИЦИНСКОГО УНИВЕРСИТЕТА, 2019, ТОМ 18, №2 телей может рассматриваться как элемент мета-
боломики. Результаты и обсуждение Однократное внутрижелудочное введение
таурина в дозе 500 мг/кг во все изучаемые сроки
(15 мин, 30 мин и 90 мин) увеличивало его кон-
центрацию в плазме крови в 2,7 раза, в 5,9 раза и
в 4,2 раза соответственно (рис.). Введение таурина через 15 минут в плазме
крови повышало общее количество серосодержа-
щих аминокислот (с 180±21 по 419±44 мкмоль/
мл) и соотношение аргинин/орнитин (с 2,3±0,28
до 3,3±0,12) (табл. 1), а также концентрацию глу-
тамата (в 1,2 раза). Одновременно снижались
уровни азотсодержащих метаболитов аминокис-
лот α-аминоадипиновой кислоты (на 41,7%) и ор-
нитина (на 28,2%) (табл. 2). Через 30 минут после введения таури-
на снижалось общее количество незаменимых Рисунок – Изменение содержания таурина и общего количества азотсодержащих метаболитов аминокислот
(исключая таурин) в плазме крови крыс, получавших однократно внутрижелудочно таурин (500 мг/кг),
относительно контрольных значений (контроль=100%): * – статистически значимые различия
со значениями в контрольной группе (p<0,05) Рисунок – Изменение содержания таурина и общего количества азотсодержащих метаболитов аминокислот
(исключая таурин) в плазме крови крыс, получавших однократно внутрижелудочно таурин (500 мг/кг),
относительно контрольных значений (контроль=100%): * – статистически значимые различия
со значениями в контрольной группе (p<0,05) Рисунок – Изменение содержания таурина и общего количества азотсодержащих метаболитов аминокислот
(исключая таурин) в плазме крови крыс, получавших однократно внутрижелудочно таурин (500 мг/кг),
относительно контрольных значений (контроль=100%): * – статистически значимые различия
со значениями в контрольной группе (p<0,05) 39 VESTNIK VITEBSKOGO GOSUDARSTVENNOGO MEDITSINSKOGO UNIVERSITETA, 2019, VOL. 18, N2 Таблица 1 – Структура пула свободных аминокислот и их азотсодержащих метаболитов в плазме
крови крыс после однократного внутрижелудочного введения таурина в дозе 500 мг/кг массы, мкмоль/л
(M±m)
Изучаемый показатель
контроль
Таурин,
15 мин
Таурин,
30 мин
Таурин,
90 мин
Общее количество свободных
аминокислот и их азотсодержащих
производных
3043±220
3288±115
3480±74
2154±375
Общее количество протеиногенных
аминокислот
2722±194
2732±116
2476±95
1467±246*
Общее количество азотсодержащих
производных аминокислот
375±33
615±50
1051±42*
718±143*
Общее количество азотсодержащих
производных аминокислот, исключая
таурин
233±15
236±9
214±10
127±21*
Общее количество заменимых
аминокислот
1687±125
1749±104
1686±72
949±165*
Общее количество незаменимых
аминокислот
1035±90
984±56
791±35*
517±82*
Общее количество серосодержащих
аминокислот
180±21
419±44*
871±45*
614±129*
Общее количество серосодержащих
аминокислот, исключая таурин
38±4
39±6
34±1
23±4*
Общее количество АРУЦ
307±21
276±10
240±15*
150±24*
Общее количество ароматических
аминокислот
170±13
158±8
165±9
105±20*
АРУЦ/ароматические аминокислоты
1,8±0,08
1,8±0,06
1,5±0,09*
1,5±0,11*
Аргинин/орнитин
2,3±0,28
3,3±0,12*
2,8±0,18
2,9±0,09
Заменимые/незаменимые аминокислоты
1,7±0,11
1,8±0,15
2,1±0,09*
1,8±0,06
Примечание: * – статистически значимые различия со значениями в контрольной группе (p<0,05). Примечание: * – статистически значимые различия со значениями в контрольной группе (p<0,05). Результаты и обсуждение аминокислот (с 1035±90 до 791±35 мкмоль/л),
аминокислот с разветвленной углеродной цепью
(изолейцин, лейцин, валин) (307±21 до 240±15
мкмоль/л) и соотношение аргинин/цитруллин (с
1,5±0,11 до 1,2±0,07) (табл. 1). В результате зна-
чительного увеличения концентрации таурина
в плазме крови увеличивалось общее количе-
ство серосодержащих аминокислот (с 180±21 до
871±45 мкмоль/л) (табл. 1). Среди индивидуаль-
ных концентраций аминокислот уменьшались
концентрации
протеиногенных
аминокислот:
тирозина (на 21,9%), лейцина (на 33,9%) и ли-
зина (на 40,3%); азотсодержащих метаболитов
аминокислот: α-аминоадипиновой кислоты (на
43,8%), орнитина (на 24,3%) и α-аминомасляной
кислоты (на 79,2%). Увеличивалась концентра-
ция β-аланина (в 1,3 раза) (табл. 2). держащих метаболитов аминокислот (с 375±33
до 127±21 мкмоль/л) (рисунок), незаменимых
аминокислот (с 1035±90 до 517±82 мкмоль/л) и
АРУЦ (с 307±21 до 150±24 мкмоль/л), общее ко-
личество ароматических аминокислот (с 170±13
до 105±20 мкмоль/л), сумма метионин+циста
тионин+цистеиновая кислота (с 38±4 до 23±4
мкмоль/л) (табл. 1). Снижение общего количе-
ства аминокислот и их азотсодержащих метабо-
литов обусловлено падением индивидуальных
концентраций-заменимых аминокислот: аспар-
тата (на 36,7%), глутамата (на 33,2%), аспараги-
на (на 44,7%), серина (на 52,6%), глутамина (на
46,3%), глицина (на 40,0%), треонина (на 43,5%),
аргинина (на 36,3%), аланина (на 44,5%); незаме-
нимых аминокислот: тирозина (на 52,4%), валина
(на 51,3%), метионина (на 35,7%), фенилаланина
(на 36,4%), изолейцин (на 46,5%), лейцина (на
54,8%) и лизина (на 58,9%). Результаты и обсуждение Очевидно, что осморегуляторные,
антиоксидантные и гормональные эффекты тау-
рина в наибольшей степени будут обусловлены
концентрация-зависимыми изменениями амино-
кислот в плазме крови и внеклеточной жидкости,
тогда как длительное его введение в небольших
(близких к физиологическим) дозах в большей
степени предполагает более тонкое воздействие
на сигнальные/регуляторные механизмы. Заключение Таким образом, однократное внутрижелу-
дочное введение таурина (500 мг/кг) приводит к
увеличению концентрации этой аминокислоты
в плазме крови, максимальный уровень кото-
рой регистрировали через 30 мин (837,5±45,89
мкмоль/л, тогда как в контроле 142,0±18,95
мкмоль/л), однако наиболее выраженные измене-
ния аминокислотного пула плазмы крови наблю-
дали через 90 мин после его введения. Однократ-
ное введение таурина снижало общее количество
аминокислот и их азотсодержащих метаболитов Результаты и обсуждение Среди азотсодержа-
щих метаболитов уменьшаются концентрации
α-аминоадипиновой кислоты (на 37,5%), фосфо- В плазме крови через 90 минут после одно-
кратного введения таурина уменьшалось общее
количество протеиногенных аминокислот (с
2722±194 до 1467±246 мкмоль/л) и их азотсо- 40 ЕСТНИК ВИТЕБСКОГО ГОСУДАРСТВЕННОГО МЕДИЦИНСКОГО УНИВЕРСИТЕТА, 2019, ТОМ 18, № Таблица 2 – Изменения концентраций свободных аминокислот и их азотсодержащих метаболи-
тов в плазме крови крыс после однократного внутрижелудочного введения таурина в дозе 500 мг/кг
массы, мкмоль/л (M±m) Таблица 2 – Изменения концентраций свободных аминокислот и их азотсодержащих метаболи-
тов в плазме крови крыс после однократного внутрижелудочного введения таурина в дозе 500 мг/кг
массы мкмоль/л (M±m) Аминокислоты
Контроль
Таурин,
15 мин
Таурин,
30 мин
Таурин,
90 мин
Аспартат
25,1±2,54
31,3±3,37
28,3±1,50
15,9±2,45*
Глутамат
121,2±6,83
149,5±10,36*
134,4±3,95
81,0±12,53*
Аспарагин
57,3±5,59
60,7±3,10
49,4±2,38
31,7±5,03*
Серин
179,2±14,55
169,1±19,78
150,6±5,94
85,0±12,94*
Глутамин
508,4±44,00
546,4±50,16
476,1±46,89
273,2±40,76*
Глицин
175,0±13,20
167,6±19,78
198,2±4,83
105,0±18,14*
Треонин
166,3±21,23
165,1±10,23
141,4±11,10
93,9±17,40*
Аргинин
106,2±9,72
113,9±7,41
99,2±2,41
67,6±12,25*
Аланин
464,8±35,08
472,6±41,44
502,5±25,91
257,9±58,43*
Тирозин
62,0±4,91
55,9±4,19
48,4±2,31*
29,5±5,10*
Валин
132,3±8,89
118,1±3,87
109,2±6,28
64,4±10,88*
Метионин
34,5±3,68
37,2±5,44
31,4±1,60
22,2±4,02*
Фенилаланин
47,2±4,07
39,3±1,99
44,0±2,46
30,0±5,22*
Изолейцин
71,4±3,53
75,7±3,13
61,8±3,85
38,2±5,59*
Лейцин
103,8±9,38
82,2±5,03
68,6±5,72*
46,9±7,34*
Лизин
357,6±51,10
347,5±36,59
213,5±20,06*
147,1±23,85*
α-аминоадипиновая
кислота
0,48±0,06
0,28±0,05*
0,27±0,05*
0,3±0,05*
Фосфоэтаноламин
2,1±0,45
1,3±0,24
1,4±0,43
0,6±0,17*
1-метилгистидин
2,8±0,33
3,5±0,27
2,2±0,21
1,4±0,27*
Цитруллин
70,3±5,63
82,6±3,52
82,6±5,00
45,1±8,21*
Таурин
142,0±18,95
379,3±45,44*
837,5±45,89*
590,4±125,84*
α-аминомасляная кислота
18,3±3,43
16,0±2,19
5,6±0,55*
3,6±0,38*
Этаноламин
16,2±1,57
18,0±1,11
17,8±0,70
9,8±1,70*
Гидроксилизин
5,7±0,66
6,4±0,51
5,9±0,23
3,2±0,56*
Орнитин
47,8±3,91
34,3± 1,74*
36,2±3,01*
23,2±4,35*
Примечание: * – статистически значимые различия со значениями в контрольной группе (p<0,05). римечание: * – статистически значимые различия со значениями в контрольной группе (p<0,05). в плазме крови, что, учитывая многочисленные
функции таурина в организме, вероятно, свиде-
тельствует о неспецифической стимуляции син-
теза белка. Очевидно, что осморегуляторные,
антиоксидантные и гормональные эффекты тау-
рина в наибольшей степени будут обусловлены
концентрация-зависимыми изменениями амино-
кислот в плазме крови и внеклеточной жидкости,
тогда как длительное его введение в небольших
(близких к физиологическим) дозах в большей
степени предполагает более тонкое воздействие
на сигнальные/регуляторные механизмы. этаноламина (на 71,4%), 1-метилгистидина (на
50%), цитруллина (на 35,8%), α-аминомасляной
кислоты (на 80,3%), этаноламина (на 39,5%), ги-
дроксилизина (на 43,9%), орнитина (на 51,5%)
(табл. 2). в плазме крови, что, учитывая многочисленные
функции таурина в организме, вероятно, свиде-
тельствует о неспецифической стимуляции син-
теза белка. References hypertensive rats. Hypertens Res. 2016 Dec;39(12):848-
856. doi: 10.1038/hr.2016.86 1. Sheybak VM, Sheybak LN. Biological role of taurine in
mammals. Med Novosti. 2005;(10):15-8. (In Russ.) 10. L’Amoreaux WJ, Marsillo A, Idrissi A. Pharmacological
characterization of GABAA receptors in taurine-fed mice. J Biomed Sci. 2010;17( Suppl 1):S14. doi: 10.1186/1423-
0127-17-S1-S14 2. Sheybak VM, Sheybak LN. Biosynthesis and taurine
metabolism. Zhurn GrGMU. 2005;(1):9-12. (In Russ.) 3. Schaffer S, Kim W. Effects and Mechanisms of Taurine
as a Therapeutic Agent. Biomol Ther (Seoul). 2018
May;26(3):225-241. doi: 10.4062/biomolther.2017.251 11. Chan CY, Sun HS, Shah SM, Agovic MS, Ho I, Friedman
E, et al. Direct interaction of taurine with the NMDA
glutamate receptor subtype via multiple mechanisms. Adv
Exp Med Biol. 2013;775:45-52. doi: 10.1007/978-1-4614-
6130-2_4 4. Wang J, Qi C, Liu L, Zhao L, Cui W, Tian Y, et al. Taurine
Protects Primary Neonatal Cardiomyocytes Against
Apoptosis Induced by Hydrogen Peroxide. Int Heart J. 2018
Jan;59(1):190-196. doi: 10.1536/ihj.16-372 12. Ito T, Miyazaki N, Schaffer S, Azuma J. Potential antiaging
role of taurine via proper protein folding: a study from
taurine transporter knockout mouse. Adv Exp Med Biol. 2015;803:481-7. doi: 10.1007/978-3-319-15126-7_38 5. Wu H, Jin Y, Wei J, Jin H, Sha D, Wu JY. Mode of action
of taurine as a neuroprotector. Brain Res. 2005 Mar
21;1038(2):123-31. 13. Ghosh S, Chowdhury S, Das AK, Sil PC. Taurine
ameliorates oxidative stress induced inflammation and ER
stress mediated testicular damage in STZ-induced diabetic
Wistar rats. Food Chem Toxicol. 2019 Feb;124:64-80. doi:
10.1016/j.fct.2018.11.055 6. Das J, Sil PC. Taurine ameliorates alloxan-induced diabetic
renal injury, oxidative stress-related signaling pathways and
apoptosis in rats. Amino Acids. 2012 Oct;43(4):1509-23. 7. Marcinkiewicz J, Kontny E. Taurine and inflammatory
diseases. Amino Acids. 2014 Jan;46(1):7-20. doi: 10.1007/
s00726-012-1361-4 14. Shetewy A, Shimada-Takaura K, Warner D, Jong CJ,
Mehdi AB, Alexeyev M, et al. Mitochondrial defects
associated with B-alanine toxicity: relevance to hyper-beta-
alaninemia. Mol Cell Biochem. 2016 May;416(1-2):11-22. doi: 10.1007/s11010-016-2688-z 8. Schaffer SW, Shimada-Takaura K, Jong CJ, Ito T, Takahashi
K. Impaired energy metabolism of the taurine-deficient
heart. Amino Acids. 2016 Feb;48(2):549-58. doi: 10.1007/
s00726-015-2110-2 15. Ulrich-Merzenich G, Zeitler H, Vetter H, Bhonde RR. Protective effects of taurine on endothelial cells impaired
by high glucose and oxidized low density lipoproteins. Eur
J Nutr. 2007 Dec;46(8):431-8 9. Katakawa M, Fukuda N, Tsunemi A, Mori M, Maruyama T,
Matsumoto T, et al. Литература 1. Шейбак, В. М. Биологическая роль таурина в организме
млекопитающих / В. М. Шейбак, Л. Н. Шейбак // Мед. новости. – 2005. – № 10. – C. 15–18. 2. Шейбак, В. М. Биосинтез и обмен таурина / В. М. Шей-
бак, Л. Н. Шейбак // Журн. ГрГМУ. – 2005. – № 1. – С. 41 VESTNIK VITEBSKOGO GOSUDARSTVENNOGO MEDITSINSKOGO UNIVERSITETA, 2019, VOL. 18, N2 3. Schaffer, S. Effects and Mechanisms of Taurine as a
Therapeutic Agent / S. Schaffer, H. W. Kim // Biomol. Ther. (Seoul). – 2018 May. – Vol. 26, N 3. – P. 225–241. 10. L’Amoreaux, W. J. Pharmacological characterization of
GABAA receptors in taurine-fed mice / W. J. L’Amoreaux,
A. Marsillo, A. Idrissi // J. Biomed. Sci. – 2010. – Vol. 17,
suppl. 1. – P. S14. 4. Taurine Protects Primary Neonatal Cardiomyocytes Against
Apoptosis Induced by Hydrogen Peroxide / J. Wang [et al.] //
Int. Heart J. – 2018 Jan. – Vol. 59, N 1. – P. 190–196. 11. Direct interaction of taurine with the NMDA glutamate
receptor subtype via multiple mechanisms / C. Y. Chan [et
al.] // Adv. Exp. Med. Biol. – 2013. – Vol. 775. – P. 45–52. 5. Mode of action of taurine as a neuroprotector / H. Wu [et. al.] // Brain Res. – 2005 Mar. – Vol. 1038, N 2. – P. 123–131. 5. Mode of action of taurine as a neuroprotector / H. Wu [et. al.] // Brain Res. – 2005 Mar. – Vol. 1038, N 2. – P. 123–131. 6. Das, J. Taurine ameliorates alloxan-induced diabetic renal
injury, oxidative stress-related signaling pathways and
apoptosis in rats / J. Das, P. C. Sil // Amino Acids. – 2012
Oct. – Vol. 43, N 4. – P. 1509–1523. 12. Potential antiaging role of taurine via proper protein folding:
a study from taurine transporter knockout mouse / T. Ito [et
al.] //Adv. Exp. Med. Biol. – 2015. – Vol. 803. – P. 481–487. 6. Das, J. Taurine ameliorates alloxan-induced diabetic renal
injury, oxidative stress-related signaling pathways and
apoptosis in rats / J. Das, P. C. Sil // Amino Acids. – 2012
Oct. – Vol. 43, N 4. – P. 1509–1523. 13. Taurine ameliorates oxidative stress induced inflammation
and ER stress mediated testicular damage in STZ-induced
diabetic Wistar rats / S. Ghosh [et al.] // Food. Chem. Toxicol. – 2019 Feb. – Vol. 124. – P. 64–80. 7. Marcinkiewicz, J. Литература Taurine and inflammatory diseases / J. Marcinkiewicz, E. Kontny // Amino Acids. – 2014 Jan. – Vol. 46, N 1. – P. 7–20. 14. Mitochondrial defects associated with B-alanine toxicity:
relevance to hyper-beta-alaninemia / A. Shetewy [et al.] //
Mol. Cell. Biochem. – 2016 May. – Vol. 416, N 1/2. – P. 11–22. 8. Impaired energy metabolism of the taurine-deficient heart /
S. W. Schaffer [et al.] // Amino Acids. – 2016 Feb. – Vol. 48,
N 2. – P. 549–558. 15. Protective effects of taurine on endothelial cells impaired
by high glucose and oxidized low density lipoproteins / G. Ulrich-Merzenich [et al.] // Eur. J. Nutr. – 2007 Dec. – Vol. 46, N 8. – P. 431–438. 9. Taurine and magnesium supplementation enhances the
function of endothelial progenitor cells through antioxidation
in healthy men and spontaneously hypertensive rats / M. Katakawa [et al.] // Hypertens Res. – 2016 Dec. – Vol. 39, N Поступила 11.02.2019 г. Принята в печать 25.03.2019 г. Сведения об авторах: Сведения об авторах:
Шейбак В.М. – д.м.н., профессор кафедры биологической химии, Гродненский государственный медицинский
университет;
Павлюковец А Ю – к б н доцент доцент кафедры микробиологии вирусологии и иммунологии им С И Гель- Шейбак В.М. – д.м.н., профессор кафедры биологической химии, Гродненский государственный медицинский
университет; Шейбак В.М. – д.м.н., профессор кафедры биологической химии, Гродненский государственный медицинский
университет;
Павлюковец А.Ю. – к.б.н., доцент, доцент кафедры микробиологии, вирусологии и иммунологии им. С.И. Гель-
бера, Гродненский государственный медицинский университет; Павлюковец А.Ю. – к.б.н., доцент, доцент кафедры микробиологии, вирусологии и иммунологии им. С.И. Гель-
бера, Гродненский государственный медицинский университет; р
р
у
р
у
р
Дорошенко Е.М. – к.б.н., доцент, ведущий научный сотрудник научно-исследовательской лаборатории, Гроднен-
ский государственный медицинский университет; Дорошенко Е.М. – к.б.н., доцент, ведущий научный сотрудник научно-исследовательской лаборатории, Гроднен-
ский государственный медицинский университет; Олехнович Е.А. – студентка 5 курса лечебного факультета, Гродненский государственный медицинский универ-
ситет. Олехнович Е.А. – студентка 5 курса лечебного факультета, Гродненский государственный медицинский универ-
ситет. References Taurine and magnesium supplementation
enhances the function of endothelial progenitor cells
through antioxidation in healthy men and spontaneously Submitted 11.02.2019
Accepted 25.03.2019 42 ЕСТНИК ВИТЕБСКОГО ГОСУДАРСТВЕННОГО МЕДИЦИНСКОГО УНИВЕРСИТЕТА, 2019, ТОМ 18, № Information about authors: f
Sheibak V.M. – Doctor of Medical Sciences, professor of the Chair of Biologic Chemistry, Grodno State Medical University;
Pauliukavets A.Y. – Candidate of Biological Sciences, associate professor of the Chair of Microbiology, Virology and
Immunology named after S.I. Gelberg, Grodno State Medical University; f
Sheibak V.M. – Doctor of Medical Sciences, professor of the Chair of Biologic Chemistry, Grodno State Medical University;
Pauliukavets A.Y. – Candidate of Biological Sciences, associate professor of the Chair of Microbiology, Virology and
Immunology named after S.I. Gelberg, Grodno State Medical University; Sheibak V.M. – Doctor of Medical Sciences, professor of the Chair of Biologic Chemistry, Grodno State Medical University;
Pauliukavets A.Y. – Candidate of Biological Sciences, associate professor of the Chair of Microbiology, Virology and
Immunology named after S.I. Gelberg, Grodno State Medical University;
Doroshenko E.M. – Candidate of Biological Sciences, associate professor, leading research offi cer of the Scientifi c-
Research Laboratory, Grodno State Medical University;
Olekhnovich E.A. – the fi fth-year medical student, Grodno State Medical University. Doroshenko E.M. – Candidate of Biological Sciences, associate professor, leading research offi cer of the Scientifi c-
Research Laboratory, Grodno State Medical University; Doroshenko E.M. – Candidate of Biological Sciences, associate professor, leading research offi cer of the Scientifi c-
Research Laboratory, Grodno State Medical University; Адрес для корреспонденции: Республика Беларусь, 230023, г. Гродно, ул. Виленская, 19, Гродненский государ-
ственный медицинский университет, кафедра микробиологии, вирусологии и иммунологии им. С.И. Гельбера. E-mail: anastasiayk@mail.ru – Павлюковец Анастасия Юрьевна. Correspondence address: Republic of Belarus, 230023, Grodno, 19 Vilenskaya str., Grodno State Medical University,
Chair of Microbiology, Virology and Immunology named after S.I. Gelberg. E-mail: anastasiayk@mail.ru – Anastasiya
Y. Pauliukavets. Y. Pauliukavets. 43
|
https://openalex.org/W4200473552
|
https://jdsr.se/ojs/index.php/jdsr/article/download/95/51
|
English
| null |
Memetic Moments
|
Journal of digital social research
| 2,021
|
cc-by-sa
| 10,267
|
JOURNAL DIGITAL
SOCIAL RESEARCH
O F JOURNAL DIGITAL
SOCIAL RESEARCH
O F JOURNAL DIGITAL
SOCIAL RESEARCH
O F ISSN: 2003-1998 WWW.JDSR.IO VOL. 4, NO. 1, 2022, 23–48 This article is published under a CC BY-SA license y
b Australian National University, Australia a Federation University Australia, Australia Keywords: memes; Twitter; speed; rhythmanalysis. a Federation University Australia, Australia
b Australian National University, Australia INTRODUCTION 1 While much research has focused on the various affordances of online platforms,
researchers are also increasingly focusing on the logic of speed at the core of online
life and its effect on social media platforms' emotional tenor and tone. Speed is
increasingly revealing itself as central to online life, with digital platforms requiring
constant attention from users to survive economically. Speed drives what some have
described as the ‘attention’ economy (Terranova, 2012; Venturini, 2019), in which
platforms seek ways to keep users’ attention in order to drive profits. Speed is
integral to this process, with its logic dominating both technical and social aspects
of digital platforms. Technically, for digital platforms to provide an enticing user experience, they
need to load quickly and reliably. Facebook, for example, famously restricted its
initial rollout to ensure it had the server capacity to meet user demand (Losse,
2012). Without server capacity to meet each new influx of users, the site could
crash, load slowly or unreliably and, as a result, increase user attrition. Site design,
focusing on speed and the sensation of smoothness it brings, is key to one of the
most successful digital companies to date, specifically Facebook. This logic is also
modelled across a range of other online platforms. The reliance of online platforms
on speed is facilitated by the implementation of high-speed internet infrastructure,
such as broadband, that encouraged more data-heavy use, including photo and
video sharing. Speed is not just about user experience or the temporal rhythm at which
things happen, but also the scale and spread of content. As a consequence, issues
like moderation on social platforms become increasingly complex. The volume of
content is vast and continues to grow daily. While the scale, growth and production
of content is the focus of much scholarly discussion of the problem of online
moderation (Gillespie, 2019; Roberts, 2019), we argue so too is the factor of speed. Speed is an important part of how information flows and circulates online. The
logic of speed even permeates the rapid cycle nature of policy change in companies
like Facebook, which often change policy on a fortnightly basis. Working towards
a sociology of digital temporality is an important part of analysing the contemporary
digital landscape. Current analysis of the dynamics of speed frequently assumes it to be one of
the more destructive elements of digital platforms. ABSTRACT This paper examines how speed shapes internet culture. To do so, it analyses ‘memetic
moments’ on Twitter, short-lived and rapidly circulated memes that quickly reach
saturation. The paper examines two ‘memetic moments’ on Twitter in 2018 and 2019
to assess how they develop over time. Each case study comprises a week’s worth of
relevant tweets that were analysed for temporal patterns. We analyse these ‘memetic
moments’ through Lefebvre’s (2004) work on rhythmanalysis, arguing that the
temporal patterns of memes on Twitter can be understood through his concepts of
repetition, presence and dialogue. While seemingly trivial, memetic moments
underscore the didactic relationship between social media and news media while also
providing a way to approach complex social issues. SMITH & COPLAND — MEMETIC MOMENTS SMITH & COPLAND — MEMETIC MOMENTS JOURNAL OF DIGITAL SOCIAL RESEARCH — VOL. 4, NO. 1, 2022 function on Reddit (Cook, 2019). These functions play an integral role in boosting
and in turn spreading popular content. Recently, in the lead up to the 2020 US
Presidential election, Twitter changed the retweet button to default to the “quote
tweet” function, turning a one click process into a two click one, and hopefully
adding a moment of consideration (or friction) into the retweeting process
(Hatmaker, 2020). It is the intersection of speed, memes and social media that we examine in
this paper. Specifically, we examine two memetic moments on Twitter to
interrogate the role of speed in the creation and circulation of memes on Twitter. In this paper, we contend that speed can also create a sense of joy, entertainment
and even important discussion online. We examine the spread of two memes across
2018 and 2019 on Twitter. These memes focused on political issues in the United
States and began after initial tweets, which were then parodied by thousands of
users across quick timeframes. These copies and remixes created what we describe
as ‘emergent memes’ or ‘memetic moments’, which not only function as
entertainment, but also facilitate discussion of important political issues. In this paper, we do two things. First, we analyse the dynamics of the spread
of the two identified memes. We do this by conducting a temporal analysis of the
development of these memes, conducting an in-depth reading of how these memes
developed over time. In doing so, we explore the role of speed in the spread of this
material and the dialectical relationship between Twitter memes and the mass
media. INTRODUCTION Large social media platforms
have been associated with or blamed for the rapid spread of ‘fake’ or ‘junk news’
(Venturini, 2019), disinformation (O’Neil & Jensen, 2020), conspiracy theories
(Tuters, Jokubauskaité & Bach, 2018) and hateful content (e.g. Massanari, 2017;
Salter, 2018). Speed plays an integral role in the spread of this material, with some,
in turn, arguing for digital platforms to attempt to ‘slow down’ discourse online. Suggestions have included the removal or limiting of ‘sharing’ functions such as the
sharing button on Facebook or Retweets on Twitter (Mirani, 2020), as well as the
removal of voting functions, such as the like button on Facebook or the voting 24 JOURNAL OF DIGITAL SOCIAL RESEARCH — VOL. 4, NO. 1, 2022 JOURNAL OF DIGITAL SOCIAL RESEARCH — VOL. 4, NO. 1, 2022 SMITH & COPLAND — MEMETIC MOMENTS requiring insider knowledge to be correctly read and interpreted. At their core,
memes are media-rich jokes that sprawl across form and content (Miltner, 2014). They are commonly conceptualised as a visual, not textual genre, although image-
macros memes rely on both visual and text elements for interpretation. Research
investigating the social implications of memes is a broad field, covering everything
from rhetoric to meme warfare (Bapna & Lokhande 2021). Bapna and Lokhande
identify memes as part of what they argue is the ‘aestheticisation’ of politics within
India and beyond (2021, p. 201). They also argue that memes are a way of
‘microdosing’ political content amongst more general, humorous content on, for
example, Instagram meme accounts. While not directly addressing speed, they
highlight how messaging services like WhatsApp can quickly spread fake news
using memes to millions of users. This means that less attention is placed on emergent memes, or memetic
moments, that are directly responsive to shared cultural or social events and do not
utilise image macros. However, these types of memes are increasingly dominant,
including on new applications like TikTok, in which sounds and videos fuel its
meme-heavy visual culture. Abidin and Kaye (2021, p. 58) argue that audio memes
are the “next frontier of meme cultures on the internet, presenting an ‘aural turn’ in
meme ecologies”. Audio use may use songs and original audio produced on the app
to tell stories and as a way to organise content within the app. In this context,
memes function as a way of cataloguing and finding similar information. While we focus on Twitter as a case study, our analysis can be applied to
memes on other platforms, which do not fit the typical image macro-based
structure. In expanding the category of a meme in this way, we follow Phillips and
Milner’s (2017) concept of the ambivalent internet. They identify a number of
‘memetic moments’ both on Twitter and on other internet platforms, which they
categorise as part of the ‘ambivalent’ internet (Phillips & Milner, 2017). They argue
that these moments are often miscategorised as ‘trolling’ or part of the ‘weird’
internet. We agree that these types of behaviour are misunderstood and perhaps
overlooked. These types of moments should also be understood as memes - in
which an initial tweet provides a meme ‘template’ that is both funny and shareable
and allows for a high degree of intertextuality. 2
MEMES AND/AS DIGITAL CULTURE Memes are a distinctive hallmark of digital culture. Memes are generally jokes that
emerge as a by-product of digital culture. Internet memes, Nissenbaum and
Shifman (2015) argue, are digital objects that follow a shared pattern or
characteristics that allow them to be identified as a unified group. Scholars have
argued for the importance of memes as a “prism for shedding light on aspects of
contemporary digital culture” (Shifman, 2012, p. 190). Most scholarship
examining memes focuses on template style image-macro memes such as the ‘first
world problems’ or ‘lol cat’ memes that rely on a text/image combination, also called
image macro memes (Wiggins & Bowers, 2015). Specifically, memes are a part of
participatory digital culture, which Wiggins and Bowers (2012) argue possess a
virtual physicality that renders them observable social artifacts in internet culture. The memes share unifying aesthetic and linguistic cues; this includes the font
choice and the central image. Some scholars have argued that other forms of viral
content creation as also memes, such as the ‘It Gets Better’ videos produced in
response to homophobic bullying (Gal, Shifman & Kampf, 2015). However, the
majority of research that considers memes identifies their humour as a quality that
makes them distinct from other forms of viral content. Template or image-based
memes are often complex and intertextual artefacts of internet subcultures, 25 SMITH & COPLAND — MEMETIC MOMENTS SMITH & COPLAND — MEMETIC MOMENTS What distinguishes these memes
from the template-based memes, which are the objects of previous research, is their
reliance on text. The emergent memes on Twitter are not usually formed around
an image, but rely on a readers’ sophisticated knowledge of online communication
norms and jokes to be read successfully. Emergent memes are the product of a
particular subset of online cultures, one that is ‘always on’ and constantly consuming
news and events. It is a culture that prioritises speed. Part of hacking the attention
economy of the ‘always on’ internet culture is to be first, first to spot the joke, first
to make the joke. In the following sections, we refer to these memes as ‘memetic
moments’ due to their ephemeral nature. They are less persistent and recursive than
image macro memes but point to a broader understanding of memes as social
artifacts in participatory digital culture. 26 JOURNAL OF DIGITAL SOCIAL RESEARCH — VOL. 4, NO. 1, 2022 SPEED AND THE ATTENTION ECONOMY 3 Speed is increasingly being addressed as a core component of the function of the
digital economy, described by theorists as the “attention economy” (Terranova,
2012; Venturini, 2019). However, speed has long been of attention to social
theorists. Speed, discussed through the lens of time-space compression, occurs
across the work of geographers (e.g. Harvey 1989; Thrift 1996) and sociologists
(Beck 2000; Giddens 1990). These accounts grapple with the effects of a range of
technologies, from the industrial to the personal, on the social world. In the work
of Giddens (1990) and Beck (2000), this is also accompanied by an attendant focus
on what they argue is the increasingly rapid pace of social and cultural change;
technology has altered the rhythms of life, speeding them up, alongside the rate of
change itself. Specific to social relations, Lash (2002) argues that speed means that
previously fixed (in place) social bonds have given way to more transient
communication bonds that are immediate yet distanciated. Virilio (2000) also
identifies similar processes, observing that the ‘speeding up’ characteristic of late
modernity means that social relations requiring time are no longer wholly
dependent on a body located in space. While Virilio and Lash are writing before
the advent of Twitter, the accelerated sociality he described, which is geographically
distant yet incredibly fast, is characteristic of the type of engagement that Twitter
encourages. Speed is increasingly being addressed as a core component of the function of the
digital economy, described by theorists as the “attention economy” (Terranova,
2012; Venturini, 2019). However, speed has long been of attention to social
theorists. Speed, discussed through the lens of time-space compression, occurs
across the work of geographers (e.g. Harvey 1989; Thrift 1996) and sociologists
(Beck 2000; Giddens 1990). These accounts grapple with the effects of a range of
technologies, from the industrial to the personal, on the social world. In the work
of Giddens (1990) and Beck (2000), this is also accompanied by an attendant focus
on what they argue is the increasingly rapid pace of social and cultural change;
technology has altered the rhythms of life, speeding them up, alongside the rate of
change itself. Specific to social relations, Lash (2002) argues that speed means that
previously fixed (in place) social bonds have given way to more transient
communication bonds that are immediate yet distanciated. SMITH & COPLAND — MEMETIC MOMENTS In the same way that DeLuca identifies speed as having a deleterious effect
on the public sphere, similar arguments are made regarding the influence of speed
on social media. Speed has been part of the analysis of social media dynamics,
particularly as it concerns fake news. For example, Vosoughi, Roy and Aral (2018)
found that fake news, being more novel than true news, spreads more quickly, and
this is seen as key to social problems born of misinformation. Venturi (2019) argues
that fake news is not new, mimicking longstanding propaganda techniques. The
present difference is the speed at which this news can spread, facilitated by digital
platforms, which provide the architectural scaffolding for its spread (Venturini,
2019). In sum, Venturini argues that social media prioritise attention and speed in
both social and technical design, as well as encouraging users to develop affective
relationships to news through liking and sharing. This allows news to travel at great
speed through online social networks, becoming ‘viral’. Venturini analyses the operation of speed on digital platforms - with the
attention economy being driven by economic, communicative, technological,
cultural and political reasons. We see similar driving factors in the speed of memes
within our own analysis. The varied literature on memes often references virality
and popularity when discussing the spread of memes. These terms, we argue, can
also be understood as proxies for speed. However, there is a dearth of research that
addresses speed directly as part of the circulation of memes. In Jenkin’s (2014)
rhetorical analysis of image-macro memes, he argues that accounting for the speed
at which memes circulate online is central to understanding the affects they
produce. He argues that it is the circulation of memes across digital media
platforms, which produce the structures that enable the ongoing remaking of
memes. The cross-platform circulation of memes produces their plasticity. The
pattern of circulation described by Jenkins is similar to the modes of circulation
described by Lefebvre (2004) in his text Rhythmanalysis. Lefebvre (2004, p. 7)
argues that “not only does repetition not exclude differences, it also gives birth to
them; it produces them.” Repetition creating difference is also what distinguishes
memes from viral content. As Miltner explains, “This is the key difference between
‘memetic’ content and ‘viral’ content; if a piece of content is passed along intact and
unaltered, it is considered to be viral. SPEED AND THE ATTENTION ECONOMY Virilio (2000) also
identifies similar processes, observing that the ‘speeding up’ characteristic of late
modernity means that social relations requiring time are no longer wholly
dependent on a body located in space. While Virilio and Lash are writing before
the advent of Twitter, the accelerated sociality he described, which is geographically
distant yet incredibly fast, is characteristic of the type of engagement that Twitter
encourages. g
Lash (2002) also identified an emphasis on information over society as part
of the transition to late modernity. This is consistent with early political economy
analyses identified ‘information’ as the core commodity being sold and traded on
the internet. However, more recent analyses have argued that attention has replaced
information as the new commodity (Terranova, 2012). Digital platforms base their
economic model on collecting and selling user data, primarily to advertisers
(Srnicek, 2016). Platforms collect the specified data of their users -- including
demographic information, interests, and the ways in which they participate -- and
sell that data to advertisers to allow them to target ads to specific audiences on the
site. Platforms require the attention of their users to maximise this model
(Terranova, 2012; Venturini, 2019). Attention provides platforms with more data
to sell and ensures greater usage of platforms, which allows for ads to be sold at a
higher level. However, this attention is scarce. Users only have so much time in the
day and often engage with different platforms at different times. In this context,
the speed and spread of material have become integral to the operations of digital
platforms. Speed provides users access to new material, keeping people’s attention
on a platform over time. However, speed is also seen as impoverishing social life. DeLuca, when discussing speed in the circulation of images, describes that in “the
ceaseless circulation of images, speed annihilates contemplation, surface flattens
depth, flow drowns moments, distraction disrupts attention, affect eclipses
meaning, the glance replaces the gaze, reiteration erases originals, and the public
screen displaces the public sphere” (2006, p. 87). 27 SMITH & COPLAND — MEMETIC MOMENTS 4
CASE STUDIES: MEMETIC MOMENTS This paper is specifically focused on the role that speed plays in the creation and
spread of emergent text-based memes on Twitter. To investigate this, we selected
two ‘memetic moments’ that took place on Twitter in 2018 and 2019 to judge the
ways in which they develop over time. We selected one moment centred on US
Congresswoman Alexandria Ocasio-Cortez (AOC) as well as the 30-50 feral hogs
meme that was a response to back to back mass shootings in the United States in
2019 in El Paso, Texas, and Dayton, Ohio, which in total killed 31 people. In the
case studies that follow, we have identified the authors of each initial tweet by name. These tweets have been widely reported on (Benwell & Paul, 2019; Rosenberg,
2018), and in one instance, the author of the original tweet has openly engaged with
reporting on their viral moment (Reply All, 2019). We have chosen these case
studies as they were widely reported outside of Twitter both during the initial
memetic moment and after it had subsided. We will further detail our methods
later in the article, but first, we want to outline the substantive events from which
these memes emerged to better situate them in the current social and political
context. Alexandria Ocasio-Cortez, also known as AOC, is the Representative for
New York’s 14th congressional district. Ocasio-Cortez is a high-profile politician
elected initially in the United States midterm elections of 2018. Ocasio-Cortez first
drew interest due to her substantial social media platform, her relative youth (29
years of age at the time of election) and her progressive policies. She was considered
to be a ‘surprise’ winner over a more established Democratic candidate in the
Primary elections. Ocasio-Cortez is also relatively unique among politicians, having
held a number of low-wage service jobs, such as bartending prior to her election. These factors, combined with Ocasio-Cortez’s viral appearances in bill reading
hearings, have meant that she has become a rising political star. It is in the context
of this additional attention, and her working-class background, that Ocasio-Cortez
became the centre of our first ‘memetic moment’. In the lead up to taking up her
position in Washington (a two-month delay between being elected and taking up
one’s seat), Ocasio-Cortez was public about her financial struggles - having no
income to bridge the gap between election and commencement. JOURNAL OF DIGITAL SOCIAL RESEARCH — VOL. 4, NO. 1, 2022 of attention economy we can see in our data set, driven by speed. In the next section,
we examine our two case studies, examining two text-based memes. SMITH & COPLAND — MEMETIC MOMENTS If a piece is altered or changed as it is passed
along, it is considered to be a meme” (2017, p. 414). Mapping the rate at which
memes circulate is difficult because, as Jenkins (2014, p. 446) highlights, “they often
emerge in numerous places simultaneously and frequently spread like wildfire.” In the case studies that follow, we begin to assess how speed factors into
emergent memes, or memetic moments on Twitter. In addition to the empirical
analysis presented, we also draw on Lefebvre’s theory of rhythmanalysis to
understand the temporal structure, or the speed of memes. In addition to drawing
on Lefebvre’s work, we also extend Venturini’s focus on ‘junk’. While the
mechanism Venturini identifies in the viral spread of information can cause
problems, it also brings joy. The internet has always been a home for the weird,
absurd and funny, alongside the loftier utopian ideals. This, we argue, is the kind 28 JOURNAL OF DIGITAL SOCIAL RESEARCH — VOL. 4, NO. 1, 2022 4
CASE STUDIES: MEMETIC MOMENTS At her 2018
congressional
orientation
in
Washington,
Ocasio-Cortez
was
covertly
photographed by an unnamed Hill staffer. This photograph was sent to
conservative political writer Eddie Scarry, who then posted it to Twitter with the
caption “Hill Staffer sent me this pic of Ocasio-Cortez they took just now. I’ll tell
you something: that jacket and coat don’t look like a girl who struggles.” (see Figure 1) 29 SMITH & COPLAND — MEMETIC MOMENTS Figure 1: @eScarry tweet Figure 1: @eScarry tweet Figure 1: @eScarry tweet The response from Twitter users was quick, funny and revelled in the absurdity of
the situation. Ocasio-Cortez responded sharply and quickly, and mainstream media
outlets were quick to point to the tweet as an example of sexism in politics (Bell,
2018). As will be discussed below, the hill staffer sent me this pic is a variation of
template memes (Wiggins & Bowers, 2015), and in turn, a valuable example to
examine how speed affects memes on Twitter. The response from Twitter users was quick, funny and revelled in the absurdity of
the situation. Ocasio-Cortez responded sharply and quickly, and mainstream media
outlets were quick to point to the tweet as an example of sexism in politics (Bell,
2018). As will be discussed below, the hill staffer sent me this pic is a variation of
template memes (Wiggins & Bowers, 2015), and in turn, a valuable example to
examine how speed affects memes on Twitter. The second memetic moment selected for analysis is the 30-50 feral hog
meme. The 30-50 feral hogs meme started as a response to a weekend of mass
shootings in the United States. On August 3, 2019, in a Walmart in El Paso, Texas,
a gunman shot and killed 22 shoppers while injuring 24 others. This was followed
by another mass shooting on August 4, 2019, in Dayton, Ohio, where the shooter,
using an automatic assault rifle, shot and killed nine victims and wounded 17 others. Two mass shootings so close together renewed debates about gun control in the
United States, particularly around automatic, assault-style weapons used in both
shootings. Subsequently, musician Jason Isbell tweeted, “If you’re on here arguing
the definition of “assault weapon” today you are part of the problem. You know
what an assault weapon is, and you know you don’t need one.” The tweet was popular, being liked and retweeted thousands of times. Amongst the replies, one stood out. JOURNAL OF DIGITAL SOCIAL RESEARCH — VOL. 4, NO. 1, 2022 user at the time, responded, “Legit question for rural Americans – How do I kill
the 30-50 feral hogs that run into my yard within 3-5 mins while my small kids
play?” (see Figure 2) Figure 2: @JasonIsbell and @WillieMcNabb tweets Figure 2: @JasonIsbell and @WillieMcNabb tweets The next day the jokes began, and the absurdity of the response, along with the
format of the original tweet, made it ideal meme material. The meme quickly
became intertextual - remixed as song lyrics, as alternative movie titles, and with
already existing memes. This meme was also followed by serious political and policy
discussion, with stories following about the feral hogs problem that exists in much
of rural America (see, for example, Reply All, 2019). While these memetic moments are focused on humour, they are also political
in nature, as well as touching on broader questions of class, gender and violence in
contemporary American life. In the latter parts of this paper, we will examine how
these memes develop on Twitter, with a particular focus on how quickly they
emerge and disappear. 4
CASE STUDIES: MEMETIC MOMENTS William McNabb, who was not a high-profile 30 JOURNAL OF DIGITAL SOCIAL RESEARCH — VOL. 4, NO. 1, 2022 JOURNAL OF DIGITAL SOCIAL RESEARCH — VOL. 4, NO. 1, 2022 SMITH & COPLAND — MEMETIC MOMENTS two terms, while the 30-50 feral hogs meme only included one term. The terms
and dates of the data analysis were as follows: two terms, while the 30-50 feral hogs meme only included one term. The terms
and dates of the data analysis were as follows: 1.) “Hill staffer sent me this pic” and "girl who struggles": 15th November 2018
– 22nd November 2018 2.) “30 - 50 feral hogs”: 4th August 2019 – 11th August 2019 These searches did not include hashtags, although hashtagged material does
appear within each data set. We chose these search terms as they represented the
core ‘memeable’ aspects of the original tweet, or the parts of the original tweet
required for subsequent memes to be recognised and referencing the original. After
we collected the data, we then cleaned each data set, removing duplicate tweets,
tweets in foreign languages and any tweets not associated with the relevant meme. In total, this created a dataset of 8,585 and 54,086 tweets, respectively. There are some limitations in these data sets. First, our data set only contains
original tweets associated with the terms and does not include retweets. While the
data set contains the number of retweets (at time of collection) for each tweet, we
cannot see who has retweeted these tweets, and this makes further analysis of these
retweets impossible. This limits our understanding of who was engaged in the
relevant memes, particularly those who participated by sharing other people’s
material. Second, our data set is limited to tweets containing the exact terms
searched for each meme. This means that tweets that contain spelling mistakes, or,
more importantly, tweets that may include the creative twists on the original tweet,
are not included in the data set. This likely represents tweets that are memes of the
original meme. The search terms also limit our examination of conversations
associated with the memes. If a user replies to a tweet about the meme, for example,
but does not include the relevant words, these tweets are not included in the data
set. However, as our focus in this research is on the spread of memes connected to
these specific terms, this does not necessarily limit our examination of conversations
and engagement around the content too heavily. We have analysed the data using the analysis tool Tableau. Tableau provides
quick and clean analysis options for digital data. 5
METHODOLOGY We collected data associated with two internet memes across 2018 and 2019. Data
was collected from Twitter using the package Twint (2021). For each meme, we
collected one week’s worth of data, searching for terms associated with the meme,
with data collection beginning on the date of the initial tweet that sparked the
memetic moment. Data were initially collected for a period longer than one week,
but our analysis has been limited to this time period as very little engagement
occurred outside it. Data collected included raw tweets and associated metadata,
such as retweet count, favourite count, and mentions of other users. The meme
associated with the American Congresswoman Alexandra Occasio Cortez included 31 SMITH & COPLAND — MEMETIC MOMENTS SMITH & COPLAND — MEMETIC MOMENTS Using this tool, we have been able
to analyse the development of each meme across time, as well as study the key users
associated with each meme. We have been able to use this tool to then conduct an
in-depth analysis of the development of each meme over time, with a specific focus
on the starting points of the meme - i.e. the moments in which memes were
sparked. As we detail below, we argue that the moments that spark these memes
are essential to understanding both the context and their temporal structure. 32 32 JOURNAL OF DIGITAL SOCIAL RESEARCH — VOL. 4, NO. 1, 2022 6
DATA ANALYSIS 6 The first clear observation associated with these datasets relates to the size of each. As already noted, the datasets consisted of 8,585 (Hill staffer sent me this pic/girl
who struggles)and 54,086 (“30-50 feral hogs) tweets, respectively. We can see a
large discrepancy between the meme associated with the American politician
Alexandria Ocasio-Cortez (AOC) and the 30-50 feral hogs tweet. As we will
discuss later, this likely represents that the meme associated with AOC only reached
a relatively small and niche audience of more politically minded tweeters, while the
30-50 feral hogs meme was able to go beyond this smaller community. This, we
theorise, is in part due to gun violence affecting far more people in the United
States, resulting in the meme having greater spread across the country. Our first key question in our data analysis was to ask how quickly did the relevant
memes spread online and what were the key moments associated with this spread? To answer this question, we created temporal graphs of the two datasets, examining
the number of tweets per hour following the initial tweet, which started the
memetic moments. Figure 3: "hill staffer sent me this pic” and "girl who struggles" tweets over time Figure 3: "hill staffer sent me this pic” and "girl who struggles" tweets over time As is observable in Figures 3 (above) and 4 (below), the two memes developed in a
similar fashion. On the left-hand side of each graph is the initial tweet, which is
then followed by a very large spike in activity. These spikes consist of 1,270 an hour
for the AOC meme, and 3,715 tweets in the case of the 30-50 feral hogs meme. 33 SMITH & COPLAND — MEMETIC MOMENTS This spike can last as little as two to three hours, as in the meme associated with
AOC, or approximately 12 hours, as in the case of the 30-50 feral hogs memes. Spikes in activity all occur during the daytime hours in the United States, where This spike can last as little as two to three hours, as in the meme associated with
AOC, or approximately 12 hours, as in the case of the 30-50 feral hogs memes. AOC, or approximately 12 hours, as in the case of the 30-50 feral hogs memes. Spikes in activity all occur during the daytime hours in the United States, where
each meme originated. We then see a sharp drop off of activity as Americans go to
sleep, although some other parts of the world pick up on the discussion during their
working hours. We then see a small bump of activity on the following day in the
United States, although it never reaches peaks anywhere close to the original period
of activity. Activity associated with each meme then trails off sharply. Figure 4: 30-50 Feral Hogs tweets over time Figure 4: 30-50 Feral Hogs tweets over time What we can see is a very quick spread of these memes, followed by an almost as
quick decline. Similar to Venturini’s (2019) notion of ‘junk’, these memetic
moments represent bubbles - like sugar highs - that burst quickly and then almost
vanish entirely. Extending Lefebvre’s work on cyclical and linear analyses of
rhythm, we can understand these memes as part of the cyclical nature of Twitter as
they exist in opposition to the linear or to the “monotonous, tiring and even
intolerable” (2004, p. 76) As part of our analysis of the role of speed in these memetic moments, we
want to analyse how the memes develop over time. We are particularly interested
in the early stages – the moments that result in the immediate spikes of popularity
for each of these three memetic moments. As Jenkins (2014) notes, mapping the 34 JOURNAL OF DIGITAL SOCIAL RESEARCH — VOL. 4, NO. 1, 2022 JOURNAL OF DIGITAL SOCIAL RESEARCH — VOL. 4, NO. 1, 2022 rate of circulation of memes is often difficult due to their emergence in numerous
spaces simultaneously. We note these challenges through our initial temporal
analysis with Tableau, which, while giving us a sense of spread over time, does not
analyse links between each tweet, simply positioning each along a timeline as
individual nodes. We cannot see, for example, if and how one tweet is inspired by
the one before it. Tableau also allows us to examine each tweet in closer depth along
the timeline, providing opportunities for a qualitative analysis of tweets at particular
moments. Using this capability, we have studied the most popular tweets (in terms
of both likes and retweets) for each meme that occur at the base of each of these
spikes – i.e. the tweets that are the first spark of the creation and spread of the
meme. We have done this to identify themes and structures associated with each
meme. This provides us with insight into the temporal patterns of these memes. Through this analysis, we identify two ways these memes spread at the speed
they do. The first is viewable in the “hill staffer sent me this pic” and "girl who
struggles" meme. This meme has the fastest incline of the two memes, with an
immediate escalation with virtually no lead-in time. This is driven primarily by a
small number of high-profile ‘verified’ Twitter users, who provided a template for
the meme. Figure 5: Bob Schooley tweet Figure 5: Bob Schooley tweet Figure 5: Bob Schooley tweet 35 1 https://twitter.com/Rschooley/status/1063188511589752833
2 https://twitter.com/ndrew_lawrence/status/1063189114323910656 p
y
2 https://twitter.com/ndrew_lawrence/status/1063189114323910656 1 https://twitter.com/Rschooley/status/1063188511589752833 SMITH & COPLAND — MEMETIC MOMENTS The most popular tweet1 in the early stages of the meme development is from Bob
Schooley, an American screenwriter, who, at the time of writing, is a verified
Twitter user with over 100,000 followers. Schooley tweeted a direct copy of the
initial tweet, with a replaced image (see Figure 5), collecting 3,193 likes and 320
retweets. Here we see the first iteration of the meme, in which Schooley replaces
the image of Ocasio-Cortez with a woman on a fashion show runway wearing an
oversized hat and large jacket. The extravagance of the runway look implicitly pokes
fun at the original tweet. 3 https://twitter.com/swin24/status/1063189546677035008 Figure 6: Andrew Lawrence tweet Figure 6: Andrew Lawrence tweet Figure 6: Andrew Lawrence tweet Other popular tweets at this stage follow a similar pattern. The two other most
popular tweets in this early stage of the meme are tweeted mere minutes after
Schooley’s tweet. Andrew Lawrence (the Deputy Director of Rapid Response at
Media Matters who has over 50,000 followers) and Asawin Suebsaeng (a senior
political reporter at The Daily Beast with over 80,000 followers), tweeted content
with 240 and 160 likes, respectively. Lawrence takes a slightly different approach2,
featuring an image of a woman wearing a barrel (see Figure 6), and replacing the 36 JOURNAL OF DIGITAL SOCIAL RESEARCH — VOL. 4, NO. 1, 2022 JOURNAL OF DIGITAL SOCIAL RESEARCH — VOL. 4, NO. 1, 2022 JOURNAL OF DIGITAL SOCIAL RESEARCH — VOL. 4, NO. 1, 2022 JOURNAL OF DIGITAL SOCIAL RESEARCH — VOL. 4, NO. 1, 2022 word ‘jacket and coat’ with ‘barrel and suspenders’. Meanwhile, Suebsaeng’s tweet3
(see Figure 7), features a scene from the popular TV show The Simpsons, in which
character Milhouse photographs Lisa Simpson in a future scene in which she is
popular, wealthy and cheered on by an adoring crowd. Figure 7: Asawin Suebsaeng tweet
This meme acts as a variation of a ‘template meme’ (Wiggins & Bowers, 2015). However, in this instance, the text becomes the original template, with the image
varied across each version of the meme. This meme acts in a similar way to other
image-based memes, such as distracted boyfriend meme or the image of Bernie
Sanders at the Joe Biden inauguration, in which users shared the same image
repeatedly for comedic sake (and at times to make social and political points). What
is different in this meme is that it is the text that stays the same, while it is the
image that varies with each share. In doing so, it could be read as a version of junk
(Venturini, 2019), with the meme presenting a repetitive, simplified, critical
response to the initial tweet from @eScarry. However, as Lefebvre (2004) notes,
repetition in the meme does not exclude difference, but instead gives birth to it. Each user chooses a unique image to accompany the text of the meme, creating a
meme that is both repetitive and similar at the same time, making it easily
recognisable and copied from other users. Figure 7: Asawin Suebsaeng tweet Figure 7: Asawin Suebsaeng tweet Figure 7: Asawin Suebsaeng tweet Figure 7: Asawin Suebsaeng tweet This meme acts as a variation of a ‘template meme’ (Wiggins & Bowers, 2015). However, in this instance, the text becomes the original template, with the image
varied across each version of the meme. This meme acts in a similar way to other
image-based memes, such as distracted boyfriend meme or the image of Bernie
Sanders at the Joe Biden inauguration, in which users shared the same image
repeatedly for comedic sake (and at times to make social and political points). What
is different in this meme is that it is the text that stays the same, while it is the
image that varies with each share. In doing so, it could be read as a version of junk
(Venturini, 2019), with the meme presenting a repetitive, simplified, critical
response to the initial tweet from @eScarry. However, as Lefebvre (2004) notes,
repetition in the meme does not exclude difference, but instead gives birth to it. Each user chooses a unique image to accompany the text of the meme, creating a
meme that is both repetitive and similar at the same time, making it easily
recognisable and copied from other users. 37 4 https://twitter.com/IWriteAllDay_/status/1063199471520047105 p
y
5 https://twitter.com/davidmackau/status/1063200091664732161 JOURNAL OF DIGITAL SOCIAL RESEARCH — VOL. 4, NO. 1, 2022 up quickly because of the ease of this template, with other users able to quickly grab
other images from the web to participate. It becomes, in turn, a quick way to both
have some fun online, while also making a political point, mixing the absurd with
genuine political critique. Notably, the trend of the early stages of this meme follows throughout its
development. The dataset contains three other highly prominent tweets from
@iwriteallday, @davidmackau and @matthewacherry. Each three of these tweeters
posted a series of tweets parodying the original tweet. @iwriteallday, for example,
posted a tweet4 copying the original text with an image of a man standing in his
bedroom wearing ugg boots, a hoodie and holding a handbag (see Figure 8). @davidmackau, a prominent Australian Twitter user, parodied the tweet 5 by
copying the original text followed by a picture of Anne Hathaway in the movie
Devil Wears Prada (see Figure 9). SMITH & COPLAND — MEMETIC MOMENTS This repetition gives birth to a range of political critiques around the
treatment of women in politics. Initially, these users used the meme to make a
political point. For example, Schooley follows up his initial tweet with a ‘thread’
criticising conservative media and their obsession with criticising Ocasio-Cortez
appearance. The involvement of Lawrence and Suebsaena also indicates the
political nature of the meme, as both are closely embedded in the US political scene,
from which the first tweet originated. The meme, in turn, becomes a political point
in and of itself, with the ridiculousness of the images chosen highlighting the
ridiculous standards to which women such as Ocasio-Cortez are expected to live up
to. to. Figure 8: @iwriteallday tweet
This meme was sparked by a small number of highly politically engaged Twitter
users, who saw the original tweet, aimed to make a political point about it, and in
doing so, provided a template for other users to follow. The meme can be picked Figure 8: @iwriteallday tweet Figure 8: @iwriteallday tweet This meme was sparked by a small number of highly politically engaged Twitter
users, who saw the original tweet, aimed to make a political point about it, and in
doing so, provided a template for other users to follow. The meme can be picked 38 JOURNAL OF DIGITAL SOCIAL RESEARCH — VOL. 4, NO. 1, 2022 JOURNAL OF DIGITAL SOCIAL RESEARCH — VOL. 4, NO. 1, 2022 JOURNAL OF DIGITAL SOCIAL RESEARCH — VOL. 4, NO. 1, 2022 6 https://twitter.com/MatthewACherry/status/1063207962410434560 Figure 9: @davidmackau tweet Figure 9: @davidmackau tweet 39 7 https://twitter.com/BarbiturateCat/status/1158450759576502273
8 https://twitter.com/tinybaby/status/1158418843393331200 SMITH & COPLAND — MEMETIC MOMENTS Finally, @matthewacherry posted several parody tweets, the most popular of which
included a photo of one of the main characters from the Netflix show Big Mouth,
Nick, a teenage boy going through puberty; in this image wearing a fancy jacket at
school6 (see Figure 10). Finally, @matthewacherry posted several parody tweets, the most popular of which
included a photo of one of the main characters from the Netflix show Big Mouth,
Nick, a teenage boy going through puberty; in this image wearing a fancy jacket at
school6 (see Figure 10). Figure 10: @MatthewACherry tweet Figure 10: @MatthewACherry tweet
These users are relatively prominent journalists and political operatives, with high Figure 10: @MatthewACherry tweet These users are relatively prominent journalists and political operatives, with high
follower numbers and verification on Twitter. The meme is developed and spread
primarily by a small number of politically engaged users. These users are “in the
know” and each take and spread the template across a short period of time. The second way in which we see the memes develop is an inverse of the
approach from the “hill staffer sent me this pic” and "girl who struggles" meme. Instead of being driven by a small number of high-profile verified users, the 30-50
feral hogs meme is instead driven by a high number of users who are not ‘high
profile’, but who have still posted extremely popular tweets. We theorise that the
ability of non-verified users to lead this meme development is in part due to the
high-profile nature of the original conversation. Specifically, @WillieMcNabb’s
initial tweet was a reply to a tweet from the very popular Jason Isbell, providing
visibility. Isbell’s initial tweet gained over 50,000 likes and 7,000 retweets, giving it
@WillieMcNabb’s reply an immediately large audience. 40 7 https://twitter.com/BarbiturateCat/status/1158450759576502273
8 h
//
i
/ i
b b /
/1158418843393331200 p
8 https://twitter.com/tinybaby/status/1158418843393331200 9 https://twitter.com/JasonArnopp/status/1158494270170501120 JOURNAL OF DIGITAL SOCIAL RESEARCH — VOL. 4, NO. 1, 2022 What we see following the initial conversation is a range of users riffing off
@WillieMcNabb’s tweet, with some being popular and others falling flat. This
meme took some time to build, with users tweeting the term for several hours before
we see the main spike of the meme. Similar to the “hill staffer sent me this pic” and
"girl who struggles" meme, the 30-50 feral hogs meme then takes off sharply. This
is driven by several different riffs on the content, with users coming up with a range
of different ways to poke fun at the original tweet. The most popular of these at the
early stage of the meme is from @BarbiturateCate7 (see Figure 11), which replicates
a template from a dating site, in which users can pick between being ‘male’, ‘female’
or ’30-50 feral hogs’, and seeking ‘male’, ‘female’ or ‘a yard with unsupervised small
children to run in to within 3-5 minutes’. @BarbiturateCat notes the immense
popularity of her tweet herself, replying to it some hours later “Ok I think this is
my best tweet I must now retire from Twitter dot com forever.” Figure 11: @BarbiturateCat Figure 11: @BarbiturateCat Figure 11: @BarbiturateCat Other users take this in different directions. For example, @kamilumin tweets “30-
50 FERAL HOGS IN YOUR AREA CHAT NOW”, satirising common pop-up
advertisements, which advertise sex work services by stating that ‘women are in your
area’ and encouraging men to ‘chat now!’. Others just make fun of the scenario that
@WillieMcNabb sets up. For example, @tinybaby uses a popular tweet8 format to
ridicule McNabb (see Figure 12). This format of tweet mocks people for saying
things that no one else is thinking – starting with the word “Nobody” followed by
silence and then repeating that with “No one” and “Not a single person” until the 41 SMITH & COPLAND — MEMETIC MOMENTS SMITH & COPLAND — MEMETIC MOMENTS initial content is thrown in. This highlights the seemingly ridiculous nature of the
initial tweet, poking fun at McNabb for saying something that no one else would
even contemplate. initial content is thrown in. This highlights the seemingly ridiculous nature of the
initial tweet, poking fun at McNabb for saying something that no one else would
even contemplate. Figure 13: @JasonArnopp tweet The 30-50 feral hogs meme operated in an inverse nature to the “hill staffer sent
me this pic” and "girl who struggles" meme, with ‘regular’ users driving the content
in a non-template form. Unlike the previous example, 30-50 feral hogs develops in
a range of different directions, with users using their creativity to riff off the meme. However, we theorise that the very potential of the meme is driven by the initial
(unwitting) involvement of Jason Isbell, whose initial tweet provided significant
exposure to Willie McNabb’s reply. The speed of this meme is driven by this initial
exposure, with users riffing off this and taking it in a range of different directions. After initial slowness, the meme then takes a life of its own, spreading quickly across
Twitter, becoming incorporated in other social media memes. Figure 12: @tinybaby tweet Notably, unlike the “hill staffer sent me this pic” and "girl who struggles" meme,
these tweets are largely non-political in nature. While many replies to
@WillieMcNabb’s tweet criticise his position around gun control, the meme itself
heads off in a non-political route. Instead, users take the seeming ridiculousness of
the initial tweet into further ridiculous directions. One of the most popular tweets
is from the user @nomiddlesliders and reads, “take me down to the paradise city
where the hogs are feral and there's 30-50”. This tweet replaces the lyrics of the
popular song Paradise City by Guns N’ Roses, but in doing so, injects a moment of
humour around a difficult debate. At the time of writing, @nomiddlesliders has
only 439 followers, yet this tweet amassed over 59,000 likes and 10,000 retweets,
once again highlighting the capacity of ‘regular users’ to spread this meme. In comparison, the hill staffer sent me this pic/girl who struggles meme followed
a template approach, in which the same text was replaced by different images. This
meme takes off in a range of different directions. While the term “30-50 feral hogs”
remains central, each tweet has a different theme - from jokes about the dating
industry, attempts to mock the initial tweeter, and riffs off song titles. The flexible
format of this memetic moment allows for a wide variety of creative responses. While initial tweets do not follow a specific template, the 30-50 feral hogs
meme then gets adapted to popular memes on social media platforms such as
Twitter. The author @jasonarnopp tweeted a call out for people to tweet in ‘30-50
feral hogs movies’9 (see Figure 13). This is part of a recurring meme that occurs on
Twitter where users replace movie or other popular culture titles with a suggested 42 JOURNAL OF DIGITAL SOCIAL RESEARCH — VOL. 4, NO. 1, 2022 JOURNAL OF DIGITAL SOCIAL RESEARCH — VOL. 4, NO. 1, 2022 Figure 13: @JasonArnopp tweet JOURNAL OF DIGITAL SOCIAL RESEARCH — VOL. 4, NO. 1, 2022 word. @jasonarnopp’s tweet received over 1,000 replies with people sending in
movie titles ranging from “When Harry Met 30-50 Feral Hogs” to “How to lose
30-50 feral hogs in ten days”. Other prominent users in the data were also associated
with similar tweets. @wildheartreads, for example, started a ’30-50 feral hogs book
titles’ thread - i.e. “Interview with 30-50 Feral Hogs”, which received over 100
replies, while @four4thefire also tweeted a separate ’30-50 feral hogs movie titles’
thread, which also received over 100 tweets. word. @jasonarnopp’s tweet received over 1,000 replies with people sending in
movie titles ranging from “When Harry Met 30-50 Feral Hogs” to “How to lose
30-50 feral hogs in ten days”. Other prominent users in the data were also associated
with similar tweets. @wildheartreads, for example, started a ’30-50 feral hogs book
titles’ thread - i.e. “Interview with 30-50 Feral Hogs”, which received over 100
replies, while @four4thefire also tweeted a separate ’30-50 feral hogs movie titles’
thread, which also received over 100 tweets. SMITH & COPLAND — MEMETIC MOMENTS divisive issue. Because of this, we were not expecting to see the data comprising
primarily good-natured involvement. The speed at which the memes grew and ‘burst’ seemed to create a space for
low-stakes, non-serious and wholesome engagement. The memetic moments also
reinforced the standing of social media spaces like Twitter as networked publics
(Ito, 2008), ones that have their own rules and ways of surfacing and responding to
content outside of Twitter’s attempts at algorithmic curation. It can be easier to dismiss memes as artefacts of internet vernacular, but
memetic moments are also additionally generative outside of the confines of Twitter
itself. Lefebvre’s (2004) work is again useful here, specifically his distinction
between the present and presence. He defines the difference between these two
states as follows: “The present is a fact and effect of commerce; while presence
situates itself in the poetic: value, creation, situation in the world and not only in
the relations of exchange” (Lefebvre 2004, p. 47). More simply, in a digital sense,
the present belongs to the commercialised attention economy, to algorithms that
attempt to hack user engagement. Presence is the ambivalent or weird internet
(Philips & Milner 2017), the emergence of memetic moments that are creative,
sometimes poetic and fundamentally linked to the social world, not just relations of
exchange. The generation of dialogue also categorises Lefebvre’s concept of
presence; this is evident in the two memes chosen as case studies for this paper. For
example, the 30-50 wild hogs created further discussions about gun control and
environmental management. Reply All’s (2019) deep dive into the meme
highlighted how wild hog populations are exploding and all but unmanageable for
many farmers in the south of America. Wild hogs not only destroy farmed crops
but can also have a devastating impact on local ecosystems. Effective management
of wild hogs is mired in bureaucracy and politics, leaving landowners to manage the
problem on their own, usually with the help of high-powered automatic weapons. While the meme hinges on the ridiculousness of citing rampaging wild hogs as a
reason for not advancing restrictions on automatic weapons, the memetic moment
also has a ‘long tail’ of discussion and analysis long after the initial meme bubble
has burst. The creativity evident in the production of memes on Twitter affords a
social presence. 7
DISCUSSION As part of our data collection process, we manually cleaned the tweets collected,
removing irrelevant content, and content in languages other than English. To do
this, we read each tweet to determine its relevance and suitability for analysis. However, this task also sensitised us to a surprising aspect of the data. The data
contained almost no angry, hateful, sexists, racist or otherwise abusive tweets. This
was unexpected. Twitter receives a lot of attention for the abusive behaviour on its
platform, which appears to be widespread and part of a systemic failure of design
and governance. The topics of the memetic moments we collected data for, Ocasio-
Cortez and gun control, are political hot button issues. Ocasio-Cortez is abused for
her gender, race and political beliefs. Gun control and gun violence is a particularly 43 CONCLUSION 8 In this paper, we identified memetic moments as primarily text-based and as an
emergent property of activity on social media. Previous research has focused on
social network analysis of Twitter and memes. These studies (Dang et al. 2019;
Jafari Asbagh 2014; Segev et al. 2015; Xu et al. 2016) have largely focused on
political discussion and the spread of hateful material or disinformation on the
platform. Despite their similarity, the memetic moments in this paper demonstrate how
humour functions as part of the speed of online platforms. Speed is not just a
destructive function of digital platforms, but can also create moments of creative
engagement and dialogue in online spaces. The creative, repetitious and quickly-
paced circulation of memes on Twitter creates a sense of ‘presence’ (Lefebvre,
2004). Repetition creates enough difference for engagement, while still producing
identifiable content to create a clear ‘moment’. This facilitates a quick uptake, with
users building off the success of previous tweets in order to turn these moments into
memes. As we noted, this occurs in two primary ways - one driven by high profile
users, and another by ‘regular’ users, who build off a high-profile conversation to
turn it into a meme. While our analysis is limited to only two memes, we argue that
this provides insight into the temporal rhythms of memes on Twitter, extending
our understanding of how aspects of time, such as speed, function in digital cultures. As Segev et al. (2015) highlight, memes are often considered trivial and/or
mundane parts of the online environment. Memes are playful, silly and sometimes
fleeting. Aside from their political resonance, these memetic moments remind us
of the positive promise of the internet that often seems like a lost dream,
collaborative, fun, intimate. The condition of speed that produces virality can be
largely blamed for many of the more toxic elements of social media, including the
persistent spread of misinformation. Still, speed can also work to knit publics
together. Alongside understanding the harms of social media and the internet more
broadly, it also seems important to focus on how social media can produce afford
presence, and as an extension, dialogue. Despite the ongoing critique of speed in
scholarly literature, speed does not straightforwardly impoverish communication,
but can be part of the joyful, creative and poetic rhythms of the social world
(Lefebvre, 2004). SMITH & COPLAND — MEMETIC MOMENTS Similar is true for the memetic moments centring on Ocasio-
Cortez, which generated a number of opinion pieces about gender (Bell, 2018),
dress and class (De Valle, 2018). While memetic moments are spontaneously
generated sites of engagement, they also serve as a springboard for deeper discussion
across other forms of media, underscoring the political and cultural power of
memes. The ‘long tail’ of memetic moments also illustrates how intertwined
‘traditional’ media has become with social media. Traditional media outlets rely
heavily on Twitter to both generate and editorialise content. Each memetic
moment/bubble detailed in this paper generated a slew of media think pieces (e.g. Bell, 2018; De Valle, 2018). Likewise, comments on Twitter about these events are
also used to illustrate journalistic content, like a digital vox pops. 44 JOURNAL OF DIGITAL SOCIAL RESEARCH — VOL. 4, NO. 1, 2022 JOURNAL OF DIGITAL SOCIAL RESEARCH — VOL. 4, NO. 1, 2022 FUNDING STATEMENT AND ACKNOWLEDGMENTS The author(s) received no financial support for the research, authorship, and/or
publication of this article. The authors would like to acknowledge the two
anonymous reviewers for their helpful feedback and Dr Clare Southerton for advice
and feedback. 45 SMITH & COPLAND — MEMETIC MOMENTS SMITH & COPLAND — MEMETIC MOMENTS REFERENCES Abidin C. and Kaye, B. (2021). Audio Memes, Earworms, and Templatability:
The ‘Aural Turn’ Of Memes” in eds. Arkenbout, C, Jack Wilson and
Daniel de Zeeuw, Critical Meme Reader: Global Mutations of the Viral Image (
pp. 58-68). Amsterdam: Institute of Network Cultures. Ackland, R. (2013). Web social science: concepts, data and tools for social scientists in
the digital age. London: SAGE. g
g
Bapna A., and Lokhande, A. (2021).“Post-Truth Design Tactics in Memes in
India: 198 Decoding Malicious Design Practices in Right-Wing Meme
Factories” in eds. Arkenbout, C, Jack Wilson and Daniel de Zeeuw,
Critical Meme Reader: Global Mutations of the Viral Image (pp. 198-216). Amsterdam: Institute of Network Cultures. Beck, U. (2000). The Brave New World of Work. Cambridge: Polity Press. Bell, C. (2018, November 16). Journalist criticised for ‘sexists’ Ocasio-Cortez
tweet. BBC News. Retrieved from: https://www.bbc.com/news/blogs-
trending-46233260. Benwell, M. and Paul, K. (2019, August 6). What about the 30-50 feral hogs? Man's defence of assault weapons goes viral. The Guardian. Retrieved from:
https://www.theguardian.com/us-news/2019/aug/05/feral-hogs-memes-
twitter-30-50-running-into-my-yard-small-kids. Cook, H. (2019, October 9). Facebook removing likes is good for social media. Mumbrella. Retrieved from: https://mumbrella.com.au/facebook-removing-
likes-is-good-for-social-media-601557. Dang, L., Chen, Z., Lee, J., Tsou, M.-H. & Ye, X. (2019). Simulating the spatial
diffusion of memes on social media networks. International Journal of
Geographical Information Science, 33(8), 1545–1568,
https://doi.org/10.1080/13658816.2019.1591414 DeLuca, K. M. (2006). “The Speed of Immanent Images,” in ed. Hope, D. Visual Communication: Perception, Rhetoric, and Technology (pp. 79-90). New
York, NY: Hampton Press. De Valle, G. (2018, November 16). The real reason conservative critics love
talking about Alexandria Ocasio-Cortez’s clothes, Vox. Retrieved from:
https://www.vox.com/the-goods/2018/11/16/18099074/alexandria-ocasio-
cortez-clothes-eddie-scarry. Gal, N. Shifman, L. & Kampf, Z. (2016). “It Gets Better”: Internet memes and
the construction of collective identity. New Media & Society, 18(8), 1698–
1714, https://doi.org/10.1177/1461444814568784. Giddens, A. (1990). The Consequences of Modernity. Cambridge: Polity Press. q
f
y
Harvey, D. (1989). The Condition of Postmodernity. Oxford: Blackwell. Hatmaker, T. (2020, October 10). Changing how retweets work, Twitter seeks to
slow down election misinformation. TechCrunch. Retrieved from: 46 JOURNAL OF DIGITAL SOCIAL RESEARCH — VOL. 4, NO. 1, 2022 JOURNAL OF DIGITAL SOCIAL RESEARCH — VOL. 4, NO. 1, 2022 https://social.techcrunch.com/2020/10/09/twitter-retweet-changes-quote-
tweet-election-misinformation/. JafariAsbagh, M., Ferrara, E., Varol, O., Menczer, F., & Flammini, A. (2014). Clustering memes in social media streams. Social Network Analysis and
Mining 4(1), 237, https://doi.org/10.1007/s13278-014-0237-x. Jenkins, E. S. (2014). ‘The Modes of Visual Rhetoric: Circulating Memes as
Expressions’, Quarterly Journal of Speech, 100(4), 442–466,
doi:10.1080/00335630.2014.989258. Lash, S. (2002). Critique of Information. London: Sage. Lears, R. (2019). Knock Down the House. Netflix. Lears, R. (2019). Knock Down the House. Netflix. Lefebvre, H. (2004). Rythmanalysis: Space, Time and Everyday Life. New York:
Continuum. Losse, K. (2012). The Boy Kings: A Journey into the Heart of the Social Network. New York: Simon and Schuster. Massanari, A. (2017). #Gamergate and The Fappening: How Reddit’s algorithm,
governance, and culture support toxic technocultures. New Media and
Society, 19(3), 329–346, https://doi.org/10.1177/1461444815608807. y
Miltner, K. (2014). “There’s no place for lulz on LOLCats”: The role of genre,
gender, and group identity in the interpretation and enjoyment of an
Internet meme. First Monday 19(8), https://doi.org/10.5210/fm.v19i8.5391. Miltner, K. (2018). Internet memes. In J. Burgess A. Marwick, & T. Poell The
sage handbook of social media (pp. 412-428). London: SAGE. https://www.doi.org/10.4135/9781473984066.n23 Mirani, L. (2020, June 29). No more going viral: why not apply social distancing
to social media? The Guardian. Retrieved from:
https://www.theguardian.com/commentisfree/2020/jun/29/social-
distancing-social-media-facebook-misinformation. Nissenbaum, A. & Shifman, L. (2017). Internet memes as contested cultural
capital: The case of 4chan’s /b/ board. New Media & Society, 19(4), 483–
501, https://doi.org/10.1177/1461444815609313 O’Neil, M. & Jensen M. (2020). Australian Perspectives on Misinformation. News and Media Research Centre: University of Canberra. Retrieved from:
https://www.canberra.edu.au/research/faculty-research-
centres/nmrc/research/australian-perspectives-on-misinformation. Phillips, W. & Milner, R. M. (2017). The Ambivalent Internet: Mischief, Oddity,
and Antagonism Online. Cambridge, UK: Polity Press. Reply All (2019, October 10). “#149 30-50 Feral Hogs”. Gimlet Media. Retrieved
from: https://gimletmedia.com/shows/reply-all/n8hw3d. Rosenberg, E. (2018, November 16). Alexandria Ocasio-Cortez wore clothing, a
journalist tweeted a photo, and the internet pounced. The Washington Post. Retrieved from:
https://www.washingtonpost.com/nation/2018/11/16/eddie-scarry-tweet-
about-alexandria-ocasio-cortezs-clothing-is-latest-critique-about-class/. 47 SMITH & COPLAND — MEMETIC MOMENTS Salter, M. (2018). From geek masculinity to Gamergate: the technological
rationality of online abuse Crime, Media, Culture, 14(2), 247–264,
https://doi.org/10.1177/1741659017690893. Segev, E., Nissenbaum, A., Stolero, N., & Shifman, L. (2015). Families and
Networks of Internet Memes: The Relationship Between Cohesiveness,
Uniqueness, and Quiddity Concreteness. Journal of Computer-Mediated
Communication, 20(4), 417–433, https://doi.org/10.1111/jcc4.12120. Souroush V., Roy D. & Aral, S. (2018). The Spread of True and False News
Online. Science, 359(6380), 1146-1151, doi: 10.1126/science.aap9559. Srnicek, N. (2017). Platform Capitalism. Cambridge, UK: Polity Press. Terranova, T. (2012). Attention, economy and the brain. Culture Machine, 13, 1–
19. Retrieved from: https://culturemachine.net/wp-
content/uploads/2019/01/465-973-1-PB.pdf. Tuters M., Jokubauskaité E. & Bach D. (2018). Post-truth protest: how 4chan
cooked up the Pizzagate bullshit. M/C Journal, 21(3),
https://doi.org/10.5204/mcj.1422. Thrift, N. (1996). New Urban Eras and Old Technological Fears: Reconfiguring
the Goodwill of Electronic Things’, Urban Studies, 33(8), 1463– 93,
https://doi.org/10.1080/0042098966754. Twint (2021). Twint Project. Retrieved from:
https://github.com/twintproject/twint. Venturini, T., Jacomy, M. & Pereira D. (2016). Visual Network Analysis. Retrieved from: http://www.tommasoventurini.it/wp/wp-
content/uploads/2014/08/Venturini-Jacomy_Visual-Network-
Analysis_WorkingPaper.pdf. Venturini, T. (2019). From fake to junk news, the data politics of online virality. In D. Bigo, E. Isin and Ruppert E. (eds), Data politics: worlds, subjects, rights
(pp. 123-144). London: Routledge. Venturini, T. (2019). From fake to junk news, the data politics of online virality. In D. Bigo, E. Isin and Ruppert E. (eds), Data politics: worlds, subjects, rights
(pp. 123-144). London: Routledge. Virilio, P. (2000). Open Sky. London: Verso. Wiggins, B.E., & Bowers, G.B. (2015) Memes as genre: A structurational
analysis of the memescape. New Media & Society, 17(11), 1886–1906,
https://doi.org/10.1177/1461444814535194. 48
|
https://openalex.org/W2477219651
|
https://www.adv-radio-sci.net/13/169/2015/ars-13-169-2015.pdf
|
English
| null |
Fast algorithm for radio propagation modeling in realistic 3-D urban environment
|
Advances in radio science
| 2,015
|
cc-by
| 3,244
|
2.1
3-D Scenario In order to evaluate the performance of next generation mo-
bile communication systems in realistic deployment scenar-
ios, system level simulation tools must be capable of inte-
grating various models that reflect network deployment (e.g.,
antenna locations, elevation, orientation, transmit powers),
user mobility and communication behavior, as well as ser-
vice characteristics. The need for modeling system aspects
and use cases of next generation mobile networks in a more
realistic manner system is already acknowledged by Euro-
pean 5G research project ICT-317669 METIS (2013). In this
paper we present a fast method for computing the path loss
of micro and macro cells in a realistic, three dimensional ur-
ban environment scenario for moving and stationary users. Therefore, we use a simplified recursive ray tracing algo-
rithm, which only takes into account one ray for each user
per base station. Channel fading will be simulated by using
Rayleigh and Rician distributed fading. To achieve a realistic
Rician distributed fading, we use a variable k factor that is
randomly created taking into account the respective distance The three-dimensional scenario used in this paper is defined
in ICT-317669 METIS (2013) and is called Madrid Grid sce-
nario. This scenario consists of 15 buildings with different
heights, one park area and corresponding streets between
the buildings. The size of the scenario is 387 by 552 m. An
overview about the scenario is given in Fig. 1. As shown in Fig. 1, there are 13 base stations considered
namely 12 microcells (yellow) and one macrocell (red). The
carrier frequency is 2.6 GHz and 10 MHz have been chosen
as the system bandwidth. Each road consists of two lanes for driving and two park-
ing lanes. The road called “Gran Via” exhibits three lanes for
driving in each direction has no parking lanes. Fast algorithm for radio propagation modeling in realistic 3-D
urban environment A. Rauch, J. Lianghai, A. Klein, and H. D. Schotten
University of Kaiserslautern, Paul-Ehrlich-Straße 11, 67663 Kaiserslautern, Germany ecember 2014 – Revised: 6 April 2015 – Accepted: 26 April 2015 – Published: 3 November 2015 Received: 15 December 2014 – Revised: 6 April 2015 – Accepted: 26 April 2015 – Published: 3 N Abstract. Next generation wireless communication systems
will consist of a large number of mobile or static terminals
and should be able to fulfill multiple requirements depend-
ing on the current situation. Low latency and high packet
success transmission rates should be mentioned in this con-
text and can be summarized as ultra-reliable communications
(URC). Especially for domains like mobile gaming, mobile
video services but also for security relevant scenarios like
traffic safety, traffic control systems and emergency manage-
ment URC will be more and more required to guarantee a
working communication between the terminals all the time. between transmitter and receiver. Both Rayleigh and Rician
fading will be precomputed for the correspondent situations
and just added to the path loss to achieve shorter computa-
tion times. Further, it is demonstrated that the employed al-
gorithm is able to outperform the state-of-the-art approach
described in Sect. 2. The remainder of this paper is organized as follows: in
Sect. 2 we give a short overview about the scenario layout
and the user mobility. In Sect. 3 we present the path loss mod-
els and discuss the commonly used parameters. The compu-
tation of fast fading is presented in Sect. 4 while the simula-
tion results are shown in Sect. 5. 2.2
User mobility The mobility model for vehicular user is adopted from ICT-
317669 METIS (2013). Here, every car moves witch the Fast algorithm for radio propagation modeling in realistic 3-D
urban environment
A. Rauch, J. Lianghai, A. Klein, and H. D. Schotten
University of Kaiserslautern, Paul-Ehrlich-Straße 11, 67663 Kaiserslautern, Germany
Correspondence to: J. Lianghai (ji@eit.uni-kl.de)
Received: 15 December 2014 – Revised: 6 April 2015 – Accepted: 26 April 2015 – Published: 3 November 2015 Published by Copernicus Publications on behalf of the URSI Landesausschuss in der Bundesrepublik Deutschland e.V. Published by Copernicus Publications on behalf of the URSI Landesausschuss in der Bundesrepublik Deutschland e.V. 3.2
Computation of microcell path loss In case of micro cells, which are installed below rooftop level
in an urban environment, most of the signal energy reaches
the user due to reflection between buildings, if the user is
not in line of sight (LoS). Since tracing multiple signal re-
flections is computionally expensive, we choose two separate
models to compute the overall microcell path loss. L = Lfs + Lrts. same constant speed of 50 kmph. Changing directions is only
possible at crossroads. The probability of a right or left turn
is 25 % respectively, while the probability of moving straight
ahead is 50 % at each crossroad. Further details related to this model are shown in ICT-
317669 METIS (2013). A. Rauch et al.: Fast algorithm for radio propagation modeling 170 Tx
Rx
j1
θ1
θ2
j2
S0
S1
S2
j0
j3
d1
d2
d3
Figure 2. Street orientation according to Berg (1995). BS0
BS1
BS2
BS3
BS4
BS5
BS6
BS7
BS8
BS
9
BS
10
BS
11
BS
12
Figure 1. Madrid grid layout according to ICT-317669 METIS
(2013). Tx
Rx
j1
θ1
θ2
j2
S0
S1
S2
j0
j3
d1
d2
d3 Rx j
d Figure 2. Street orientation according to Berg (1995). where where 2 = tan−11h
x
,
and
r =
p
(1h)2 + x2, 2 = tan−11h
x
, and 1h represents the difference of the building height and the
mobile antenna height and x the horizontal distance between
the diffracting edge and the user. The resulting path loss is
calculated as: Figure 1. Madrid grid layout according to ICT-317669 METIS
(2013). Lfs + Lrts. (3) (3) L = Lfs + Lrts. 3.1
Computation of macrocell path loss In order to determine the path loss of macro cells, which are
usually installed on rooftops, we use a path loss model that
is proposed in ICT-317669 METIS (2013). This model takes
into account the diffraction effects that radio signals experi-
ence and that is responsible for most of the signal energy re-
ceived on ground level in urban environments. In this model,
the calculation is divided in two parts. The first part Lfs de-
scribes the free space loss of signal from transmit antenna to
the edge of the rooftop (dr) and can be calculated as: The first model is used, if the user is in line of sight and
the path loss is calculated as shown in is shown in Eq. (4). LdB = −10log10
λ
4πd
,2,
(4) (4) where λ is the wavelength and d is the distance between
transmitter and receiver. The second model is employed when the user is not in
line of sight (nLoS). Here, we use Berg’s recursive path loss
model (Berg, 1995), which takes into account geometrical
conditions, such as angle of crossroads and distance between
two crossroads. In principle, an imaginary distance is calcu-
lated that depends on the real distances between transmitter
and receiver and the angle between two streets. Lfs = −10log10
λ
4πdr
2
. (1) (1) The second part describes the diffraction loss of the signal
from the rooftop to the receiver on street and is calculated as
shown in Eq. (2). As shown in Fig. 2, s1 is the distance between crossing
j1 and j2. s0 is the distance between transmitter j0 and j1,
whereas s2 is the distance between j2 and receiver j3. For Lrts = −10log10
h
λ
2π2r
1
2 −
1
2π + 2
i
,
(2) (2) Adv. Radio Sci., 13, 169–173, 2015 Adv. Radio Sci., 13, 169–173, 2015 www.adv-radio-sci.net/13/169/2015/ A. Rauch et al.: Fast algorithm for radio propagation modeling
1
Figure 3. Calculated path loss for moving user. Figure 4. Movement of user. A. Rauch et al.: Fast algorithm for radio propagation modeling 171 g
p
p g
Figure 3. Calculated path loss for moving user. Figure 4. Movement of user. Figure 4. Movement of user. Figure 3. Calculated path loss for moving user. reflection, diffraction and scattering would lead to a com-
plex model that could not be computed in reasonable time. 4
Computation of fast fading Simulating fast fading in detail according to the three di-
mensional scenario including multipath propagation due to To simulate multipath propagation where at least one path
is much stronger than the others (line of sight), we employ 3.1
Computation of macrocell path loss Since we simulate numerous moving users, the fast fading
will be calculated by using some statistical approaches. As
suggested in Tang and Hongbo (2003), Rayleigh and Rician
distributions are used to generate fast fading. In the following
sections, the utilization of both models is described. the considered dense urban scenario angle 21 and 22 are set
to 90◦. The imaginary distances are calculated according to
Eq. (5) as described in Berg (1995). kj = kj−1 + dj−1 qj−1,
dj = kj sj−1 + dj−1,
(5) (5) where k0 = 1 and d0 = 0. The variable q represents the angle
dependence of the path loss according to Eq. (6). where k0 = 1 and d0 = 0. The variable q represents the angle
dependence of the path loss according to Eq. (6). 4.1
Rayleigh distributed fast fading If there is no dominant propagation path between transmit-
ter and receiver (no line of sight) Rayleigh distributed ran-
dom variables are used to describe fast fading, according to
Eq. (8). qj =
2j
q90
90
v
. (6) (6) In this case, q90 is set to 0.5 and v is set to 1.5 as recom-
mended in Berg (1995). The path loss is now calculated as
follows: PRayleigh(r) = r
σ 2 e
−r2
2σ2 ,
(8) (8) L(n)
dB = −10log10
λ
4 π dn
2
,
(7) (7) where σ 2 represents the energy of the signal and r the magni-
tude. To calculate the Doppler power spectral density, we use
Eq. (9), which is proposed in Kostov (2003), for the Rayleigh
and Rician channel: where dn is the imaginary distance between the respective
transmitter and point n. Figure 3 shows the corresponding path loss characteris-
tic of a user moving with constant speed from point P0 to
point P2 in Fig. 4. The dotted line shows the path loss of the
reference model, which is further described in ICT-317669
METIS (2013) and ITU-R M.2135 (2009). The continuous
line depicts the path loss of the bergs model, that is used in
our approach. This approach leads to a smoother curve that
may be adjusted by using the parameters v and q90 in Eq. (6). S(f ) =
1
πfm
r
1 −
f
fm
2 ,
(9) (9) where fm is the Doppler frequency shift that is calculated as: The corresponding base station is depicted in yellow in
Fig. 4 as well. fm = vfc
c ,
(10) (10) As illustrated in Fig. 4 the user is in line of sight of BS4
until he reaches P2. After turning left, the user is in nLoS
and the corresponding path loss model for a 90◦crossroad is
chosen. v = 50 kmph represents the user speed, fc is the carrier fre-
quency, which is 2.6 GHz in this case and c = 3 × 108 m s−1
is the speed of light. The Rayleigh fading will be pre-
calculated, stored and simply added to the path loss during
the simulation. Fading is not considered here. with K = β2
2σ 2 (Jeruchim et al., 2000; Iskander, 2008). 2σ
For K = 0 and I0
h
rβ
σ 2
i
= 1 Eq. (11) leads again to the
Rayleigh fading calculated in Eq. (8) as proposed in Kostov
(2003). The variable K is called Ricean k factor and rather
describes the relation between the signal of the line of sight
path and the non line of sight paths. As proposed in Thiele
et al. (2006), the non line of sight signal components need to
be summed up to determine the k factor. To avoid simulating
every single signal path we use the statistical approaches to
estimate the k factor, that are described in Thiele et al. (2006),
Medawar et al. (2013) and Greenstein et al. (1999). In Thiele
et al. (2006) and Medawar et al. (2013) functions are pro-
vided that take into account the urban environment and the
distance between transmitter and receiver resulting in a cu-
mulative distribution function (CDF) of the Rician k factor. Figure 7. Ricean fading for k = 15 dB. 5
Simulation results In this section, we will present some simulation results re-
garding the path loss of a moving user in the described sce-
nario. Figure 8 shows the path loss between transmitter and
a moving user that follows the same route as described in
Fig. 4. The Rician k factor for nLoS conditions changes due
to the distance as shown in Fig. 5. The appropriate models for
fading and propagation will be chosen automatically using a
single ray to detect NLoS or LoS conditions. The simulated
path loss from P0 to P2 is shown in Fig. 8. The k factor is now calculated using the line of sight dis-
tance from transmitter to receiver and a random value, which
is generated according to the corresponding probability dis-
tribution function. Figure 5 shows the probability distribution
function (PDF) of the k factor over the distance, where the
y axis represents the distance in meters, the z axis stands for
the probability, and the x axis shows the k factor in dB. Since the main objective of this approach was to find a
method that is fast enough to simulate numerous randomly
distributed user in a realistic, 3-D-environment, we compare
this method to a reference method, that is further described
in ICT-317669 METIS (2013) and ITU-R M.2135 (2009). This method does not precompute the fading and does not
use the recursive method to estimate the path loss. Further the
computation of numerous log-operations is necessary here,
which takes some computing time. In Fig. 9 we compare the
averaged frame refresh rate during a simulation set over an
increasing number of simulated users in the same scenario
and the same user positions. As shown in Fig. 5, the probability of a high k factor value
decreases with an increasing distance between receiver and
transmitter. In our simulation tool the k factor values are
stored separately and used to generate Rician fading for the
relevant distance between transmitter and receiver according
to Eq. (11) (Jeruchim et al., 2000; Iskander, 2008). Figures 6 and 7 show the fast fading for a k factor of 0 and
15 dB respectively. Figures 6 and 7 show the fast fading for a k factor of 0 and
15 dB respectively. For k = 0 the Ricean fading transfers into Rayleigh fading
as shown in Eqs. (11) and (8). www.adv-radio-sci.net/13/169/2015/ www.adv-radio-sci.net/13/169/2015/ Adv. Radio Sci., 13, 169–173, 2015 172
A. Rauch et al.: Fast algorithm for radio propagation modeling
Figure 5. PDF of the ricean k factor. Figure 6. Ricean fading for k = 0 dB. Rauch et al.: Fast algorithm for radio propagation modeling
Figure 6. Ricean fading for k = 0 dB. Figure 7. Ricean fading for k = 15 dB. A. Rauch et al.: Fast algorithm for radio propagation modeling Rauch et al.: Fast algorithm for radio propagation modeling
Figure 6. Ricean fading for k = 0 dB. 172 Figure 5. PDF of the ricean k factor. Figure 5. PDF of the ricean k factor. Figure 6 Figure 5. PDF of the ricean k factor. Figure 6. Ricean fading for k = 0 dB. Figure 7. Ricean fading for k = 15 dB. Rician distributed random variables. Rician distributed fast
fading can be described as: PRice(r) = r
σ 2 e−r2
2σ2 +K I0
h rβ
σ 2
i
,∀r ≥0
(11) (11) with K = β2
2σ 2 (Jeruchim et al., 2000; Iskander, 2008). 5
Simulation results For k = 0 the Ricean fading transfers into Rayleigh fading
as shown in Eqs. (11) and (8). Adv. Radio Sci., 13, 169–173, 2015 www.adv-radio-sci.net/13/169/2015/ A. Rauch et al.: Fast algorithm for radio propagation mo
Figure 8. Path loss with fast fading. Figure 9. Refresh rate vs. number of users. 173 A. Rauch et al.: Fast algorithm for radio propagation modeling Acknowledgements. Part of this work has been performed in
the framework of the FP7 project ICT-317669 METIS – Mobile
and wireless communications Enablers for the 2020 Information
Society, which is partly funded by the European Commission. This
paper expresses the authors views, which are not necessarily those
of the METIS consortium. The authors alone are responsible for
the content of this paper. Edited by: M. Chandra
Reviewed by: two anonymous referees Edited by: M. Chandra
Reviewed by: two anonymous referees Edited by: M. Chandra
Reviewed by: two anonymous referees References Figure 8. Path loss with fast fading. Berg, J.-E.: A Recursive Method For Street Microcell Path Loss
Calculation, Proc. IEEE International Symposium on Personal,
Indoor and Mobile Radio Communications, 1, 140–143, 1995. Greenstein, L., Michelson, D., and Erceg, V.: Moment-Method Es-
timation of the Rician K-Factor, IEEE Comm. Lett., 3, 175–176,
doi:10.1109/4234.769521, 1999. ICT-317669 METIS: Deliverable 6.1, ICT-317669, available at
https://www.metis2020.com, 2013. Iskander, C.-D.: A MATLAB-based Object-Oriented Approach to
Multipath Fading Channel Simulation, white paper, February
2008. ITU-R
M.2135:
Guidelines
for
evaluation
of
radio
inter-
face
technologies
for
IMT-Advanced,
M.2135,
available
at:
https://www.itu.int/dms_pub/itu-r/opb/rep/R-REP-M. 2135-1-2009-PDF-E.pdf, 2009. Figure 9. Refresh rate vs. number of users. Jeruchim, M. C., Balaban, P., and Shanmugan, K. S.: Simulation
of Communication Systems, 2nd Edn., New York, Kluwer Aca-
demic/Plenum, 2000. As illustrated in Fig. 9 we can simulate up to 80 more users
with the method proposed in this paper at a frame refresh rate
of 30 fps than with the reference method. Kostov, N.: Mobile Radio Channels Modeling in MATLAB, De-
partment of Radio Engineering, Technical University of Varna,
RadioEngineering, 12, 12–16, 2003. Medawar, S., Haendel, P., and Zetterberg, P: Approximate Maxi-
mum Likelihood Estimation of Rician K-Factor and Investiga-
tion of Urban Wireless Measurements, Trans. Wireless Comm.,
12, 2545–2555, doi:10.1109/TWC.2013.042413.111734, 2013. 6
Conclusions In this work, we use a simplified model for efficiently com-
puting the path loss between receiver and transmitter in line
of sight conditions as well as in non line of sight conditions
using Berg’s recursive model. To achieve realistic simulation
results we combine this model with fast fading models, such
as Rayleigh distributed fading for non line of sight propaga-
tion. To consider multipath fading in line of sight conditions,
we employ Rician distributed fading with a variable k factor,
which depends on the distance between transmitter and re-
ceiver. The proposed model was compared with a reference
model that does not use pre-computation and recursive algo-
rithms. The model presented in this paper leads to a better
performance that allows to simulate more moving users in a
realistic 3-D-environment. Saunders, S. R.: Antennas and propagation for wireless communi-
cation systems, John Wily & Sons, Ltd., New York, 1999. Tang, L. and Hongbo, Z.: Analysis and Simulation of Nakagami
Fading Channel with MATLAB, Asia-Pacific Conference on En-
vironmental Electromagnetics Hangzhou, 2003. Thiele, L., Peter, M., and Jungnickel, V.: Statistics of the rician k-
factor at 5.2 GHz in an urban macro-cell scenario, IEEE Interna-
tional Symposium on Personal, Indoor and Mobile Radio Com-
munications, 1, 2006. Adv. Radio Sci., 13, 169–173, 2015 www.adv-radio-sci.net/13/169/2015/ www.adv-radio-sci.net/13/169/2015/ www.adv-radio-sci.net/13/169/2015/
|
https://openalex.org/W4282570091
|
https://zenodo.org/records/6636343/files/192.pdf
|
Russian
| null |
DEVELOP CREATIVE COLLABORATION SKILLS THROUGH VERTICAL INTEGRATION OF HIGHER TECHNICAL EDUCATION STUDENTS
|
Zenodo (CERN European Organization for Nuclear Research)
| 2,022
|
cc-by
| 3,532
|
ISSN: 2181-3337 SCIENCE AND INNOVATION
INTERNATIONAL SCIENTIFIC JOURNAL Аннотация. Мақолада олий техник таълим талабаларининг ижодий ҳамкорлик
кўникмаларини ривожлантиришга таъсир кўрсатувчи ижтимоий-сиёсий, миллий-
маънавий, психологик-педагогик омилларнинг ўзаро алоқадорлигини аниқлаш, вертикал
педагогика асосида олий техник таълим муассасаалри талабаларида ижодий ҳамакорлик
кўникмалари таркибий компанентлари бўлган интилувчанлик, изланувчанлик ва
яратувчанлик қобилияти, янгиликка мойиллик, мантиқий, аналитик ва синтетик
фикрлаш қобилияти ва юксак касбий интеллектни ривожлантиришнинг назарий асослари
таҳлил қилинган. Калит сўзлар: техника, олий таълим, талаба, техник таълим, интеграция,
вертикал интеграция, лойиҳалаш, ҳамкорлик, ҳамкорлик педагогикаси. РАЗВИВАТЬ НАВЫКИ ТВОРЧЕСКОГО СОТРУДНИЧЕСТВА ПОСРЕДСТВОМ
ВЕРТИКАЛЬНОЙ ИНТЕГРАЦИИ СТУДЕНТОВ ВЫСШИХ ТЕХНИЧЕСКИХ
УЧЕБНЫХ ЗАВЕДЕНИЙ Аннотация. В статье анализируется взаимодействие социально-политических,
национально-духовных, психолого-педагогических факторов, влияющих на развитие
навыков творческого сотрудничества у студентов высшего технического образования,
стремления, вопрошаемости и творческих способностей, склонности к инновациям,
способности мыслить логически, аналитически и синтетически, а также теоретической
основы для развитие высоких профессиональных навыков в высших технических учебных
заведениях на основе вертикальной педагогики. Ключевые
слова:
техника,
высшее
образование,
студент,
техническое
образование, интеграция, вертикальная интеграция, дизайн, сотрудничество, педагогика
сотрудничества. DEVELOP CREATIVE COLLABORATION SKILLS THROUGH VERTICAL
INTEGRATION OF HIGHER TECHNICAL EDUCATION STUDENTS Annotation. The article analyzes the interaction of socio-political, national-spiritual,
psychological-pedagogical factors influencing the development of creative cooperation skills of
students of Higher Technical Education, the aspiration, questionability and creative ability,
tendency to innovation, the ability to think logically, analytically and synthetically, and the
theoretical basis for the development of high professional skills in higher technical institutions
on the basis of vertical pedagogy. Keywords: technical, higher education, student, technical education, integration, vertical
integration, design, cooperation, collaboration pedagogy. КИРИШ Ўзбекистон таълим тизими ривожининг ҳозирги босқичида ҳамкорлик
педагогикаси муҳим ўрин тутади. Педагог олимлар ҳам-корлик педагогикасига 446 446 ISSN: 2181-3337 SCIENCE AND INNOVATION
INTERNATIONAL SCIENTIFIC JOURNAL «таълим берувчилар ва таълим олувчиларнинг ўзаро мулоқотига асосланган шаклда
таълим беришга йўналтирилган ижодий фаолият жараёнидир»[1], дея таъриф
беришади. Бу мулоқотда таълим берувчи ва таълим олувчи ўзаро масъулият сезади,
томонларнинг ҳамкорлиги таълим сифатининг ошишига хизмат қилади. Ўқитувчининг
ҳамкорликдаги педагогик жараённи ташкил этишга оид фаолияти биринчи навбатда
талабаларнинг жамоада асосий қадрият сифатида ўзларини намоён қилишларини
таъминлашга йўналтирилган ўқув-тарбия жараёнини ташкил этиш йўллари ва
усулларини излашга қаратилиши лозим. Мана шундай ҳамкорлик педагогикаси асосида
вертикал интегрциялашган тизимни яратиш ва бу асосида ўзаро ҳамкорликдаги,
алоқадоликдаги педагогик жараённни ташкил этиш мумкин бўлади. «таълим берувчилар ва таълим олувчиларнинг ўзаро мулоқотига асосланган шаклда
таълим беришга йўналтирилган ижодий фаолият жараёнидир»[1], дея таъриф
беришади. Бу мулоқотда таълим берувчи ва таълим олувчи ўзаро масъулият сезади,
томонларнинг ҳамкорлиги таълим сифатининг ошишига хизмат қилади. Ўқитувчининг
ҳамкорликдаги педагогик жараённи ташкил этишга оид фаолияти биринчи навбатда
талабаларнинг жамоада асосий қадрият сифатида ўзларини намоён қилишларини
таъминлашга йўналтирилган ўқув-тарбия жараёнини ташкил этиш йўллари ва
усулларини излашга қаратилиши лозим. Мана шундай ҳамкорлик педагогикаси асосида
вертикал интегрциялашган тизимни яратиш ва бу асосида ўзаро ҳамкорликдаги,
алоқадоликдаги педагогик жараённни ташкил этиш мумкин бўлади. Айниқса, олий техник таълимдаги ўзига хос ўқитиш тизими ва таълим хусусияти
вертикал интеграциялашган таълимни шакллантиришни талаб этади. Зеро, «биз
замонавий билим ва технологияларни эгаллаган авлодни тарбиялашимиз керак. Табиий-
техник ва ва ихтисослик фанлари бундай мутахассисларни тайёрлаш учун асосий омил
экани ҳам шубҳасиз. Аммо янги авлодни халқ ва жамият, миллат ва давлат манфаатларини
кўзлайдиган, ўзини юрт манфаатлари билан бирга кўрадиган ва ислоҳотларга
дахлдорликни, ижтимоий масъулиятни сезадиган, етук инсон қилиб тарбиялаш»[2]
устуворлик касб этмоқда. Бунинг учун олий техника таълим муассасалари талабаларида
ижодий ҳамкорлик кўникмалрини ривожлантириш ва бу орқали уларда ижодкорлик,
инновацияларни яратишга тайёрлик, жамоа бўлиб технологик тадқиқотларни олиб
боришга доир билим, кўникма ва малакаларни шакллантириш мумкин бўлади. ISSN: 2181-3337 SCIENCE AND INNOVATION
INTERNATIONAL SCIENTIFIC JOURNAL муносабатни шакллантириш, техникани яратиш ва ундан фойдаланишда юксак
инноваторлик позициясида туриш билан боғлиқ бўлган серқирра ва мураккаб жараёндир. Айнан ижодий ҳамкорлик талаба томонидан техника оламига масъулиятли ёндашувни ва
креативликни янги босқичга кўтаради. Натижада техника тўғрисида талабада нафақат
билимлар, балки техник қадриятлар ҳақидаги тасаввурлар ҳам ўзига хос тарзда шакллана
бошлайди. Шу билан бирга, техника йўналиши талабаларида яратувчанлик, бунёдкорлик
фазилатлари ривожланади, аниқроғи ижодий ҳамкорлик ана шундай сифатларни
ривожлантириш учун руҳлантирувчи, илҳомлантирувчи, даъват этувчи восита ролини
бажаради. Шунингдек, фанларни ўқитишда вертикал интеграциялаш тизимини амалга
ошириш, қуйи босқичларда ўтилган фанлардаги таянч билим ва кўникмалар асосида
юқори босқичларда ўтиладиган мутахассислик фанлари интегрция қилиш, бир бирини
такомиллаштириш асосида ўқитиш, янгича билим ва кўникмларни шакллантиришга
йўналтириш самара беради. муносабатни шакллантириш, техникани яратиш ва ундан фойдаланишда юксак
инноваторлик позициясида туриш билан боғлиқ бўлган серқирра ва мураккаб жараёндир. Айнан ижодий ҳамкорлик талаба томонидан техника оламига масъулиятли ёндашувни ва
креативликни янги босқичга кўтаради. Натижада техника тўғрисида талабада нафақат
билимлар, балки техник қадриятлар ҳақидаги тасаввурлар ҳам ўзига хос тарзда шакллана
бошлайди. Шу билан бирга, техника йўналиши талабаларида яратувчанлик, бунёдкорлик
фазилатлари ривожланади, аниқроғи ижодий ҳамкорлик ана шундай сифатларни
ривожлантириш учун руҳлантирувчи, илҳомлантирувчи, даъват этувчи восита ролини
бажаради. Шунингдек, фанларни ўқитишда вертикал интеграциялаш тизимини амалга
ошириш, қуйи босқичларда ўтилган фанлардаги таянч билим ва кўникмалар асосида
юқори босқичларда ўтиладиган мутахассислик фанлари интегрция қилиш, бир бирини
такомиллаштириш асосида ўқитиш, янгича билим ва кўникмларни шакллантиришга
йўналтириш самара беради. ТАДҚИҚОТ МАТЕРИАЛЛАРИ ВА МЕТОДОЛОГИЯСИ ТАДҚИҚОТ МАТЕРИАЛЛАРИ ВА МЕТОДОЛОГИЯСИ Ғарб мамлакатларида вертикал интеграцияш орқали олий техник таълим муассасаси
талабаларида ижодий ҳамкорлик кўникмаларини
ривожлантиришнинг педагогик,
психологик ва фалсафий жиҳатлари кўплаб тадқиқотчилар томонидан ўрганилган. Хусусан, хорижлик олмилар Ж.Мартин, Х.Бек, Ф.Дессауэр, С.Хокинг, Н.Виг, Ж.Эллюль,
М.Хайдеггер, К.Ясперс, Д.Медоузларнинг тадқиқот ишларида мазкур муаммолар таҳлил
қилинган. Мустақил Давлатлар Ҳамдўстлиги олимларидан В.С.Степанов, B.Степин,
М.И.Щадов, Ю.А.Чернегов, Н.Ю.Чернегов ва бошқаларнинг асарларида талабаларда
ижодий ҳамкорлик кўникмаларини ривожлантириш ижтимоий-педагогик жиҳатлари ҳам
ёритилган. Ўзбекистонда ҳам олий техник таълим муассасалари бўлажак мутахассисларида
техник билимларни ошириш, инновацион ёндашувлар асосида ўқитиш, ижодий
ҳамкорлик кўникмаларини ривожлантиришнинг педагогик асосларини А.Авлакулов,
Н.Шермухамедова, О.Файзиев, Э.Усмонов а ва бошқалар томонидан ўрганилган. Бироқ
юқорида келтирилган олимларнинг тадқиқот ишларида вертикал интеграциялаш орқали
олий техник таълим муассасаси талабаларида ижодий ҳамкорлик кўникмаларини
ривожлантиришнинг инновацион педагогик асослари махсус ўрганилмаганлиги ҳам
мазкур тадқиқот ишининг аҳамиятини оширади. Олий техник таълим талабаларининг вертикал интеграциялаш орқали ижодий
ҳамкорлик кўникмаларини ривожлантиришга эҳтиёжни кучайтирувчи қатор асослар
мавжуд. Бу бевосита инсониятнинг техникага нисбатан ижодий ва ҳамкорлик 447 447 ТАДҚИҚОТ НАТИЖАЛАРИ Олий техника таълим муассасаларида юқори босқич талаблари билан қуйи босқич
талабалари ўртасида интеграциялашган ўқитиш, ҳамкорликда ўрганиш, бир бирига ёрдам
беришга асосланган таълимни шакллантириш ривожлангна давлатлар тажрибасида ўтган
методика ҳисобланади. Ривожланган халқаро олий таълим муассасаларида таълим олишда
талабани мустақил ўрганишга кўпроқ йўналтириш, уни доимий изланишда бўлишини
таъминлаш, аудиторияда пассив маъруза тинглашдан бахс-мунозараларга асосланган
таълимни ривожлантириш, шу билан бирга мустақил таълимда талабаларнинг ўзаро
ҳамкорликка
киришишлари,
биргаликда
илмий
изланишларни
олиб
боришга
йўналтирилган, ўз қизиқшлари доираисда жамоа бўлиб илмий фаолиятни йўлга қўиш
имкониятини берадиган таълим технологиясини яратиш юқори самарадорликни намоён
қилади. Бу орқали қуйи ва юқори босқич талабаларида қизиқишлари бўйича ҳамкорлик
қилиш, бир бирига ёрдам кўрсатиш, қуйи босқич талабаларини қуйи босқич талабаларига
ёрдам кўрсатиш, йўналиш бериш, қизиқишларини йўналтиришга қаратилган дарсдан
ташқари педагогик жараённи ташкил этиш ҳам муҳим аҳамият касб этади. Шу маънода олий техник таълимда вертикал интеграция муҳитини ташкил қилиш
ва таълимнинг биринчи йилида вертикал интнеграцияни қай даражада самарали бўлиши,
қизиқшлар бўйича ўз жамоаси ва илмий тадқиқот йўналишини танлаш учун йўналиш
беришга қаратилган таълим модулини ўқитиш самарали бўлади. Бунда талабанинг
ижодий ҳамкорлик кўникмаларини ривожлантиришда “Техник таълимда вертикал
инеграциялаш асослари” модулини ўқитиш ва мазкур фанга доир билим, дунёқараш,
малака ва кўникмаларни шакллантириш муҳим аҳамият касб этади. Мазкур фан орқали олий таълим муассасасига янги қабул қилинган талабани
қизиқшлари, иқтидори ва салоҳиятига қараб йўналтириш, бун жараёнда юқори босқич
талабалри имкониятлари ва салоҳиятидан фойдаланиш назарда тутилади. Яъни биринчи
курс талабаларини фан ва соҳаларга қараб қизиқшлари ва иқтидорини аниқлаш мақсадида
юқори курс талабалари томонидан олиб борилаётган илмий изланишлар, фанларни
ўрганишдаги таълимий фаолият ва бошқа турли тадқиқотларга жалб қилиш, таълимнинг
биринчи босқичининг илк давридаёқ уларни ўз қизиқиши ва иқтидори бўйича
йўналтиришга имконият беради. Муаммо шундаки, биринчи босқичга қабул қилинган 448 448 ISSN: 2181-3337 SCIENCE AND INNOVATION
INTERNATIONAL SCIENTIFIC JOURNAL талаба базавий фанларни ўқиш билан бир ўқув йилини ўтказиб юбориши, уни ўз соҳаси ва
турдош соҳаларга йўналтириувчи, қизиқтирувчи, иқтидорини намоён қилувчи илмий-
педагогик муҳитнинг йўқлиги, уларни одатий таълимга мослашишларини мажбур қилади. Шу боис олий техника таълимида техника ва технологиялар яратиш ва фойдаланишга
қратилган таълим муҳитида ўзи қизиқан ва кичик бўлса ҳам ўзгариш қила оладиган
мутахасисни тарбиялаш муҳим аҳамият касб этади. талаба базавий фанларни ўқиш билан бир ўқув йилини ўтказиб юбориши, уни ўз соҳаси ва
турдош соҳаларга йўналтириувчи, қизиқтирувчи, иқтидорини намоён қилувчи илмий-
педагогик муҳитнинг йўқлиги, уларни одатий таълимга мослашишларини мажбур қилади. Шу боис олий техника таълимида техника ва технологиялар яратиш ва фойдаланишга
қратилган таълим муҳитида ўзи қизиқан ва кичик бўлса ҳам ўзгариш қила оладиган
мутахасисни тарбиялаш муҳим аҳамият касб этади. “Техник таълимда вертикал инеграциялаш асослари” модулини ўқитиш ва таҳлил
этиш орқали талабада ижодий техник фаолият, унинг имкониятлари хусусидаги
хулосалари кенгайишига ёрдаб беради. Талаба таълимдги ижодий ҳамкорлик асосида
ижодкор ва инноваторлик сифатлари ривожланиб боради. Бу орқали қуйи босқич
талабаларида вертикал интеграцияни ташкил этиш ва бу орқали юқори босқичда таълим
олаётган иқтидорли, ўз изланишларини йўлга қўйган, салоҳиятли талабалар билан маълум
бир муддат ҳамкорликда ишлаш, уларга ёрдам бериш ва бу орқали уларнинг ҳам
қизиқишини шаакллантириш мақсад қилинади. Олий техник таълим муассасалари талабаларида вертикал интеграциялашган
таълим муҳитида ижодий ҳамкорлик кўникмаларини шакллантириш ва бу асосида фаол
илмий тадқиқотлар олиб бора олдаиган, қизиқшлари ва салоҳиятини эркин намоён қила
олдаиган, киришимли кадрларни тарбиялаш ва бу асосда олий техник таълимда ижодий
ҳамкорлик муҳитини қарор топтириш, шу билан бирга талабаларида ижодкорлик
кўникмаларини ривожлантириш мумкин бўлади. Бунинг учун мазкур модулни ўқитиш
технологясини таснифий белгиларини аниқлаб олиш керак бўлади. ISSN: 2181-3337 SCIENCE AND INNOVATION
INTERNATIONAL SCIENTIFIC JOURNAL тезкорликда амалга ошириладиган тизимга келтиришдан иборат. Бунинг учун эса,
технологияни ижтимоий ва педагогик талабларидан келиб чиққан ҳолда мақсадга
йўналтирувчи жиҳаталрини аниқлаб олиш мақсадга мувофиқ. тезкорликда амалга ошириладиган тизимга келтиришдан иборат. Бунинг учун эса,
технологияни ижтимоий ва педагогик талабларидан келиб чиққан ҳолда мақсадга
йўналтирувчи жиҳаталрини аниқлаб олиш мақсадга мувофиқ. Технологиянинг мақсадга йўналганлик жиҳатлари: жамият тараққиёти билан
боғлиқликда технологиянинг мақсадга йўналтирилганлиги қуйидагиларда ёрқин намоён
бўлади: олий техник таълим муассасалари талабаларида илмий-ижодий ҳамкорлик
кўникмалрини ривожлантириш орқали уларда бидрамлик кўникмаларини шакллантириш,
амалга оширилаётган ислоҳотлар моҳиятини чуқурроқ англай олиш; масъулият, кишилик
жамияти тақдирига тақдирдошлик, масъуллик туйғусини таркиб топтириш; мантиқий
мушоҳада юритиш кўникма ва малакаларини таркиб топтириш; индивидуал ёндашув нуқтаи назаридан эса: талабаларда ижодий ҳамкорлик
кўникмаларини асосида илмий фаолият муҳитини ҳосил қилиш; ижодий ҳамкорлик
кўникмаларини ривожлантириш жараёнида талабанинг ўз-ўзини англаш, ўз-ўзини
ривожлантириш, ўз-ўзини такомиллаштириш, ўз-ўзини намоён этишига ёрдамлашиш;
унга ўз муносабатини билдира олишга йўналтириш ва эркин фикр юритиш кўникма ва
малакаларини
ривожлантириш. Демак,
технологиянинг
мақсадга
йўналтирилган
жиҳатлари асосида талабаларда ижтимоий масъулиятни ривожлантириш ва индивидуал
сифатларни мустаҳкамлаш мумкин бўлади. Шунингдек, ҳар бир фан соҳасидаги илмий
тадқиқот ишининг умумий мақсади бўлган ижтимоий наф ва фан ривожига ҳисса қўшиш
мақсади ҳам мавжуд. Шу билан бирга олий техник таълим муассасалари талабаларида вертикал
интеграция асосида ижодий ҳамкорлик ғоясини ва ишлаб чиқишга қаратилган таълим
тамойилларини ҳам аниқлаб олиш керак. Бунинг учун мазкур технологиянинг диологик,
ўзаро мослик ва креативлий тамойилларини алоҳида ажратиб кўрсатиш мумкин. Диалоглик тамойили. Ижодий ҳамкорлик талабалар жамоасининг ўзаро биргаликда
ҳамкорликдаги фаолияти бўлиб, улар томонидан яратилган моддий ва маънавий маданият
кўринишида уларни ягона маконга бирлаштиради; Ўзаро мослик тамойили. Ижодий ҳамкорликни ривожлантиришнинг турли шакл,
метод ва воситаларининг инсонпарварликка, юксак креативликка йўналганлиги
шароитида зарурий ва муҳим фазилат сифатида воқеликнинг моҳиятини ва ўзаро бир-
бирини тушуниш муҳим ўрин тутади; Ўзаро мослик тамойили. Ижодий ҳамкорликни ривожлантиришнинг турли шакл,
метод ва воситаларининг инсонпарварликка, юксак креативликка йўналганлиги
шароитида зарурий ва муҳим фазилат сифатида воқеликнинг моҳиятини ва ўзаро бир-
бирини тушуниш муҳим ўрин тутади; Креативлик тамойили. Мазкур тамойил ижодий ҳамкорликни ривожлантириш
жараёнида ижодий ёндашувга асосланиш зарурлиги ҳамда талабаларинг ижодий ўз-ўзини
намоён этишлари учун шароит яратишга имкон берувчи шакл, метод ва воситалардан
фойдаланишни талаб этади. Мазкур тамойиллар ва уларнинг хусусиятларидан келиб чиқиб, олий техник таълим
муассасалари
талабаларида
ижодий
ҳамкорлик
кўникмаларини
ривожлантириш
технологяисни амалга ошириш ва самрадорликка эришиш мумкин бўлади. Бунинг учун
эса, ушбу таълим тамойилларидан ўз ўрнида ва аниқ вазиятларда фойдаланиш талаба
этилади. ISSN: 2181-3337 SCIENCE AND INNOVATION
INTERNATIONAL SCIENTIFIC JOURNAL “Техник таълимда вертикал инеграциялаш асослари” модулини ўқитиш
технологиясининг таснифий белгилари қуйидагича белгилаб олинди: қўлланилиши
даражаси ва тавсифи: локал даражада қўлланилиш тавсифига эга мезотехнология;
фалсафий асоси:1) антропологик; 2) онтологик; 3) диалогик; методологик ёндашуви:
инсонпарвар, алгоритмли, амалий йўналтирилган, қадриятга йўналтирилган ёндашув;
ривожланишнинг
етакчи
омиллари:
1)
социоген;
2)
психоген;
тажрибани
ўзлаштиришнинг
илмий
концепцияси:
ассосациатив-рефлексив
+
фаолиятли
+
ривожлантирувчи + бихевиористик; мазмунига кўра: умумтаълимий; ижтимоий-
педагогик
фаолият
тури:
таълимий
(дидактик),
тарбиявий,
ривожлантирувчи;
қўлланиладиган методлари: 1) диалогли; 2) вазиятли; 3) бир-бирига тушунтириш; 4) бир-
бирига ўргатиш; 3) мустақил иш; 4) ижодий ҳамкорлик; 5) ўйин; 6) изланишли-
тадқиқотчилик; ташкилий шакллари: 1) индивидуал; 2) гуруҳли; 3) жуфтликларда ишлаш;
4) жамоавий; қўлланиладиган воситалари: 1) манбалар; 2) ҳужжатли фильмлар; 3) илмий
адабиётлар; 4) мультимедиа; талабанинг илмий фаолиятига ижодий ёндашув ва
тарбиявий ўзаро ҳаракат тавсифи: шахсга йўналтирилган; модернизацион йўналиши ва
анъанавий таълим тизимга муносабати: ижтимоий-тарбиявий функцияни кучайтириш
асосидаги педагогик технология; қўлланилиш тоифаси: олий таълим муассасалари
талабалари учун. “Техник таълимда вертикал инеграциялаш асослари” модулини ўқитиш
технологиясининг таснифий белгилари қуйидагича белгилаб олинди: қўлланилиши
даражаси ва тавсифи: локал даражада қўлланилиш тавсифига эга мезотехнология;
фалсафий асоси:1) антропологик; 2) онтологик; 3) диалогик; методологик ёндашуви:
инсонпарвар, алгоритмли, амалий йўналтирилган, қадриятга йўналтирилган ёндашув;
ривожланишнинг
етакчи
омиллари:
1)
социоген;
2)
психоген;
тажрибани
ўзлаштиришнинг
илмий
концепцияси:
ассосациатив-рефлексив
+
фаолиятли
+
ривожлантирувчи + бихевиористик; мазмунига кўра: умумтаълимий; ижтимоий-
педагогик
фаолият
тури:
таълимий
(дидактик),
тарбиявий,
ривожлантирувчи;
қўлланиладиган методлари: 1) диалогли; 2) вазиятли; 3) бир-бирига тушунтириш; 4) бир-
бирига ўргатиш; 3) мустақил иш; 4) ижодий ҳамкорлик; 5) ўйин; 6) изланишли-
тадқиқотчилик; ташкилий шакллари: 1) индивидуал; 2) гуруҳли; 3) жуфтликларда ишлаш;
4) жамоавий; қўлланиладиган воситалари: 1) манбалар; 2) ҳужжатли фильмлар; 3) илмий
адабиётлар; 4) мультимедиа; талабанинг илмий фаолиятига ижодий ёндашув ва
тарбиявий ўзаро ҳаракат тавсифи: шахсга йўналтирилган; модернизацион йўналиши ва
анъанавий таълим тизимга муносабати: ижтимоий-тарбиявий функцияни кучайтириш
асосидаги педагогик технология; қўлланилиш тоифаси: олий таълим муассасалари
талабалари учун. Вертикал интеграция асосида талабаларга ёндашиш уларда ижодий ҳамкорликни
ривожлантиришга таъсир кўрсатиши керак. Демак, тадқиқот ишининг бош мақсади олий
техник таълим муассасалари талабаларида ижодий ҳамкорлик кўникмаларини
ривожлантириш ва унга йўналтирилган педагогик жараён технологиясини аниқ ва 449 ISSN: 2181-3337 SCIENCE AND INNOVATION
INTERNATIONAL SCIENTIFIC JOURNAL муассасалари талабалрида вертикал интеграция муносабатини яратиш ва бу орқали
рефлекция ҳамкорлик амалиётини юзага келтириш мумкрн бўлади. Бу орқали талабаларда
ижодий ҳамкорликка рефлексив ёндашув юзага келади. Олий техника таълим
муассасалари талабларида ижодий ҳамкорликни ривожлантириш технологияси рефлексив
ёндашувнинг концептуал асосларига таянган ҳолда лойиҳалаштирилиб, мақсадга
йўналтирилган, тизимли тавсифга эга қуйидаги босқичларни ўз ичига қамраб олди:
аналитик – техника йўналиши талабаларида вертикал интеграциялашган таълим асосида
ижодий ҳамкорликни ривожлантиришга доир билим, малака ва кўникмларини ошириш ва
уларни баҳолаш; ташкилий-тайёрлов – технологияни амалга ошириш истиқболини
аниқлаштириш, педагогик жараён шакл, метод ва воситаларини танлаш, профессор-
ўқитувчиларнинг методик тайёргарлигини такомиллаштириш; фаолиятга доир – олий
техника таълим муассасалари талабларида ижодий ҳамкорликни ривожлантириш
технологиясини босқичма-босқич амалга ошириш; мониторинг – олий техника таълим
муассасалари талабларида ижодий ҳамкорликни ривожлантириш динамикасини баҳолаш,
зарурият юзага келганда педагогик жараёнга тузатишлар киритиш; натижавий –
натижаларни баҳолаш ва таълимий-тарбиявий ишларнинг истиқболдаги вазифаларини
аниқлаштириш. Бу орқали талабаларда ижодий ҳамкорли кўникмаларини ривожланиш
динамикаси, унга таъсир кўрсатувчи омилар ҳам ўрганилади. муассасалари талабалрида вертикал интеграция муносабатини яратиш ва бу орқали
рефлекция ҳамкорлик амалиётини юзага келтириш мумкрн бўлади. Бу орқали талабаларда
ижодий ҳамкорликка рефлексив ёндашув юзага келади. Олий техника таълим
муассасалари талабларида ижодий ҳамкорликни ривожлантириш технологияси рефлексив
ёндашувнинг концептуал асосларига таянган ҳолда лойиҳалаштирилиб, мақсадга
йўналтирилган, тизимли тавсифга эга қуйидаги босқичларни ўз ичига қамраб олди:
аналитик – техника йўналиши талабаларида вертикал интеграциялашган таълим асосида
ижодий ҳамкорликни ривожлантиришга доир билим, малака ва кўникмларини ошириш ва
уларни баҳолаш; ташкилий-тайёрлов – технологияни амалга ошириш истиқболини
аниқлаштириш, педагогик жараён шакл, метод ва воситаларини танлаш, профессор-
ўқитувчиларнинг методик тайёргарлигини такомиллаштириш; фаолиятга доир – олий
техника таълим муассасалари талабларида ижодий ҳамкорликни ривожлантириш
технологиясини босқичма-босқич амалга ошириш; мониторинг – олий техника таълим
муассасалари талабларида ижодий ҳамкорликни ривожлантириш динамикасини баҳолаш,
зарурият юзага келганда педагогик жараёнга тузатишлар киритиш; натижавий –
натижаларни баҳолаш ва таълимий-тарбиявий ишларнинг истиқболдаги вазифаларини
аниқлаштириш. Бу орқали талабаларда ижодий ҳамкорли кўникмаларини ривожланиш
динамикаси, унга таъсир кўрсатувчи омилар ҳам ўрганилади. Натижада талабаларда ижодий ҳамкорлик кўникмаларини ривожлантиришнинг
изчил ва тизимли технологяисни ишлаб чиқиш икоони юзага келади. Олий техник таълим муассасалари талабаларида вертикал интегрция асосида
ижодий ҳамкорлик кўникмаларини ривожлантириш технологяси мазмунли-процессуал
ва ташкилий-педагогик қисмларда ташкил топган. Технологиянинг мазмуни ўзида когнитив – амалий фаолият – рефлексия бирлиги ва
ана шу босқичларни ҳар бирида ўзига хос тарзда намоён бўлувчи ижодий ҳамкорликни
ривожлантириш даражалари (англаш, тушуниш, муносабат ва фикрлаш)ни акс эттиради. Демак, талабалардага ижодий ҳамкорлик кўникмалари уларда илмий билишга, амалий
фаолиятга ва рефлекцияга асосланган фаолиятга таняниши келиб чиқади. ISSN: 2181-3337 SCIENCE AND INNOVATION
INTERNATIONAL SCIENTIFIC JOURNAL Олий техника таълим муассасаларида вертикал интеграциялашган таълим асосида
ижодий ҳамкорлик кўникмаларини ривожлантириш асосларини алоҳида фан сифатида
ўқитилмаслигини ҳисобга олган ҳолда, тўгарак машғулотларида фойдаланиш учун 1-4
курсларга мўлжалланган “Техник таълимда вертикал интеграциялаш асослари” модулли
курс дастури тақвим-мавзу режаси ишлаб чиқилди. Мазкур тақвим-режада ижодий
вертикал интегрция таълимнинг самрадорлиги, уни йўлга қўйиш шартлари ва педагогик
талабалари ва бошқа жиҳатларга доир мавзулар киритилган. Шунингдек, талабаларда
ижодий ҳамкорликни ўрнитишга қаратилган топшириқлар ва муаммоли вазияларни ҳосил
қилиш йўллари игшлаб чиқилган. Тақвим-режа мавзуларини кетма-кетликда ўтилиши
талабаларда ижодий ҳамкорликка вертикал интегрция асосида ривожлантириш имконияти
ва лаёқати юзага келади. ХУЛОСА МУҲОКАМА Талабаларда ижодий ҳамкорлик кўникмаларини ривожлантиришда рефлексив
ёндашувдан ҳам оқилона фойдаланиш зарур. Унга кўра олий техника таълим 450 450 ISSN: 2181-3337 SCIENCE AND INNOVATION
INTERNATIONAL SCIENTIFIC JOURNAL таълим технологясининг мазмунли-процессуал ва ташкилий-педагогик қисмларини тўғри
ташкил этиш ва олиб бориш юқори самара бериши келиб чиқди. таълим технологясининг мазмунли-процессуал ва ташкилий-педагогик қисмларини тўғри
ташкил этиш ва олиб бориш юқори самара бериши келиб чиқди. 1. Вертикал интеграциялаш орқали ташкил этилган педагогик жараёнда тизимли-
жараёнли ёндашув, аксиологик ва рефлексив ёндашув, шахсий-фаолиятга йўналтирилган
ёндашувни комплекс қўллаш ва уйғунликда олиб бориш керак бўлади. Бу орқали олий
техник
таълим
муассасалари
талабаларида
ижодий
ҳамкорлик
кўникмаларини
ривожлантиришда энг кичик жиҳатларгача эътибор қаратиш имконини беради. 2. Ўқув лойиҳаларини ишлаб чиқиш ва амалга ошириш орқали талабаларда
ижодий ҳамкорлик кўникмлари лойиҳаларни ҳамкорликда яратиш ва унинг концептуал
асосларини ишлаб чиқиш ва амалга ошириш имконини беради. Олий техник таълимда
лойҳилашга доир илмий топширифқларни бериш ва мустақил таълимда лойиҳалашдан
фойдаланиш самарали бўлдаи. Чунки лойиҳалаш ҳамкорликдагина самара беради ва
талабаларга уни бажаришда ижодий ёндашувни талаб қилади. ХУЛОСА Хулоса қилганда, олий техник таълим муассасалари талабаларида вертикал
интеграция асосида ижодий ҳамкорлик кўникмаларини ривожлантиришга йўналтирилган 451 451 ФОЙДАЛАНИЛГАН АДАБИЁТЛАР ОЙДАЛАНИЛГАН АДАБИЁТЛАР
1. Ҳасанбоев Ж., Тўрақулов Х. Педагогика назарияси фанидан изоҳли луғат. -
Тошкент.: «Фан», 2008. –Б.182. 1. Ҳасанбоев Ж., Тўрақулов Х. Педагогика назарияси фанидан изоҳли луғат. -
Тошкент.: «Фан», 2008. –Б.182. 2. Мирзиёев Ш.М. Янги Ўзбекистон стратегияси. –Тошкент.:Ўзбекистон. 2021. –Б. 232 2. Мирзиёев Ш.М. Янги Ўзбекистон стратегияси. –Тошкент.:Ўзбекистон. 2021. –Б. 232 3. Авлакулов А.М. Техноген цивилизация ва таълим-тарбия жараёнлари. -Тошкент.:
2016. –Б. 12. 3. Авлакулов А.М. Техноген цивилизация ва таълим-тарбия жараёнлари. -Тошкент.:
2016. –Б. 12. 4. Ибрагимов
С. Таълим
олувчиларнинг
креатив
қобилиятларини
ривожлантиришнинг аҳамияти. // Аcademic research in educational sciences. VOLUME 1. ISSUE 4. 2020. –Р. 408. 4. Ибрагимов
С. Таълим
олувчиларнинг
креатив
қобилиятларини
ривожлантиришнинг аҳамияти. // Аcademic research in educational sciences. VOLUME 1. ISSUE 4. 2020. –Р. 408. 5. Bakhromovich SI. The impact of managerial professional development on the
effectiveness of Higher Education institution management. Academicia: an international
multidisciplinary research journal. 2020;10(12):1014-20. 5. Bakhromovich SI. The impact of managerial professional development on the
effectiveness of Higher Education institution management. Academicia: an international
multidisciplinary research journal. 2020;10(12):1014-20. 6. Siddikov,
I. B. (2019). Философско-педагогические
аспекты
развития
интеллектуальной культуры студентов. Вестник Ошского государственного
университета, (3), 38-42. 6. Siddikov,
I. B. (2019). Философско-педагогические
аспекты
развития
интеллектуальной культуры студентов. Вестник Ошского государственного
университета, (3), 38-42. 7. Bakhromovich, S. I. Development trends and transformation processes in academic
mobility in higher education in Uzbekistan and the world. 7. Bakhromovich, S. I. Development trends and transformation processes in academic
mobility in higher education in Uzbekistan and the world. 8. Bakhromovich, S. I. (2021). A comparative induction of the epistemological and
theological views of medieval Islamic oriental scholars. Berlin Studies Transnational
Journal of Science and Humanities, 1(1.7 Philosophical sciences). 8. Bakhromovich, S. I. (2021). A comparative induction of the epistemological and
theological views of medieval Islamic oriental scholars. Berlin Studies Transnational
Journal of Science and Humanities, 1(1.7 Philosophical sciences). 9. Bakhromovich, S. I. (2021). Views on the role of science in human and society life in
islamic teaching. International Journal of Philosophical Studies and Social Sciences,
1(3), 79-86. 9. Bakhromovich, S. I. (2021). Views on the role of science in human and society life in
islamic teaching. International Journal of Philosophical Studies and Social Sciences,
1(3), 79-86. 10. Bakhromovich, S. I. (2020). Effects of Objective and Subjective Factors to Develop
Intellectual Culture of Youth. Canadian Social Science, 16 (2), 55-59 p. 11. Bakhromovich, S. I. (2018). Social and philisophical performance of making youth’s
intellectual culture. ФОЙДАЛАНИЛГАН АДАБИЁТЛАР European science review, (7-8). 452 452 ISSN: 2181-3337 SCIENCE AND INNOVATION
INTERNATIONAL SCIENTIFIC JOURNAL 12. Bakhromovich, S. I. (2020). Analysis Of Modern Approaches To Ensuring The
Effectiveness Of Management In Higher Education Institutions. The American Journal
of Social Science and Education Innovations, 2(12), 364-369. 12. Bakhromovich, S. I. (2020). Analysis Of Modern Approaches To Ensuring The
Effectiveness Of Management In Higher Education Institutions. The American Journal
of Social Science and Education Innovations, 2(12), 364-369. 13. Bakhromovich, S. I. (2021). Philosophical comparative studies of the epistemological
and theological views of medieval eastern scholars. Turkish Journal of Physiotherapy
and Rehabilitation. Turkey, 2021. №32 (3), (Scopus) – P. 30338-30355 13. Bakhromovich, S. I. (2021). Philosophical comparative studies of the epistemological
and theological views of medieval eastern scholars. Turkish Journal of Physiotherapy
and Rehabilitation. Turkey, 2021. №32 (3), (Scopus) – P. 30338-30355 14. Bakhromovich, S. I. (2022).Dialectical and synergetic features of the development of
theological and epistemological views in medieval eastern islam. European Journal of
Humanities and Educational Advancements (EJHEA), - Las Palmas, Spain, Volume 3,
Issue 2 February, 2022. – P. 79-83 15. Nurullayevna, N. M., & Muhlisa, T. (2021, December). SOME PROBLEMS AND
SHORTCOMINGS IN THE HIGHER EDUCATION SYSTEM. In Archive of
Conferences (pp. 61-62). 15. Nurullayevna, N. M., & Muhlisa, T. (2021, December). SOME PROBLEMS AND
SHORTCOMINGS IN THE HIGHER EDUCATION SYSTEM. In Archive of
Conferences (pp. 61-62). 16. Nasritdinova, M., & Khokimova, K. (2021, October). LANGUAGE IS THE MIRROR
OF THE NATION. In Archive of Conferences (pp. 92-94). 16. Nasritdinova, M., & Khokimova, K. (2021, October). LANGUAGE IS THE MIRROR
OF THE NATION. In Archive of Conferences (pp. 92-94). 453
|
https://openalex.org/W2130805800
|
https://journals.plos.org/plosone/article/file?id=10.1371/journal.pone.0084079&type=printable
|
English
| null |
Breaking of Icosahedral Symmetry: C60 to C70
|
PloS one
| 2,014
|
cc-by
| 3,677
|
Abstract * E-mail: mbodner@mindresearch.net (MB); patera@crm.umontreal.ca (JP); m.szajewska@math.uwb.edu.pl (MS) * E-mail: mbodner@mindresearch.net (MB); patera@crm.umontreal.ca (JP); m.szajewska@math.uwb.edu.pl (MS) ¤ Current address: Institute of Mathematics, University of Bialystok, Bialystok, Poland ¤ Current address: Institute of Mathematics, University of Bialystok, Bialystok, Poland In the present work we consider the existence and structure of
higher fullerenes as a symmetry breaking problem, starting from
the C60 molecule which possesses the highest degree of symmetry. Guided by the common practices in particle physics, we consider
the description specifically of the C70 molecule as a symmetry
breaking problem, with the additional twist that the usual
branching rule for the icosahedral symmetry group H3 to the
dihedral symmetry group H2 is enhanced by adding to it one more
decagonal term. The group H2 is the lowest noncrystallographic
finite reflection group. We consider the icosahedral symmetry
group H3 of order 120 of certain carbon molecules as the exact
symmetry that is broken to its subgroup H2 or order 10 dihedral
symmetries. We also suggest within this framework how higher
order structures such as nanotubes may naturally arise. This
provides a framework for understanding of the observed even
carbon number rule and for predicting higher order structures
which may be assembled. Introduction Fullerenes are molecules composed entirely of carbon, taking
the form of a cage or tube. The family of cage cluster fullerenes is
also commonly referred to as buckyballs. The most stable and
commonly occurring member of this family is the molecule C60,
which consists of 60 carbon atoms arranged in a structure of
truncated icosahedrons, made of hexagons and pentagons, with
carbon atoms at the corners of each hexagon and a bond along
each edge (creating the well-known soccer ball structure - Fig. 1). This structure has been investigated and determined experimen-
tally in both the solid state [1] and in the gas phase [2]. The
second most commonly occurring cage structure fullerene is the
molecule C70, composed of 70 carbon atoms. Electron diffraction
and theoretical studies have verified that this molecule possesses a
‘‘rugby ball’’ structure with a pinching of the waist as the bond
lengths follow a simple pattern determined by their relationship to
the 5- and 6-membered rings [2,3] - Fig. 2. The paper is unique in providing exact coordinates of the
vertices of the fullerenes thus eliminating any additional number-
ing conventions used elsewhere [16]. This opens the possibility of
defining special functions of 3 variables generated by the vertices
(see Example 3), to study their possible orthogonality, and
conceivably even the corresponding orthogonal polynomials
defined by the fullerene structures. Much effort has been directed to answer why the fullerenes C60
and C70 are so stable and which other higher fullerenes with
different sizes and shapes can be formed as stable entities [2,4–6]. With respect to the C60 molecule, its stability and its position as the
most commonly occurring buckyball can be attributed, at least in
part, to its high degree of symmetry [7]. In addition to C60 and
C70, many other larger fullerenes have been observed, while
theoretical calculations have indicated that all fullerenes with an
even number of carbon atoms can exist [8]. Significant amounts of
work have gone towards elucidating the structure of these higher
fullerenes. For example, the structures of C84, as well as C60 and
C70 have been identified through spectroscopy studies [9–12], and
by chemical derivatization [13,14], while others have been
proposed theoretically [15]. Breaking of Icosahedral Symmetry: C60 to C70 D Research Institute, Irvine, California, United States of America, 2 Centre de Recherches Mathe´matiques, Universite´ de Montre´al, Mo rvine, California, United States of America, 2 Centre de Recherches Mathe´matiques, Universite´ de Montre´al, Montre´al, Que´bec, Canada Abstract We describe the existence and structure of large fullerenes in terms of symmetry breaking of the C60 molecule. Specifically,
we describe the existence of C70 in terms of breaking of the icosahedral symmetry of C60 by the insertion into its middle of
an additional H2 decagon. The surface of C70 is formed by 12 regular pentagons and 25 regular hexagons. All 105 edges of
C70 are of the same length. It should be noted that the structure of the molecules is described in exact coordinates relative
to the non-orthogonal icosahedral bases. This symmetry breaking process can be readily applied, and could account for and
describe other larger cage cluster fullerene molecules, as well as more complex higher structures such as nanotubes. Citation: Bodner M, Patera J, Szajewska M (2014) Breaking of Icosahedral Symmetry: C60 to C70. PLoS ONE 9(3): e84079. doi:10.1371/journal.pone.0084079 Citation: Bodner M, Patera J, Szajewska M (2014) Breaking of Icosahedral Symmetry: C60 to C70. PLoS ONE 9(3): e84079. doi:10.1371/journal.pone.0084079
Editor: Guido Germano University College London United Kingdom ation: Bodner M, Patera J, Szajewska M (2014) Breaking of Icosahedral Symmetry: C60 to C70. PLoS ONE 9(3): e84079. doi:10.1371/jou Editor: Guido Germano, University College London, United Kingdom Received May 3, 2013; Accepted November 20, 2013; Published March 5, 2014 Received May 3, 2013; Accepted November 20, 2013; Published March 5, 2014 Received May 3, 2013; Accepted November 20, 2013; Published March 5, 2014 Copyright: 2014 Bodner et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits
unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. nding: Partial support from the Natural Sciences and Engineering Research Council of Canada and MITACS. The funders had no role
lection and analysis, decision to publish, or preparation of the manuscript. t from the Natural Sciences and Engineering Research Council of Canada and MITACS. The funders had no role in study design, data
ecision to publish, or preparation of the manuscript. Competing Interests: The authors acknowledge the partial support of a commercial funder, OODA Technologies, for this work. This does not alter the authors’
adherence to all the PLOS ONE policies on sharing data and materials. Icosahedral bases in R3 doi:10.1371/journal.pone.0084079.g001 C~(Cjk)~(Saj,akT)~
2
{1
0
{1
2
{t
0
{t
2
0
B
@
1
C
A,
t~1
2(1z
ffiffiffi
5
p
): ð1Þ
The dual or reciprical v-basis is defined by
a1~2v1{v2
v1~(1z1
2t)a1z(1zt)a2z(1
2zt)a3
a2~{v1z2v2{tv3 v2~(1zt)a1z(2z2t)a2z(1z2t)a3
a3~{tv2z2v3
v3~(1
2zt)a1z(1z2t)a2z(3
2z3
2t)a3
ð4Þ a1~2v1{v2
v1~(1z1
2t)a1z(1zt)a2z(1
2zt)a3
a2~{v1z2v2{tv3 v2~(1zt)a1z(2z2t)a2z(1z2t)a3
a3~{tv2z2v3
v3~(1
2zt)a1z(1z2t)a2z(3
2z3
2t)a3
ð4Þ The dual or reciprical v-basis is defined by The dual or reciprical v-basis is defined by a3
ð4Þ Saj,vkT~djk,
j,k~1,2,3:
ð2Þ ð2Þ ð2Þ Saj,vkT~djk,
j,k~1,2,3: For the v-basis of H2 we choose v2 and v3 of H3. By Eq.(2),
the direction orthogonal to the plane spanned by v2 and v3, is
that of a1. The inverse matrix to C is calculated as follows, The inverse matrix to C is calculated as follows, C{1~(Svj,vkT)~
1
det C
3{t
2
t
2
4
2t
t
2t
3
0
B
B
@
1
C
C
A
~ 1
2
2zt
2z2t
1z2t
2z2t
4z4t
2z4t
1z2t
2z4t
3z3t
0
B
B
@
1
C
C
A:
ð3Þ The reflections r1, r2, and r3 in R3 generate the icosahedral
group of order 120. Their action on any vector x[R3 is given by rkx~x{Sx,akTak,
k~1,2,3,
x[R3:
ð5Þ
In particular, rk0~0 and rkak~{ak, and also rkvj~vj{djkak. ð3Þ rkx~x{Sx,akTak,
k~1,2,3,
x[R3:
ð5Þ ð5Þ In particular, rk0~0 and rkak~{ak, and also rkvj~vj{djkak. C60 Repeated application of the three reflections to the seed point
(1,1,0)~v1zv2 of C60, according to Eq.(5), yields the 60 vertices
of C60 in the v-basis: It follows from Eq.(2) that the a- and v-bases are related by the
matrix equality a~Cv, and v~C{1a. Explicitly we have Figure 2. The structure of the C70 molecule. (a) The polytope C70 has 105 edges and 12 pentagonal and 25 hexagonal faces. (b) C70 viewed from
a direction almost parallel to the plane of v2 and v3. (c) The H2 structure of C70 viewed from a direction parallel to the plane of v2 and v3. The
column of numbers shows the a1-coordinate of the H2 orbits of vertices of C70. The inserted decagon has the a1-coordinate equal to 0. Surface edges
are omitted in (b) and (c). doi:10.1371/journal.pone.0084079.g002 Figure 2. The structure of the C70 molecule. (a) The polytope C70 has 105 edges and 12 pentagonal and 25 hexagonal faces. (b) C70 viewed from
a direction almost parallel to the plane of v2 and v3. Icosahedral bases in R3 In order to get exact coordinates of polytopes related to
icosahedral symmetry, one has to use bases in the real 3-
dimensional space R3 that reflects the symmetry, namely the
simple roots a1, a2, a3 of the icosahedral group and their duals
[17]. The geometric relations of the vectors in the a-basis are
described by the matrix of scalar products March 2014 | Volume 9 | Issue 3 | e84079 PLOS ONE | www.plosone.org 1 C60 to C70 C60 to C70 Figure 1. The structure of the C60 molecule. (a) The polytope C60 is formed by 60 vertices equidistant from its center. Its surface consists of 12
regular pentagons and 20 regular hexagons. All 90 edges are of the same length. (b) The C60 polytope viewed in the direction almost parallel to the
plane spanned by v2 and v3, which makes the H2 orbits (pentagons and decagons) easy to identify. (c) The C60 polytope viewed in the direction
parallel to the H2 plane spanned by v2 and v3. Surface edges of C60 are omitted in (b) and (c). Horizontal segments are projections of the H2 orbits. The number in a row shows the a1 coordinate of the H2 orbit. The vertical direction is that of a1. doi:10.1371/journal.pone.0084079.g001 Figure 1. The structure of the C60 molecule. (a) The polytope C60 is formed by 60 vertices equidistant from its center. Its surface consists of 12
regular pentagons and 20 regular hexagons. All 90 edges are of the same length. (b) The C60 polytope viewed in the direction almost parallel to the
plane spanned by v2 and v3, which makes the H2 orbits (pentagons and decagons) easy to identify. (c) The C60 polytope viewed in the direction
parallel to the H2 plane spanned by v2 and v3. Surface edges of C60 are omitted in (b) and (c). Horizontal segments are projections of the H2 orbits. The number in a row shows the a1 coordinate of the H2 orbit. The vertical direction is that of a1. Example 1 The hexagon faces of C60 come up naturally from the classification of its 2-
faces [18] as one orbit of the seed hexagon. The symmetry group of the seed
hexagon is generated by the reflections r1 and r2. Similarly the pentagon faces
of C60 come up naturally from the classification of its 2-faces as one orbit of
the seed pentagon. The symmetry group of the seed pentagon is generated by the
reflections r2 and r3. C70 The general idea, we pursue here for the modification
C60?C70, is first to decompose C60 into the sum of 8 orbits of
H2, then to insert another H2 decagon into its middle. In Fig. 3 it
can be seen that the upper and lower half of C60 are connected by
a ring of 5 hexagons. Replacing that ring by a larger one that is Figure 3. The 12 pentagons of the surface of C60 are shown
without the hexagons. The 60 dots are the vertices of C60. The
polytope is oriented as in Fig. 1. doi:10.1371/journal.pone.0084079.g003 The vertices of a hexagon and pentagon of the surface of C60 adjacent to the
dominant point of C60 are shown here in the v-basis. Icosahedral bases in R3 (c) The H2 structure of C70 viewed from a direction parallel to the plane of v2 and v3. The
column of numbers shows the a1-coordinate of the H2 orbits of vertices of C70. The inserted decagon has the a1-coordinate equal to 0. Surface edges
are omitted in (b) and (c). doi:10.1371/journal.pone.0084079.g002 March 2014 | Volume 9 | Issue 3 | e84079 PLOS ONE | www.plosone.org 2 C60 to C70 C60 to C70 C60 to C70 (6) (6) The points in which both the second and third coordinates are
non-negative are the dominant points of the H2 orbits (pentagons
and decagons). Boxed points in which the second and third
coordinates are positive dominate the decagons. The boxed points,
which have 0 as their second or third coordinate, are the dominant
points of H2 pentagons. a2*(1,1,0)z({1,2,0)~(0,3,0) (iii) The straight line containing +a3 passes through the center of opposite
hexagons on the surface of C60. a3*(1,1,0)z({1,2,0)z(2,{1,t)
z(1,{2,2t)z({2,1,t)z({1,{1,2t)~(0,0,6t) Example 3 In this example let us view each point l of Eq.(6) as an exponential function,
l?e2piSl,xT, where x[R3, and then add up all 60 such exponentials. Call
such a sum C(l,x). Since each l comes with both signs in Eq.(6), we have
C(l,x) equal to the sum of 30 cosines cos(2pSl,xT). Properties of C(l,x)
deserves further study. The function ‘remembers’ the structure of C60 in the entire
3-space and shows intricate interferences of the cosines with a clear maximum
when x is at the origin of R3. On the spherical surface of the C60 shell, the
function C(l,x) depends periodically on the radius of C60. Let us illustrate the construction of the seed hexagon and of the seed
pentagon, starting from the dominant point of C60: Symmetry breaking Summarizing, below are the exact coordinate of the 70 vertices
of C70 in the basis fa1,v2,v3g: The H3 symmetry gets broken when the additional decagon is
inserted into the middle of its decomposition into H2-orbits. The
H2 symmetry remains exact. C70 :
5z5z10z10z%
10z10z10z5z5,
ð7Þ (8) (8) According to [19] the carbon polytope C70 is slimmer in the
middle. Such data can be matched by choosing a smaller radius for
the decagon (0,1,2) in Eq.(8) in the middle of C70. Also the edges
leading to that decagon may have to be changed correspondingly. The boxed points in Eq.(8) are the dominant points. The enlarged structure is C70 which has lost the spherical
symmetry of C60. It has 70 vertices, and in the middle of it there
are 5 consecutive parallel decagons centered at the a1-axis. There are still two questions to be answered however before
one can call it C70. The answers to these questions must assure
that the exterior surface of C70 is composed of pentagons and
hexagons of the same size as it is for C60. From Fig. 3 we see that
the upper and lower halves of C60 are bound by a ring of
hexagons. (i) What are the distances between the five decagons,
and (ii) what is the orientation of the inserted decagon in the
v2,v3 plane? In the present work we have described the existence of the
molecule C70 in terms of a symmetry breaking process of the
insertion of an H2 decagon (or equivalently inserting a ring of
surface hexagons), thus breaking the icosahedral symmetry of C60. There is nothing to prevent however, the insertion of three or
more rings of hexagons into the C60 structure creating ones that
are larger and more complex. Thus the mechanism enables the
creation from C60, the fullerenes C70,C80,C90 and so on. From the
continuation of the process of the insertion of hexagon rings in this
fashion, it can also readily be seen that it enables the creation of
nanotubes of any length. p
The answers to the questions are found by making two
observations from Fig. 4, where the additional decagon is placed
in the middle, so that its a1 coordinate is zero. Acknowledgments M.S. is grateful to CRM, Universite´ de Montre´al, for the hospitality
extended to her during her postdoctoral fellowship. The authors would like
to thank Dr. M. Angelova for stimulating discussions and comments. (ii)
The first row of hexagons in Fig. 4 (right). belonged to the
upper half of C60. The second row in Fig. 4 (right) is
situated as was the second row in Fig. 4 (left). There it was
the top row for the lower part of C60. In Fig. 4 (right) it is
the inserted middle row of C70. The third row of hexagons
in Fig.4 (right) is the top row of the lower half of C70. The
third row of hexagons in Fig. 4 (right) is the top row of the Symmetry breaking (i)
In order to keep the distances between the five decagons of
C70 equal to what they are in C60, we have to shift
correspondingly the upper and the lower halves of what used
to be C60. Their a1 coordinates are increased and decreased
by 0:5t respectively. Example 2 The three simple roots, a1, a2, a3, of the icosahedral group form a special
basis in R3. Their relative angles and lengths are read from the matrix C
(Eq.(1)). Those values define the icosahedral group H3. Therefore they take
special positions also in C60. Let us show that (i) the straight line containing +a1 passes through the
center of opposite pentagons on the surface of C60. To show it, one needs to take
the hexagon generated by r1 and r2, and add its vertices to verify that
coordinates of the sum are zero in the plane spanned by v2 and v3. a1*(1,1,0)z(2,{1,t)z(2,t,{t)
z(2zt,{t,1)z(2zt,0,{1)~(9z2t,0,0) (ii) The straight line containing +a2 passes through the center of opposite
edges on the surface of C60 that separate two hexagons. Figure 3. The 12 pentagons of the surface of C60 are shown
without the hexagons. The 60 dots are the vertices of C60. The
polytope is oriented as in Fig. 1. doi:10.1371/journal.pone.0084079.g003 PLOS ONE | www.plosone.org PLOS ONE | www.plosone.org March 2014 | Volume 9 | Issue 3 | e84079 3 C60 to C70 Figure 4. A ring of hexagons from the middle of the surface of C60 and C70 unwrapped into the plane. Horizontal lines indicate positions
of the four H2 decagons in C60 and five H2 decagons in C70. The dominant points identify the H2 decagons relative to the basis fa1,v2,v3g. Dashed
lines are the boundaries of the ring of five inserted hexagons. doi:10.1371/journal.pone.0084079.g004 Figure 4. A ring of hexagons from the middle of the surface of C60 and C70 unwrapped into the plane. Horizontal lines indicate positions
of the four H2 decagons in C60 and five H2 decagons in C70. The dominant points identify the H2 decagons relative to the basis fa1,v2,v3g. Dashed
lines are the boundaries of the ring of five inserted hexagons. doi:10.1371/journal.pone.0084079.g004 lower half of C60. However, its position matches the
hexagons of the first row. Hence the dominant points of
the first and third rows differ by the sign of the first
coordinate only. made out of 10 hexagons (see Fig. 4), one gets the polytope C70. It
is shown in Fig. 2 in three different views analogous to the
presentation of C60 in Fig. 1. References 12. Weaver JH, Poirier DM (1994) Solid state properties of fullerenes and fullerene
based materials. In: Erenreih H, Spaepen F, editors. Solid State Physics:
Advances in Research and Applications 48. San Diego: Academic Press. pp. 1–
107. 1. Curl RF, Smalley RE (1988) Probing C60 Science 242: 1017–102 2. Kroto H (1987) The stability of the fullerenes Cn, with n = 24, 28, 32, 36, 50, 60
and 70. Nature 329: 529–531. 3. Terrones H, Terrones M (1997) The transformation of polyhedral particles into
graphitic onions. J Phys Chem Solids 58: 1789–1796. 13. Ajie H, Alvarez MM, Anz AJ, Beck RD, Diederich F, et al. (1990)
Characterization of the soluble all-carbon molecules C60 an C70. J Phys Chem
94: 8630–8633. g
p
J
y
4. Gao T, Nikolaev P, Rinzler AG, Tomanek D, Colbert DT, et al. (1995) Self
assembly of tubular fullerenes. J Phys Chem 99: 10694–10697. 4. Gao T, Nikolaev P, Rinzler AG, Tomanek D, Colbert DT, et assembly of tubular fullerenes. J Phys Chem 99: 10694–10697. 5. Kerner R (1994) Nuleation and growth of fullerenes. J Comp Mater Sci 2: 500–
508. 14. Taylor R, Hare JP, Abdul-Sada AK, Kroto HW (1990) Isolation, separation and
characterisation of the fullerenes C60 and C70; the third form of carbon. J Chem
Soc Chem Commun 1423–1425. 6. Terrones M, Terrones G, Terrones H (2002) Structure, chirality, and formation
of giant icosahedral fullerenes and spherical graphitic onions. Struct Chem 13
(3/4): 373–374. 15. Smalley RE (1992) Self-assembly of the fullerenes. Acc Chem Res 25: 98–105. 15. Smalley RE (1992) Self-assembly of the fullerenes. Ac 16. Powell WH, Cozzi F, Moss GP, Thilgen C, Hwu JR, et al. (2002) Nomenclature
for the C60-Ih and C70-D5h(6) fullerenes. Pure and Appl Chem 74: 629–695. (
)
7. Fowler PW, Manolopoulos DE (1992) Magic numbers and stable structures for
fullerenes, fullerides and fullerenium ions. Nature 355: 428–430. 17. Chen L, Moody RV, Patera J (1998) Non-crystallographic root systems. In:
Patera J, editor. Quasicrystals and Discrete Geometry. Fields Institute
Monograph Series 10. Providence: Amer. Math. Soc. pp. 135–178. 8. Pisanski T (1977) On planar graphs with 12 vertices of degree five. Glas Mat 12:
233–235. 9. Behune DS, Jeijer G, Tang WC, Rosen HJ (1990) The vibrational raman
spectra of purified solid films of C60 and C70. Chem Phys Lett 174: 219–222. 18. Author Contributions Conceived and designed the experiments: MB JP MS. Performed the
experiments: MB JP MS. Analyzed the data: MB JP MS. Wrote the paper:
MB JP MS. March 2014 | Volume 9 | Issue 3 | e84079 PLOS ONE | www.plosone.org 4 C60 to C70 C60 to C70 C60 to C70 References Patera J, Sharp RT, Champagne, Kjiri M (1995) Description of reflection
generated polytopes using decorated Coxeter diagrams. Can J Phys 73: 566–
584. p
p
y
10. McElvany SW, Ross MM, Callahan JH (1992) Characterization of fullerenes by
mass spectrometry. Acc Chem Res 25: 162–168. 10. McElvany SW, Ross MM, Callahan JH (1992) C
mass spectrometry. Acc Chem Res 25: 162–168 mass spectrometry. Acc Chem Res 25: 162–168. 19. McKenzie DR, Davis CA, Cockayne DJH, Muller D, Vassallo AM (1992) The
structure of the C70 molecule. Nature 355: 622–624. 11. Raghavachari K, Roohlfing CM (1991) Structures and vibrational frequencies of
carbon molecules (C60, C70, and C84). J Phys Chem 95: 5768–5733. March 2014 | Volume 9 | Issue 3 | e84079 PLOS ONE | www.plosone.org PLOS ONE | www.plosone.org 5
|
https://openalex.org/W2783533281
|
https://diposit.ub.edu/dspace/bitstream/2445/165602/1/680667.pdf
|
English
| null |
Assessing <i>GW</i> Approaches for Predicting Core Level Binding Energies
|
Journal of chemical theory and computation
| 2,018
|
public-domain
| 6,710
|
Assessing GW approaches for predicting core level binding energies Michiel J. van Setten,1,* Ramon Costa,2 Francesc Viñes,3 Francesc Illas3,† 1) Nanoscopic Physics, Institute of Condensed Matter and Nanosciences, Université
catholique de Louvain, 1348 Louvain-la-Neuve, Belgium. 1) Nanoscopic Physics, Institute of Condensed Matter and Nanosciences, Université
catholique de Louvain, 1348 Louvain-la-Neuve, Belgium. 2) Departament de Química Inorgànica i Orgànica & Institut de Química Teòrica i
Computacional (IQTCUB), Universitat de Barcelona, Martí i Franqués 1, 08028
Barcelona, Spain. 2) Departament de Química Inorgànica i Orgànica & Institut de Química Teòrica i
Computacional (IQTCUB), Universitat de Barcelona, Martí i Franqués 1, 08028
Barcelona, Spain. 3) Departament de Ciència de Materials i Química Fisica & Institut de Química
Teòrica i Computacional (IQTCUB), Universitat de Barcelona, Martí i Franqués 1,
08028 Barcelona, Spain. 3) Departament de Ciència de Materials i Química Fisica & Institut de Química
Teòrica i Computacional (IQTCUB), Universitat de Barcelona, Martí i Franqués 1,
08028 Barcelona, Spain. Abstract Here we present a systematic study on the performance of different GW approaches:
G0W0, G0W0 with linearized quasi particle equation (lin-G0W0), and quasiparticle self-
consistent GW (qsGW), in predicting core level binding energies (CLBEs) on a series of
representative molecules comparing to Kohn-Sham (KS) orbital energy based results. KS orbital energies obtained using the PBE functional are 20-30 eV lower in energy
than experimental values obtained from X-ray photoemission spectroscopy (XPS),
showing that any Koopmans-like interpretation of KS core level orbitals fails
dramatically. Results from qsGW lead to CLBEs that are closer to experimental values
from XPS, yet too large. For the qsGW method, the mean absolute error is about 2 eV,
an order of magnitude better than plain KS PBE orbital energies and quite close to
predictions from ΔSCF calculations with the same functional, which are accurate within
~1 eV. Smaller errors of ~0.6 eV are found for qsGW CLBE shifts, again similar to
those obtained using ΔSCF PBE. The computationally more affordable G0W0
approximation leads to results less accurate than qsGW, with an error of ~9 eV for
CLBEs, and ~0.9 eV for their shifts. Interestingly, starting G0W0 from PBE0 reduces
this error to ~4 eV with a slight improvement on the shifts as well (~0.4 eV). The
validity of the G0W0 results is however questionable since only linearized quasi particle
equation results can be obtained. The present results pave the way to estimate CLBEs in
periodic systems where ΔSCF calculations are not straightforward although further
improvement is clearly needed. * mjvansetten@gmail.com, orcid.org/0000-0003-0557-5260 † Corresponding author: francesc.illas@ub.edu, orcid.org/0000-0003-2104-6123 1 Graphic for TOC Graphic for TOC Graphic for TOC 2 2 Introduction Electron correlation
effects can be explicitly included through post-Hartree-Fock methods such as
configuration interaction or coupled cluster, or implicitly through the exchange- 3 3 correlation potential in density functional theory (DFT) based methods. The latter has
the additional advantage of being easily implemented for periodic systems, which is
very appealing when dealing with solid materials. In practice, DFT based methods rely on the Kohn-Sham implementation, a
formalism closely resembling Hartree-Fock to the point that upon solving the so-called
Kohn-Sham equations one obtains a set of single particle orbitals (KS orbitals) and
energies (KS orbital energies or KS-εs). However, there are important differences
between HF-εs and KS-εs since the later do not fulfill the Koopmans’ theorem and, in
the case of core states, the numerical values are smaller than the experimental CLBEs. Although not rigorously supported from theory, it has been claimed that conceptually
HF and KS orbitals are similar.9,10 From this perspective, taking the KS-εs as a measure
of initial state CLBEs will lead to an unphysical positive orbital relaxation energy.11,12
The alternative view, to interpret KS orbitals as an approximation to Dyson orbitals,13,14
which, as such, would contain orbital relaxation, hence seems more physical. This point
of view is developed in detail in the review papers by Ortiz15 and by Ortiz and Öhrn.16 Yet, KS-εs become very useful when aiming at analyzing CLBE shifts (ΔCLBE),
that is, the difference between CLBEs for a given element in the system of interest and a
reference value.9 For the 1s core level of carbon atoms for instance, the energy ranges
form 290 to about 300 eV. To distinguish between carbon atoms in similar surroundings
however, an accuracy of the order of 0.1 eV is often needed. Accurate predictions of
CLBEs by means of DFT methods are well possible within a ΔSCF formalism, and
recent work shows that, depending on the functional used, the calculated values are of a
quality similar or superior to those predicted by Hartree-Fock.17,18 However, apart from
the inconvenient dependence on the choice of the functional, it is necessary to point out
that ΔSCF in periodic systems is problematic since it involves the use of charged unit
cells and different, more or less accurate, ways to counteract this have been proposed. Introduction Core level binding energies (CLBEs), experimentally accessible through X-Ray
photoelectron spectroscopy (XPS),1 provide direct insight into the elemental chemistry
of molecules, nanostructures, and solid materials. They contain information about the
chemical environment, oxidation state, hybridization, and coordination of a given type
of atom in a given sample.2,3 XPS applied to solid samples is, in addition, surface
sensitive and, hence, XPS is broadly used in surface science and catalysis. Here, the
assignment of XPS features is not always straightforward and assistance from
theoretical calculations becomes necessary as recognized in several reviews.2,4 For
instance, XPS of atomic O on Al(111) exhibit two well defined peaks that one would
naively assign to O above and below the surface. However, theoretical analysis reveals
that this hypothesis is not consistent with results from ab initio cluster model
calculations, and these offer indeed an alternative explanation.5,6 The landmark paper of Bagus7 on the application of the Hartree-Fock method
defines the variational determination of CLBE as a difference of the total energy
between the molecule’s ground state and the cation with a core hole. The resulting
procedure is generally known as ΔSCF since it implies the difference of two self-
consistently determined energies. Because the involved energies are obtained from a
variational calculation, it turns out that, for a given method used to obtain the total
energy of atoms, molecules or solids, ΔSCF constitutes the best possible estimate of
CLBEs. The use of Hartree-Fock (HF) to predict CLBEs has an additional advantage
since the orbital energies (HF orbital energies or HF-εs) fulfill the Koopman`s theorem8
and, hence, the core orbital energies provide an estimate of the so-called initial state
effects, those already present in the neutral molecule. The difference between the
CLBEs estimated from the HF orbital energy and the ΔSCF calculation is usually
referred to as orbital relaxation energy, and essentially contains the response of the
electron density to the presence of the core hole, usually denoted as a final state
effect.2,3 Note, however, that the distinction between initial and final state is applicable
to XPS main peaks only, i.e., excluding satellites and multiplets. Nevertheless, one has
to point out that Hartree-Fock lacks electron correlation effects, which, although
modestly contributing to the CLBEs, would need to be considered. Introduction These range from the use of a uniform negative background that may distort the
electron density, to various pseudopotential based approaches 19 among which a
common choice is to make use of a pseudopotential extracted from a core ionized
atom20 with concomitant limitations due to constraints in the extent of core electron
density relaxation. Recent work shows that these procedures are useful on determining
ΔCLBEs but, unfortunately, fail to predict accurate absolute values of CLBEs.21 4 A possible way to overcome the difficulties arising from the use of DFT
methods to approach CLBEs in periodic systems is to make use of the GW formulation,
earlier introduced by Hedin.22 The GW method includes many body effects beyond the
mean-field description of the electron-electron interaction in DFT via the so-called self-
energy, which is non-local and energy dependent and, in a sense, replaces the exchange
correlation potential in DFT.23,24 Reviewing the GW formalism here is beyond the scope
of this letter and the interested reader is addressed to the excellent review of
Aryasetiawan and Gunnarsson for detailed information 25 All available practical
implementations of GW have a common feature, the need for large computational
resources. This has led to different levels of approximations resulting in various flavors
denoted as G0W0, GW0, eigenvalue self-consistent GW, or quasi-particle self-consistent
GW.26 All these approximations convert the KS particles into quasiparticles (qp) with a
well-defined physical meaning. In the case of occupied states, the qp energies
effectively represent ionization potentials whereas in the case of unoccupied states, they
represent electron affinities. In the case of periodic insulators and semiconductors qp
from GW calculations provide an accurate estimate of the fundamental band gap of
these materials27,28 and also of states associated to point defects such as F-centers in
simple oxides.29,30 Apart from the high computational cost of the GW calculations, one
must also point out the dependence of the results with respect to the starting density,31
when the results are not obtained from a fully self-consistent approach as is the case in
G0W0, lowest level of this theoretical framework. Introduction In particular, it has been suggested
that hybrid functionals should provide a better starting point.32 The GW method has also been applied to molecular systems and small
clusters.33-42 In particular, the G0W0 level has been applied to 100 molecules (the
GW100 database)43 with excellent performance for the vertical ionization potential; the
same database has more recently been used to benchmark different implementations and
levels of self-consistency of the theory.44,45 The rather good success of GW in predicting
the ionization potential of molecular systems strongly suggest that it may as well
provide an estimate of the CLBEs in molecules and solids where, as above commented,
ΔSCF calculations are cumbersome. The calculation of CLBEs obviously requires the
presence of actual core levels in the calculation. Note, however, that many solid-state
codes use pseudopotentials with a frozen core approximation. These make the direct
calculation of CLBEs impossible although different approximations have been proposed 5 5 to predict ΔCLBEs.14 Moreover, one must point out that GW calculations for core levels
may be very tricky since describing the self-energy at deep energies requires a full
frequency method and solving the quasiparticle equation can be very complicated as
discussed later on. The goal of the present work is precisely to investigate the
performance of GW on predicting CLBEs using a series of simple molecules, where
experimental data and ΔSCF results for several exchange-correlation functionals are
available, as a convenient benchmark. Computational details. The ground-state calculations are based on density functional theory within the
Perdew-Burke-Ernzerhof (PBE)46 generalized-gradient approximation (GGA) for the
exchange-correlation functional. The molecular data set has been taken from previous
work.17 The molecular structures have been fully relaxed using the PBE functional and
a tight Tier 2 numerical atom-centered orbitals (NAO) basis set47-49 using the FHI-aims
code.47 The quasi-particle calculations are performed with the Turbomole package using
the def2-TZVP and def2-QZVP Gaussian basis sets50 and extrapolated to the complete
basis-set limit.51 In addition to electron correlation, relativistic effects also play a role in
determining the final value of calculated CLBEs.2,12,17 Within the GW method, these
can be introduced through various formalisms. 52 , 53 Nevertheless, for sake of
comparison, relativistic effects have been omitted. We note, however, that the
relativistic contribution to the CLBEs is essentially atomic in nature and it increases
with the atomic number. For the core levels studied in the present work contribution of
relativistic effects to the CLBE vary from 0.13 eV for C to 0.75 eV for F,17 increasing
along the C−F series. Different levels of GW calculations have been carried out including the so-called
one shot G0W0 approach, where both the Green’s function G and the screened Coulomb
potential W are obtained from an initial electron density arising from a given density
functional approach, and the quasi particle self-consistent GW (qsGW), where both G
and W are iterated until self consistency is achieved. Clearly, qsGW results do not
depend on the starting exchange-correlation functional and this has been numerically
verified here. Consequently, qsGW results are taken here as benchmark.54,55 Note also
that two different implementations of the simpler G0W0 level are used. Nevertheless,
both implementations of G0W0 and qsGW approaches use the full analytic expression 6 from the reducible response function for the self-energy and hence the results are
obtained from a full frequency method. The difference in the two G0W0 approaches used
lies only in how the quasi particle equation is solved. In one case, the quasi particle
equation is iteratively solved, in the other, denoted as lin-G0W0, the quasi particle
equation is linearized. For the HOMO level the linearized and solved version often give
very similar results. For core levels this is not the case. Computational details. Due to the much larger
corrections for core levels, in all molecules studied here the G0W0@PBE self-energy has
poles in the region where the quasi particle equation (QPE) needs to be solved to
resolve the core-levels. A unique solution hence becomes impossible. The final G0W0
results reported in this work are all obtained from the linearized version, which
circumvents this problem. Besides the PBE functional we also report lin-G0W0 starting
from PBE0. For the PBE0 starting point the solving the QPE is also not uniquely
possible. Only for a HF starting point this becomes feasible, see supplementary
information. For a complete description of the GW implementations including the two
G0W0 approaches mentioned above, the reader is referred to the original work on the
implementation of GW in Turbomole56 reporting also result for valence ionization
potentials of a broad set of molecules for G0W0, and in Kaplan et al. for qsGW.57 Results and discussion The set of molecules used to assess the reliability of the different methods
considered here consists of CH4, CF4, CO2, HCN, H3COCH3, H3COH, H2O, NH3,
pyridine, and pyrrole. The calculations are carried out for the C, N, O, and F 1s core
levels for which experimental data is available. For reproducibility, all molecular
geometries are reported in Table S1 of the supporting information. In addition, Table S2
reports the PBE total energy at the optimized geometry corresponding to the tight Tier 2
NAO basis set. To provide insight into the quality of this basis set Table S2 also reports
the PBE total energy, at the same geometry, obtained with Gaussian type orbital (GTO)
basis sets of aug-cc-pCVTZ and aug-cc-pCV5Z level of quality,58,59 computed using
Gaussian09.60 Inspection of Table S2 clearly shows that the tight Tier 2 NAO basis set
quality is even higher than that of aug-cc-pCVTZ and only slightly below aug-cc-
pCV5Z. Table 1 collects the CLBEs for the whole set of molecules as obtained from the
different levels of theory together with experimental values taken from the
literature.61,62 Nevertheless, one must advert that with the current implementation, fully 7 7 analytic frequency treatment and with only a smp parallelism in TURBOMOLE, the
calculation for pyridine at the def2-QZVP is not feasible and the corresponding
extrapolated result could not be obtained; here the def2-TZVP value is listed instead. A
quick inspection of Table 1 shows that, on average, going from KS to lin-G0W0 to
qsGW, the calculated results are progressively approaching experiment with best results
attained for the qsGW method. A deeper insight into the relative errors is provided in
Table 2 showing that the error for each core level is different with a mean absolute error
(MAE) larger than 25 eV for the KS-εs estimates, about 3-9 eV for G0W0, and slightly
less than 2 eV at the highest qsGW level. These results illustrate that qsGW significantly
improves the agreement with experiment as compared to the KS eigenvalues, a result
which is in line with previous findings concerning the first ionization potential (HOMO)
in molecules.43,63 Nevertheless, the accuracy reached by qsGW may not be sufficient to
properly interpret XPS experiments. The present results illustrate the difficulties of the
GW methods on describing CLBEs. We stress that describing the self-energy at deep
energies requires a full frequency method, which makes the calculations quite costly. Results and discussion Also, for CLBEs, G0W0 starting from the PBE functional faces fundamental problems
since the self-energy has poles in the region where the quasiparticle equation is solved
giving rise to a multitude of possible solutions. Besides the obvious numerical issues
this grossly complicates the physical interpretation. Using the linearized quasiparticle
equation solves at least the numerical part of the problem, solutions become numerically
stable and unique but accuracy remains poor. In practice, it is more suitable to predict
the quasiparticle energies of valence states.30 Using a hybrid functional as starting point
may remediate this deficiency as shown in valence states comparing to experiment31
and also to scGW calculations.64 This is because the core levels are already deeper
causing the poles of the self-energy to be deeper in energy as well. Alternatively one
can start from the HF density but here the lin-G0W0 leads to values, which, not
surprisingly, are overestimating. The advantage is that in this case the QPE can be
solved, the original poles are further away and come up in energy during the self-
consistency process. The average over overestimation is however roughly 6 eV (see
Table S4); a significant improvement over HF-εs again with the correct physical
ingredients where final state effects are introduced by the many body terms of the GW
approach but a deterioration with respect to G0W0@PBE0. In the case of the qsGW, the
poles move deeper as well and, in addition, the contribution of the off-diagonal terms of
h
lf
i
l
k i l
h
l
k d 8 For CLBEs calculated using the PBE functional and employing a ΔSCF
approach the MAE is of ~1 eV, this is larger than the 0.3-0.4 eV corresponding values
for calculations at the Hartree-Fock level or using the meta-GGA TPSS functional.17 In
this sense, the MAE of ~2 eV for the qsGW results is really remarkable and puts this
method as a good choice to estimate CLBEs, especially in periodic solids. Here one
must point out that the qsGW result does not depend on the starting point, yet the fact
that the accuracy reached is lower than ΔSCF with PBE0 or TPSS functionals17
indicates that higher order terms are likely to be needed in the expansion of the screened
potential W and the self-energy. Results and discussion Finally, it is also important to note that, in most practical cases, one is not
interested in the absolute CLBEs but in their shifts with respect to a given reference
(ΔCLBE). Taking CH4, H2O, and NH3 as references for the C(1s), O(1s), and N(1s), the
values for the ΔCLBEs arising from KS-εs, lin-G0W0@PBE, lin-G0W0@PBE0, and
qsGW are reported in Table 3 and the statistic analysis of errors in Table 4. The later
shows that MAE values of 0.7, 0.9, 0.4 and 0.6 eV are found for KS-εs, lin-G0W0@PBE,
lin-G0W0@PBE0, and qsGW respectively. Note, however, that the experimental ΔCLBE
values seldom exceed 10 eV, which makes the relative errors still rather large. Another
important feature emerges when comparing calculated and experimental ΔCLBE values. Figure 1 shows that while the trends are well reproduced, there is a significant
dispersion with the best values corresponding to lin-G0W0@PBE0, and qsGW
respectively. Conclusions A study has been carried out to assess the performance of different GW
approaches in predicting core level binding energies (CLBEs). Results presented for a
series of representative molecules allows us to establish some first conclusions. First, the present calculations agree with previous studies9,10 showing that the
Kohn-Sham orbital energies cannot be taken as a measure of the CLBEs and they do not
represent an estimate of initial state effects. The Kohn-Sham orbital energies are always
smaller than experimental CLBEs with differences with respect to experiment in the 20-
30 eV range. Interpreting them as approximate initial states would hence imply
unphysical positive relaxation energies. In the framework of DFT, initial state effects
can be recovered by computing the molecular system with a core hole with the fixed 9 9 electron density of neutral molecule as discussed in previous works10 and will not be
further commented here. The interpretation of the KS-orbitals as approximate Dyson
orbitals hence seems more physical especially for core levels. The most accurate results, as compared to experiment, are obtained at the
quasiparticle self-consistent GW level (qsGW). Nevertheless, to achieve good agreement
with experiment, extrapolation to the complete basis set limit is needed, especially for
the absolute CLBE although it may be less problematic for the shifts. In any case, the
rather good agreement with experiment evidences that GW is able to largely correct the
failure of the KS and HF orbital energies covering a large part of relaxation energy, thus
going in the direction of the right-answer for the right-reason. We find that the qsGW calculated CLBEs are always larger than the
experimental values, as expected due to underscreening. The mean absolute error with
respect to experiment is reduced to about 2 eV; one order of magnitude smaller than for
KS values and close to predictions from variational ΔSCF calculations which, for the
PBE functional, are in the 1 eV range. Here it is important to point out that an absolute
error of 2 eV on a quantity in the 300-500 eV range implies a percent error of 0.6%
only, which is quite remarkable, and, actually, very similar to the percent error of 0.5%
achieved when considering the qsGW prediction of the first ionization potential.43 a
somehow larger error (6.5%) is obtained in the case of lin-G0W0@HF. Conclusions Note also that
adding contribution from relativistic effects will further decrease the ΔSCF error to less
than 1 eV. Yet, the error bar of XPS, especially when synchrotron radiation is used, can
be as small as 0.1 eV, meaning that further developments are needed. Smaller errors of
roughly 0.6 eV are found for the qsGW calculated core level binding energy shifts
(ΔCLBE) although these are surprisingly larger than those arising from Kohn-Sham
orbital energies. This is an indication that qsGW effects on same core level in different
molecules are not equally taken into account. The simpler G0W0 and computationally more affordable level of the theory also
leads to a significant overall improvement with respect to KS orbital energies. However,
starting from the PBE density, the calculated CLBEs are still smaller than experiment
evidencing limitations inherent to this level of approximation as commented in the
previous section. The absolute mean errors (8.98 and 3.76 eV for the PBE and PBE0
starting points respectively) are significantly larger than the corresponding value for 10 qsGW but smaller than the one arising from the direct use of KS orbital energies. A
paired T-test indeed proves that the differences are significant. To summarize, CLBEs derived from GW approaches represent a considerable
improvement with respect to KS energies although the accuracy reached with the
present implementations, even for the qsGW, is still lower than the one obtained from
ΔSCF calculations. The simpler G0W0 method applied on top of the PBE density also
lead to results improved with respect to KS predictions but with too large errors. These
errors are considerably reduced when starting the G0W0 calculations from a density
obtained from a hybrid functional. Similar considerations apply to the core level
binding energy shifts with rather satisfactory results for the lin-G0W0@PBE0 and qsGW
methods. For practical applications in computational materials science, G0W0 on top of
single point PBE0 density at the PBE optimized structure may provide a practical
approach since this will be computationally less demanding than going to the qsGW
level and lead to similar accuracy. Acknowledgements This research was supported by the Spanish MINECO/FEDER CTQ2015-64618-R
grant and, in part, by Generalitat de Catalunya (grants 2014SGR97 and XRQTC) and
by the NOMAD Center of Excellence project, which received funding from the
European Union’s Horizon 2020 research and innovation programme under grant
agreement No 676580. F.V. thanks MINECO for a postdoctoral Ramón y Cajal (RyC)
research contract (RYC-2012-10129) and F.I. acknowledges additional support from the
2015 ICREA Academia Award for Excellence in University Research. We also thank
the Céci facilities funded by F.R.S-FNRS (Grant No. 2.5020.1) and Tier-1
supercomputer of the Fédération Wallonie-Bruxelles funded by the Walloon Region
(Grant No. 1117545). The authors are very grateful to Dr. Fabio Caruso of Humboldt-Universität zu Berlin
(Germany) and Prof. Patrick Rinke of Aalto University in Helsinki (Finland) for their
many helpful discussions regarding the G0W0 methods and their implementations in
FHI-AIMS computer program package. Mrs. Helena Muñoz Galan is also
acknowledged for helping to carry out some of the preliminary, not reported here,
calculations. 11 Supporting information The Supporting Information is available free of charge on the ACS Publications website
at DOI: xxxxxxxxxxxxxx Table S4. CLBEs as obtained from lin-G0W0@HF. The Supporting Information is available free of charge on the ACS Publications website
at DOI: xxxxxxxxxxxxxx Table S1. Cartesian coordinates of the molecules studied in the present work as
optimized FHI-AIMS tight Tier 2 PBE. Table S2. Total PBE energy (eV) of the molecules studied in the present work with
geometries as in Table S1 and corresponding to numerical atomic orbital (NAO) tight
Tier 2 basis set using the aims code and using the FHI-AIMS code and to the aug-cc-
pCVTZ and aug-cc-pCV5Z gaussian type orbitals (GTO) basis sets as obtained with the
Gaussian09 code. Table S3. Total PBE energy (eV) of the molecules studied in the present work with
geometries as in Table S1 as obtained with the def2-VPTZ GTO basis and different
codes. Table S4. CLBEs as obtained from lin-G0W0@HF. Table S4. CLBEs as obtained from lin-G0W0@HF. 12 12 Table 1. Core level binding energies as obtained from KS, lin-G0W0@PBE, lin-
G0W0@PBE0, and quasi particle self-consistent (qsGW) methods starting from PBE. All
values correspond to a linear extrapolation of the def2-TZVP and def2-QZVP results to
the complete basis set limit. Experimental values (positive as they correspond to
ionization processes) are provided for comparison. Core level Molecule
KS
lin-G0W0
@PBE
lin-G0W0
@PBE0
qsGW
Experiment
C(1s)
CH4
-268.83
-284.10
-287.82
-293.54
290.90
CF4
-277.89
-292.18
-299.39
-302.84
301.80
CO2
-273.69
-287.48
-294.40
-298.78
297.91
HCN
-270.41
-286.13
-290.45
-294.67
293.50
H3COCH3
-269.96
-284.99
-289.14
-294.13
292.30
H3COH
-269.96
-286.15
-291.14
-294.69
292.30
Pyrrole
-268.87
-282.88
-286.78
-293.02
290.80
O(1s)
H2O
-510.09
-529.45
-534.90
-542.13
539.70
CO2
-512.42
-530.41
-536.28
-542.46
541.30
H3COH
-510.13
-528.72
-534.28
-541.12
538.90
H3COCH3
-510.35
-528.15
-533.24
-540.77
539.03
N(1s)
NH3
-379.82
-397.04
-401.75
-407.84
405.60
HCN
-381.18
-397.38
-402.21
-408.78
406.80
Pyridine
-380.51
-395.24
-400.66
-407.4*
404.90
Pyrrole
-381.74
-397.10
-402.31
-408.11
406.10
F(1s)
CF4
-661.79
-683.19
-689.26
-696.15
692.40
*) def2-TZVP value; see text 13 Table 2. Mean absolute error (MAE) for the core level binding energies reported in
Table 1 and obtained from KS, lin-G0W0@PBE, lin-G0W0@PBE0, and quasi particle
self-consistent (qsGW) methods MAE
Core level
KS
lin-G0W0
@PBE
lin-G0W0
@PBE0
qsGW
C(1s)
22.84
7.94
2.91
1.74
O(1s)
28.99
10.55
5.06
1.89
N(1s)
25.04
9.16
4.12
2.08
TOTAL
25.41
8.98
3.76
1.98 14 14 Table 3. ΔCLBEs as obtained from KS, lin-G0W0@PBE, lin-G0W0@PBE0, and quasi
particle self-consistent (qsGW) methods starting from PBE, using CH4, H2O, and NH3
as reference for C(1s), O(1s), and N(1s) core levels. The Supporting Information is available free of charge on the ACS Publications website
at DOI: xxxxxxxxxxxxxx All values are given in eV Core level Molecule
KS
lin-G0W0
@PBE
lin-G0W0
@PBE0
qsGW
Experiment
C(1s)
CH4
0.00
0.00
0.00
0.00
0.00
CF4
-9.06
-8.08
-11.57
-9.30
-10.90
CO2
-4.86
-3.38
-6.58
-5.24
-7.01
HCN
-1.58
-2.03
-2.63
-1.13
-2.60
H3COCH3
-1.13
-0.89
-1.32
-0.59
-1.40
H3COH
-1.13
-2.05
-3.32
-1.15
-1.40
Pyrrole
-0.04
1.22
1.04
0.52
0.10
O(1s)
H2O
0.00
0.00
0.00
0.00
0.00
CO2
-2.33
-0.96
-1.38
-0.33
-1.60
H3COH
-0.04
0.73
0.62
1.01
0.80
H3COCH3
-0.26
1.30
1.66
1.36
0.67
N(1s)
NH3
0.00
0.00
0.00
0.00
0.00
HCN
-1.36
-0.34
-0.46
-0.94
-1.20
Pyridine
-0.69
1.80
1.09
—
0.70
Pyrrole
-1.92
-0.06
-0.56
-0.27
-0.50 15 Table 4. Mean absolute error (MAE) for the ΔCLBEs reported in Table 3 and obtained
from KS, lin-G0W0@PBE, lin-G0W0@PBE0, and quasi particle self-consistent (qsGW)
methods. MAE
Core level
KS
lin-G0W0
@PBE
lin-G0W0
@PBE0
qsGW
C(1s)
0.81
1.33
0.58
0.90
O(1s)
0.63
0.34
0.35
0.54
N(1s)
0.74
0.60
0.30
0.16
TOTAL
0.74
0.87
0.44
0.64 16 Figure 1.- Calculated KS, lin-G0W0@PBE, lin-G0W0@PBE0, and quasi particle self-
consistent (qsGW) versus experimental core level binding energy shifts (ΔCLBE). The
dashed line corresponds to a perfect agreement. Figure 1.- Calculated KS, lin-G0W0@PBE, lin-G0W0@PBE0, and quasi particle self-
consistent (qsGW) versus experimental core level binding energy shifts (ΔCLBE). The
dashed line corresponds to a perfect agreement. 17 17 References 1
Barr, T.L. The Principles and Practice of X-Ray Photoelectron Spectroscopy. Taylor and Francis, London, 1994. 2
Bagus, P.S.; Ilton, E.S.; Nelin C.J. The interpretation of XPS spectra: Insights into
materials properties. Surf. Sci. Rep. 2013, 68, 273-304. 3
Bagus, P.S.; Illas, F.; Pacchioni, G.; Parmigiani, F. Mechanisms responsible for
chemical shifts of core-level binding energies and their relationship to chemical
bonding. J. Electr. Spectrosc. Relat. Phenom. 1999, 100, 215-236. 4
Egelhoff, W.F. Core-level binding-energy shifts at surfaces and in solids. Surf. Sci. Rep. 1987, 6, 253-415. 5
Bagus, P.S.; Parmigiani, F.; Polzonetti, G..; Illas, F.; Brundle, C. R. X-ray
photoelectron spectroscopy of oxygen adsorbates on Al(111) Theory and
experiment. J. Vac. Sci. Technol. A 1991, 9, 1747-1748. 6
Bagus, P.S.; Brundle, C.R.; Illas, F.; Parmigiani, F.; Polzonetti, G. Evidence for
oxygen island formation on Al(111) Cluster model theory and X-ray photoelectron
spectroscopy. Phys. Rev. B 1991, 44, 9025-9034. 7
Bagus, P.S. Self-Consistent-Field Wave Functions for Hole States of Some Ne-
Like and Ar-Like Ions. Phys. Rev. 1965, 139, 619–634. 8
Koopmans, T. The Supporting Information is available free of charge on the ACS Publications website
at DOI: xxxxxxxxxxxxxx Über die Zuordnung von Wellenfunktionen und Eigenwerten zu den
Einzelnen Elektronen eines Atoms. Physica 1933, 1, 104-113. 9
Baerends, E.J.; Gritsenko, O.V. A Quantum Chemical View of Density Functional
Theory. J. Phys. Chem. A 1997, 101, 5383–5403. 10 Stowasser, R.; Hoffmann, R. What Do the Kohn−Sham Orbitals and Eigenvalues
Mean? J. Am. Chem. Soc. 1999, 121, 3414–3420. 11 Pueyo Bellafont, N.; Illas, F.; Bagus, P. S. Validation of Koopmans' theorem for
density functional theory binding energies. Phys. Chem. Chem. Phys. 2015, 17,
4015-4019 18 18 12 Pueyo Bellafont, N.; Bagus, P. S.; Illas, F. Prediction of core level binding energies
in density functional theory: Rigorous definition of initial and final state
contributions and implications on the physical meaning of Kohn-Sham energies. J. Chem. Phys. 2015, 142, 214102 (1-12). 13 Duffy P.; Chong D.P.; Casida, M.E.; Salahub, D.R. Assessment of Kohn-Sham
density-functional orbitals as approximate Dyson orbitals for the calculation of
electron-momentum-spectroscopy scattering cross sections. Phys. Rev. A 1994, 50,
4707-4728. 14 Gritsenko, O.V.; Baerends, E.J. The analog of Koopmans’ theorem in spin-density
functional theory. J. Chem. Phys. 2002, 117, 9154-9159. 15 Ortiz, J.V. Brueckner Orbitals, Dyson Orbitals, and Correlation Potentials. Int. J. Quantum Chem. 2004, 100, 1131-1135. 16 Öhrn, Y.; Ortiz, J. V. The electron propagator. Mol. Phys. 2010, 108, 2871-2875
and references therein 17 Pueyo Bellafont, N.; Viñes, F.; Illas, F. Performance of the TPSS Functional on
Predicting Core Level Binding Energies of Main Group Elements Containing
Molecules: A Good Choice for Molecules Adsorbed on Metal Surfaces. J. Chem. Theory and Comput. 2016, 12, 324-331. 18 Pueyo Bellafont, N.; Álvarez Saiz, G.; Viñes, F.; Illas, F. Performance of
Minnesota functionals on predicting core-level binding energies of molecules
containing main-group elements. Theor. Chem. Acc. 2016, 135, 35-43. 19 Pehlke, E; Scheffler, M. Evidence for Site-Sensitive Screening of Core Holes at the
Si And Ge (001) Surface. Phys. Rev. Lett. 1993, 71, 2338-2341. 20 Köhler, L.; Kresse, G. Density functional study of CO on Rh(111). Phys. Rev. B
2004, 70, 165405. 21 Pueyo Bellafont, N.; Viñes, F.; Hieringer, W.; Illas, F. Accurate Prediction of Core
Level Binding Energies Shifts: Suitability of Projector Augmented Wave Approach
as implemented in VASP. J. Comput. Chem. 2017, 38, 518-522. 19 19 22 Hedin, L. New Method for Calculating the One-Particle Green's Function with
Application to the Electron-Gas Problem. Phys. Rev. 1965, 139, A796-A822. 23 Strinati, G.; Mattausch, H. J.; Hanke, W. The Supporting Information is available free of charge on the ACS Publications website
at DOI: xxxxxxxxxxxxxx Dynamical Correlation Effects on the
Quasiparticle Bloch States of a Covalent Crystal. Phys. Rev. Lett. 1980, 45, 290−
294. 24 Hybertsen, M. S.; Louie, S. G. Electron Correlation in Semiconductors and
Insulators: Band Gaps and Quasiparticle Energies. Phys. Rev. B 1986, 34, 5390−
5413. 25 Aryasetiawan, F.; Gunnarsson, O. The GW method. Rep. Prog. Phys. 1998, 61,
237-312. 26 Faleev, S.; van Schilfgaarde, M.; Kotani, T. All-electron self-consistent GW
approximation: Application to Si, MnO, and NiO. Phys. Rev. Lett. 2004, 93,
126406 (1-4). 27 Kang, W.; Hybertsen, M. S. Quasiparticle and Optical Properties of Rutile and
Anatase TiO2. Phys. Rev. B 2010, 82, 085203 (1-11). 28 Zhu, T.; Gao, S. P. The Stability, Electronic Structure, and Optical Property of TiO2
Polymorphs. J. Phys. Chem. C, 2014, 118, 11385-11396. 29
Rinke, P.; Schleife, A.; Kioupakis, E.; Janotti, A.; Rödl, C.; Bechstedt, F.;
Scheffler, M.; Van de Walle, C.G. First-principles optical spectra for F centers in
MgO. Phys. Rev. Lett. 2012, 108, 126404 (1-5). 30
Tosoni, S.; Fernandez Hevia, D.; Perez Peña, J.; Illas, F. Accurate description of
optical absorption of F centers from quasiparticle band structure calculations. Phys. Rev. B 2012, 85, 115114 (1-4). 31 Liao P.; Carter, E.A. Testing variations of the GW approximation on strongly
correlated transition metal oxides: hematite (α-Fe2O3) as a benchmark. Phys. Chem. Chem. Phys. 2011, 13, 15189-15199. 20 32 Ren, X.; Rinke, P.; Scheffler, M. Exploring the Random Phase Approximation:
Application to CO adsorbed on Cu(111). Phys. Rev. B 2009, 80, 045402 (1-8). 33 Ishii, S.; Ohno, K.; Kawazoe, Y.; Louie, S. G. Ab Initio GW Quasiparticle Energies
of Small Sodium Clusters by an All-electron Mixed-basis Approach Phys. Rev. B
2001, 63, 155104 (1-6). 34 Tiago, M. L.; Chelikowsky, J. R. Optical Excitations in Organic Molecules,
Clusters, and Defects Studied by First-Principles Green’s Function Methods. Phys. Rev. B 2006, 73, 205334 (1-19) 35 Bruneval, F. GW Approximation of the Many-Body Problem and Changes in the
Particle Number. Phys. Rev. Lett. 2009, 103, 176403 (1-4). 36 Rostgaard, C.; Jacobsen, K. W.; Thygesen, K. S. Fully Self-consistent GW
Calculations for Molecules. Phys. Rev. B 2010, 81, 085103 (1-10). 37 Baumeier, B.; Andrienko, D.; Ma, Y.; Rohlfing, M. Excited States of
Dicyanovinyl-Substituted Oligothiophenes from Many-Body Green’s Functions
Theory. J. Chem. Theory Comput. 2012, 8, 997-1002. 38 Caruso, F.; Rinke, P.; Ren, X.; Scheffler, M.; Rubio, A. The Supporting Information is available free of charge on the ACS Publications website
at DOI: xxxxxxxxxxxxxx Unified Description of
Ground and Excited States of Finite Systems: The Self-consistent GW Approach. Phys. Rev. B 2012, 86, 081102 (1-5). 39 Faber, C.; Boulanger, P.; Duchemin, I.; Attaccalite, C.; Blase, X. Many-body
Green's Function GW and Bethe-Salpeter Study of the Optical Excitations in a
Paradigmatic Model Dipeptide. J. Chem. Phys. 2013, 139, 194308 (1-10). 40 Govoni, M.; Galli, G. Large Scale GW Calculations. J. Chem. Theory Comput. 2015, 11, 2680−2696. 41 Laflamme Janssen, J.; Rousseau, B.; Cote, M. Efficient Dielectric Matrix
Calculations Using the Lanczos Algorithm for Fast Many-body G0W0
Implementations. Phys. Rev. B 2015, 91, 125120 (1-17). 21 42 Maggio, E.; Kresse, G. GW Vertex Corrected Calculations for Molecular Systems. J Chem Theory Comput. 2017, 13, 4765-4778. 43 van Setten, M. J.; Caruso, F.; Sharifzadeh, S.; Ren, X.; Scheffler, M.; Liu, F.;
Lischner, J.; Lin, L.; Deslippe, J. R.; Louie, S. G.; Yang, C.; Weigend, F.; Neaton,
J. B.; Evers, F.; Rinke, P. GW100: Benchmarking G0W0 for Molecular Systems. J. Chem. Theory Comput. 2015, 11, 5665−5687. J. B.; Evers, F.; Rinke, P. GW100: Benchmarking G0W0 for Molecular Systems. J. Chem. Theory Comput. 2015, 11, 5665−5687. 44 Caruso, F.; Dauth, M.; van Setten, M.J.; Rinke, P. Benchmark of GW Approaches
for the GW100 Test Set. J. Chem. Theory Comput. 2016, 12, 5076−5087. 45 Maggio, E.; Liu, P.; van Setten, M.J.; Kresse, G. GW100: A Plane Wave
Perspective for Small Molecules. J. Chem. Theory Comput. 2017, 13, 635−648. 46 Perdew, J.P.; Burke, K.; Ernzerhof, M. Generalized Gradient Approximation Made
Simple. Phys. Rev. Lett. 1996, 77, 3865-3868. 47 Blum, V.; Gehrke, R.; Hanke, F.; Havu, P.; Havu, V.; Ren, X.; Reuter, K.;
Scheffler, M. Ab Initio Molecular Simulations with Numeric Atom-centered
Orbitals. Comput. Phys. Commun. 2009, 180, 2175–2196. 48 Havu, V.; Blum, V.; Havu, P.; Matthias S. Efficient O(N) Integration for All-
electron Electronic Structure using Numerically Tabulated Basis Functions. J. Comput. Phys. 2009, 228, 8367-8379. 49 Ren, X.; Rinke, P.; Blum, V.; Wieferink, J.; Tkatchenko, A.; Sanfilippo, A.; Reuter,
K.; Scheffler, M. Resolution-of-identity Approach to Hartree-Fock, Hybrid Density
Functionals, RPA, MP2, and GW with Numeric Atom-centered Orbital Basis
Functions. New J. Phys. 2012, 14, 053020 (1-55). 50 Weigend, F.; Ahlrichs, R. Balanced basis sets of split valence, triple zeta valence
and quadruple zeta valence quality for H to Rn: Design and assessment of accuracy. Phys. Chem. Chem. Phys. 2005, 7, 3297−3305. The Supporting Information is available free of charge on the ACS Publications website
at DOI: xxxxxxxxxxxxxx 51 TURBOMOLE V7.0 2015, a development of University of Karlsruhe and
Forschungszentrum Karlsruhe GmbH, 1989−2007, TURBOMOLE GmbH, since
2007. Available from http://www.turbomole.com 22 52 Kühn, M.; Weigend, F. One-Electron Energies from the Two-Component GW
Method. J. Chem. Theory Comput. 2015, 11, 969− 979. 53 Scherpelz, P.; Govoni, M.; Hamada, I.; Galli, G. Implementation and Validation of
Fully Relativistic GW Calculations: Spin–Orbit Coupling in Molecules,
Nanocrystals, and Solids. J. Chem. Theory Comput. 2016, 12, 3523-3544. 54 van Schilfgaarde, M.; Kotani, T.; Faleev, S. Quasiparticle Self-Consistent GW
Theory. Phys. Rev. Lett. 2006, 96, 226402 (1-4). 55 Kotani, T.; Van Schilfgaarde, M.; Faleev, S. V. Quasiparticle self-consistent GW
method: A basis for the independent-particle approximation. Phys. Rev. B 2007, 76,
165106 (1−24). 56 van Setten, M. J.; Weigend, F.; Evers, F. The GW-Method for Quantum Chemistry
Applications: Theory and Implementation. J. Chem. Theory Comput. 2013, 9, 232-
246. 57 Kaplan, F.; Seiler, C.; Weigend, F.; Evers, F.; van Setten, M.J. Quasiparticle-
selfconsistent GW for molecules. J. Chem. Theory Comput. 2016, 12, 2528–2541. 58 Feller, D. The Role of Databases in Support of Computational, Chemistry
Calculations. J. Comp. Chem. 1996, 17, 1571-1586. 59 Schuchardt, K.L.; Didier, B.T.; Elsethagen, T.; Sun, L.; Gurumoorthi, V.; Chase, J.;
Li, J.; Windus, T.L. Basis set exchange: a community database for computational
sciences. J. Chem. Inf. Model. 2007, 47, 1045-1052. 60 Gaussian 09, Revision A.02. Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.;
Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.;
Mennucci, B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H. P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.; Ehara, M.;
Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao,
O.; Nakai, H.; Vreven, T.; Montgomery, J. A., J.; Peralta, J. E.; Ogliaro, F.;
Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin, K. N.; Staroverov, V. N.;
Kobayashi, R.; Normand, J.; Raghavachari, K.; Rendell, A.; Burant, J. C.; Iyengar,
S. S.; Tomasi, J.; Cossi, M.; Rega, N.; Millam, J. M.; Klene, M.; Knox, J. E.; Cross,
J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; 23 Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Martin, R. L.;
Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.;
Dapprich, S.; Daniels, A. D.; Farkas, Ö.; Foresman, J. B.; Ortiz, J. The Supporting Information is available free of charge on the ACS Publications website
at DOI: xxxxxxxxxxxxxx V.; Cioslowski,
J.; Fox, D. J. Wallingford CT 2009. 61 Bakke, A.A.; Chen, H.W.; Jolly, W.L. A table of absolute core-electron binding-
energies for gaseous atoms and molecules. J. Electron Spectrosc. Relat. Phenom. 1980, 20, 333-366. 62 Myrseth, V.; Saethre, L.J.; Borve, K.J.; Darrah Thomas, T. The Substituent Effect
of the Methyl Group. Carbon 1s Ionization Energies, Proton Affinities, and
Reactivities of the Methylbenzene. J. Org. Chem. 2007, 72, 5715-5723. 63 Rostgaard, C.; Jacobsen, K. W.; Thygesen, K. S. Fully Self-consistent GW
Calculations for Molecules. Phys. Rev. B 2010, 81, 085103. 64 Gao, S. P. Cubic, Wurtzite, and 4H-BN Band Structures Calculated using GW
Methods and Maximally Localized Wannier Functions Interpolation. Comput. Mater. Sci. 2012, 61, 266-269. 24
|
https://openalex.org/W4250546024
|
https://www.jour.fnisc.ru/index.php/vlast/article/download/8381/8185
|
Russian
| null |
Стратегия национальной безопасности Российской Федерации
|
Vlastʹ
| 2,021
|
cc-by
| 1,909
|
GEOPOLITICAL CHALLENGES
AND RISKS TO THE NATIONAL SECURITY
OF THE RUSSIAN FEDERATION IN THE FAR EAST Abstract. The article identifies the main geopolitical challenges and threats to Russia's national security in the Far East. The author determines external threats and especially internal problems of the macroregion and proposes the directions of
overcoming the challenges and risks to the national security in the Far East. Keywords: national security, macroregion, military security, economic security АЗИМОВ Фуркат Абдухамидович, аспирант кафедры международных отношений и политологии
Нижегородского государственного лингвистического университета им. Н.А. Добролюбова (603155,
Россия, г. Нижний Новгород, ул. Минина, 31а; furkat72@inbox.ru) 79 79 2021’04 ВЛАСТЬ AVDOI Joni Titalovich, Applicant at the Institute of Socio-Political Studies, Federal Center of Theoretical and Applied
Sociology of the Russian Academy of Sciences (bld. 1, 6 Fotievoj St, Moscow, Russia, 119333; adt81@mail.ru) /
/
_
_
_
/ ( р
р
)
Конституция РФ. Доступ: http://www.kremlin.ru/acts/constitution (проверено 19.07.2021 1 Федеральный закон «О безопасности» от 28.12.2010 № 390-ФЗ. Доступ: http://www.
consultant.ru/document/cons_doc_LAW_108546/ (проверено 19.07.2021).
2 1 Федеральный закон «О безопасности» от 28.12.2010 № 390-ФЗ. Доступ: http://www.
consultant.ru/document/cons_doc_LAW_108546/ (проверено 19.07.2021).
2 Конституция РФ. Доступ: http://www.kremlin.ru/acts/constitution (проверено 19.07.2021). СТРАТЕГИЯ НАЦИОНАЛЬНОЙ БЕЗОПАСНОСТИ
РОССИЙСКОЙ ФЕДЕРАЦИИ Аннотация. Статья посвящена изучению внутренней и внешней политики, направленной на укрепление
национальной безопасности Российской Федерации и обеспечение устойчивого развития страны на
долгосрочную перспективу. Ключевые слова: национализм, сепаратизм, многополярные международные отношения, националь-
ная безопасность, национальные интересы, угроза национальной безопасности, стратегические нацио-
нальные приоритеты, система обеспечения национальной безопасности, силы обеспечения националь-
ной безопасности, полицентричная модель, мобилизационная готовность, гражданская оборона Р Р
оссия преодолела последствия системного политического и социально-
экономического кризиса конца XX в. – остановила падение уровня и
качества жизни российских граждан, устояла под напором национализма,
сепаратизма и международного терроризма, предотвратила дискредитацию
конституционного строя, сохранила суверенитет и территориальную целост-
ность1. Обеспечение национальной безопасности (далее – НБ) всегда являлось
важнейшим направлением деятельности любого государства, в т.ч. в рам-
ках гибкого и соответственного учета тенденций развития данного инсти-
тута, проявляющихся во внутригосударственной и внешней сферах, что
было также закреплено в пункте м) ст. 71 Конституции РФ2. Употребление
в данном пункте указанной статьи Конституции России термина «безопас-
ность» указывает на понимание данного феномена в едином, комплексном, 8 0 2 0 21 ’ 0 4 2 0 21 ’ 0 4 ВЛ А С Т Ь «интегрированном» виде, что создает основы для учета правового опыта в
данной сфере иных государств, способствующего прогрессивному развитию
конституционного права в целом в духе современных международных тен-
денций. С учетом глобализации государство в рамках актов, посвященных
НБ, не может отражать исключительно внутренние аспекты данного явления
и ограничиваться лишь общими указаниями на приверженность традициям
поддержания мира и безопасности. Государственная политика в области национальной обороны, государст-
венной и общественной безопасности, устойчивого развития России реа-
лизуется в соответствии с внутренними и внешними условиями. Решаются
первоочередные задачи в экономической сфере, растет инвестиционная
привлекательность национальной экономики, возрождаются исконно рос-
сийские идеалы, духовность, достойное отношение к исторической памяти. Укрепляется общественное согласие на основе общих ценностей – свободы
и независимости Российского государства, гуманизма, межнационального
мира и единства культур многонационального народа РФ, уважения семейных
традиций, патриотизма. В целом сформированы предпосылки для надежного
предотвращения внутренних и внешних угроз НБ, динамичного развития и
превращения РФ в одну из лидирующих держав по уровню технического про-
гресса, качеству жизни населения, влиянию на мировые процессы. В усло-
виях глобализации процессов мирового развития, международных полити-
ческих и экономических отношений, формирующих новые угрозы и риски
для развития личности, общества и государства, Россия в качестве гаранта
благополучного национального развития переходит к новой государственной
политике в области НБ. 1 ФЗ от 28.06.2014 г. № 172-ФЗ «О стратегическом планировании в РФ». СТРАТЕГИЯ НАЦИОНАЛЬНОЙ БЕЗОПАСНОСТИ
РОССИЙСКОЙ ФЕДЕРАЦИИ Основными направлениями обеспечения НБ РФ являются стратегические наци-
ональные приоритеты, которыми определяются задачи важнейших социаль-
ных, политических и экономических преобразований для создания безопасных
условий реализации конституционных прав и свобод граждан РФ, осуществле-
ния устойчивого развития страны, сохранения территориальной целостности
и суверенитета государства. Основные понятия, используемые в Стратегии: национальная безопасность,
национальные интересы РФ, угроза национальной безопасности, стратегиче-
ские национальные приоритеты, система обеспечения НБ, силы обеспечения
национальной безопасности, средства обеспечения НБ. Россия продемонстрировала способность к обеспечению суверенитета, неза-
висимости, государственной и территориальной целостности, защиты прав
соотечественников за рубежом. Возросла роль РФ в решении важнейших
международных проблем, урегулировании военных конфликтов, обеспечении
стратегической стабильности и верховенства международного права в межго-
сударственных отношениях1. Укрепление России происходит на фоне новых угроз национальной безо-
пасности, имеющих комплексный взаимосвязанный характер. Проведение
РФ самостоятельной внешней и внутренней политики вызывает противодей-
ствие со стороны США и их союзников, стремящихся сохранить свое доми-
нирование в мировых делах. Реализуемая ими политика сдерживания России
предусматривает оказание на нее политического, экономического, военного и
информационного давления. Процесс формирования новой полицентричной модели мироустрой-
ства сопровождается ростом глобальной и региональной нестабильности. 1 ФЗ от 28.06.2014 г. № 172-ФЗ «О стратегическом планировании в РФ». 81 2021’04 2021’04 ВЛАСТЬ Наращивание силового потенциала Организации Североатлантического дого-
вора (НАТО) и наделение ее глобальными функциями, реализуемыми в нару-
шение норм международного права, активизация военной деятельности стран
блока, дальнейшее расширение Альянса, приближение его военной инфра-
структуры к российским границам создают угрозу НБ. При размещении в
Европе, Азиатско-Тихоокеанском регионе и на Ближнем Востоке компонен-
тов системы противоракетной обороны США в условиях практической реа-
лизации концепции «глобального удара», развертывания стратегических не-
ядерных систем высокоточного оружия, а также в случае размещения оружия
в космосе возможности поддержания глобальной и региональной стабильно-
сти существенно снижаются. В условиях конкурентной борьбы за ресурсы не
исключено решение возникающих проблем с применением военной силы, что
может привести к нарушению сложившегося баланса сил вблизи границ РФ и
границ ее союзников. р
РФ выступает за всемерное укрепление механизмов взаимодействия с
Европейским союзом, включая последовательное формирование общих про-
странств в сферах экономики, внешней и внутренней безопасности, обра-
зования, науки, культуры. Долгосрочным национальным интересам России
отвечает формирование в Евро-Атлантике открытой системы коллективной
безопасности на четкой договорно-правовой основе. СТРАТЕГИЯ НАЦИОНАЛЬНОЙ БЕЗОПАСНОСТИ
РОССИЙСКОЙ ФЕДЕРАЦИИ д
р
р
Для предотвращения угроз НБ необходимо обеспечить социальную стабиль-
ность, этническое и конфессиональное согласие, повысить мобилизационный
потенциал и рост национальной экономики, поднять качество работы органов
государственной власти и сформировать действенные механизмы их взаимо-
действия с гражданским обществом в целях реализации гражданами РФ права
на жизнь, безопасность, труд, жилье, здоровье и здоровый образ жизни, на
доступное образование и культурное развитие. Россия проводит открытую, рациональную и прагматичную внешнюю поли-
тику, исключающую затратную конфронтацию. В области международной бе-
зопасности Россия сохраняет приверженность использованию прежде всего
политических и правовых инструментов, механизмов дипломатии и миро-
творчества. Применение военной силы для защиты национальных интересов
возможно только в том случае, если все принятые меры ненасильственного
характера оказались неэффективными. Обеспечение национальных интересов осуществляется посредством реали-
зации таких стратегических национальных приоритетов, как оборона страны,
государственная и общественная безопасность, повышение качества жизни
российских граждан, экономический рост, наука, технологии и образование,
здравоохранение, культура, экология живых систем и рациональное приро-
допользование, стратегическая стабильность и равноправное стратегическое
партнерство. Достижение стратегических целей обороны страны осуществляется в рамках
реализации военной политики путем стратегического сдерживания и предот-
вращения военных конфликтов, совершенствования военной организации го-
сударства, форм и способов применения ВС РФ, других войск, воинских фор-
мирований и органов, повышения мобилизационной готовности РФ, готов-
ности сил и средств гражданской обороны. Главной задачей укрепления национальной обороны в среднесрочной пер-
спективе является переход к качественно новому облику ВС РФ с сохранением
потенциала стратегических ядерных сил за счет совершенствования организа-
ционно-штатной структуры и системы территориального базирования войск
и сил, наращивания числа частей постоянной готовности, а также совершен- 8 2 2 0 21 ’ 0 4 2 0 21 ’ 0 4 ВЛ А С Т Ь ствования оперативной и боевой подготовки, организации межвидового вза*
имодействия войск и сил. Стратегическими целями государственной и общественной безопасности
являются защита конституционного строя, суверенитета, государственной и
территориальной целостности РФ, основных прав и свобод человека и граж-
данина, сохранение гражданского мира, политической и социальной стабиль-
ности в обществе, защита населения и территорий от чрезвычайных ситуаций
природного и техногенного характера. Обеспечение обороны страны осуществляется на основании принципов
рациональной достаточности и эффективности, в том числе путем примене-
ния методов и средств невоенного реагирования, механизмов дипломатии и
миротворчества, расширения международного военного и военно-техниче-
ского сотрудничества, контроля над вооружением и использования других
международно-правовых инструментов. Реализация Стратегии призвана способствовать развитию национальной
экономики, улучшению качества жизни граждан, укреплению политической
стабильности в обществе, обеспечению обороны страны, государственной и
общественной безопасности, повышению конкурентоспособности и между-
народного престижа Российской Федерации. СТРАТЕГИЯ НАЦИОНАЛЬНОЙ БЕЗОПАСНОСТИ
РОССИЙСКОЙ ФЕДЕРАЦИИ В пункте 3 Стратегии НБ РФ
2015 г. содержится тезис о консолидации усилий федеральных органов го-
сударственной власти, других государственных органов, органов государст-
венной власти субъектов РФ, органов местного самоуправления (фактически
– это просто декларация намерений, и только). Да и в законе «О безопасно-
сти» о правовой основе обеспечения безопасности говорится лишь в общих
чертах, как и о функциях органов власти субъектов РФ и органов местного
самоуправления. Об
б Обзор содержания отдельных научных работ позволяет говорить и о приня-
тии в обозримом будущем федерального (федерального конституционного)
закона «О национальной безопасности РФ», который должен прийти на смену
действующему закону «О безопасности»; в нем должны быть учтены основные
положения Стратегии НБ, многочисленные авторские пожелания, предложе-
ния, рекомендации и т.п. Изложение материала о роли и значении законодательства применительно
к оптимизации обеспечительного механизма НБ закончим вот таким, пола-
гаем, правильным и заслуживающим поддержки суждением М.Ю. Зеленкова:
«Дальнейшее совершенствование законодательства должно проводиться в
неразрывном взаимодействии всех ветвей власти с научными организаци-
ями и экспертами гражданского общества в направлении создания единой
целостной правовой системы. Это позволит регулировать весь спектр право-
отношений в данной области. Чиновники и бюрократия не способны создать
единую, универсальную и внутренне согласованную нормативную правовую
основу национальной безопасности РФ. Какими бы грамотными они ни были,
на первом месте у них всегда будут корпоративный ведомственный интерес и
лоббистские устремления» [Беляев, Чапчиков 2018]. Определение понятия «национальная безопасность» в политической науке
на данный момент представляется весьма некорректным и чересчур широким. Его трудность заключается также и в том, что существующие парадигмальные
подходы к данному явлению весьма широки и происходят из сферы таких наук,
как социология, политология, экономика, военная наука. Предпринимаются
попытки определения феномена национальной безопасности также со сто-
роны исследователей, реализующих комплексное изучение данного понятия 2021’04 83 2021’04 ВЛАСТЬ с прикладной целью разрешения проблем совершенствования системы обе-
спечения НБ в России. с прикладной целью разрешения проблем совершенствования системы обе-
спечения НБ в России. Национальная безопасность РФ состоит в применении синтезирующего
подхода к характеристике отдельных компонентов данного явления, последо-
вательный анализ которых приводит к конструированию наиболее целостного,
комплексного определения явления НБ. С.В. Список литературы
А
Ф А 2020 С Список литературы Азимов Ф.А. 2020. Современные парадигмы изучения НБ РФ. – Социально-
гуманитарный вестник: всероссийский сборник научных трудов. № 26. С. 103-
111. Беляев В.П., Чапчиков С.Ю. 2018. Оптимизация механизма обеспечения
НБ: некоторые направления. – Актуальные проблемы российского права. № 9. С. 31-41. ASIMOV Furkat Abdukhamidovich, postgraduate student of the Chair of International Relations and Political Science,
Dobroljubov State Linguistics University of Nizhny Novgorod (31a Minina St, Nizhny Novgorod, Russia, 603155;
furkat72@inbox.ru) СТРАТЕГИЯ НАЦИОНАЛЬНОЙ БЕЗОПАСНОСТИ
РОССИЙСКОЙ ФЕДЕРАЦИИ Степашин так характеризует
явление НБ с позиций данного генерализованного подхода: национальная бе-
зопасность – это «такое состояние защищенности личности, а также общества
и государства от явлений внутренней и внешней угрозы, при котором сохра-
няется высокий уровень реализации конституционных прав и свобод граждан
РФ, что находит свое отражение в обеспечении достойного качества и уровня
жизни, сохранении суверенитета, независимости, государственной и террито-
риальной целостности, которые являются залогом для устойчивого социально-
экономического развития всей страны» [Азимов 2020: 103]. STRATEGY OF NATIONAL SECURITY
OF THE RUSSIAN FEDERATION Abstract. The article is devoted to the study of domestic and foreign policies aimed at strengthening the national security of
the Russian Federation and ensuring sustainable development of the country in the long term. Keywords: nationalism, separatism, multipolar international relations, national security, national interests, threat to national
security, strategic national priorities, national security system, national security forces, polycentric model, mobilization readi-
ness, civil defense
|
W3198023737.txt
|
https://repository.ubn.ru.nl//bitstream/handle/2066/237383/237383.pdf
|
de
|
Innenrücktitelbild: Origin of the α‐Effect in S<sub>N</sub>2 Reactions (Angew. Chem. 38/2021)
|
Angewandte Chemie
| 2,021
|
cc-by
| 6,877
|
PDF hosted at the Radboud Repository of the Radboud University
Nijmegen
The following full text is a publisher's version.
For additional information about this publication click this link.
https://repository.ubn.ru.nl/handle/2066/237383
Please be advised that this information was generated on 2024-07-13 and may be subject to
change.
Research Articles
Organic Chemistry
Angewandte
Chemie
How to cite: Angew. Chem. Int. Ed. 2021, 60, 20840 – 20848
International Edition:
doi.org/10.1002/anie.202106053
German Edition:
doi.org/10.1002/ange.202106053
Origin of the a-Effect in SN2 Reactions
Thomas Hansen+, Pascal Vermeeren+, F. Matthias Bickelhaupt, and Trevor A. Hamlin*
Abstract: The a-effect is a term used to explain the dramatically enhanced reactivity of a-nucleophiles (R@Y@X:@) compared to their parent normal nucleophile (R@X:@) by deviating
from the classical Brønsted-type reactivity-basicity relationship. The exact origin of this effect is, however, still heavily
under debate. In this work, we have quantum chemically
analyzed the a-effect of a set of anionic nucleophiles, including
O-, N- and S-based normal and a-nucleophiles, participating in
an SN2 reaction with ethyl chloride using relativistic density
functional theory at ZORA-OLYP/QZ4P. Our activation
strain and Kohn–Sham molecular orbital analyses identified
two criteria an a-nucleophile needs to fulfill in order to show aeffect: (i) a small HOMO lobe on the nucleophilic center,
pointing towards the substrate, to reduce the repulsive occupied–occupied orbital overlap and hence (steric) Pauli repulsion with the substrate; and (ii) a sufficiently high energy
HOMO to overcome the loss of favorable HOMO–LUMO
orbital overlap with the substrate, as a consequence of the first
criterion, by reducing the HOMO–LUMO orbital energy gap.
If one of these two criteria is not fulfilled, one can expect no aeffect or inverse a-effect.
Introduction
The a-effect is a fundamental phenomenon in organic
chemistry that refers to the dramatically enhanced reactivity
of a nucleophile featuring a lone pair-bearing heteroatom
adjacent to the nucleophilic center (i.e., the a-position).[1] In
1962, Pearson and Edwards introduced the term a-effect[1b] to
denote a downward deviation from the Brønsted-type correlation (reaction barrier versus proton affinity, see Figure 1)
found for normal nucleophiles.[2] The a-effect has been found
in a myriad of reactions in which the magnitude of this effect
is highly dependent on both the studied class of reaction and
type of a-nucleophiles.[3]
Despite extensive experimental and computational studies, the origin of the a-effect remains elusive and under
debate.[4] Based on the intrinsic properties of the a-nucleophile several theories on the origin of the a-effect have been
proposed, such as ground state destabilization, transition state
stabilization, or thermodynamic product stability.[3f–h, 5] Next
to that, the a-effect is also ascribed to external factors, like
solvent-induced effects.[2a] The ground state destabilization
theory proposes that the electron-electron repulsion between
the lone pair electrons of the nucleophilic center and the aatom result in a higher HOMO energy of the a-nucleophile,
making it more reactive. This explains the higher reactivity of
the a-nucleophile; however, this electronic mechanism should
also increase the basicity and, therefore, cannot fully explain
the a-effect. Several explanations, which fall in the class of the
transition state stabilization, have been proposed, e.g.,
secondary orbital interactions, electron transfer contribution,
tighter transition state, and polarizability of a-nucleophiles.
[*] Dr. T. Hansen,[+] P. Vermeeren,[+] Prof. Dr. F. M. Bickelhaupt,
Dr. T. A. Hamlin
Department of Theoretical Chemistry, Amsterdam Institute of
Molecular and Life Sciences (AIMMS), Amsterdam Center for
Multiscale Modeling (ACMM), Vrije Universiteit Amsterdam
De Boelelaan 1083, 1081 HV Amsterdam (The Netherlands)
E-mail: t.a.hamlin@vu.nl
Dr. T. Hansen[+]
Leiden Institute of Chemistry, Leiden University
Einsteinweg 55, 2333 CC Leiden (The Netherlands)
Prof. Dr. F. M. Bickelhaupt
Institute for Molecules and Materials, Radboud University
Heyendaalseweg 135, 6525 AJ Nijmegen (The Netherlands)
[+] These authors contributed equally to his work.
Supporting information and the ORCID identification number(s) for
the author(s) of this article can be found under:
https://doi.org/10.1002/anie.202106053.
T 2021 The Authors. Angewandte Chemie International Edition
published by Wiley-VCH GmbH. This is an open access article under
the terms of the Creative Commons Attribution License, which
permits use, distribution and reproduction in any medium, provided
the original work is properly cited.
20840
Figure 1. a) Normal and a-nucleophiles; b) overview of the Brønstedtype correlation between the reactivity and basicity (black dotted line),
in which a-nucleophiles with a-effect have a downward deviation from
this trend (teal arrow), showing greater reactivity than would be
expected based on their basicity (i.e., a-effect).
T 2021 The Authors. Angewandte Chemie International Edition published by Wiley-VCH GmbH
Angew. Chem. Int. Ed. 2021, 60, 20840 – 20848
Research Articles
The secondary orbital interactions theory argues that the
second lone-pair-bearing heteroatom of a-nucleophiles can
engage in an additional orbital interaction with the substrate
and thereby stabilizing the TS.[4m, 5b] The electron transfer
contribution model suggests that a-nucleophiles exhibit some
single electron transfer character in the SN2 transition state,
which can be stabilized by the adjacent lone-pair bearing
atom.[5c–e] Many also ascribe the a-effect to a tighter transition
state of the a-nucleophiles.[3j] Lastly, the higher polarizability
of a-nucleophiles also has been proposed as the driving force
behind the a-effect.[5f] Although these theories share the
requirements, rendering the system to be more reactive than
one would expect based on their basicity, only a limited
amount of quantitative data is available on these explanations, which ultimately hampers pinpointing the exact underlying mechanism which is responsible for the a-effect.
With the aim of providing a unified framework with which
to understand the a-effect in terms of the intrinsic properties
of the a-nucleophile, we disentangled the physical mechanisms and identified causal structure–reactivity relationships
controlling SN2 reactions involving normal and a-nucleophiles. We have explored and analyzed the potential energy
surface of Nu:@ + C2H5Cl, with Nu:@ being a set of anionic
nucleophiles including O-, N- and S-based normal and anucleophiles, by using relativistic density functional theory
(DFT) at ZORA-OLYP/QZ4P (Scheme 1). The nucleophile
(Nu:@) is, for the parent normal nucleophiles, defined by
Angewandte
Chemie
Results and Discussion
Main Trends in Reactivity
The first step was to quantify which of the a-nucleophiles
exhibit a-effect in the SN2 reaction with C2H5Cl. Figure 2
shows the Brønsted-type correlation diagram of the analyzed
SN2 reactions, in which the reaction barrier (i.e., DH*) is
plotted as a function of the basicity (i.e., DHPA). See SI
Tables S1 and S2 for all complete reaction profiles. In line
with previous studies,[4j–l] a good correlation emerges between
the reaction barrier and the basicity of the six normal
nucleophiles (R@X:@ = HO@ , CH3O@ , H2N@ , CH3HN@ ,
CH3S@ , HS@), which are indicated by black dots (Figure 2).
For the a-nucleophiles (R@Y@X:@), the introduction of the
electron-withdrawing heteroatom Y (i.e., O, HN, S) adjacent
to the nucleophilic center (i.e., X:@) leads in most cases to
a less basic and reactive nucleophile. In order words, most anucleophiles shift left-upwards, following the Brønsted-type
correlation diagram of the normal nucleophiles (Figure 1,
black arrow on the dotted correlation line). However, some of
these a-nucleophiles also deviate in a downward trend from
the Brønsted-type correlation, making them more reactive
than one would expect based on their basicity (Figure 1, teal
arrow below the dotted correlation line). These a-nucleophiles, therefore, exhibit a-effect.
By analyzing the computed reaction barriers and basicityQs, three distinct classes of a-nucleophiles (i.e., R@Y@X:@)
can be discerned from the Brønsted-type correlation diagram
(Figure 2): (i) a-nucleophiles with an apparent downward
deviation from the Brønsted-type correlation and hence
exhibiting a-effect, such as HOO@ , CH3OO@ , H2NO@ ,
CH3HNO@ , HOHN@ (teal dots); (ii) a-nucleophiles, mainly
S-, N-based, but also O-based with an adjacent sulfur atom,
showing a minor degree or no a-effect and thus behaving like
Scheme 1. Schematic overview of the computationally analyzed SN2
reactions of normal and a-nucleophiles with C2H5Cl.
R@X:@ while for a-nucleophiles by R@Y@X:@ , in which X,
Y = O, HN, S and R = H, CH3. The activation strain model
(ASM)[6] of reactivity in combination with Kohn–Sham
molecular orbital (KS-MO) theory[7a] and the matching
energy decomposition analysis (EDA)[7b,c] were employed to
provide quantitative insight into the factors that are responsible for the a-effect. This methodological approach enables
the analysis of the potential energy surface and, more
importantly, the reaction barrier, by decomposing the total
energy of the system into physically meaningful and chemically intuitive terms, and has shown to be valuable for
understanding the reactivity of, amongst others, nucleophilic
substitution reactions.[8]
Angew. Chem. Int. Ed. 2021, 60, 20840 – 20848
Figure 2. Brønsted-type correlation between the reaction barrier (i.e.,
DH*; kinetics) and the basicity (i.e., DHPA ; thermodynamics) for the
SN2 reaction of Nu:@ + C2H5Cl. The normal nucleophiles (i.e., R@Y:@)
are indicated in black. The a-nucleophiles (i.e., R@X@Y:@) have three
distinct classes: class I, exhibiting a-effect, teal dots; class II, having
a minor degree or no a-effect, grey dots; class III, showing inverse aeffect, red dots, where X, Y = O, HN, S and R = H, CH3. The linear
trend line (black dotted line; R2 = 0.93) is fitted to the normal
nucleophile data set. Computed at ZORA-OLYP/QZ4P.
T 2021 The Authors. Angewandte Chemie International Edition published by Wiley-VCH GmbH
www.angewandte.org
20841
Research Articles
their parent normal analog (grey dots); (iii) a-nucleophiles
featuring a degree of inverse a-effect, making them less
reactive than would be expected based on their basicity and
show an upward deviation from the Brønsted-type correlation, such as HSHN@ and CH3SHN@ (red dots).[9] In addition,
we have computed the Brønsted-type correlation diagrams of
the SN2 reactions between the herein studied normal and anucleophiles with substrates (i.e., electrophiles) varying in
size and leaving group: R@Y, where R = Me, Et, i-Pr; Y = F,
Cl (SI Figures S1–S3 and S9, and Tables S3 and S11). We are
able to derive the same conclusions as for ethyl chloride,
which shows that the origin of the a-effect in SN2 reactions is
independent of the studied substrate (see below). Notably, we
found, like Ren and Yamataka,4l that the a-effect becomes
more apparent as the size of the substrate increases.
Origin of Reactivity
In order to gain quantitative insight into the physical
factors behind the a-effect, we turn to the activation strain
model (ASM) of reactivity.[6] The ASM decomposes the
electronic energy (DE) into two distinct energy terms, namely,
the strain energy (DEstrain) and the interaction energy (DEint).
The strain energy results from the deformation of the
individual reactants and the interaction energy consists of
all mutual interactions between the deformed reactants along
the reaction coordinate, defined in this case as the IRC
projection onto the Ca···Cl distance.[8, 10] Figure 3 a shows the
activation strain diagrams (ASDs) of HO@ and HOO@ +
C2H5Cl, which are the most representative and well-known
models for a normal and a-nucleophile that exhibits strong aeffect. We found that all other a-nucleophiles that show aeffect are governed by the same underlying physical mechanism (see below). The ASDs of all other SN2 reactions
involving O-nucleophiles are provided in the Supporting
Information (see SI Figures S7 and S8). As already shown in
Figure 2, HO@ and HOO@ have similar reactivity (i.e.,
Angewandte
Chemie
reaction barriers), but have vastly differing basicity (DHPA =
394 and 380 kcal mol@1 for HO@ and HOO@ , respectively),
which was also experimentally found by Bierbaum and coworkers.[4a,b] Thus, HOO@ exhibits a strong a-effect by having
a significantly lower basicity while being equally reactive
compared to the corresponding parent normal nucleophile
HO@ and hence having a downward deviation from the
Brønsted-type correlation. The origin of the similarity in
reaction barriers, in terms of electronic energy (trends in DE*
and DH* are identical), can be traced back, by using the
ASM, to a nearly identical DEstrain and DEint for both
nucleophiles. Thus, despite the significantly lower basicity of
the a-nucleophile HOO@ , the interaction with the substrate
(i.e., DEint) is maintained equivalent to that of the normal
nucleophile HO@ .[9]
To understand why both nucleophiles interact with the
substrate with equal interaction energy, we employ the
canonical energy decomposition analysis (EDA).[7b,c] Our
canonical EDA decomposes the DEint between the reactants
into the following three physically meaningful energy terms:
electrostatic interactions (DVelstat), (steric) Pauli repulsion
(DEPauli), and orbital interaction (DEoi). Herein, DVelstat is the
classical electrostatic interaction between the unperturbed
charge distributions of the (deformed) reactants. The (steric)
Pauli repulsion, DEPauli, includes the destabilizing interaction
between the occupied orbitals of both fragments, due to the
PauliQs exclusion principle, and is a measure for steric
repulsion. The orbital interaction energy, DEoi, accounts for,
amongst others, charge transfer between the fragments, such
as HOMO–LUMO interactions. We found that, despite the
nearly equivalent DEint for both nucleophiles, the individual
energy terms (i.e., DVelstat and DEPauli) are vastly different, in
which the (steric) Pauli repulsion is significantly less destabilizing for the a-nucleophile compared to the parent normal
nucleophile (Figure 3 b; see SI Tables S8–S9 and S14 for the
ASM/EDA data on consistent geometries, which render
identical trends as our initial ASM/EDA data). In contrast,
the orbital interactions of the normal nucleophile are almost
Figure 3. a) Activation strain analysis; and b) energy decomposition analysis of the SN2 reactions between HO@ (black; normal nucleophile) and
HOO@ (red; a-nucleophile) + C2H5Cl, along the IRC projected on the Ca···Cl bond stretch. Computed at ZORA-OLYP/QZ4P.
20842 www.angewandte.org
T 2021 The Authors. Angewandte Chemie International Edition published by Wiley-VCH GmbH
Angew. Chem. Int. Ed. 2021, 60, 20840 – 20848
Research Articles
identical to those of the a-nucleophile, while the electrostatic
interaction is significantly more stabilizing for the normal
nucleophile. Thus, it can be concluded that the significant
reduction in destabilizing (steric) Pauli repulsion effectively
offsets the loss of stabilizing electrostatic interactions and
equips the a-nucleophiles with the a-effect. In all, the abovementioned features compose an a-nucleophile with a strong
a-effect. Later, we will discuss how this gives rise to a downward deviation from the classical Brønsted-type reactivitybasicity relationship.[9]
In order to find the origin of the less destabilizing (steric)
Pauli repulsion for the a-nucleophile HOO@ compared to its
parent normal analog HO@ , we perform a Kohn–Sham
molecular orbital analysis.[7a] The occupied orbitals of Nu:@
and C2H5Cl, for both HO@ and HOO@ , were quantified at
transition state-like, consistent geometries with the Ca···Cl
Angewandte
Chemie
bond stretch of 0.30 c (Figure 4 a). Analysis at this point on
the reaction coordinate (near all transition states), rather than
the transition state alone, ensured that the results are not
skewed by the position of the transition state (i.e., early or
late transition state).[6b] Of all possible computed occupied–
occupied orbital overlaps, the most important occupied
molecular orbitals (MOs) that dictate the trend in (steric)
Pauli repulsion, that is, the occupied orbitals responsible for
the differences in steric repulsion between the normal and anucleophile and the substrate, are the HOMONu:@ of the
(a-)nucleophile and HOMO@4C2 H5 Cl and HOMO@5C2 H5 Cl of
the substrate C2H5Cl. The HOMONu:@ is the lone-pair orbital
predominantly located on the nucleophilic center, whereas
HOMO@4C2 H5 Cl and HOMO@5C2 H5 Cl are the filled C@H and
C@C s-bonding orbitals that are delocalized over the
substrate. Note, that the overlap between HOMONu:@ and
Figure 4. a) Molecular orbital diagram of the most important occupied–occupied orbital overlaps of the SN2 reaction between HO@ (black; normal
nucleophile) and HOO@ (red; a-nucleophile) + C2H5Cl; b) molecular orbital diagram of the most important donor–acceptor interaction between
the HOMO of the nucleophile and the LUMO of C2H5Cl computed at consistent geometries with a Ca···Cl bond stretch of 0.30 b; c) representation
of the DFT HOMONu:@ (isovalue = 0.03 Bohr@3/2), where the gray horizontal line indicates the maximum spatial extent of the HOMOHO@ ; and
d) molecular electrostatic potential maps (at 0.03 Bohr@3/2) from @0.4 (red) to 0.0 (blue) Hartree e@1 and the Voronoi deformation density[11] of
the nucleophilic center at their equilibrium geometries. Computed at ZORA-OLYP/QZ4P.
Angew. Chem. Int. Ed. 2021, 60, 20840 – 20848
T 2021 The Authors. Angewandte Chemie International Edition published by Wiley-VCH GmbH
www.angewandte.org
20843
Research Articles
other filled s-orbitals on the substrate (e.g., HOMO@3 and
HOMO@6) also contribute, although less prominently, to the
observed trend in (steric) Pauli repulsion (see SI Table S10).
The higher-lying filled orbitals of the substrate, on the
contrary, are the lone-pair orbitals on the chloride atom
(i.e., HOMO and HOMO@1) or form the C@Cl bond sorbital (i.e., HOMO@2), and, therefore, have no occupied–
occupied orbital repulsion with the incoming nucleophile. The
a-nucleophile HOO@ engages in less occupied–occupied
orbital overlap (S = 0.00 and S = 0.06), and hence experiences
less repulsion, compared to the normal nucleophile HO@ (S =
0.03 and S = 0.08). The difference in (steric) Pauli repulsion
can be rationalized when comparing the spatial extent of the
HOMONu:@ of the normal and a-nucleophile (Figure 4 c). The
lobe of the filled orbital on the nucleophilic center of the anucleophile HOO@ is significantly smaller than the analogous
lobe of the normal nucleophile HO@ , due to the more
electronegative oxygen atom adjacent to the nucleophilic
center, which, in turn, polarizes orbital density away from the
nucleophilic center (see below for a detailed analysis). This
ultimately results in less repulsive overlap with the
HOMO@4C2 H5 Cl and HOMO@5C2 H5 Cl of the substrate and,
therefore, a less destabilizing (steric) Pauli repulsion compared to the parent normal nucleophile.
Despite the significantly smaller HOMONu:@ lobe of the anucleophile HOO@ , it can engage in similar orbital interactions (DEoi) with the substrate (i.e., C2H5Cl) as its parent
normal nucleophile HO@ . This initially counterintuitive
observation can be understood by the fact that, while the
smaller HOMONu:@ lobe leads to less overlap with the
LUMOC2 H5 Cl of the substrate (Figure 4 b), it is effectively
compensated by the higher-lying HOMO of the a-nucleophile
HOO@ . Now, why is it that HOO@ has a higher-lying HOMO
than HO@ ? This can be traced back to the repulsive occupied–
occupied orbital interaction between the filled 2pz atomic
orbitals (AOs) of the two neighboring oxygen atoms of
HOO@ (see below), which, in turn, pushes the HOMOHOO@ up
in energy and ultimately results in a smaller HOMONu:@ –
LUMOC2 H5 Cl orbital energy gap compared to the SN2 reaction
involving OH@ . Furthermore, the electrostatic interaction
(DVelstat) is less stabilizing for the a-nucleophile, which can be
directly related to the electron-withdrawing character of
adjacent heteroatom at the a-position that diminishes the
negative charge on the nucleophilic center (Figure 4 d; see SI
Table S13 for the VDD atomic charges of all nucleophiles).
By performing a numerical experiment where we artificially
constrained the HOO@···Ca bond length of the reaction
involving HOO@ to the bond length of HO@···Ca while
keeping the Ca···Cl bond stretch at 0.30 c (see SI Tables S3
and S4), we can conclude that the obtained interaction energy
terms shown in Figure 3 b are not skewed by the difference in
Nu:@···Ca distances.
The a-effect can also manifest in a-nucleophiles by having
a similar basicity but a lower reaction barrier than its normal
parent nucleophile, hence having a downward deviation from
the Brønsted-type correlation.[12] For example, CH3O@ and
CH3HNO@ are equally basic (DHPA = 379 kcal mol@1), but the
reaction barrier of CH3HNO@ is lower than of CH3O@
(DDH* = 2.4 kcal mol@1; see SI Table S2). This enhanced
20844 www.angewandte.org
Angewandte
Chemie
reactivity of the a-nucleophile can again be traced back to
the less destabilizing (steric) Pauli repulsion (see SI Figures S8c,d). The lobe of the filled orbital on the nucleophilic
center of the a-nucleophile H3CHNO@ is significantly smaller
than the analogous lobe of the normal nucleophile H3CO@
(see below), which, in turn, undergoes less repulsive overlap
with the filled orbitals of the substrate (i.e., C2H5Cl). This
significant reduction of destabilizing (steric) Pauli repulsion
effectively overcomes the loss of stabilizing electrostatic
interaction, making CH3HNO@ engage in a stronger interaction with the substrate than CH3O@ . Note that the reaction
involving CH3HNO@ also experiences less destabilizing strain
energy, along the entire reaction pathway, compared to the
reaction with CH3O@ . This difference in strain energy results
from the deformation of the (a-)nucleophile to accommodate
the newly formed covalent bond with the substrate. However,
a nearly identical difference in strain energy is also found
when comparing the decomposed energy terms of the basicity
(see below). This energy term is, therefore, not an important
factor for downward deviation from the Brønsted-type
correlation between reactivity and basicity.
To test our proposed general model, we studied an
additional set of a-nucleophiles, including the hypohalite
series (i.e., FO@ , ClO@ , BrO@ , and IO@). We found, like Ren
and Yamataka,4l that FO@ and ClO@ show a profound a-effect
(see SI Figure S4). The a-effect for these a-nucleophiles is
induced by the same intrinsic mechanism as discussed above
(see SI Figures S14, S19, and Table S12), which reinforces the
generality of our findings. Moreover, in line with the
experimental work of Bierbaum and co-workers, we found
that BrO@ does not exhibit a-effect.[4c] Taken altogether, the
criteria that must be satisfied for an a-nucleophile to exhibit
strong a-effect are as follows: (i) a HOMONu:@ that has a small
orbital lobe on the nucleophilic center to reduce (steric) Pauli
repulsion with the substrate; and (ii) a high energy
HOMONu:@ that results in a small HOMONu:@ –LUMOsubstrate
orbital energy gap to overcome the diminished favorable
HOMONu:@ –LUMOsubstrate orbital overlap. If one of these two
criteria is not met, one can expect that the corresponding anucleophile exhibits no a-effect or even inverse a-effect.
Thus, one only needs to analyze the electronic structure of an
a-nucleophile to determine if it exhibits a-effect.
To showcase the impact of the HOMONu:@ of the anucleophile on its degree of a-effect, we depict the
HOMONu:@ of all studied nucleophiles in Figure 5. The anucleophiles that exhibit large degrees of a-effect (i.e.,
HOO@ , H2NO@ , CH3OO@ , CH3HNO@ , HOHN@) all share
the required characteristics, namely, the HOMONu:@ has
a smaller lobe on the nucleophilic center and is higher in
energy compared to their parent normal nucleophile. Some anucleophiles only meet one of the criteria and, therefore, have
less or no a-effect. For example, HSO@ and CH3SO@ feature
a significantly smaller HOMONu:@ lobe at the nucleophilic
center, however, the HOMONu:@ is not sufficiently high in
energy to compensate for the considerable loss in favorable
HOMONu:@ –LUMOC2 H5 Cl orbital overlap, rendering them
significantly less reactive and basic, yielding in almost no aeffect for both a-nucleophiles. Many of the studied anucleophiles, such as most S-based a-nucleophiles, do not
T 2021 The Authors. Angewandte Chemie International Edition published by Wiley-VCH GmbH
Angew. Chem. Int. Ed. 2021, 60, 20840 – 20848
Angewandte
Research Articles
Figure 5. Key occupied orbitals (HOMONu:@ ; energies in eV; isovalue = 0.030 Bohr@3/2 for O- and
N-nucleophiles and isovalue = 0.035 Bohr@3/2 for S-nucleophiles) computed at equilibrium geometries. Computed at ZORA-OLYP/QZ4P. [a] HOMO@1 is the key occupied orbital.
Angew. Chem. Int. Ed. 2021, 60, 20840 – 20848
Chemie
even show any of the a-effect characteristics,
having
similar-sized
HOMONu:@ lobes as their parent
normal nucleophile. Moreover, the
reason for inverse a-effect (e.g.,
HSHN@ and CH3SHN@) can also be
directly related to the HOMONu:@ of
the a-nucleophiles. This class of anucleophiles have a relatively large
HOMONu:@ lobe at the nucleophilic
center compared to normal nucleophiles with a similar basicity (see SI
Figures S15), while also having a lower-lying HOMO, which results in
more destabilizing (steric) Pauli repulsion and a large HOMO–LUMO
orbital energy gap compared to their
parent normal analog.
At last, we aim to fundamentally
understand how different heteroatoms adjacent to the nucleophilic
center of an a-nucleophile yield
different degrees of a-effect. In other
words, why does HOO@ exhibit aeffect by fulfilling the above-mentioned requirements (i.e., small orbital lobe on the nucleophilic center
and high energy HOMO), while, for
example, HSO@ does not possess this
significantly enhanced reactivity
(Figure 2). As previously mentioned,
introducing an electronegative heteroatom adjacent to the nucleophilic
oxygen center polarizes the orbital
density away from the nucleophilic
oxygen center, which, ultimately,
makes the filled orbital lobe of anucleophile HOO@ on the nucleophilic oxygen center smaller than the
analogous lobe on the normal nucleophile HO@ . Figure 6 shows a schematic representation of the construction of the HOMOHOO@ (left side,
red) and HOMOHSO@ (right side,
green) from the interaction between
the filled 2p atomic orbital (AO) of
the nucleophilic oxygen center (2pO)
and the filled np and empty 3d AOs
of the adjacent oxygen and sulfur
atom (npX and 3dX, where X = O, S).
Note that a detailed analysis of the
construction of the HOMOHOO@ and
HOMOHSO@ , including the orbital
energies overlaps, and populations
can be found in Supplementary Information Figure S16. The interplay
between these two orbital interactions determines the stability and
shape of the HOMOHXO@ . The
T 2021 The Authors. Angewandte Chemie International Edition published by Wiley-VCH GmbH
www.angewandte.org
20845
Research Articles
Figure 6. Schematic representation of the construction of the HOMOHOO@ (left, red) and HOMOHSO@ (right, green) from the interaction
between the filled 2p atomic orbital of the nucleophilic oxygen center
(middle, black, 2pO) and the filled np and empty 3d atomic orbitals of
the adjacent oxygen and sulfur atom (npX and 3dX).
HOMOHXO@ is the antibonding combination of the two-center
four-electron orbital interaction between the 2pO of the
nucleophilic oxygen center and the npX of the adjacent
heteroatom. As a result, the larger the 2pO-npX orbital
overlap, the more the HOMOHXO@ becomes destabilized
(i.e., higher-lying in energy). The 3pS of HSC overlaps to
a larger extent with the relatively diffuse anionic oxygen
center than the 2pO of HOC, because they have a better match
in diffuseness of atomic orbitals (see SI Figure S17). Thus, one
would suggest that the HOMOHSO@ will be destabilized to
a larger extent than the HOMOHOO@ . This is, however, not the
case, because the HOMOHSO@ becomes stabilized by a favorable donor–acceptor interaction between the filled 2pO of the
nucleophilic oxygen center and the low-lying empty 3dS AO
of the adjacent sulfur atom. The formation of HOMOHOO@ , on
the other hand, does not benefit from a stabilizing donor–
acceptor interaction. This, ultimately, makes the HOMOHSO@
a lower-lying orbital than the HOMOHOO@ (HOMOHSO@ =
2.5 eV; HOMOHOO@ = 3.3 eV), which, therefore, engages in
a weaker orbital interaction with the substrate along the SN2
pathway (see SI Table S8).
The presence, or absence, of the stabilizing orbital
interaction not only plays a crucial role in the stability of
the HOMOHXO@ , but also determines the shape of this orbital.
As seen in Figure 5, the HOMOHSO@ has a smaller orbital lobe
on the nucleophilic center than HOMOHOO@ , which makes the
former experience both less repulsive occupied–occupied
orbital overlap (i.e., Pauli repulsion) and less stabilizing
HOMO–LUMO orbital overlap with the substrate along the
SN2 pathway (see SI Table S8). The strong 2pO–3dS donor–
acceptor orbital interaction polarizes the HOMOHSO@ orbital
density away from the nucleophilic oxygen center towards the
adjacent sulfur atom. This polarizing effect can be quantified
by computing the 2pO MO-coefficient on the nucleophilic
oxygen center contributing to the overall HOMOHXO@ orbital,
which is for HOMOHSO@ significantly smaller than for
20846 www.angewandte.org
Angewandte
Chemie
HOMOHOO@ , where the donor–acceptor interaction is nearly
absent, namely, 0.79 and 0.97, respectively. We have validated
the role of the stabilizing 2pO–3dX donor–acceptor orbital
interaction on the stability and shape of the HOMOHXO@ by
performing an additional bonding analysis where the empty
acceptor orbitals on the HX radical fragments (3dO and 3dS)
are artificially removed (see SI Figure S18). As expected, in
absence of the empty 3dX AOs, and hence without the
stabilizing 2pO–3dX donor–acceptor orbital interaction, the
HOMOHSO@ is higher energy than the HOMOHOO@
(HOMOHSO@ = 3.7 eV; HOMOHOO@ = 3.3 eV), due to the
priory discussed larger repulsive orbital overlap. In addition,
the shapes of both orbitals are nearly identical, due to the lack
of polarizing effect induced by the empty 3dS AO. These
results confirm the importance of the 2pO–3dX donor–
acceptor orbital interaction on both the stability and shape
of the HOMOHXO@ .
We find that bulk solvation, in general, stabilizes the
HOMO of the (a-)nucleophile by decreasing the electrondonating capabilities, rendering both a significantly less
reactive and basic (a-)nucleophile (see Figures S21, S22,
and Tables S16, S17). These results are in line with Bierbaum
and co-workers, which found that coordination of a single
water molecule to the (a-)nucleophile resulted in a higher
reaction barrier and lower proton affinity compared to bare
(a-)nucleophiles.[4e] Interestingly, they established that the
reaction efficiency decreases faster as a function of proton
affinity for monosolvated nucleophiles than their unsolvated
counterparts. Our Brønsted-type correlation diagrams in bulk
solution (both in dichloromethane and water) reveal that the
large degree of a-effect for the strong a-nucleophiles (i.e.,
HOO@ , H2NO@), is maintained. We found that, in line with
the gas-phase results, the HOMO of the a-effect exhibiting anucleophiles (i) has a smaller lobe on the nucleophilic center
and (ii) is higher-lying in energy than the analogous HOMO
of the normal nucleophile (see SI Figures S23 and S24). These
findings indicate that the intrinsic properties of the anucleophile may also contribute to the a-effect in bulk
solvation.
Origin of Basicity
Lastly, we wish to establish why this (steric) Pauli
repulsion reduction mechanism is not manifested in the
corresponding basicity (i.e., DHPA ; proton affinity), which
ultimately leads to the observed downward deviation from
the reactivity-basicity correlation and hence the a-effect.
Table 1 shows the activation strain and energy decomposition
analyses for the interaction of HO@ and HOO@ with H+
forming HO@H and HOO@H, respectively (see SI Table S15
for the EDA data of all nucleophiles). In analogy with our
previous analysis of the reactivity (Figure 3 and SI Table S8),
the DVelstat is less stabilizing for the a-nucleophile, while the
DEoi is nearly similar for both nucleophiles. In contrast with
the SN2 reaction, there is no contribution of the (steric) Pauli
repulsion (DEPauli) in the proton affinity, because H+ does not
have any electrons which can engage in a repulsive occupied–
occupied orbital interaction with the nucleophile. This lack of
T 2021 The Authors. Angewandte Chemie International Edition published by Wiley-VCH GmbH
Angew. Chem. Int. Ed. 2021, 60, 20840 – 20848
Angewandte
Research Articles
Chemie
Strikingly, the significantly smaller HOMO lobe of
the a-nucleophile can still engage in a similar orbital
interaction as its parent normal nucleophile. This
can be traced back to the significantly higher-lying
HOMO of the a-nucleophiles, which results in
DHPA
DE
DEstrain DEint
DVelstat DEPauli
DEoi
r (O@H)
a smaller HOMOnucleophile–LUMOsubstrate orbital enHO@H 394.2 @401.5
0.0
@401.5 @224.4
0.0
@177.1
0.964
ergy gap and hence compensates for the loss in
HOO@H 379.8 @387.4
1.0
@388.4 @211.9
0.0
@176.5
0.970
overlap. In all, a-nucleophiles need to fulfill the
[a] Analyses at equilibrium geometries. Computed at ZORA-OLYP/QZ4P.
following two requirements to show strong a-effect:
1. The HOMO of the a-nucleophile should have
destabilizing (steric) Pauli repulsion in the proton affinity,
a small orbital lobe on the nucleophilic center to reduce
gives rise to the deviation from the classical Brønsted-type
(steric) Pauli repulsion with the substrate.
correlation between reactivity and basicity, because a smaller
2. The HOMO of the a-nucleophile should be sufficiently
high in energy (relative to the normal nucleophile) to engage
HOMONu:@ lobe of a-nucleophiles has a more significant
in a strong orbital interaction with the substrate, by compenimpact on the reactivity of the nucleophile than on the
basicity. In other words, the basicity of (a-)nucleophiles is
sating for the reduced favorable HOMO–LUMO overlap.
determined by the electrostatic and orbital interactions
If one of these two criteria are not fulfilled, one can expect
between the nucleophile and H+. The reactivity, on the other
no a-effect or inverse a-effect. Therefore, one only needs to
hand, is not only controlled by the prior mentioned stabilizing
analyze the electronic structure of an a-nucleophile to
energy terms but also the destabilizing (steric) Pauli repuldetermine if it will exhibit a-effect based on its intrinsic
properties.
sion. This renders a-nucleophiles with a-effect to be more
This (steric) Pauli repulsion reduction mechanism has,
reactive, based on their basicity, because the electrostatic
however, no effect on the basicity (i.e., proton affinity) of the
interactions are as always significantly less stabilizing for the
a-nucleophiles, while this is compensated in the reactivity by
a-nucleophiles since H+ has no electrons and, therefore,
the less destabilizing (steric) Paul repulsion. To compensate
cannot engage in a repulsive occupied–occupied orbital
for this intrinsic deviation, one could use the carbon basicity
interaction. This ultimately leads to the deviation from the
(e.g., ethyl cation affinity, DHEtA), instead of the proton
classical Brønsted-type correlation between reactivity and
basicity, introducing also (steric) Pauli repulsion in the
basicity. We believe that this is a finding which has an impact
basicity term.[13] Even though this significantly reduces the
on many Brønsted-type correlations, in which one can expect
downward deviation of a-nucleophiles compared to the
deviation as a result of the polarization of filled orbitals,
classical Brønsted-type correlation, the a-effect is still present
leading to less (steric) Pauli repulsion.
for the strong a-nucleophiles H2NO@ and HOO@ (see SI
Figure S20).
Table 1: Activation strain and energy decomposition analyses (in kcal mol@1) and
O@H distance (in b) for the interaction between the Nu:@ and H+ in Nu@H, where
Nu:@ = HO@ and HOO@ .[a]
Acknowledgements
Conclusion
The present computational study on a series of O-, N- and
S-based normal and a-nucleophiles participating in an SN2
reaction with ethyl chloride identifies three distinct groups of
a-nucleophiles: (i) a-nucleophiles with a downward deviation
from the classical Brønsted-type reactivity-basicity correlation and hence exhibiting a-effect, i.e., acceleration (e.g.,
HOO@ , H2NO@ , CH3OO@ , CH3HNO@ , HOHN@); (ii) anucleophiles, primarily S- and N-based, showing a minor or
no degree of a-effect and thus behaving like their parent
normal counterpart; (iii) a-nucleophiles showing a degree of
inverse a-effect, rendering them less reactive than their
parent normal analog.
Our activation strain and Kohn–Sham molecular orbital
analyses elucidate the underlying electronic mechanism
behind the a-effect. In contrast to the current rationales, we
found that a-nucleophiles exhibiting a-effect are more
reactive than their normal analogs due to less repulsive
occupied–occupied orbital overlap between the nucleophile
and substrate. The adjacent electronegative atom of anucleophiles can polarize orbital density away from the
nucleophilic center, resulting in a smaller HOMO lobe, and
thus less (steric) Pauli repulsion between the reactants.
Angew. Chem. Int. Ed. 2021, 60, 20840 – 20848
This work was supported by the Netherlands Organization for
Scientific Research (NWO) and the Dutch Astrochemistry
Network (DAN) for financial support. We thank Pieter C. M.
Laan for exploring the viability of this project.
Conflict of Interest
The authors declare no conflict of interest.
Keywords: activation strain model · basicity ·
density functional calculations · nucleophilicity · a-effect
[1] a) W. P. Jencks, J. J. Carriuolo, J. Am. Chem. Soc. 1960, 82, 1778;
b) J. O. Edwards, R. G. Pearson, J. Am. Chem. Soc. 1962, 84, 16;
c) N. J. Fina, J. O. Edwards, Int. J. Chem. Kinet. 1973, 5, 1.
[2] a) E. Buncel, I. H. Um, Tetrahedron 2004, 60, 7801; b) S. Hoz, E.
Buncel, Isr. J. Chem. 1985, 26, 313.
[3] a) J. E. Dixon, T. C. Bruice, J. Am. Chem. Soc. 1972, 94, 2052;
b) J. E. McIsaac, Jr., L. R. Subbaraman, J. Subbaraman, H. A.
Mulhausen, E. J. Behrman, J. Org. Chem. 1972, 37, 1037; c) M. J.
Colthurst, A. J. S. S. Kanagasooriam, M. S. O. Wong, C. Contini,
A. Williams, Can. J. Chem. 1998, 76, 678; d) A. R. Fersht, Y.
Requena, J. Am. Chem. Soc. 1971, 93, 3499; e) G. M. Blackburn,
T 2021 The Authors. Angewandte Chemie International Edition published by Wiley-VCH GmbH
www.angewandte.org
20847
Angewandte
Research Articles
W. P. Jencks, J. Am. Chem. Soc. 1968, 90, 2638; f) M. Laloi-Diard,
J.-F. Verchere, P. Gosselin, R. Terrier, Tetrahedron Lett. 1984, 25,
1267; g) I. H. Um, E. J. Lee, E. Buncel, J. Org. Chem. 2001, 66,
4859; h) I. H. Um, E. Buncel, J. Org. Chem. 2000, 65, 577; i) Y.
Nomura, T. Kubozono, M. Hidaka, M. Horibe, N. Mizushima, N.
Yamamoto, T. Takahashi, M. Komiyama, Bioorg. Chem. 2004,
32, 26; j) K. R. Fountain, C. J. Felkerson, J. D. Driskell, B. D.
Lamp, J. Org. Chem. 2003, 68, 1810; k) I.-H. Um, E. J. Lee, J.-A.
Seok, K.-H. Kim, J. Org. Chem. 2005, 70, 7530; l) G. Moutiers,
E. L. Gu8vel, C. Cannes, F. Terrier, E. Buncel, Eur. J. Org. Chem.
2001, 3279.
[4] a) J. M. Garver, S. Gronert, V. M. Bierbaum, J. Am. Chem. Soc.
2011, 133, 13894; b) C. H. DePuy, E. W. Della, J. Filley, J. J.
Grabowski, V. M. Bierbaum, J. Am. Chem. Soc. 1983, 105, 2481;
c) S. M. Villano, N. Eyet, W. C. Lineberger, V. M. Bierbaum, J.
Am. Chem. Soc. 2009, 131, 8227; d) G. Klopman, K. Tsuda, J. B.
Louis, R. E. Davis, Tetrahedron 1970, 26, 4549; e) D. L. Thomsen, J. N. Reece, C. M. Nichols, S. Hammerum, V. M. Bierbaum,
J. Am. Chem. Soc. 2013, 135, 15508; f) D. L. Thomsen, J. N.
Reece, C. M. Nichols, S. Hammerum, V. M. Bierbaum, J. Phys.
Chem. A 2014, 118, 8060; g) D. L. Thomsen, C. M. Nichols, J. N.
Reece, S. Hammerum, V. M. Bierbaum, J. Am. Soc. Mass
Spectrom. 2014, 25, 159; h) W.-Y. Zhao, J. Yu, S.-J. Ren, X.-G.
Wei, F.-Z. Qiu, P.-H. Li, H. Li, Y.-P. Zhou, C.-Z. Yin, A.-P. Chen,
H. Li, L. Zhang, J. Zhu, Y. Ren, K.-C. Lau, J. Comput. Chem.
2015, 36, 844; i) N. Singh, Y. Karpichev, R. Sharma, B. Gupta,
A. K. Sahu, M. L. Satnami, K. K. Ghosh, Org. Biomol. Chem.
2015, 13, 2827; j) Y. Ren, H. Yamataka, Org. Lett. 2006, 8, 119;
k) Y. Ren, H. Yamataka, Chem. Eur. J. 2007, 13, 677; l) Y. Ren,
H. Yamataka, J. Org. Chem. 2007, 72, 5660; m) E. Juaristi, G.
dos Passos Gomes, A. O. Terent’ev, R. Notario, I. V. Alabugin, J.
Am. Chem. Soc. 2017, 139, 10799; n) A. J. Kirby, A. M.
Manfredi, B. S. Souza, M. Medeiros, J. P. Priebe, T. A. Brand¼o,
F. Nome, ARKIVOC 2009, 3, 28; o) Y. Ren, X. G. Wei, S. J. Ren,
K. C. Lau, N. B. Wong, W. K. Li, J. Comput. Chem. 2013, 34,
1997; p) E. V. Patterson, K. R. Fountain, J. Org. Chem. 2006, 71,
8121; q) A. M. McAnoy, M. R. L. Paine, S. J. Blanksby, Org.
Biomol. Chem. 2008, 6, 2316.
[5] a) E. Buncel, S. Hoz, Tetrahedron Lett. 1983, 24, 4777; b) J. F.
Liebman, R. M. Pollack, J. Org. Chem. 1973, 38, 3444; c) S. S.
Shaik, J. Am. Chem. Soc. 1981, 103, 3692; d) A. Pross, S. S. Shaik,
J. Am. Chem. Soc. 1981, 103, 3702; e) S. Hoz, J. Org. Chem. 1982,
47, 3545; f) M. M. Heaton, J. Am. Chem. Soc. 1978, 100, 2004;
g) G. Klopman, M. R. Frierson, Croat. Chem. Acta 1984, 57,
1411.
[6] a) P. Vermeeren, S. C. C. van der Lubbe, C. Fonseca Guerra,
F. M. Bickelhaupt, T. A. Hamlin, Nat. Protoc. 2020, 15, 649; b) I.
20848 www.angewandte.org
Chemie
Fern#ndez, F. M. Bickelhaupt, Chem. Soc. Rev. 2014, 43, 4953;
c) F. M. Bickelhaupt, K. N. Houk, Angew. Chem. Int. Ed. 2017,
56, 10070; Angew. Chem. 2017, 129, 10204; d) P. Vermeeren,
T. A. Hamlin, F. M. Bickelhaupt, Chem. Commun. 2021, 57,
5880 – 5896.
[7] a) R. van Meer, O. V. Gritsenko, E. J. Baerends, J. Chem. Theory
Comput. 2014, 10, 4432; b) F. M. Bickelhaupt, E. J. Baerends,
Reviews in Computational Chemistry, Vol. 15 (Eds.: K. B.
Lipkowitz, D. B. Boyd), Wiley-VCH, Weinheim, 2000, pp. 1 –
86; c) T. A. Hamlin, P. Vermeeren, C. Fonseca Guerra, F. M.
Bickelhaupt, in Complementary Bonding Analysis (Ed.: S.
Grabowsky), De Gruyter, Berlin, 2021, pp. 199 – 212.
[8] a) T. Hansen, P. Vermeeren, A. Haim, M. J. H. van Dorp, J. D. C.
Cod8e, F. M. Bickelhaupt, T. A. Hamlin, Eur. J. Org. Chem.
2020, 3822; b) P. Vermeeren, T. Hansen, P. Jansen, M. Swart,
T. A. Hamlin, F. M. Bickelhaupt, Chem. Eur. J. 2020, 26, 15538;
c) P. Vermeeren, T. Hansen, M. Grasser, D. R. Silva, T. A.
Hamlin, F. M. Bickelhaupt, J. Org. Chem. 2020, 85, 14087; d) T.
Hansen, P. Vermeeren, R. Yoshisada, D. V. Filippov, G. A.
van der Marel, J. D. C. Cod8e, T. A. Hamlin, J. Org. Chem. 2021,
86, 3565.
[9] The same general conclusions are obtained at ZORA-M06-2X/
QZ4P, ZORA-M06-2X-D3/QZ4P, ZORA-OLYP-D3(BJ)/
QZ4P, and ZORA-B3LYP-D3(BJ)/QZ4P, see SI Figures S5,
S6, S10–S13, Tables S4–S7.
[10] W.-J. van Zeist, A. H. Koers, L. P. Wolters, F. M. Bickelhaupt, J.
Chem. Theory Comput. 2008, 4, 920.
[11] C. Fonseca Guerra, J. W. Handgraaf, E. J. Baerends, F. M.
Bickelhaupt, J. Comput. Chem. 2004, 25, 189.
[12] The direct comparison between CH3O@ and HOO@ is a “popular” approach in the literature because these nucleophiles
exhibit a very similar proton affinity while displaying vastly
different SN2 reaction barriers. However, this comparison may
not be ideal because these nucleophiles belong to different
“chemical families” (i.e., R = CH3 and H). Therefore, it is
recommended to exercise caution when directly comparing
these nucleophiles (for an extensive discussion on this matter,
please see Figure S25 and accompanying text in the SI).
[13] J. Hine, R. D. Weimar, Jr., J. Am. Chem. Soc. 1965, 87, 3387.
Manuscript received: May 5, 2021
Revised manuscript received: May 27, 2021
Accepted manuscript online: June 4, 2021
Version of record online: July 26, 2021
T 2021 The Authors. Angewandte Chemie International Edition published by Wiley-VCH GmbH
Angew. Chem. Int. Ed. 2021, 60, 20840 – 20848
|
|
https://openalex.org/W2859778880
|
https://journals.plos.org/plosone/article/file?id=10.1371/journal.pone.0200419&type=printable
|
English
| null |
Canine filariasis in the Amazon: Species diversity and epidemiology of these emergent and neglected zoonoses
|
PloS one
| 2,018
|
cc-by
| 5,376
|
RESEARCH ARTICLE OPEN ACCESS Citation: de Argoˆlo EGG, Reis T, Fontes DAT,
Gonc¸alves EC, Giese EG, Melo FTdV, et al. (2018)
Canine filariasis in the Amazon: Species diversity
and epidemiology of these emergent and neglected
zoonoses. PLoS ONE 13(7): e0200419. https://doi. org/10.1371/journal.pone.0200419 Editor: Carlos E. Ambro´sio, Faculty of Animal
Sciences and Food Engineering, University of São
Paulo, BRAZIL Received: August 30, 2017
Accepted: June 26, 2018
Published: July 11, 2018
Copyright: © 2018 de Argoˆlo et al. This is an open
access article distributed under the terms of the
Creative Commons Attribution License, which
permits unrestricted use, distribution, and
reproduction in any medium, provided the original
author and source are credited. Received: August 30, 2017
Accepted: June 26, 2018
Published: July 11, 2018
Copyright: © 2018 de Argoˆlo et al. This is an open
access article distributed under the terms of the
Creative Commons Attribution License, which
permits unrestricted use, distribution, and
reproduction in any medium, provided the original
author and source are credited. Received: August 30, 2017
Accepted: June 26, 2018
Published: July 11, 2018 Copyright: © 2018 de Argoˆlo et al. This is an open
access article distributed under the terms of the
Creative Commons Attribution License, which
permits unrestricted use, distribution, and
reproduction in any medium, provided the original
author and source are credited. Canine filariasis in the Amazon: Species
diversity and epidemiology of these emergent
and neglected zoonoses Emanuelle Gabriela Gualberto de Argoˆlo1*, Thais Reis1, Diego Ari Teixeira Fontes1,
Evonnildo Costa Gonc¸alves2, Elane Guerreiro Giese3, Francisco Tiago de
Vasconcelos Melo1, Jeannie Nascimento dos Santos1, Adriano Penha Furtado1 Emanuelle Gabriela Gualberto de Argoˆlo1*, Thais Reis1, Diego Ari Teixeira Fontes1,
Evonnildo Costa Gonc¸alves2, Elane Guerreiro Giese3, Francisco Tiago de
Vasconcelos Melo1, Jeannie Nascimento dos Santos1, Adriano Penha Furtado1 1 Laborato´rio de Biologia Celular e Helmintologia, Instituto de Ciências Biolo´gicas, Universidade Federal do
Para´, Bele´m – Para´ – Brazil, 2 Laborato´rio de Tecnologia Biomolecular, Instituto de Ciências Biolo´gicas,
Universidade Federal do Para´, Bele´m – Para´ – Brazil, 3 Laborato´rio de Histologia e Embriologia Animal,
Universidade Federal Rural da Amazoˆnia, Instituto de Sau´de e Produc¸ão Animal, Bele´m – Para´ – Brazil a1111111111
a1111111111
a1111111111
a1111111111
a1111111111 a1111111111
a1111111111
a1111111111
a1111111111
a1111111111 * emanu_bielle@yahoo.com.br Data Availability Statement: All relevant data are
within the paper. Funding: This research has received several
funding sources. All authors received investments
from Fundac¸ão Amazoˆnia de Amparo a Estudos e
Pesquisas (http://www.fapespa.pa.gov.br),
Ministe´rio da Sau´de (http://portalsaude.saude.gov. br) and Conselho Nacional de Desenvolvimento
Cientı´fico e Tecnolo´gico (http://cnpq.br) from Edital
023/2009 PPSUS-Marajo´, FAPESPA/MS/CNPq in Abstract Dirofilaria immitis and Acanthocheilonema reconditum are common parasites in dogs but
have also been reported parasitizing humans. The differential diagnosis and epidemiological
evaluation of these zoonoses are important to the development of efficient public health poli-
cies and control strategies. The purpose of this study was to carry out an epidemiological
survey by using molecular methods for the specific identification of filarid parasites of
domestic dogs in the Marajo´ mesoregion, State of Para´ (PA), Brazil. A total of 418 canine
blood samples from Marajo´ mesoregion (Northern Brazil) were collected, submitted to DNA
extraction, polymerase chain reaction (PCR) with “pan filarial” primer, subsequent sequenc-
ing and sequence analysis using BLASTn software comparison with previously deposited
sequences in GenBank. After that, a phylogenetic analysis by Maximum Parsimony was
performed to aid the specific diagnosis. The obtained sequences showed the occurrence of
9 (2.15%) dogs infected with D. immitis and 30 (7.18%) by A. reconditum, with a confidence
interval of 95%, there were no cases of co-infection. We observed that male dogs were
more likely to D. immits and A. reconditum infection. However, age was not significant to
both infections. This study reports for the first time the occurrence of A. reconditum in
the northern region of Brazil and confirmed the presence of D. immitis in the Marajo´
mesoregion. Citation: de Argoˆlo EGG, Reis T, Fontes DAT,
Gonc¸alves EC, Giese EG, Melo FTdV, et al. (2018)
Canine filariasis in the Amazon: Species diversity
and epidemiology of these emergent and neglected
zoonoses. PLoS ONE 13(7): e0200419. https://doi.
org/10.1371/journal.pone.0200419 * emanu_bielle@yahoo.com.br Molecular epidemiology of canine filariasis in Marajo´ mesoregion, Brazil Acanthocheilonema reconditum, Cercoptifilaria bainae, Dipetalonema grassi, Dirofilaria immitis
and Dirofilaria repens [3]. Marcadores Epidemiolo´gicos em Sau´de no
Arquipe´lago do Marajo´ project, for sample
collection and equipment acquisition. JNS, EGG,
FTVM, APF, EGGA, TR and DATF received
investment from Coordenac¸ão de Aperfeic¸oamento
de Pessoal de Nı´vel Superior (www.capes.gov.br)
from Programa Pro´-Amazoˆnia – Biodiversidade e
Sustentabilidade (Edital 047/2012) and Programa
Nacional de Incentivo à Pesquisa em Parasitologia
Ba´sica (Edital 032/2010), for reagents acquisition. Also, JNS receives Scholarship Conselho Nacional
de Desenvolvimento Cientı´fico e Tecnolo´gico
(http://cnpq.br) (Edital Produtividade emPesquisa–
PQ/2016) and invests additional resources to
reagents. APF received investment from Pro´-
Reitoria de Pesquisa e Po´s-Graduac¸ão de
Universidade Federal do Para´ (www.propesp.ufpa. br) from Edital 02/2017-PROPESPPAPQ 2017 for
translation and revision of this paper. This work is
part of the results of the Master’s of Science thesis
of EGGA, developed in the Biology of Infectious and
Parasitic Agents Postgraduation Program, Federal
University of Para´ (Universidade Federal do Para´
-UFPA). The Para´ State Health Department
(Secretaria de Sau´de do Estado Para´ - SESPA)
gave support to this project. The funders had no
role in study design, data collection and analysis,
decision to publish, or preparation of the
manuscript. In these hosts, the adult stages of A. reconditum, C. bainae, D. grassi, and D. repens are com-
monly found in the subcutaneous and/or intramuscular tissue and are considered to be less
pathogenic [4]. On the other hand, D. immitis adults are located in the pulmonary arteries and
right ventricle of the heart, and D. immitis is considered the most pathogenic of these filarid
species [5]. The occurrence of these infections in Brazil is often related to D. immitis [6], and A. recondi-
tum reports are limited to one case with a morphological description of circulating microfilar-
iae from dogs in Minas Gerais State (Southeast) [7], epidemiological evaluations with low
frequencies in Parana´ State (South) [8] and Alagoas (Northeast) [9]. Until now, A. reconditum
has previously not been reported previously in Brazilian Amazon, specifically in the State of
Para´ (North). Zoonotic cases of D. immitis and A. reconditum occur more frequently in areas with high
prevalences of infected canine hosts [4]. However, only D. Introduction Filariasis can be caused by several onchocercid nematodes that usually have arthropods as
intermediate hosts, and can be found parasitizing different species of vertebrates, as definitive
hosts [1]. Canine filariases have been reported in several localities around the world, mainly
in tropical regions, principally due to the greater concentration and diversity of vectors [2]. Currently, in Brazil, the following filarid species have been found to parasitize canids: Ministe´rio da Sau´de (http://portalsaude.saude.gov. br) and Conselho Nacional de Desenvolvimento
Cientı´fico e Tecnolo´gico (http://cnpq.br) from Edital
023/2009 PPSUS-Marajo´, FAPESPA/MS/CNPq in 1 / 9 PLOS ONE | https://doi.org/10.1371/journal.pone.0200419
July 11, 2018 PLOS ONE | https://doi.org/10.1371/journal.pone.0200419
July 11, 2018 immitis has been previously identi-
fied parasitizing humans in Brazil [10–12], but in The Amazon Region, which is considered
endemic for canine filariasis and has climatic characteristics that favor the transmission of the
parasite, only one human case has been reported [13]. Due to their higher sensitivity and specificity, molecular diagnostic methods, such as nucle-
otide sequence comparison, are important tools when compared with traditional methods
such as blood smear analysis, immunoenzymatic and histochemical tests [14]. The specific diagnosis of these infections in domestic dogs in the Amazon is important
because it can be used as a basis to direct health public policies to the proper treatment and
control of this infection in both the canine and human populations [15]. Previous works reported dogs infected with D. immitis in Marajo´ mesoregion with high
prevalence in some studied areas [16–17]. Thus, the purpose of this study was to carry out
an epidemiological study using molecular tools for identification of domestic dogs infected
with filarid nematodes in the Marajo´ mesoregion (Para´ State, Brazil), expanding the analyzed
area. Competing interests: The authors have declared
that no competing interests exist. Ethics statement This study was carried out in strict accordance with the recommendations in the Guide for the
Care and Use of Laboratory Animals of the National Institutes of Health. The protocol was
approved by the “Conselho Nacional de Sau´de, Comissão Nacional de E´tica em Pesquisa, Fun-
dac¸ão Hemopa, Ministe´rio da Sau´de” (Permit Number: 0003.0.324.000–10). Molecular diagnosis All blood samples were submitted to DNA extraction using phenol-chloroform method [18]. The extracted DNA was subjected to PCR with DIDR F1 and DIDR R1 “pan filarial” primer,
to amplify 5.8S–ITS2–28S filarid rDNA regions, according to a protocol established by the
authors [14]. The amplicons were visualized after electrophoresis in 1.5% agarose gel stained
with SYBR1 Green (Invitrogen, California, USA). Canine blood samples were considered posi-
tive for filariasis when one or more bands were observed. The PCR products were cloned into vector p1GEM-T (Promega, Wisconsin, USA) which in
turn was introduced into Escherichia coli TOP 10 electrocompetent cells (Invitrogen, Califor-
nia, USA) and sequenced with ABI 3130 model automatic DNA analyzer (Applied Biosystems,
California, USA), in conjunction with a BigDye1 Terminator v3.1 Cycle Sequencing Kit (Applied
Biosystems, California, USA), according to the manufacturer’s specifications. For a specific diagnosis, the obtained sequences were edited and aligned in the BioEdit pro-
gram [19]. Compared with previously deposited ones in GenBank using BLASTn program
[20] additonally a phylogenetic inference was made with all the positive sequences using Maxi-
mum Parsimony (MP) method [21] in MEGA 7 program [22] to aid in the identification. The
percentage of identical trees in which the associated taxa were grouped (1000 replicates) is
shown next to the branches [23]. Ascaris lumbricoides (AB571301.1) and Trichuris trichiura
(KC877992.1) sequences being used as an outgroups. Published sequences from Dipetalonema
reconditum, now stated as Acanthocheilonema reconditum (AF217801.2), Dipetalonema
dracunculoides (DQ018785.1), Dirofilaria repens (AY693808.1) and Dirofilaria immitis
(EU182330.1) were also used for comparison with the sequences obtained in this work and for
the construction of the phylogenetic tree. All obtained sequences in this work were submitted to GenBank (accession numbers are
shown in S1 Appendix). Material and methods
Sample collection Several expeditions to Marajo´ mesoregion were carried out in the following municipalities:
Anaja´s (00˚59’13"S, 49˚56’24"W), Chaves (00˚09’36"S, 49˚59’16"W), Portel (01˚56009@S,
50˚49015@W), São Sebastião da Boa Vista (01˚43’04"S, 49˚32’27"W) and Soure (00˚43001@S,
48˚31022@W). The expeditions occurred according to the convenience of the health authorities
between the years of 2011 to 2014 and were performed by the team of the Cell Biology and Hel-
minthology Laboratory (Institute of Biological Sciences, Federal University of Para´). A total of
418 canine blood samples from domestic dogs living in those areas were collected [Anaja´s
(n = 70), Chaves (n = 28), Portel (n = 141), São Sebastião da Boa Vista (n = 49) and Soure
(n = 130)]. The blood was collected via cephalic vein puncture, using disposable sterile needles
and stored in tubes with Ethylenediamine tetraacetic acid (EDTA) in a freezer at -20˚C, for
molecular analysis. The blood samples were collected from dogs with respective owner’s agree-
ment, who were informed about the purpose of this research. Blood samples were collected fol-
lowing the order of houses within randomly chosen neighborhoods for sampling. In houses
containing more than one dog all animals were sampled except for puppies less than 3 months
of age. 2 / 9 PLOS ONE | https://doi.org/10.1371/journal.pone.0200419
July 11, 2018 Molecular epidemiology of canine filariasis in Marajo´ mesoregion, Brazil Epidemiological and statistical analysis After specific diagnosis, we conducted an epidemiological analysis to investigate parameters
that might influence the distribution of these infections. The prevalence of infected dogs was
estimated based on specific diagnosis performed through phylogenetic inference. During sam-
pling, information regarding age and sex of all dogs were obtained for association analysis
with canine filariasis cases throughout Pearson chi-square test (X2), using the BioEstat 5.0 [24]
software, with a significance level of 5% (p<0.05). PLOS ONE | https://doi.org/10.1371/journal.pone.0200419
July 11, 2018 Molecular epidemiology of canine filariasis in Marajo´ mesoregion, Brazil that this difference in amplicon size varies according to the filarid species. However, the differ-
ences observed after electrophoresis are discrete and insufficient for specific determination. Thus, all amplicons, were subjected to nucleotide sequencing and BLASTn analysis (S1
Appendix) and 9 (2.15%) have high identity values to D. immitis sequences (between 91% to
100%), and the other 30 (7.58%) positive samples presented high identity values to A. recondi-
tum (between 80% to 100%). Interestingly, the difference between the band patterns observed
in this work was similar to that found by Rishniw et al.[14]. 500 bp for D. immitis and 400 bp
A. reconditum. To verify the relationships between sequences and to confirm the diversity of the filarid spe-
cies in dogs from the municipalities of the Marajo´ mesoregion, a phylogenetic analysis was car-
ried out in all obtained sequences using MP method to generate a cladogram (Fig 1) with two
distinct clades: 1. Clade I: formed by a politomy between the 30 sequences of this work and D. reconditum,
now stated as A. reconditum (AF217801.2), with 73% of Bootstrap. 2. Clade II: formed by a politomy between the 9 sequences of this work and D. immitis
(EU182330.1), with 83% of Bootstrap. Furtado et al. [17] performed a filariasis diagnosis based on blood smear analysis, molecular
techniques using blood samples and they considered the sequences that formed politomies in
the cladogram as the species D. immitis. However these authors did not identify A. reconditum
in their analysis. Fig 1. Phylogeny of filarid of the Marajo´ mesoregion. The cladogram was inferred using the Maximum Parsimony
method. It contains a length of 266, the consistency index is 0.614458 and the retention index is 0.872170. The
Bootstrap is indicated next to the knots of the branches. The sequence of the obtained samples were grouped in two
clades: In clade I, 30 sequences of the municipalities Soure, Chaves, Portel, Anaja´s São Sebastião da Boa Vista formed a
politomy with A. reconditum (AF217801.2), with Bootstrap of 73%. Clade II is composed by 9 sequences of Soure and
Chaves municipalities, which form a polytomy with D. immitis (EU182330.1) withal 83% of Bootsrap. The other filarid
sequences remained in isolated branches: D. dracunculoides (DQ018785.1) as a brother group of A. reconditum
(Bootstrap 96%) and D. repens (AY693808.1) most related to D. immitis (Bootstrap 96%). A. lumbricoides (AB571301.1)
and T. Results and discussion Among the 418 canine blood samples analyzed, 39 (9.33%) were considered positive for canine
filariasis based in the presence of visible bands after PCR and electrophoresis. However, of the
positive samples, we distinguish 9 samples with bands of approximately 500 bp and another 30
samples with bands of approximately 400 bp. None of the samples presented mixed bands;
Rishniw et al. [14] also observed a similar result among their samples, and the authors suggest PLOS ONE | https://doi.org/10.1371/journal.pone.0200419
July 11, 2018 3 / 9 Molecular epidemiology of canine filariasis in Marajo´ mesoregion, Brazil This is not the first description of canine filariasis in northern Brazil. Microfilaremic dogs
have been reported in several cities, example, in Bele´m metropolitan region, State of Para´,
based on the presence of microfilariae in blood smears [25]; Salvaterra (PA), located in the
Marajo´ mesoregion, using immunoenzymatic diagnosis methods [16]; Porto Velho, State of
Rondoˆnia [26] and La´brea, State of Amazonas [27], using similar methods. However the epide-
miological data presented in these studies were associated with D. immitis infection, but no
specific method for species identification was performed. Thus, it is also possible that cases of
misidentification by morphological analysis or cross-reaction in immunoenzymatic essays
occurred in those studies. However, an epidemiological study in two municipalities of Marajo´
Archipelago (Salvaterra and São Sebastião da Boa Vista), using classical and molecular
methods for diagnosis, confirmed D. immitis parasitizing dogs in these municipalities and
highlighted the endemic character of this zoonosis in this area [17]. Our results show for the first time the infection of D. immitis in Anaja´s, São Sebastião da
Boa Vista and Portel, with a lower prevalence in Soure and Chaves, additionally, A. reconditum
was detected in all municipalities studied (Table 1). Previous studies in Marajo´ mesoregion, using molecular techniques, PCR and sequencing,
have identified only D. immitis parasitizing dogs, and the prevalence described in this area dif-
fers from that found in this study, since higher percentages in Salvaterra (37.4%) and Sebastião
da Boa Vista (6.67%) [17]. The dog gender showed statistically significant association for D. immitis and A. reconditum
infection in Marajo´ mesoregion, with males having a higher probability of acquiring the infec-
tion. On the other hand, the overall age was not significant in both infections. Specifically, in
Soure municipality, only age (3 or more years old) was significant for infection by D. immitis. In contrast, in Chaves, just males showed to be more susceptible to the infection. The same
was observed in Portel, where males were more susceptible to A. reconditum infection. This findings agree with studies performed by Furtado et al. [17] in the same area but differ
from what has been reported for other regions [28–29]. Several authors have searched for asso-
ciations between host gender or age and parasitism by filarids [30–31], and their results have Table 1. https://doi.org/10.1371/journal.pone.0200419.t001 trichiura (KC877992.1) are the outgroup. https://doi org/10 1371/journal pone 0200419 g001 Fig 1. Phylogeny of filarid of the Marajo´ mesoregion. The cladogram was inferred using the Maximum Parsimony
method. It contains a length of 266, the consistency index is 0.614458 and the retention index is 0.872170. The
Bootstrap is indicated next to the knots of the branches. The sequence of the obtained samples were grouped in two
clades: In clade I, 30 sequences of the municipalities Soure, Chaves, Portel, Anaja´s São Sebastião da Boa Vista formed a
politomy with A. reconditum (AF217801.2), with Bootstrap of 73%. Clade II is composed by 9 sequences of Soure and
Chaves municipalities, which form a polytomy with D. immitis (EU182330.1) withal 83% of Bootsrap. The other filarid
sequences remained in isolated branches: D. dracunculoides (DQ018785.1) as a brother group of A. reconditum
(Bootstrap 96%) and D. repens (AY693808.1) most related to D. immitis (Bootstrap 96%). A. lumbricoides (AB571301.1)
and T. trichiura (KC877992.1) are the outgroup. 4 / 9 PLOS ONE | https://doi.org/10.1371/journal.pone.0200419
July 11, 2018 Molecular epidemiology of canine filariasis in Marajo´ mesoregion, Brazil been shown inconsistent, and no explanations were proposed by those associations. Similar to
the results observed in the present work, we did not observe environmental or behavioral char-
acteristics that justified the differences found in gender or age, so we infer that the cause of
some associations found herein may be related to the sample number. However the differences observed here and the factors associated with heartworm transmis-
sion in dogs need to be studied in the Amazon Region, since a complex system of interaction
between domestic dogs and wild animals can provide differentiated scenarios in the dispersion
of infection, as previously described [17]. The data presented in this paper describe for the first time in the Amazon Region the pres-
ence of dogs infected with A. reconditum (phylogenetic analysis). And this infection seems to
be related to gender, but not to age, when analyzing the studied municipalities. None of the
work reported the presence of this species of filarid in the Amazon, for this reason we cannot
compare our results. Previous studies conducted in Brazil about A. reconditum only described the occurrence,
without presenting consistent epidemiological data on this infection. Unfortunately, it was not
possible to evaluate if this infection is recent or has been neglected in previous studies in the
Amazon, since previous work have only used morphological and immunoenzymatic tech-
niques which may lead to an erroneous species identification. Furthermore, the distribution
patterns of D. immitis also need to be minutely studied in the Amazon, because, the correct
diagnosis and dynamic of infections in the dog population will be important to the develop-
ment of efficient public health policies, diagnosis and control strategies of zoonotic cases. Adult stages of D. immitis and A. reconditum are morphologically different and occupy dis-
tinct niches in the definitive host [5]. However, circulatory larval forms (microfilariae) may be
confused when observed by light microscopy in routine laboratory diagnosis, even though mor-
phological and biochemical differences have been described among them [32–34]. Although the
advantages of the molecular diagnosis of canine filariasis have already been highlighted in other
regions around the world [15, 17, 35–36], morphological differentiation is still necessary for tax-
onomic identification and the detection of A. reconditum in the Amazon Region opens up new
possibilities for research on morphological and molecular differentiation of this filarid species. Besides the diversity of canine filarid species in the extensive geographic area of the Ama-
zon, other factors may have influence on the dynamic of those infections in the Amazon, and
should be considered, as for example: (1) the high prevalence observed in previous studies;
(2) the close association between humans and domestic animals cohabiting both urban and
wild environments; and (3) the neglect of these infections by the public health authorities. In the same manner that sequences of A. reconditum was described for the first time in this
study, other species not yet recorded, or new species, may be detected in future studies in this
area. Supporting information S1 Appendix. Tableau with listing of PCR positive samples and their GenBank accession
numbers, as well as identities according to BLASTn and accession numbers of the corre-
sponding GenBank sequencesb. (DOCX) Distribution of filarid infection in domestic dogs by gender and age in domestic dogs from Marajo´ mesorregion and municipalities: Anaja´s, Chaves São
Sebastião da Boa Vista, Portel and Soure. D. immitis
A. reconditum
Municipality
Positive/tested (%)
p
Positive/tested (%)
p
Age
Gender
Age
Sex
Anaja´s
0/70 (0.0%)
-
-
2/70 (2.86%)
0.6201
0.9966
Chaves
3/28 (10.71%)
0.9248
0.0252
5/28 (17.86%)
0.4275
0.9602
S. S. Boa Vista
0/49 (0.0%)
-
-
7/49 (14.00%)
0.4349
0.7756
Soure
6/130 (4.62%)
0.0256
0.0675
4/130 (3.01%)
0.1960
0.9034
Portel
0/141 (0.0%)
-
-
14/141 (9.93%)
0.5784
0.0182
Gender
Male
9/418 (2.15%)
0.0166
23/418 (5.45%)
0.0254
Female
0/418 (0.00%)
9/418 (2.13%)
Age
0–1 y
0/418 (0.0%)
0.0737
5/418 (1.18%)
0.2246
>1–3 y
3/418 (0.71%)
13/418 (3.08%)
>3 y
6/418 (1.44%)
14/418 (3.32%)
Overall
9/418 (2.15%)
30/418 (7.58%)
() significance level of 5% (p<0.05) Table 1. Distribution of filarid infection in domestic dogs by gender and age in domestic dogs from Marajo´ mesorregion and municipalities: Anaja´s, Chaves São
Sebastião da Boa Vista, Portel and Soure. stribution of filarid infection in domestic dogs by gender and age in domestic dogs from Marajo´ mesorregion and munic
a Boa Vista, Portel and Soure. domestic dogs by gender and age in domestic dogs from Marajo´ mesorregion and municipalities: Anaja´s, Chaves São PLOS ONE | https://doi.org/10.1371/journal.pone.0200419
July 11, 2018 5 / 9 Acknowledgments We are grateful to Professors Dr. Antoˆnio Carlos Rosa´rio Vallinoto of the Laboratory of Virol-
ogy, Biological Sciences Institute (Instituto de Ciências Biolo´gicas—ICB/UFPA), and Dr. 6 / 9 PLOS ONE | https://doi.org/10.1371/journal.pone.0200419
July 11, 2018 Molecular epidemiology of canine filariasis in Marajo´ mesoregion, Brazil Aˆndrea Kelly Campos Ribeiro dos Santos of the Laboratory of Human and Medical Genetics—
ICB/UFPA for their technical support with regard to the molecular analyses and nucleotide
sequencing. We would like to thank Jurupytan Vianna da Silva, Lilian Cristina Macedo and
Fabrisia D’encarnac¸ão for their help in collecting samples. We are grateful to the municipalities of Soure, Salvaterra, São Sebastião da Boa Vista, Ana-
ja´s, Portel and Chaves for the assistance made available at the time of collection and the Para´
State Health Department (Secretaria de Sau´de do Estado Para´—SESPA) for the support given
to this project. Author Contributions Conceptualization: Evonnildo Costa Gonc¸alves, Elane Guerreiro Giese, Francisco Tiago de
Vasconcelos Melo, Jeannie Nascimento dos Santos, Adriano Penha Furtado. Data curation: Emanuelle Gabriela Gualberto de Argoˆlo, Thais Reis, Adriano Penha Furtado. Formal analysis: Emanuelle Gabriela Gualberto de Argoˆlo, Elane Guerreiro Giese, Francisco
Tiago de Vasconcelos Melo, Jeannie Nascimento dos Santos. Data curation: Emanuelle Gabriela Gualberto de Argoˆlo, Thais Reis, Adriano Penha Furtado. Data curation: Emanuelle Gabriela Gualberto de Argoˆlo, Thais Reis, Adriano Penha Furtado. Formal analysis: Emanuelle Gabriela Gualberto de Argoˆlo, Elane Guerreiro Giese, Francisco
Tiago de Vasconcelos Melo, Jeannie Nascimento dos Santos. Formal analysis: Emanuelle Gabriela Gualberto de Argoˆlo, Elane Guerreiro Giese, Francisco
Tiago de Vasconcelos Melo, Jeannie Nascimento dos Santos. Funding acquisition: Jeannie Nascimento dos Santos, Adriano Penha Furtado. Investigation: Emanuelle Gabriela Gualberto de Argoˆlo, Thais Reis, Diego Ari Teixeira
Fontes, Adriano Penha Furtado. Methodology: Emanuelle Gabriela Gualberto de Argoˆlo, Thais Reis, Diego Ari Teixeira
Fontes, Evonnildo Costa Gonc¸alves, Francisco Tiago de Vasconcelos Melo, Adriano Penha
Furtado. Supervision: Evonnildo Costa Gonc¸alves. Supervision: Evonnildo Costa Gonc¸alves. Writing – original draft: Emanuelle Gabriela Gualberto de Argoˆlo, Evonnildo Costa Gonc¸al-
ves, Elane Guerreiro Giese, Jeannie Nascimento dos Santos, Adriano Penha Furtado. Writing – review & editing: Emanuelle Gabriela Gualberto de Argoˆlo, Francisco Tiago de
Vasconcelos Melo, Jeannie Nascimento dos Santos, Adriano Penha Furtado. PLOS ONE | https://doi.org/10.1371/journal.pone.0200419
July 11, 2018 References 1. Anderson R. Nematode parasites of vertebrates their development and transmission. 2nd ed. CABI
Publishing; 2000; pp. 482–496. 2. Simo´n F, Siles-Lucas M, Morcho´n R, Gonza´lez-Miguel J, Mellado I, Carreto´n E, et al. Human and ani-
mal dirofilariasis: the emergence of a zoonotic mosaic. Clin Microbiol Rev. 2012; 25: 507–544. https://
doi.org/10.1128/CMR.00012-12 PMID: 22763636 3. Vicente JJ, Rodrigues HDO, Gomes DC, Pinto RM. Nemato´ides do Brasil. Parte V: Nemato´ides de
Mamı´feros. Rev Bras Zool. 1997; 14: 1–452. https://doi.org/10.1590/S0101-81751997000500001 4. Megat Abd Rani PA, Irwin PJ, Gatne M, Coleman GT, McInnes LM, Traub RJ. A survey of canine filarial
diseases of veterinary and public health significance in India. Parasit Vectors. 2010; 3: 30. https://doi. org/10.1186/1756-3305-3-30 PMID: 20377864 5. Irwin PJ, Jefferies R. Arthropod-transmitted diseases of companion animals in Southeast Asia. Trends
Parasitol. 2004; 20: 27–34. https://doi.org/10.1016/j.pt.2003.11.004 PMID: 14700587 6. Labarthe N, Guerrero J. Epidemiology of heartworm: what is happening in South America and Mexico? Vet Parasitol. 2005; 133: 149–156. https://doi.org/10.1016/j.vetpar.2005.04.006 PMID: 16198820 7. Gomes LR, Rodrigues RD, de Souza RR, Rodrigues GG, Mundim AV, Barbosa FC. Morphological
identification of Acanthocheilonema reconditum in a dog in the municipality of Uberlaˆndia-MG: case
report. Veterina´ria Notı´cias. 2014; 18: 126–130. 8. Leite LC, Luz E, Cirio SM, Leite MC, Molinari HP, Zadorosnei ACB, Oliveira TMFS, Leite SC. Infection
in domestic dogs (Canis lupus familiaris, Linnaeus, 1758) from Guaratuba, Parana´, Brazil by circulating 7 / 9 PLOS ONE | https://doi.org/10.1371/journal.pone.0200419
July 11, 2018 Molecular epidemiology of canine filariasis in Marajo´ mesoregion, Brazil microfilariae of Acanthocheilonema reconditum (Grassi, 1889). Semin: Ciênc. Agra´r. 2012; 33: 1149–
1156. https://doi.org/10.5433/1679-0359.2012v33n3p1149 microfilariae of Acanthocheilonema reconditum (Grassi, 1889). Semin: Ciênc. Agra´r. 2012; 33: 1149–
1156. https://doi.org/10.5433/1679-0359.2012v33n3p1149 9. Brito AC, Vila-Nova MC, Martins Rocha DA, Gomes Costa L, Pinheiro de Almeida WA, da Silva Viana
L. Prevalence of canine filariasis by Dirofilaria immitis and Dipetalonema reconditum in Maceio´, Alagoas
State, Brazil. Cad Saude Publica. 2001; 17: 1497–1504. 10. Milanez de Campos JR, Barbas CS, Filomeno LT, Fernandez A, Minamoto H, Filho JV. Human pulmo-
nary dirofilariasis: analysis of 24 cases from São Paulo, Brazil. Chest. 1997; 112: 729–733. https://doi. org/10.1378/chest.112.3.729 PMID: 9315807 11. Cavallazzi RS, Cavallazzi AC, Souza IV, Deus Cardoso JJ. Dirofilariose pulmonar humana: relato de
sete casos. J Pneumologia. 2002; 28: 100–102. https://doi.org/10.1590/S0102-35862002000200007 12. Rodrigues-Silva R, Guerra RJ, de Almeida FB, Machado-Silva JR, de Paiva DD. Human pulmonary dir-
ofilariasis at Rio de Janeiro, Brazil: a case report. References Rev Soc Bras Med Trop. 2004; 37: 56–59. https://doi. org/10.1590/S0037-86822004000100015 PMID: 15042186 13. Otranto D, Diniz DG, Dantas-Torres F, Casiraghi M, de Almeida INF, de Almeida LNF, et al. Human
intraocular filariasis caused by Dirofilaria sp. nematode, Brazil. Emerg Infect Dis. 2011; 17: 863–866. https://doi.org/10.3201/eid1705.100916 PMID: 21529396 14. Rishniw M, Barr SC, Simpson KW, Frongillo MF, Franz M, Alpizar JL. Discrimination between six spe-
cies of canine microfilariae by a single polymerase chain reaction. Vet Parasitol. 2006; 135: 303–314. https://doi.org/10.1016/j.vetpar.2005.10.013 PMID: 16289566 15. Casiraghi M, Bazzocchi C, Mortarino M, Ottina E, Genchi C. A simple molecular method for discriminat-
ing common filarial nematodes of dogs (Canis familiaris). Vet Parasitol. 2006; 141: 368–372. https://doi. org/10.1016/j.vetpar.2006.06.006 PMID: 16857318 16. Garcez LM, de Souza NF, Mota EF, Dickson LA, Abreu WU, Cavalcanti Vde F, et al. Focus of canine
heartworm disease in Marajo´ Island, North of Brazil: A risk factor for human health. Rev Soc Bras Med
Trop. 2006; 39: 333–336. https://doi.org/10.1590/S0037-86822006000400003 PMID: 17119746 17. Furtado AP, Do Carmo ES, Giese EG, Vallinoto AC, Lanfredi RM, Santos JN. Detection of dog filariasis
in Marajo Island, Brazil by classical and molecular methods. Parasitol Res. 2009; 105: 1509–1515. https://doi.org/10.1007/s00436-009-1584-9 PMID: 19693539 18. Sambrook J, Russell DW. Molecular cloning: A laboratory manual, Vol. 3. New York: Cold Spring Har-
bor Laboratory Press; 2001. 19. Hall TA. BioEdit: a user-friendly biological sequence alignment editor and analysis program for Windows
95/98/NT. Nucl Acids Symp. 1999; 41: 95–98. 20. Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ. Basic local alignment search tool. J Mol Biol. 1990; 215: 403–410. https://doi.org/10.1016/S0022-2836(05)80360-2 PMID: 2231712 21. Nei M, Kumar S. Molecular Evolution and Phylogenetics, New YorK: Oxford University Press; 2000. . Nei M, Kumar S. Molecular Evolution and Phylogenetic 22. Kumar S, Stecher G, Tamura K. MEGA7: Molecular Evolutionary Genetics Analysis version 7.0 for big-
ger datasets. Mol Biol Evol. 2016; 33: 1870–1874. https://doi.org/10.1093/molbev/msw054 PMID:
27004904 23. Felsenstein J. Confidence limits on phylogenies: An approach using the bootstrap. Evolution. 1985; 39:
783–791. https://doi.org/10.1111/j.1558-5646.1985.tb00420.x PMID: 28561359 24. Ayres M, Ayres Junior M, Ayres DL, Santos AS. Bioestat 5.0 –Aplicac¸ões estatı´sticas nas a´reas das
ciências biome´dicas. Bele´m: ONG Mamiraua. 2007. 25. Souza NF, Benigno RNM, Figueiredo M, Salim SK, Silva D, Goncalves R, et al. Prevalência de Dirofi-
laria immitis em cães no municı´pio de Bele´m—PA, com base na microfilaremia. Rev Bras Parasitol Vet. 1997; 6: 83–86. 26. PLOS ONE | https://doi.org/10.1371/journal.pone.0200419
July 11, 2018 References Ogawa GM, Da Cruz EN, Cunha PN, Camargo LM. Canine heartworm disease in Porto Velho: first
record, distribution map and occurrence of positive mosquitoes. Rev Bras Parasitol Vet. 2013; 22: 559–
564. https://doi.org/10.1590/S1984-29612013000400017 PMID: 24473882 27. Soares HS, Camargo LM, Gennari SM, Labruna MB. Survey of canine tick-borne diseases in La´brea,
Brazilian Amazon: ‘accidental’ findings of Dirofilaria immitis infection. Rev Bras Parasitol Vet. 2014; 23:
473–480. https://doi.org/10.1590/S1984-29612014093 PMID: 25517525 28. Nzabanita E, Priester W, Ferver T. Distribution of canine filariasis. Calif Vet. 1982; 36: 24–28. 29. Oge H, Doğanay A, Oge S, Yildirim A. Prevalence and distribution of Dirofilaria immitis in domestic dogs
from Ankara and vicinity in Turkey. Dtsch Tierarztl Wochenschr. 2003; 110: 69–72. PMID: 12666502 30. Montoya JA, Morales M, Juste MC, Bañares A, Simon F, Genchi C. Seroprevalence of canine heart-
worm disease (Dirofilaria immitis) on Tenerife Island: an epidemiological update. Parasitol Res. 2006;
100: 103–105. https://doi.org/10.1007/s00436-006-0257-1 PMID: 16821033 8 / 9 PLOS ONE | https://doi.org/10.1371/journal.pone.0200419
July 11, 2018 Molecular epidemiology of canine filariasis in Marajo´ mesoregion, Brazil 31. Boonyapakorn C, Srikitjakarn L, Morakote N, Hoerchner F. The epidemiology of Dirofilaria immitis infec-
tion in outpatient dogs at Chiang Mai University Small Animal Hospital, Thailand. Southeast Asian J
Trop Med Public Health. 2008; 39: 33–38. PMID: 18567441 32. Anderson RC, Bain O. Keys to Genera of the Order Spirurida. In: Anderson RC, Chabaud AG, Willmott
S, editors. Keys to the Nematode Parasites of Vertebrates. England: C.A.B. Commonwealth Agricul-
tural Bureaux; 1976; pp. 59–116. 33. Song KH, Hayasaki M, Cho KW, Lee SE, Kim DH. Cross-reactivity between sera from dogs experimen-
tally infected with Dirofilaria immitis and crude extract of Toxocara canis. Korean J Parasitol. 2002; 40:
195–198. https://doi.org/10.3347/kjp.2002.40.4.195 PMID: 12509105 34. Magnis J, Lorentz S, Guardone L, Grimm F, Magi M, Naucke TJ. Morphometric analyses of canine
blood microfilariae isolated by the Knott’s test enables Dirofilaria immitis and D. repens species-specific
and Acanthocheilonema (syn. Dipetalonema) genus-specific diagnosis. Parasit Vectors. 2013; 48: 25. 35. Xie H, Bain O, Williams SA. Molecular phylogenetic studies on filarial parasites based on 5S ribosomal
spacer sequences. Parasite. 1994; 1: 141–151. https://doi.org/10.1051/parasite/1994012141 PMID:
9140481 36. Nuchprayoon S, Junpee A, Poovorawan Y, Scott AL. Detection and differentiation of filarial parasites by
universal primers and polymerase chain reaction-restriction fragment length polymorphism analysis. Am J Trop Med Hyg. 2005; 73: 895–900. PMID: 16282300 36. Nuchprayoon S, Junpee A, Poovorawan Y, Scott AL. References Detection and differentiation of filarial parasites by
universal primers and polymerase chain reaction-restriction fragment length polymorphism analysis. Am J Trop Med Hyg. 2005; 73: 895–900. PMID: 16282300 9 / 9 PLOS ONE | https://doi.org/10.1371/journal.pone.0200419
July 11, 2018
|
https://openalex.org/W2042653501
|
https://europepmc.org/articles/pmc2688136?pdf=render
|
English
| null |
The effect of diabetes mellitus on organ dysfunction with sepsis: an epidemiological study
|
Critical care
| 2,009
|
cc-by
| 5,471
|
Abstract People with DM were less likely to develop acute respiratory
failure (9% vs. 14%, p < 0.05) and more likely to develop acute
renal failure (13% vs. 7%, p < 0.05). Of people with DM and
sepsis, 27% had a respiratory source of infection compared
with 34% in people with no DM (p < 0.05). Among patients with
a pulmonary source of sepsis, 16% of those with DM and 23%
of those with no DM developed acute respiratory failure (p <
0.05); in non-pulmonary sepsis acute respiratory failure
occurred in 6% of people with DM and 10% in those with no DM
(p < 0.05). People with DM were less likely to develop acute respiratory
failure (9% vs. 14%, p < 0.05) and more likely to develop acute
renal failure (13% vs. 7%, p < 0.05). Of people with DM and
sepsis, 27% had a respiratory source of infection compared
with 34% in people with no DM (p < 0.05). Among patients with
a pulmonary source of sepsis, 16% of those with DM and 23%
of those with no DM developed acute respiratory failure (p <
0.05); in non-pulmonary sepsis acute respiratory failure
occurred in 6% of people with DM and 10% in those with no DM
(p < 0.05). Introduction Diabetes mellitus (DM) is one of the most common
chronic co-morbid medical conditions in the USA and is
frequently present in patients with sepsis. Previous studies
reported that people with DM and severe sepsis are less likely
to develop acute lung injury (ALI). We sought to determine
whether organ dysfunction differed between people with and
without DM and sepsis. Methods Using the National Hospital Discharge Survey US,
sepsis cases from 1979 to 2003 were integrated with DM
prevalence from the Centers for Disease Control and Prevention
(CDC) Diabetes Surveillance System. Conclusions In sepsis, people with diabetes are less likely to
develop acute respiratory failure, irrespective of source of
infection. Future studies should determine the relationship of
these findings to reduced risk of ALI in people with DM and
causative mechanisms. Results During the study period 930 million acute-care
hospitalisations and 14.3 million people with DM were
identified. Sepsis occurred in 12.5 million hospitalisations and
DM was present in 17% of patients with sepsis. In the
population, acute respiratory failure was the most common
organ dysfunction (13%) followed by acute renal failure (6%). p
g
y
Annette M Esper1, Marc Moss2 and Greg S Martin1 1Division of Pulmonary, Allergy and Critical Care Medicine, Department of Medicine, Emory University, 201 Dowman Drive, Atlanta, Georgia 30322
USA
2Division of Pulmonary Sciences and Critical Care Medicine, Department of Medicine, University of Colorado at Denver and Health Sciences Center,
4200 E. Ninth Avenue, Denver, Colorado 80262 USA 1Division of Pulmonary, Allergy and Critical Care Medicine, Department of Medicine, Emory University, 201 Dowman Drive, Atlanta, Georgia 30322
USA
2Division of Pulmonary Sciences and Critical Care Medicine, Department of Medicine, University of Colorado at Denver and Health Sciences Center,
4200 E. Ninth Avenue, Denver, Colorado 80262 USA Division of Pulmonary Sciences and Critical Care Medicine, Department of Medicine, University of Colorado at Denver a
200 E. Ninth Avenue, Denver, Colorado 80262 USA Corresponding author: Greg S Martin, greg.martin@emory.edu Corresponding author: Greg S Martin, greg.martin@emory.edu Received: 8 Oct 2008 Revisions requested: 30 Oct 2008 Revisions received: 19 Nov 2008 Accepted: 13 Feb 2009 Published: 13 Feb 2009
Critical Care 2009, 13:R18 (doi:10.1186/cc7717)
This article is online at: http://ccforum.com/content/13/1/R18 Critical Care 2009, 13:R18 (doi:10.1186/cc7717) This article is online at: http://ccforum.com/content/13/1/R18
© 2009 E
l li
Bi M d C
l L d This article is online at: http://ccforum.com/content/13/1/R18
© 2009 Esper et al.; licensee BioMed Central Ltd. © 2009 Esper et al.; licensee BioMed Central Ltd. p
This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0),
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. See related commentary by Yende and van der Poll, http://ccforum.com/content/13/1/117 Available online http://ccforum.com/content/13/1/R18 Available online http://ccforum.com/content/13/1/R18 Open Access
Vol 13 No 1Research
The effect of diabetes mellitus on organ dysfunction with sepsis:
an epidemiological study p
g
y
Annette M Esper1, Marc Moss2 and Greg S Martin1 Page 1 of 6
(page number not for citation purposes) ALI: acute lung injury; ARDS: acute respiratory distress syndrome; CDC: Centers for Disease Control and Prevention; CNS: central nervous system;
DM: diabetes mellitus; GU: genitourinary; ICD-9-CM: Clinical Modification of the International Classification of Diseases, 9th Revision; IL: interleukin;
NF: nuclear factor; NHDS: National Hospital Discharge Survey; RSE: relative standard error; TNF: tumour necrosis factor; TZD: thiazolidinediones. Introduction tors, including age and chronic co-morbid medical conditions,
such as chronic liver disease, HIV infection and cancer, have
been identified that predispose patients to sepsis or severe
sepsis [5-9]. The ability of chronic co-morbid medical condi-
tions, such as diabetes mellitus (DM), to influence the risk of
sepsis or sepsis-related organ dysfunction remains unclear. Sepsis is a common disease that continues to increase in inci-
dence in the USA [1]. Severe sepsis, sepsis associated with
acute organ system dysfunction, is frequently encountered in
the intensive care unit (ICU) population and is associated with
a high morbidity and mortality [2]. Of the disorders commonly
associated with acute lung injury (ALI), sepsis carries the high-
est risk of progression at about 40% [3,4]. Specific risk fac- ALI: acute lung injury; ARDS: acute respiratory distress syndrome; CDC: Centers for Disease Control and Prevention; CNS: central nervous system;
DM: diabetes mellitus; GU: genitourinary; ICD-9-CM: Clinical Modification of the International Classification of Diseases, 9th Revision; IL: interleukin;
NF: nuclear factor; NHDS: National Hospital Discharge Survey; RSE: relative standard error; TNF: tumour necrosis factor; TZD: thiazolidinediones. ALI: acute lung injury; ARDS: acute respiratory distress syndrome; CDC: Centers for Disease Control and Prevention; CNS: central nervous system;
DM: diabetes mellitus; GU: genitourinary; ICD-9-CM: Clinical Modification of the International Classification of Diseases, 9th Revision; IL: interleukin;
NF: nuclear factor; NHDS: National Hospital Discharge Survey; RSE: relative standard error; TNF: tumour necrosis factor; TZD: thiazolidinediones. Page 1 of 6
(page number not for citation purposes) Page 1 of 6
(page number not for citation purposes) Critical Care Vol 13 No 1 Esper et al. Critical Care Vol 13 No 1 Esper et al. DM is one of the most common diseases in the USA, with sta-
tistics from the Centers for Disease Control and Prevention
(CDC) reporting that in 2005 almost 21 million people in the
USA (7% of the population) have DM. The prevalence of DM
is rising [10] and from 2002 to 2005, 2.6 million individuals
have been newly diagnosed with DM. Patients with DM are at
increased risk of developing infections and are frequently part
of both epidemiological studies and clinical trials in critically ill
patients. For example, two studies have suggested that
patients with DM and septic shock are less likely to develop
ALI [11,12]. This may be due to differences in the inflammatory
response between people with and without DM. Materials and methods
Dataset The National Center for Health Statistics has conducted the
National Hospital Discharge Survey (NHDS) continuously
since 1965 [13]. Since 1979, the NHDS has conformed to
the guidelines of the Uniform Hospital Discharge Data Set for
consistency of reporting in records. The NHDS is composed
of a sample of all nonfederal acute care hospitals in the USA,
including about 500 hospitals. Discharge records from inpa-
tients are surveyed from each hospital, representing about 1%
of all hospitalisations in the USA. The database includes
patient-specific information such as age, sex, self-reported
racial category, seven diagnostic and four procedural codes
(from the Clinical Modification of the International Classifica-
tion of Diseases, 9th Revision (ICD-9-CM)), sources of pay-
ment and discharge disposition. DM prevalence from 1979 to
2003 was obtained from the CDC Diabetes Surveillance Sys-
tem which collects, analyses and disseminates data on DM
and its complications [14]. Introduction However, at
this time the aetiology for this association remains unclear, and
may represent differential risk for organ dysfunction as a
whole. In order to further understand differences between crit-
ically ill patients with and without DM, we sought to identify dif-
ferences in organ dysfunction between patients with and
without DM and sepsis. Accurate identification of populations
at risk for acute organ dysfunction is crucial to improve our
understanding of the mechanisms involved and to develop
novel therapies for these patients. 730.9, skin/soft tissue 003.24, central nervous system (CNS)
013, cardiovascular 036.45 to 036.43; DM 250.x; organ dys-
function: respiratory 96.7× to ventilator management, 518.x to
acute respiratory failure, acute respiratory distress syndrome
(ARDS), ARDS after shock or trauma, cardiovascular 458.x,
785.5, renal 39.95, 580.x, 584.x, hepatic 570, haematological
287.4, 286.9, metabolic 276.2, CNS 780.01, 780.09, 348.x. 730.9, skin/soft tissue 003.24, central nervous system (CNS)
013, cardiovascular 036.45 to 036.43; DM 250.x; organ dys-
function: respiratory 96.7× to ventilator management, 518.x to
acute respiratory failure, acute respiratory distress syndrome
(ARDS), ARDS after shock or trauma, cardiovascular 458.x,
785.5, renal 39.95, 580.x, 584.x, hepatic 570, haematological
287.4, 286.9, metabolic 276.2, CNS 780.01, 780.09, 348.x. Chronic co-morbid medical conditions were also cumulatively
quantified by an established co-morbidity index (Charlson-
Deyo score) [16-18]. Other outcome variables such as mortal-
ity, length of stay and discharge status were collected. All data
collected represent data available during hospitalisation,
therefore long-term outcome data is not available. This project
was exempt from the requirement for informed consent
according to federal regulations of human subjects protection
45 CFR § 46.101(b). The Emory Institutional Review Board
approved the study as exempt from the requirement for
consent. Identification of cases Cases of patients with sepsis were identified from discharge
records in the NHDS during the 25-year period from 1979 to
2003 that included an ICD-9-CM code for sepsis as previ-
ously validated [1,15]: 038.x (septicaemia), 790.7 (bacterae-
mia), 117.9 (disseminated fungal infection), 112.5 (systemic
candidiasis) and 112.81 (disseminated fungal endocarditis). Type of infection refers to the causative organism for sepsis;
source of infection refers to the anatomical site of infection. Type and source of infection, DM and acute organ dysfunction
were identified using ICD-9 groupings, as previously pub-
lished. Source of infection: Respiratory 010.0 to 011.9, geni-
tourinary 098.17, gastrointestinal 001.9 to 009.9, bone/joint Page 2 of 6
(page number not for citation purposes) Statistical analysis All estimates are presented according to accepted guidelines
for the accuracy of NHDS data, restricting use to absolute,
unweighted samples of more than 60 patients with relative
standard error (RSE) measures of less than 30% for data anal-
ysis. The RSE was calculated as a first-order Taylor-series
approximation, as outlined in the RSE tables of the 2000
NHDS documentation. The standard error was calculated by
multiplying the RSE by the estimated incidence or mortality
rate, and 95% confidence intervals (CI) were calculated from
these standard errors with the use of Excel software (Microsoft
Corporation, Redmond, Washington). Data for continuous var-
iables were compared by analysis of variance and data for cat-
egorical variables were compared by the chi-squared test, with
the use of SAS software (SAS 9.1 for Windows; SAS Insti-
tute, Cary, North Carolina). When stated race was missing for
a given observation (ranging from 1 to 20% for any given year),
these persons were excluded from the calculations of race
specific rates but were included in all other calculations of
rates. An a priori stratified analysis between pulmonary and
non-pulmonary sources of sepsis was conducted to differenti-
ate the risk of acute respiratory failure in patients with a pulmo-
nary source of infection. Differences were considered
significant when the 95% CIs did not overlap and/or when
two-sided p-values were less than 0.05. Available online http://ccforum.com/content/13/1/R18 Available online http://ccforum.com/content/13/1/R18 Demographics and causes of sepsis Demographic data for populations with and without DM and
sepsis are shown in Table 1. Forty-three percent (892,230) of
people with DM and sepsis were male and 57% (1,178,229)
were female (p < 0.001). Among the patients with sepsis and
DM, 64% were white, 17% black and 5% other race. The
mean number of co-morbidities was greater in people with DM
and sepsis (2.07) compared with people with sepsis and no
DM (0.88, p < 0.0001). To account for differences between people with and without
DM, the frequency and type of organ dysfunction was exam-
ined within strata of infection sources. Among patients with
respiratory source of sepsis, 16% of those with DM developed
acute respiratory failure compared with 23% in people with no
DM (p < 0.05). Among patients with a non-pulmonary source
of sepsis, those with DM were still less likely to develop acute
respiratory failure when compared with those with no DM (6%
vs. 10%, p < 0.05). People with DM and sepsis were more Figure 1 represents the sources of infections in patients with
and without DM and sepsis. A respiratory source of infection
was present in 27% of patients with DM and sepsis compared
with 34% in patients with no DM (p < 0.05). People with DM
had an increased frequency of genitourinary (GU; 28% vs. 22%), skin soft tissue (4% vs. 2%) and bone (3% vs. 2%)
infections in comparison to people with no DM (p < 0.05). Organ dysfunction from 11% (17,249 cases) in 1979 to 18% (122,824 cases) in
2003. In the overall population of patients with sepsis, 13% devel-
oped acute respiratory failure, 8% acute renal failure, 4% car-
diovascular
failure
and
6%
developed
other
organ
dysfunctions. People with DM and sepsis were more likely to
develop acute renal failure compared with people with no DM
(13% vs. 7%, p < 0.05) and are less likely to develop acute
respiratory failure (9% vs. 14%, p < 0.05). There were no dif-
ferences in the occurrence of other organ dysfunctions
between the two groups or in the overall mean number of
organ dysfunctions. Results Using the NHDS, from the years 1979 to 2003, there were
12.5 million cases of sepsis identified and DM was present in
17% (2,070,459) of the cases. Based on the CDC Diabetes
Surveillance System, the number of persons with DM in the
USA increased from 5,762,000 persons in 1980 to
14,275,000 in 2003. Among general hospitalised patients the
frequency of DM increased from 5.2% in 1979 to 13.3% in
2003. The frequency of DM among septic patients increased Page 2 of 6
(page number not for citation purposes) Available online http://ccforum.com/content/13/1/R18 Page 3 of 6
(page number not for citation purposes) Table 1 Demographic characteristics of patients with sepsis from 1979 to 2003. Patients with diabetes
Patients with no diabetes
Number of patients (n)
2,070,459
10,430,000
Age (mean, years)
68
59
Race
Caucasian
64%
67%
Black
17%
14%
Other
5%
4%
Unknown
14%
15%
Gender
Male (%)
43%
49%
Female (%)
57%
51%
Pathogens (%)
Gram negatives
25%
22%
Gram positives
29%
28%
Anaerobic
1%
1%
Fungal
2.4%
2.2%
Unknown
43%
49%
Number of co-morbidities (mean)
2.07
0.88
Number of dysfunctional organ systems (mean)
0.36
0.41
Mean hospital length of stay (days)
13
14
Case fatality (%)
18.5%
20.6% Demographic characteristics of patients with sepsis from 1979 to 2003. (page number not for citation purposes) Critical Care Vol 13 No 1 Esper et al. Figure 2
Frequency of acute organ dysfunction
Frequency of acute organ dysfunction. Frequency of acute organ dys-
function in patients with diabetes mellitus (DM) and those with no dia-
betes mellitus (non-DM) with a respiratory source of sepsis. CV = cardiovascular; Heme = haematological; Resp = respiratory. * p
< 0.05. Figure 1
Frequency of sepsis cases
Frequency of sepsis cases. Frequency of sepsis cases among patients
with diabetes mellitus (DM) and those with no diabetes mellitus (non-
DM) with a source of infection identified. CV = cardiovascular; GI = gastrointestinal; GU = genitourinal; Resp =
respiratory; SST = skin and soft tissue. * p < 0.05. Figure 2
Frequency of acute organ dysfunction
Frequency of acute organ dysfunction. Frequency of acute organ dys-
function in patients with diabetes mellitus (DM) and those with no dia-
betes mellitus (non-DM) with a respiratory source of sepsis. CV = cardiovascular; Heme = haematological; Resp = respiratory. * p
< 0.05. Figure 2
Frequency of acute organ dysfunction
Frequency of acute organ dysfunction. Frequency of acute organ dys-
function in patients with diabetes mellitus (DM) and those with no dia-
betes mellitus (non-DM) with a respiratory source of sepsis. CV = cardiovascular; Heme = haematological; Resp = respiratory. * p
< 0.05. Figure 1 Frequency of sepsis cases
Frequency of sepsis cases. Frequency of sepsis cases among patients
with diabetes mellitus (DM) and those with no diabetes mellitus (non-
DM) with a source of infection identified. CV = cardiovascular; GI = gastrointestinal; GU = genitourinal; Resp =
respiratory; SST = skin and soft tissue. * p < 0.05. Frequency of acute organ dysfunction
Frequency of acute organ dysfunction. Discussion The current epidemiological study allows us to further charac-
terise the impact of DM on the development of organ dysfunc-
tion among patients with sepsis. When compared with
patients with severe sepsis and no DM, people with DM are
less likely to develop acute respiratory failure. The lower risk of
acute respiratory failure among patients with severe sepsis
and DM was irrespective of whether the primary source of
infection was pulmonary or non-pulmonary. With respect to
other organ dysfunctions, people with DM were more likely to
develop acute renal failure. The presence of a GU source of
infection did not affect the development of acute renal failure
among those with DM. The decrease in the frequency of res-
piratory failure in people with DM was associated with a signif-
icant difference in case fatality. Outcomes Among patients with sepsis, the case-fatality rate was lower
for those with DM at 18.5% versus 20.6% in those with no DM
(p < 0.05). No other significant differences were found in case
fatality with respect to source of infection except for GU sep-
sis at 9% for those with DM vs. 12% in patients with no DM (p
< 0.05). People with DM and sepsis who developed acute res-
piratory failure had a case fatality rate of 52% versus 48% in
those with no DM (p = NS). These data are consistent with previous observations made by
our group in a study evaluating the impact of DM on the devel-
opment of ARDS in patients with septic shock [11]. In that pro-
spective multi-centre ICU study, 28% of the patients with
septic shock had a history of DM. Patients in the ICU with no
DM were more likely to have pneumonia, urinary tract and
abdominal infections. Only 25% of patients with DM devel-
oped ARDS compared with 47% of those with no DM (p =
0.03, relative risk = 0.53, 95% CI = 0.28 to 0.98). In a multi-
variable model, the protective association between DM and
the development of ARDS remained significant. Our novel
observation was confirmed in another prospective cohort
study of 688 heterogeneous patients in the ICU [12]. After
multivariate adjustment, DM was again associated with a
decreased risk of ARDS, with a similar odds ratio of 0.58
(95% CI = 0.36 to 0.92). In agreement with the current study,
the above data suggest that people with DM and a variety of Table 1 Frequency of acute organ dys-
function in patients with diabetes mellitus (DM) and those with no dia-
betes mellitus (non-DM) with a respiratory source of sepsis. CV = cardiovascular; Heme = haematological; Resp = respiratory. * p
< 0.05. likely to develop acute renal failure than those with no DM irre-
spective of the source of infection (10% vs. 6% for pulmonary
sepsis, 14% vs. 8% for non-pulmonary sepsis; both p < 0.05). When a GU source of infection was compared with a non-GU
source of infection, people with DM and sepsis were still more
likely to develop acute renal failure than those with no DM. Among people with DM and sepsis, 46% with a non-GU
source of infection developed acute renal failure, compared
with 44% with a GU source of infection. The only other signif-
icant difference in organ dysfunction was observed in non-pul-
monary sepsis: haematological failure occurred in 1.6% of
patients with DM and in 3.1% of those with no DM (p < 0.05)
(Figures 2 and 3). Page 4 of 6
(page number not for citation purposes) Available online http://ccforum.com/content/13/1/R18 Available online http://ccforum.com/content/13/1/R18 other conditions are less likely to develop both acute respira-
tory failure and ARDS
Figure 3
Frequency of organ dysfunction
Frequency of organ dysfunction. Frequency of organ dysfunction in
patients with diabetes mellitus (DM) and those with no diabetes melli-
tus (non-DM) with a non-respiratory source of sepsis. CV = cardiovascular; Heme = haematological; Resp = respiratory. * p
< 0.05. Figure 3 stood if haemoglobin A1C levels were available as a marker of
previous glycaemic control, in addition to serial glucose levels
during the patient's stay in the ICU. Another possible explana-
tion for the association between DM and the risk of ARDS may
relate to increased medical care among patients with DM. Patients with DM may be hospitalised earlier than those with
no DM in the course of their illness because they learn to be
aware of specific signs of infection. Information on timing of
presentation and onset of symptoms, however, may be difficult
to obtain in many patients. Pharmacological aspects of DM care may also influence the
development of organ dysfunction, because many medica-
tions administered to patients with DM, including insulin and
thiazolidinediones (TZDs), are known to have anti-inflamma-
tory effects in addition to lowering blood glucose. Although the
role of intensive insulin therapy in patients with severe sepsis
remains uncertain, insulin may have other beneficial effects in
this patient population. A key feature of ARDS is the systemic
production of pro-inflammatory mediators and cytokines, such
as tumour necrosis factor (TNF) , interleukin (IL) 1, IL-6 and
IL-8, which have been found in the bronchoalveolar lavage
fluid and plasma of patients with ARDS; and elevated concen-
trations have been associated with an unfavourable outcome
[22-24]. A critical mediator of this inflammatory cascade is the
transcriptional regulatory factor nuclear factor (NF) B, which
may be suppressed by insulin administration. Insulin adminis-
tration to animals challenged with lipopolysaccharide inhibits
TNF production in a dose-dependent manner [25] and pre-
vents the development of ALI [26]. Similarly, TZDs may modu-
late the inflammatory response through the peroxisome-
proliferator-activated receptor gamma and at the transcrip-
tional level through inhibition of NF-B activity [27-31]. Further
investigations on the role of insulin and TZDs on the inflamma-
tory response are necessary to identify a possible mechanism
for affecting the development of ALI. Frequency of organ dysfunction
Frequency of organ dysfunction. Available online http://ccforum.com/content/13/1/R18 Frequency of organ dysfunction in
patients with diabetes mellitus (DM) and those with no diabetes melli-
tus (non-DM) with a non-respiratory source of sepsis. CV = cardiovascular; Heme = haematological; Resp = respiratory. * p
< 0.05. Frequency of organ dysfunction
Frequency of organ dysfunction. Frequency of organ dysfunction in
patients with diabetes mellitus (DM) and those with no diabetes melli-
tus (non-DM) with a non-respiratory source of sepsis. CV = cardiovascular; Heme = haematological; Resp = respiratory. * p
< 0.05. other conditions are less likely to develop both acute respira-
tory failure and ARDS. Our present study has limitations related to the use of hospital
administrative data. Although the use of ICD-9 codes to iden-
tify specific medical conditions is not ideal, it has been vali-
dated for sepsis as having a positive predictive value of 88.9%
and a sensitivity of 87.7% [1,15]. Individual patient-level data,
such as medications, haemoglobin A1C levels and glucose lev-
els would be difficult to obtain from such data sets. The study
is further limited by the lack of data on severity of organ dys-
function, which may have implications on other outcomes. Dis-
crimination between patients with type 1 and type 2 DM may
also provide clues to mechanisms for differential organ dys-
function. However, the large sample size of patients obtained
from utilising a national data base may offset some of these
limitations. Conclusions This study confirms previous observations that a history of DM
is associated with a lower incidence of acute respiratory failure
in patients with severe sepsis. The information obtained moves
us a step closer to better understanding the pathogenesis of
sepsis and sepsis-related organ dysfunction, such as ALI. Identifying conditions that have an effect on the propensity to
develop organ dysfunction in sepsis will allow for the expan-
sion of studies on interventions for this disease. Further pro-
spective data need to be collected in this cohort of patients to
identify the factors that contribute to this protective effect of
DM. The mechanisms responsible for this epidemiological associa-
tion between DM and ARDS are unclear. The effect of DM on
the immune system and inflammatory response is thought to
play a role [19], and perhaps a blunted inflammatory response
effects the development of organ dysfunction in sepsis. Possi-
ble mechanisms of protection in patients with DM may be
impaired neutrophil function or altered neutrophil-endothelial
interactions [20,21]. Obtaining data on specific inflammatory
markers that may play a role in the differences in response to
an infectious insult may clarify the association as well. Discharge status Over the study period, the hospital length of stay for people
with DM was 12.8 days versus 14.1 days in those with no DM
(p < 0.001). Discharge status during the study period was dif-
ferent between patients with and without diabetes: patients
with no DM were more likely to be discharged home (65% vs. 58%, respectively, p < 0.05) whereas patients with DM were
more likely to be discharged to an outside health care facility
(32% vs. 28%, respectively, p < 0.05). Page 4 of 6
(page number not for citation purposes) References 25. Satomi N, Sakurai A, Haranaka K: Relationship of hypoglycemia
to tumor necrosis factor production and antitumor activity: role
of glucose, insulin, and macrophages. J Natl Cancer Inst 1985,
74:1255-1260. 1. Martin GS, Mannino DM, Eaton S, Moss M: The epidemiology of
sepsis in the United States from 1979 through 2000. N Engl J
Med 2003, 348:1546-1554. 26. Fraker DL, Merino MJ, Norton JA: Reversal of the toxic effects of
cachectin by concurrent insulin administration. Am J Physiol
1989, 256:E725-E731. ,
2. Bone RC, Sprung CL, Sibbald WJ: Definitions for sepsis and
organ failure. Crit Care Med 1992, 20:724-726. 2. Bone RC, Sprung CL, Sibbald WJ: Definitions for sepsis and
organ failure. Crit Care Med 1992, 20:724-726. 3. Fowler AA, Hamman RF, Good JT, Benson KN, Baird M, Eberle DJ,
Petty TL, Hyers TM: Adult respiratory distress syndrome: risk
with common predispositions. Ann Intern Med 1983,
98:593-597. 27. Fahmi H, Di Battista JA, Pelletier JP, Mineau F, Ranger P, Martel-
Pelletier J: Peroxisome proliferator – activated receptor gamma
activators inhibit interleukin-1beta-induced nitric oxide and
matrix
metalloproteinase
13
production
in
human
chondrocytes. Arthritis Rheum 2001, 44:595-607. 4. Pepe PE, Potkin RT, Reus DH, Hudson LD, Carrico CJ: Clinical
predictors of the adult respiratory distress syndrome. Am J
Surg 1982, 144:124-130. 28. Guyton K, Zingarelli B, Ashton S, Teti G, Tempel G, Reilly C, Gilke-
son G, Halushka P, Cook J: Peroxisome proliferator-activated
receptor-gamma agonists modulate macrophage activation
by gram-negative and gram-positive bacterial stimuli. Shock
2003, 20:56-62. g
5. Sands KE, Bates DW, Lanken PN, Graman PS, Hibberd PL, Kahn
KL, Parsonnet J, Panzer R, Orav EJ, Snydman DR, Black E,
Schwartz JS, Moore R, Johnson BL Jr, Platt R, Academic Medical
Center Consortium Sepsis Project Working Group: Epidemiol-
ogy of sepsis syndrome in 8 academic medical centers. JAMA
1997, 278:234-240. 29. Aljada A, Garg R, Ghanim H, Mohanty P, Hamouda W, Assian E,
Dandona P: Nuclear factor-kappaB suppressive and inhibitor-
kappaB stimulatory effects of troglitazone in obese patients
with type 2 diabetes: evidence of an antiinflammatory action? J Clin Endocrinol Metab 2001, 86:3250-3256. 6. Angus DC, Linde-Zwirble WT, Lidicker J, Clermont G, Carcillo J,
Pinsky MR: Epidemiology of severe sepsis in the United
States: analysis of incidence, outcome, and associated costs
of care. Crit Care Med 2001, 29:1303-1310. 30. Authors' contributions 20. Goetz MB, Proctor RA: Normalization of intracellular calcium: a
sweet solution to neutrophil dysfunction in diabetes? Ann
Intern Med 1995, 123:952-954. AE carried out the main analysis and interpretation of the data,
in addition to preparing the manuscript. MM and GM both con-
tributed to the design of the study, interpretation of the data,
statistical analysis and manuscript revision. All authors read
and approved the final manuscript. 21. Delamaire M, Maugendre D, Moreno M, Le Goff MC, Allannic H,
Genetet B: Impaired leucocyte functions in diabetic patients. Diabet Med 1997, 14:29-34. 22. Piantadosi CA, Schwartz DA: The acute respiratory distress
syndrome. Ann Intern Med 2004, 141:460-470. 23. Bellingan GJ: The pulmonary physician in critical care * 6: The
pathogenesis of ALI/ARDS. Thorax 2002, 57:540-546. Acknowledgements 24. Meduri GU, Kohler G, Headley S, Tolley E, Stentz F, Postlethwaite
A: Inflammatory cytokines in the BAL of patients with ARDS. Persistent elevation over time predicts poor outcome. Chest
1995, 108:1303-1314. g
DSM was supported by grant HL K23-067739 and MM by grant AA
R01-014435 from the National Institutes of Health. DSM was supported by grant HL K23-067739 and MM by grant AA
R01-014435 from the National Institutes of Health. Key messages 15. Eaton S, Burnham E, Martin GS, Moss M: The ICD-9 code for
septicemia maintains a high positive predictive value for clini-
cal sepsis. Am J Respir Crit Care Med 2002, 165:A471-A471. •
Patients with DM and severe sepsis are less likely to
develop acute respiratory failure than patients with no
DM, irrespective of source of infection. •
Patients with DM and severe sepsis are less likely to
develop acute respiratory failure than patients with no
DM, irrespective of source of infection. p
p
,
16. Charlson ME, Pompei P, Ales KL, MacKenzie CR: A new method
of classifying prognostic comorbidity in longitudinal studies:
development and validation. J Chronic Dis 1987, 40:373-383. p
,
17. Deyo RA, Cherkin DC, Ciol MA: Adapting a clinical comorbidity
index for use with ICD-9-CM administrative databases. J Clin
Epidemiol 1992, 45:613-619. •
Patients with DM and severe sepsis are more likely to
develop acute renal failure than patients with no DM. p
,
18. Charlson M, Szatrowski TP, Peterson J, Gold J: Validation of a
combined comorbidity index. J Clin Epidemiol 1994,
47:1245-1251. •
The decreased frequency of acute respiratory failure in
patients with DM and severe sepsis did not translate
into a significant difference in case fatality. 19. Frank JA, Nuckton TJ, Matthay MA: Diabetes mellitus: a negative
predictor for the development of acute respiratory distress
syndrome from septic shock. Crit Care Med 2000,
28:2645-2646. Competing interests Hyperglycaemia may be another factor that influences the
development of ARDS. In our previous prospective study,
there was a trend towards a lower incidence of ARDS in
hyperglycemic patients with no DM; however, this was based
on admission glucose values. This effect may be better under- p
g
The authors declare that they have no competing interests. Page 5 of 6
(page number not for citation purposes) Page 5 of 6 Critical Care Vol 13 No 1 Esper et al. 14. CDC Diabetes Program [http://www.cdc.gov/diabetes/NDEP/] References Ghanim H, Garg R, Aljada A, Mohanty P, Kumbkarni Y, Assian E,
Hamouda W, Dandona P: Suppression of nuclear factor-kap-
paB and stimulation of inhibitor kappaB by troglitazone: evi-
dence for an anti-inflammatory effect and a potential
antiatherosclerotic effect in the obese. J Clin Endocrinol Metab
2001, 86:1306-1312. ,
7. Martin GS, Mannino DM, Moss M: The effect of age on the devel-
opment and outcome of adult sepsis. Crit Care Med 2006,
34:15-21. 8. Danai PA, Moss M, Mannino DM, Martin GS: The epidemiology
of sepsis in patients with malignancy. Chest 2006,
129:1432-1440. 31. Liu D, Zeng BX, Zhang SH, Wang YL, Zeng L, Geng ZL, Zhang SF:
Rosiglitazone, a peroxisome proliferator-activated receptor-
gamma agonist, reduces acute lung injury in endotoxemic
rats. Crit Care Med 2005, 33:2309-2316. 9. Mrus JM, Braun L, Yi MS, Linde-Zwirble WT, Johnston JA: Impact
of HIV/AIDS on care and outcomes of severe sepsis. Crit Care
2005, 9:R623-R630. ,
10. Mokdad AH, Ford ES, Bowman BA, Dietz WH, Vinicor F, Bales VS,
Marks JS: Prevalence of obesity, diabetes, and obesity-related
health risk factors, 2001. JAMA 2003, 289:76-79. 11. Moss M, Guidot DM, Steinberg KP, Duhon GF, Treece P, Wolken
R, Hudson LD, Parsons PE: Diabetic patients have a decreased
incidence of acute respiratory distress syndrome. Crit Care
Med 2000, 28:2187-2192. 12. Gong MN, Thompson BT, Williams P, Pothier L, Boyce PD, Chris-
tiani DC: Clinical predictors of and mortality in acute respira-
tory distress syndrome: potential role of red cell transfusion. Crit Care Med 2005, 33:1191-1198. Crit Care Med 2005, 33:1191-1198. 13. National Hospital Discharge Survey Description [http://
www.cdc.gov/nchs/about/major/hdasd/nhdsdes.htm] Page 6 of 6
(page number not for citation purposes)
|
https://openalex.org/W2948203008
|
http://pure-oai.bham.ac.uk/ws/files/67767949/published_1.pdf
|
English
| null |
Whales, fish and Alaskan bears: interest-relative taxonomy and kind pluralism in biology
|
Synthese
| 2,019
|
cc-by
| 11,838
|
Whales, fish and Alaskan bears
Taylor, John DOI:
10.1007/s11229-019-02284-9
License:
Creative Commons: Attribution (CC BY) Citation for published version (Harvard):
Taylor, J 2019, 'Whales, fish and Alaskan bears: interest-relative taxonomy and kind pluralism in biology',
Synthese. https://doi.org/10.1007/s11229-019-02284-9 Link to publication on Research at Birmingham portal Publisher Rights Statement:
Checked for eligibility: 24/06/2019 Publisher Rights Statement:
Checked for eligibility: 24/06/2019 Publisher Rights Statement:
Checked for eligibility: 24/06/2019 General rights
U l
li General rights
Unless a licence is specified above, all rights (including copyright and moral rights) in this document are retained by the authors and/or the
copyright holders. The express permission of the copyright holder must be obtained for any use of this material other than for purposes
permitted by law. •Users may freely distribute the URL that is used to identify this publication. d
l
d
d/
i
f h
bli
i
f
h
i
i •Users may freely distribute the URL that is used to identify this publication. •Users may download and/or print one copy of the publication from the University of Birmingham research portal for the purpose of private
study or non commercial research y
y
y
p
•Users may download and/or print one copy of the publication from the University of Birmingham research portal for the purpose of private
study or non-commercial research. y
•User may use extracts from the document in line with the concept of ‘fair dealing’ under the Copyright, Designs and Patents Act 1988 (?)
•Users may not further distribute the material nor use it for the purposes of commercial gain. here a licence is displayed above, please note the terms and conditions of the licence govern your use of this document Where a licence is displayed above, please note the terms an When citing, please reference the published version. Take down policy Take down policy
While the University of Birmingham exercises care and attention in making items available there are rare occasions when an item has been
uploaded in error or has been deemed to be commercially or otherwise sensitive. If you believe that this is the case for this document, please contact UBIRA@lists.bham.ac.uk providing details and we will remove access to
the work immediately and investigate. Download date: 24. Oct. 2024 Synthese
https://doi.org/10.1007/s11229-019-02284-9 Synthese y
https://doi.org/10.1007/s11229-019-02284-9 BIOBEHAVIOUR Abstract This paper uses two case studies to explore an interest-relative view of taxonomy and
how it complements kind pluralism in biology. First, I consider the ABC island bear,
which can be correctly classified into more than one species. I argue that this classifi-
catory pluralism can be explained by reference to the range of alternative explanatory
interests in biology. In the second half of the paper, I pursue an interest-relative view
of classification more generally. I then apply the resultant view to a second case study:
whether whales are fish. I argue that this question is not one about scientific vs folk
usage, as has been assumed. I also develop a new view: that Fish should be rejected
as a category, both from the point of view of biological science, and from the point of
view of folk taxonomy. Along the way, I use the interest-relative view to shed light on
the circumstances under which higher taxa should be accepted as legitimate categories
for biological science. Keywords Interest-relative taxonomy · Kind pluralism · Species · Higher Taxa Keywords Interest-relative taxonomy · Kind pluralism · Species · Higher Taxa Whales, fish and Alaskan bears: interest-relative taxonomy
and kind pluralism in biology Henry Taylor1
Received: 24 May 2018 / Accepted: 4 June 2019
© The Author(s) 2019 Henry Taylor1
Received: 24 May 2018 / Accepted: 4 June 2019
© The Author(s) 2019 Henry Taylor1 Received: 24 May 2018 / Accepted: 4 June 2019
© The Author(s) 2019 1
Department of Philosophy, University of Birmingham, ERI Building, Edgbaston,
Birmingham B15 2TT, UK B Henry Taylor
j.h.taylor.1@bham.ac.uk B Henry Taylor
j.h.taylor.1@bham.ac.uk
1
Department of Philosophy, University of Birmingham, ERI Building, Edgbaston,
Birmingham B15 2TT, UK 1 Introduction Imagine that you are a school photographer, and I am a school cook. You want to take
a good photo of the children, and I want to cook them lunch. For your task, you need to
arrange the children into groups by height, because you need to put the shortest children
at the front, and the tallest ones at the back. For me, things are different. I should
categorise them based on their dietary requirements. I will use groups such as meat-
eater, vegetarian, vegan, child with food allergies, child with religious commitments,
etc. It makes sense for me to group the children this way because it is most useful to
my purpose. Needless to say, our two systems will diverge in their groupings. Suppose B Henry Taylor
j.h.taylor.1@bham.ac.uk 123 12 Synthese Synthese Anita and Polly are meat eaters (with no allergies or religious commitments), and that
Anita is short and Polly is tall. They will fall into the same kind for me but not for you. Our kinds cross-classify each other. It is clear that neither one of us has a uniquely
privileged taxonomy of the children. Your taxonomy is good for your purposes and
mine is good for mine. Anita is short and Polly is tall. They will fall into the same kind for me but not for you. Our kinds cross-classify each other. It is clear that neither one of us has a uniquely
privileged taxonomy of the children. Your taxonomy is good for your purposes and
mine is good for mine. This paper is about how this sort of classificatory taxonomic divergence occurs in
biology, and what we should do when it does. We must start with the notion of kind
pluralism. By this, I mean all and only cases when we have more than one classification
system, such that: (1) The systems do not agree in all of their groupings (2) For at least two of the systems, it is not the case that they should be rejected as
illegitimate (2) For at least two of the systems, it is not the case that they should be rejected as
illegitimate (3) No one of the systems is the ‘best’, ‘privileged’ or ‘objectively correct’ system The core terms in (1–3) will be explained and elaborated as we go. The photog-
rapher/cook case is a small example of kind pluralism. 1 As Dasgupta (2018, pp. 279–280) points out, Goodman thought that linguistic history (not usefulness)
determines which taxonomic practices we engage in. 2 The ABC island bear It is well-known that there are several distinct concepts of ‘species’ in biology (Kitcher
1984; Dupré 1981, 1993; Ereskefsky 1992, 1998, 2001; Stanford 1995; Zachos 2016;
Wilkins 2018). The phylogenetic concept sorts organisms into species based on their
ancestry. The ecological concept classifies organisms into species based on their being
lineages that share the same ecological niche. Very roughly, an ecological niche is
composed of the environmental pressures that are placed on an organism, and the
properties and behaviours it has developed to cope in that environment (Ereshefsky
2001). The biological concept claims that species are relatively isolated gene pools. On
this picture, species are groups of actually or potentially interbreeding populations, and
these populations must be reproductively isolated from other populations (Mayr 1957). A species (on the biological species concept) will include populations of organisms
that can (actually or potentially) exchange genetic material in a relatively free manner,
and will exclude those populations where this kind of gene exchange is either not
possible, or is highly restricted. These are only three species concepts, and they only scratch the surface of diver-
gence in species concepts within biology. Zachos (2016, ch. 4) and Wilkins (2018,
appendix B) each list 32 species concepts, and both acknowledge that their lists aren’t
exhaustive.2 With at least 32 concepts to pick from, why choose these three? The first
reason is that this paper is concerned with how the interest-relative view deals with
cases of kind pluralism, and the three concepts already outlined are sufficient to gen-
erate cases of kind pluralism (as we shall soon see). For this reason, they are sufficient
to make the points with which this paper is concerned. Extension of the arguments to
cases where more species concepts are at work will be straightforward. I will make some further comments about the other species concepts operative in
biology in order to justify setting them aside. Many of them simply aren’t relevant to
this paper. For example, the phylo-phenetic species concept was primarily designed for
microbiology (Rosselló-Mora and Amann 2001), which is not the focus of this paper. A further important point is that it is misleading to see this long list of 32 species
concepts as distinct from the three ones that I will primarily be working with. 1 Introduction The case also embodies a
particular view of taxonomic divergence, which we can call the interest-relative view. On this view, alternative categorisation systems are assessed relative to certain inter-
ests. Three core points make up the substance of the interest-relative view of taxonomic
divergence.First,differentinterestsarewhatleadtothetaxonomicdivergence.Second,
what vindicates the alternative systems as legitimate is the fact that each taxonomy is
indeed useful for that particular interest. Third, because each is legitimate for its own
purpose, no one of them is objectively ‘the best’. Rather, each is useful for the interest
it is intended for. The interest-relative view just summarised is a general picture of
how divergent taxonomies arise, and how we should judge them as legitimate. g
j
g
g
Views similar to the interest-relative view are defended by Rorty (2000), Taylor
and Vickers (2017) and Goodman (1955).1 In philosophy of biology, a view similar
to this is associated with Dupré (1993), and I have recently developed a similar view
for a case in psychology Taylor (2017). The purpose of this paper is not to argue that
the interest-relative view is better than any other view of taxonomy. Rather, it is to
introduce an interest-relative view of taxonomy, clarify some of its relations with kind
pluralism, and lend it initial plausibility by showing how it can correctly deem some
taxonomies legitimate and others illegitimate. For these reasons, the interest-relative
view will largely be assumed in what follows. Even opponents of the view will be
interested in how the view copes with concrete case studies from biology, so this
paper will still be important for the unconvinced. In Sect. 2, I examine the various alternative concepts of ‘species’ within biology. I
argue that the interest-relative view can make sense of this divergence in terms of the
variety of explanatory interests within biology. I do this by way of a case study: the
ABC island bear. In Sect. 3, I impose two criteria that taxonomic systems must fulfil
in order to be deemed legitimate. I then briefly consider how the view relates to the
issue of natural kinds (Sect. 4). Finally, I consider what the interest-relative view has
to say about the issue of whether whales are fish. I argue that the Fish category fails to
be a useful category, both for the purposes of biological science, and from the point of 12 3 Synthese view of folk taxonomy (Sect. 5). 1 Introduction In a sense I argue that the question of whether whales
are fish is a misplaced one because it assumes that Fish is a respectable category into
which whales should either be placed or not placed. But the truth is different. Fish is
not a legitimate category in the first place. 2 Thanks to an anonymous referee for pushing me on this.
3 Isaythatmost ofthealternativespeciesconceptscontainelementsofthethreementionedabove.Arguably,
the pheneticist one does not (I will return to this species concept in Sect. 3). 3 Isaythatmost ofthealternativespeciesconceptscontainelementsofthethreementionedabove.Arguabl
the pheneticist one does not (I will return to this species concept in Sect. 3). 2 Thanks to an anonymous referee for pushing me on this. 2 The ABC island bear Rather,
most of these alternative species concepts contain elements of the three concepts out-
lined above.3 There are various ways that this occurs. The phylogenetic, biological and
ecological species concepts will themselves subdivide into other subconcepts. Indeed,
it is probably more accurate to say that each one represents an approach to biological
taxonomy, rather than a single concept. As an example of this, take the biological 12 3 3 Synthese species concept. In animal biology, interbreeding is often invoked with relation to the
biological species approach, because interbreeding is the property that animals typi-
cally use to exchange genetic information. However, distinct tools are needed to apply
the biological species concept to the microbiological world, because of course micro-
biological organisms do not interbreed (Redfield 2001). It’s not even clear whether
the biological species concept can apply to microbiology at all (Ereshefsky 2010). Indeed, even if we restrict our attention to animal biology, there are many different
versions of the biological species concept, which differ in the extent to which they
emphasise mate recognition behaviours in understanding gene flow between organism
populations (Mayr 1970; Paterson 1985). Another way that elements of the three approaches given above might filter down
into other species concepts is when species concepts are developed that try to draw
together the insights of other species concepts in order to form a unified one (Stamos
2003; Van Valen 1976; de Querioz 2007). Indeed, even authors who claim to reject
the species category altogether typically replace it with concepts that draw heavily on
one, two or all of the three approaches summarised above. For example, Pleijel (1999)
aims to replace ‘species’ with the notion of a ‘Least Inclusive Taxonomic Unit’, but
this concept makes crucial use of the notion of monophyly, which is an idea at the core
of the phylogenetic species concept. Yet another way that we can see the influence of
the phylogenetic, biological and ecological approaches is in the integrative taxonomy
movement, which applies several distinct approaches to the same group of organisms,
in the hope of finding convergence (Padial et al. 2010; Schlick-Steiner et al. 2010). In
summary, though I do select three specific approaches from the panoply of available
species concepts, they are the three approaches that most prominently breathe through
the contemporary work on biological taxonomy, and so much that I say will have
general application. Consider the ABC island bear. 4 There is some debate over whether polar bear fur is ‘white’, strictly speaking, because of its unusual
radiative properties (Preciado et al. 2002), but I set this aside.
5 The ecological and interbreeding concepts represent the clearest example of kind divergence, and are
sufficient to establish my conclusions. Similar things will apply to the phylogenetic concept. 2 The ABC island bear There is a population of bears that live on the Alaskan
‘ABC’ islands. They are the result of successful interbreeding between female polar
bears and male brown bears, and they can mate with both (Cahill et al. 2013). Because
they are capable of this interbreeding, they constitute a population which can exchange
genetic material with brown bears (Ursus arctos) and polar bears (Ursus maritimus). By the biological species concept, then, they are conspecific with these two groups. However, they have behavioural adaptations to their environment that are shared by
brown bears, but not polar bears. So by the ecological concept, they should be classified
as conspecific with brown bears, but not polar bears. Herewehaveaninstanceofkindpluralism.Inowshowhowaninterest-relativeview
of biological taxonomy can explain the emergence of this kind pluralism. Amongst
the range of explanatory goals in biology is assessing the extent to which a particular
behaviour or phenotypic trait helps the organism in a specific setting, and thus what
fitness advantage it confers. This helps us explain how it evolved. For example, perhaps
we are interested in the explaining the evolution of bears that live in cold environments
where there is a lot of ocean life, but very little land or plant life. If this is our interest,
then we will be interested in particular traits of bears that live in such an environment. Given this interest, it is useful for us to taxonomise together those bears that live
in the particular cold environment we are interested in, and which have developed a
similar cluster of traits specifically to deal with them, such as white fur, the building of 123 Synthese especiallywellinsulateddens,orthehuntingofveryfattyoceananimals.4 Thisrestricts
the grouping to one particular environmental setting, and to the particular properties
thatwewishtoexplain.Thisrestrictionisusefulbecauseitsetstheexplanatorytargetof
the enquiry (we want to explain how these particular features help in this particular
setting). It would not make sense for our group to include bears that live in other
environments (and have different adaptations to their environments), because that is
not our focus. If we included those bears in the group that requires explanation, then our
explanations would be ranging over the wrong set of organisms. This grouping would
of course include polar bears that live in the Arctic, and it would exclude ABC bears,
because they do not live in such an environment, and they do not have the properties
that are distinctive of animals that fulfil such an ecological niche. 2 The ABC island bear This would be the
ecological species concept at work. Other biologists may be interested in other explanatory goals, such as explaining
how certain traits entered the population. For example, we may be interested in why
ABC bears have a mitochondrial DNA profile like that of polar bears, whilst having
the fur colour of brown bears (Cahill et al. 2013). For this task, we will be interested
in grouping together those populations of bears that can successfully pass around
genetic information with each other. Recognising these populations is essential for our
explanation. For example, once we know that female polar bears engaged in breeding
behaviour with male brown bears, we can explain how the ABC bears got their brown
fur: they got it from the male brown bears, through the exchange of genetic material, by
mating behaviour. Likewise, they have the mitochondrial DNA profile of polar bears
because mitochondrial DNA is passed down the female line, from the female polar
bears that they are descended from (Cahill et al. 2013). On this taxonomy, brown bears,
ABC bears and polar bears are all grouped together inclusively. Without including both
brown bears and polars into the populations of organisms that can exchange genetic
information, we will not have our explanation of how polar bear mitochondrial DNA
and brown fur is present in the same population of ABC bears. In this way, the interest-relative view neatly explains this case of kind pluralism. One taxonomy (the ecological one), does not group together ABC bears with polar
bears, but does group them with brown ones. On the other (the interbreeding one),
ABC bears are grouped together with polar bears and brown ones. For the interest-
relative view, these distinct systems arose because of distinct explanatory interests in
biology. They are worthy of preservation because each one is useful for the interest
that it was meant for.5 Notice that sometimes we will require different species concepts for different
explanatory targets within the same overall project. If we are interested in how the
brown fur trait got into the population, we must talk about interbreeding behaviours. When we are interested in how the brown fur became propagated within the popula-
tion, and became the dominant fur colour, we will have to talk about ecological niche
(and also common ancestry). The explanatory goals may not be explicitly separated,
and the distinctions between them may be extremely fine-grained. 3 Anything goes As just stated, the interest-relative view is open to the possibility that a difference of
interests (explanatory or not) can give rise to taxonomic divergence. Some will object
that this is too liberal (e.g. Ereshefsky 2001, pp. 150–154; Ghiselin 1987, p. 146). The objection is not so much that an interest-relative view vindicates all possible
taxonomies, but rather that it vindicates ones that we have some independent reason
to think should not be vindicated. This paper is not the place for a thorough analysis of this problem in all of its
forms. Rather, I will outline two criteria that we can use to dismiss some taxonomies
as illegitimate, on the interest-relative view. The aim is not to argue that these criteria
can handle all of the potential problematic cases. I rather argue that these two criteria
as plausible ways that the interest-relative view can dismiss at least some problematic
taxonomies. The main purpose of this section is to establish these criteria, such that
they can be deployed later in the paper when we consider whether whales are fish. First, consider that the interest-relative view says that alternative systems should
be accepted insofar as they fulfil a certain interest. So, the following criterion is rea-
sonable: (i) Taxonomic systems should be rejected as illegitimate if they fail to fulfil the
interest that they are intended for. (i) Taxonomic systems should be rejected as illegitimate if they fail to fulfil the
interest that they are intended for. How does this work in practice? Consider astrology. Astrology recognises the Sun
and Moon as planets. Is this a legitimate classification system, by the lights of the
interest-relative view? No. Astrology has a particular set of interests, which is predict-
ing future events in humans’ lives and predicting personality traits. But the grouping it
produces is not useful for that interest. The system that astrology uses doesn’t actually
help predict events or personality traits (no grouping of celestial bodies could). So it
isn’t useful by its own lights, and it fails by the first criterion. It can be rejected. y
g
y
j
One of the worries that fuels the ‘anything goes’ objection is that we should avoid
views which legitimise pseudoscientific systems. Demonstrating how the interest-
relative view can do this will help lessen this worry. However, similar points apply
outside of pseudoscience as well. Consider the pheneticist school of biological taxon-
omy. 2 The ABC island bear 12 3 3 Synthese I have argued that the interest-relative view can make sense of this instance of kind
pluralism in terms of a range of alternative explanatory goals within biology. Here one
clarification is important. This does not imply that divergence in explanatory goals is
the only way that kind pluralism can arise in biology. On the interest-relative view,
any kind of divergence in interests can potentially give rise to kind pluralism, even
interests that are nothing to do with explanation. This may raise a worry: if any interest
can in principle give rise to kind pluralism, does the interest-relative view end up too
permissive? It is to this that we now turn.6 6 For some examples of non-epistemic interests that can result in different kind taxonomies, see Ludwig
(2017). 7 Dupré (1993, p. 51) also suggests something similar to the second criterion. But Dupré also claims that
we can judge taxonomies based on whether they are epistemically virtuous (1993, pp. 242–243). Ereshefsky
(2001, p. 162) criticises Dupré for not specifying which virtues he has in mind. We need not settle this
debate, since my own proposed view does not make use of these virtue claims.
8 Thanks to Will Davies for making me aware of this revision in cancer taxonomy
9 Of course, in reality, it will not always be so straightforward. There may be aggregation issues. For
example, suppose taxonomy X is useful for n purposes, whilst taxonomy Y is useful for n + 1 purposes. But
the purposes for which X is useful may be somehow more important, along some suitably defined metric
of importance (thanks to an anonymous referee for drawing my attention to this). I do not have space to
address these issues fully here, but my preference would be to say that in such cases, it is indeterminate
whether X should be preferred to Y. We should not worry if our view of classification has indeterminacy as
part of it. On the contrary, indeterminacy arises in many ways when we examine kinds in the life sciences
(e.g. Taylor 2019). 3 Anything goes Pheneticism taxonomises organisms into species based on their overall similarity
(Sneath and Sokal 1973; Sneath 1995). The purpose of this system is to group life in a 123 12 Synthese theoretically neutral manner. The generally accepted view is that this approach failed,
because the choice of which properties one picked would be theoretically driven, and
thus the hope of establishing a taxonomy in a theoretically neutral way is violated
(Ridley 1986; cf. Matthen 2009, p. 99). This received view of the pheneticist project
is controversial (Lewens 2012), but we can say that if it is accurate, then it represents
a good example of a serious taxonomy in biology that can be rejected for violating the
first criterion. Return again to the fact that the interest-relative view judges taxonomies by their
ability to fulfil certain interests. Suppose we have two proposed taxonomies, A and B,
and that A is useful for fulfilling some purpose, but B is more useful for fulfilling that
very same purpose. Then, since A is the less useful for this purpose, it can be rejected
and B can be embraced. This naturally leads to a second criterion by which to judge
rival systems: (ii) For a system to be accepted for a particular purpose, it must be the case that no
rival system is better at fulfilling the purpose in question.7 (ii) For a system to be accepted for a particular purpose, it must be the case that no
rival system is better at fulfilling the purpose in question.7 As a potential example, consider cancer taxonomy8. Standardly, cancers were tax-
onomised by their organ of origin (perhaps along with their histological profile). But
recently it has been proposed that grouping them by their epigenetic molecular profile
is more useful for cancer treatment (Cortés et al. 2014, p. 70, Prat et al. 2013, p. 1). Suppose this new taxonomy really is more useful for this purpose. In this case, the
former system is of course useful for treatment, but the new proposal is more useful,
so the new one should be preferred (at least for that purpose) and the former should
be rejected.9 Notice that these two criteria are in a sense not substantive additions to an
interest-relative view of taxonomy. Rather, the interest-relative view says that different
taxonomies flow from different interests. 10 Thanks to an anonymous referee for emphasising this. 3 Anything goes In this vein, judging different taxonomies
by their ability to fulfil different interests (in the two ways outlined above) is a natural
extension of the theory, rather than an ad hoc modification. I will noteoneissueinorder toset it aside. Usefulness canof coursecomeindegrees:
some groupings can be more or less useful for a particular purpose. Because the
current view explains legitimacy in terms of usefulness, then legitimacy will also come
in degrees. There are many interesting questions connected to spectra of legitimacy
(some of which I have touched on elsewhere, Taylor 2018). However, I do not have
space to discuss them here. 3 12 3 Synthese At this point, we can draw together the ideas of the paper developed so far to
summarise the links between the interest-relative view of taxonomy and the kind
pluralism we find in biology. The fact that different interests will deliver different
groupings explains why it is that different systems with different groupings arise in
biology(thatis,itexplainsthephenomenonthatpoint(1)isreferringtoinmydefinition
of kind pluralism from Sect. 1). Conditions (i–ii) provide an explanation of what it
means for two taxonomies to be ‘legitimate’. That is, they are legitimate when they
cannot be dismissed by either criterion. The concept of a ‘legitimate’ taxonomy is a
core part of understanding kind pluralism, as kind pluralism is all about when different
taxonomies can both be legitimate (see point (2) in the definition of kind pluralism). Finally, the two criteria (i–ii) also help to explain why we have cases where there is
no objectively correct taxonomy of a group of organisms (point (3)). 4 Natural kinds I have been talking about kinds, but to many, natural kinds are where it’s at. Natural
kinds (it is said) are part of the objective structure of the world, and when our systems
latch onto those categories, we are carving nature at the joints (Sider 2011). Such a
view may be thought to clash with the interest-relative view. But on inspection, the
difference between natural kinds theory and the interest-relative view is not great.10 In
the philosophy of science, a theory of natural kinds attempts to give a general theory
about what structure certain groups of entities have. This structure is then invoked in
order to explain what makes certain explanatory and predictive practices in science
highly successful (Magnus 2012, pp. 48–50; Boyd 1999, p. 68; Slater 2015; Franklin-
Hall 2015). The core claim of any theory of natural kinds is that there are certain
categories that are especially useful for scientific prediction and explanation, and that
these categories refer to kinds that have a certain common structure (this common
structure is spelled out differently by different theories, of course). This core claim
is completely consistent with the view I have been putting forward. Whether or not
the kinds that are especially useful for explanation and prediction do have a certain
common structure, the main claims of this paper (that alternative taxonomies should
be judged based on their usefulness) remain the same. Of course, some theorists of natural kinds claim that we should prefer the cate-
gorisation practices that match up with the natural kinds in the world. But even this
claim is not necessarily in tension with the interest-relative view. Suppose for the sake
of argument that it is true that all categories that are particularly useful for scientific
prediction and explanation have a certain structure in common, and that these are the
natural kinds. If that is true then of course we should prefer the categories that latch
onto natural kinds for the purposes of scientific explanation and prediction. There is
no conflict here between my view and a theory of natural kinds. How my account compares to theories of natural kinds can more easily be seen
by way of comparison with Ereshefsky and Reydon’s view (2015). Like myself, they
impose a set of restrictions by which we can judge some kinds as superior to oth- 123 12 Synthese ers, and like myself, these are not metaphysical criteria. 11 Ereshefsky and Reydon (2015, p. 973) assume that scientific kinds are epistemically superior to folk
kinds. They presumably mean superior for the interests of scientific explanation and prediction. Note that
it may very well be that scientific kinds are the best ones for these interests, whilst it is still the case that
scientific kinds are not superior to folk kinds in any absolute sense. 4 Natural kinds They claim that kinds must
be internally coherent, empirically testable and non-degenerative in Lakatos’ (1970)
sense. By contrast, I do not place these constraints on legitimate kinds. The constraints
I place (spelled out with criteria (i–ii)) are much more liberal. Once again, the fric-
tion here is only superficial. Ereshefsky and Reydon offer their account as a way
to judge the relative merits of scientific kinds (they sometimes say ‘natural kinds’),
but the interest-relative view is a way of testing the legitimacy of kinds in general,
including non-scientific ones. It would be inappropriate to place constraints such as
non-degeneracy and empirical testability on non-scientific taxonomies, because these
are not relevant to non-scientific interests. It may be that Ereshefsky and Reydon
are correct about what restrictions should be placed on scientific kinds, whilst the
two criteria given above provide a good account of the legitimacy of kinds generally
(including non-scientific ones).11 5 Are whales fish? Inowturntothecomplexissueofwhetherwhalesarefish,andwhattheinterest-relative
view has to say about this issue. The whales/fish question is especially important to
us, for several reasons. First, it is at the core of debates over how folk kinds relate to
scientific kinds (Dupré 1981, 1993, 1999; Phillips 2014; Sainsbury 2014). So, it must
be addressed by any view that purports to be about the legitimacy of both scientific
and non-scientific kinds. Second, I have noted the worry that the interest-relative view
is too permissive. In this section, I will deploy the two criteria I introduced above, and
argue that they can provide us with a great deal of power when it comes to dismissing
some taxonomic systems as illegitimate. Demonstrating the power of these criteria
will have an emollient effect on those still concerned about whether the view entails
‘anything goes’. y
g g
The whales/fish issue is usually seen as a clash between folk and scientific usage. For example, when discussing the issue, Dupré says: ‘The general point of the discussion so far is to insist that ordinary language and
science provide largely independent and often disjoint ways of classifying the
biological world’ (1999, p. 469) Similarly, Ereshefsky and Reydon say: Similarly, Ereshefsky and Reydon say: ‘[in the case of whale classification] folk biology and scientific biology provide
competing classificatory programs’ (2015, p. 984). ‘[in the case of whale classification] folk biology and scientific biology provide
competing classificatory programs’ (2015, p. 984). Here we clearly see Dupré, Ereshefsky and Reydon setting up the issue as one where
folk and scientific groupings clash with each other. But setting up the issue like this
is imprecise for several reasons. For one thing, folk linguistic usage gives us at least
two distinct taxonomies. At one time, folk usage had it that whales are fish (Burnett 12 3 3 Synthese Synthese 2007, ch. 2). But now, of course, folk usage has it that whales are not fish. That is
what we are all taught in school, and when you ask people if whales are fish, then they
will say no. So here we have one kind conflict: between modern and non-modern folk
classification. Scientific taxonomy is also not monolithic. Scientific taxonomy used to
claim that whales are indeed fish (Linneaus 1756, p. 39). So we have the non-modern
scientific system to contend with, as well as the modern scientific system. 5.1 The modern scientific use First, terminology: I use the italicised Fish to refer to the proposed category, and the
non-italicised word to refer to fish themselves. It is natural to think that on a modern
scientific taxonomy of life, whales are not fish. After all, biology tells us that they
are mammals. However, I shall now argue that a scientific taxonomy of life does not
classify whales as non-fish. My argument for this claim is based on the claim that Fish
is not a biological category, and that as a result science is silent on whether whales are
or are not fish. Dupré notes an idea like this, saying that ‘the notion that there is a ‘scientific’ usage
of the word ‘fish’ is a decidedly suspect one’ (1999, p. 466). However, the reasons
he gives for this claim should be resisted. He considers the proposed definition of a
fish as a ‘cold blooded-vertebrate with gills’ (this would exclude whales, which are
warm-blooded). He points out that on this definition, the category will not cover all of
its intended referents. Some tuna are warm-blooded, and so would be excluded. Dupré
also claims that on this definition, Fish would include a wide range of organisms, and
that there is ‘little or no biological rationale for a category containing just these groups’
(1999, p. 466). The fact that a good definition of ‘fish’ cannot be given which gives a set of sci-
entifically respectable properties possessed by all and only its intended referents does
not show that Fish is not a biological category. In an era of biological taxonomy that
has rejected essentialism, the idea that different members of a single taxon must all
share a set of intrinsic, scientifically respectable properties such as ‘vertebrate’ or
‘cold-blooded’ that is necessary and sufficient for membership into that taxon is now
almost dead (Sober 1980, Hull 1965). Given that we are no longer essentialists, we
shouldn’t expect a good scientific taxon to be amenable to a definition in terms of
intrinsic properties. So the failure of this particular definition of ‘fish’ (both in terms
of capturing its intended referents, and in terms of the scientific rationale given for it)
does not show that Fish isn’t a biological category. Dupré’s reasons should be rejected. So, we should set aside Dupré’s reasons to reject Fish as a good category for
biological science. 5 Are whales fish? The most
tempting misconception of all is to assume that modern scientific taxonomy claims
that whales are not fish. As I will argue below, this is incorrect. 5.1 The modern scientific use We should look to the interest-relative view for such reasons. To
summarise, my argument will be this: recall that for a category to be accepted as
legitimate, two criteria must be met. First, it must fulfil an interest. Second, it must
be the case that no other taxonomy fulfils that same interest better. So, for Fish to be
accepted as a legitimate category for contemporary biological science, it must fulfil
these two criteria with respect to some interest from contemporary biology. I will first 123 Synthese consider the ecological and interbreeding approaches considered above, and argue that
they do not show that the Fish category is useful for biological science. Then, I turn
to cladistics, and argue that we have good reason to think that cladistics is useful for
biology, but that we cannot use the usefulness of cladistics to show that Fish is useful,
because cladistics does not recognise Fish. With the conclusion that cladistics does not recognise Fish, we can engage in the
following reasoning. In order to be a legitimate category for biology, a grouping
must fulfil some biological interest such that no other grouping is better at fulfilling
that interest. So, given that cladistics is a taxonomic system that gives us groupings
alternative to the Fish one, it follows that, in order for Fish to be legitimate, it must
fulfil some biological interest, such that the groupings given by cladistics are not better
at fulfilling that interest. This is because if the categories given to us by cladistics are
better at fulfilling that interest than Fish, then Fish will be an inferior category for that
interest and can be rejected. I consider a range of biological uses that Fish may fulfil,
and argue that in all cases, either Fish is not useful for that purpose, or the groupings
given to us by cladistics are more useful. Therefore, it is not the case that Fish can
fulfil some biological interest and that no rival is better at fulfilling that interest, and
so Fish is not a legitimate biological category. I will not attempt to consider every
possible biological use that might be suggested for Fish, but rather just restrict myself
to the clearest examples. Extension to other potential uses for the category will be
straightforward. This argument demonstrates some more of the power of the interest-relative view. 5.1 The modern scientific use The idea that Fish is not a biological category has often been mentioned, but as we
shall now see, the interest-relative view can give us a principled explanation of why
it is not one, which is not susceptible to the weaknesses that other arguments for the
same conclusion face. 5.1.1 Interbreeding and ecology again The first suggestion can be dealt with very quickly. Above, I discussed the ecological
and interbreeding species concepts. These concepts pick out groups that are useful
for biology, as we have seen. But obviously, we cannot claim that such concepts will
vindicate Fish. Such concepts do not apply to Fish, because they are species concepts,
and so do not apply to a proposed taxon as high-level as Fish. It is not the case that
Fish represents a stable set of populations that can freely exchange genetic material
with each other. It is also not true that all fish occupy the same ecological niche. If
anything is going to vindicate Fish, then it will have to be suitable to apply to higher
taxa (i.e. taxa higher than ‘species’ in the Linnean hierarchy). 12 Strictly speaking, Ghiselin and Eldredge and Cracraft argue that higher taxa are not the units of evolution,
not that they are unimportant for biology. Ereshefsky (1991) replies to these arguments. 5.1.3 Cladistics Above, I dismissed the ecological and interbreeding concepts of species because they
clearly can’t apply to a taxon as high as Fish. Isn’t the same true of the phylogenetic
species concept? Well of course the phylogenetic species concept won’t apply to Fish,
but the general cladistics approach (that gave birth to the phylogenetic concept) is
an obvious place to look for a way of classifying higher taxa in a biologically useful
manner (Hennig 1950, 1966). Indeed, cladistics is by far the dominant approach to
higher taxa, in both biology and philosophy (Rieppel 2005; Boyd 2010). The success
of cladistics gives us good reason to think that it is a useful approach to higher taxa,
at least for some biological interests. So, if Fish is recognised by cladistics, then this
gives some good reason to think that it will be useful for biology. Maybe this can
legitimate Fish. But this doesn’t work. At the core of cladistics is the idea of a monophyletic taxon:
a taxon that contains an ancestor and all and only its descendants. These ‘clades’
are the groups that are recognised by cladistics. It is well known that ‘fish’ is not a
monophyletic taxon. Rather, it is paraphyletic (it includes a common ancestor and
some but not all of its descendants). The paraphyletic status of the fish category stops
it from being a clade, and thus from being recognised as a scientific category by
cladistics. The fact that Fish is not recognised by cladistics blocks us from justifying
it based on the claim that cladistic taxa are useful. I have argued that we have good reason to think that cladistics is taxonomic system
that is useful for at least some of the purposes of biology, and can apply to higher
taxa, and it excludes the Fish category. This is important for the argument that Fish
isn’t a legitimate biological category. To be a legitimate biological system, a taxonomic
system must fulfil some biological interest such that no other system fulfils that interest
better. So, in order to legitimate Fish, it is not enough that it fulfils some interest, it
must do so better than the groupings of higher taxa given to us by cladistics (since
these are rival groupings to the Fish one). 5.1.2 No higher taxa! One extreme view is that there is no useful way for biology to classify higher taxa,
perhaps because higher taxa are unimportant for biology generally (cf. Ghiselin 1987,
p. 141; Eldredge and Cracraft 1980, p. 249).12 This is an extreme view. But even if 12 3 3 Synthese Synthese we do accept it, it is perfectly consistent with my claim that Fish is not a scientific
category. If we claim that there is no useful way for biology to classify higher taxa,
then (since Fish is a higher taxon if it is a taxon at all) of course it will follow that Fish
is not a useful category for the purposes of biology. Thus, it will not pass the test of
usefulness, and (for the interest-relative view) can be rejected. we do accept it, it is perfectly consistent with my claim that Fish is not a scientific
category. If we claim that there is no useful way for biology to classify higher taxa,
then (since Fish is a higher taxon if it is a taxon at all) of course it will follow that Fish
is not a useful category for the purposes of biology. Thus, it will not pass the test of
usefulness, and (for the interest-relative view) can be rejected. 5.1.4 Descent and modification, uniformity, homeostatic cohesion and selection Ereshefsky (1991) presents various ways that higher taxa may prove useful for biology. Since we are concerned with higher taxa, I take the uses illustrated in his paper as
my model (to be clear, Ereshefsky is arguing for the legitimacy of higher taxa in
general, not defending the Fish category). I argue that for all of the uses he proposes,
Fish would either be useless, or be less useful than the monophyletic system that
is its rival. Ereshefsky comments that higher taxa display a pattern of descent with
modification, and that this makes them legitimate units of evolution. But when he says
that higher taxa have common patterns of descent, he is defending the legitimacy of
monophyletic higher taxa (he calls them ‘lineages’ but by this he means monphyletic
taxa (1991, pp. 87–88)). This is with good reason. Using a paraphyletic category
would be suboptimal for the purposes of tracking patterns of descent and modification,
because paraphyletic groups exclude certain descendants of a common ancestor, so
they would give one only a partial image of how certain traits have spread down the
generations, and how they have been modified. For example, suppose we wanted to
track the descent and modification of the mammalian forelimb. Using a paraphyletic
group that excluded bats (say) would deprive us of a complete picture of how this
forelimb has descended and been modified. The monophyletic group Mammalia would
be much more useful for this interest, and should thus be preferred. Ereshefsky (1991) considers other reasons to vindicate higher taxa as legitimate
categories. I will deal with them quickly. First, he considers the suggestion (made by
others) that a stable evolutionary unit should have ‘cohesiveness’, or ‘uniformity’. As
he points out, this notion is very unclear (1991, p. 89). Until such a notion is spelled out,
we can’t adjudicate whether Fish has cohesiveness or uniformity. Second, Ereshefsky
notes that higher taxa can share a common ancestor, and thus will share some common
genetic material. Ereshefsky suggests that this may act as a homeostatic constraint on
the characteristics of members of that higher taxon (1991, p. 92). This homeostatic
cohesion may legitimate the higher taxon in question. This is perfectly reasonable but
it doesn’t save Fish. Tetrapods share a common ancestor with fish, so we should expect
them also to have inherited the same genetic homeostatic constraints from the common
ancestor they share with fish. 13 Evolutionary taxonomy is a rival view to cladistics, which recognises paraphyletic taxa (Ereshefsky 2001,
pp. 51–53). But as Ereshefsky and Reydon say: Cladism has ‘led to the positing of countless successful
classifications, as well as contributed to the extinction of the Phenetics school, and the near extinction of 5.1.3 Cladistics Below, I consider several suggestions, and
suggest that either Fish is useless for the purpose in question, or that the groupings
given to us by cladistics are better. Note that my argument doesn’t imply that all higher taxa must be monophyletic
in order to be useful for biology. We can agree with some authors that we shouldn’t
always be wedded to monophyly (Brummitt 2002, 2003; Mayr and Brock 2002). The claim isn’t that monophyly is always the way. The claim is just that we have
available a monophyletic taxonomy which is different from Fish. Thus, if Fish is to be
vindicated, it must be the case that there is at least one biological interest such that the
monophyletic taxonomy is not better at fulfilling that interest than Fish. The claim that
there are some taxonomic groupings that are non-monophyletic, and which are better Synthese at fulfilling some interest than the monophyletic one, is not in tension with this claim. The claim is just that (for the interests I will examine) Fish is not such a category. Footnote 13 continued
Evolutionary Taxonomy’ (2015, p. 982). It would be unwise to rest our case that Fish is a biological category
on an almost extinct research programme such as evolutionary taxonomy. In any case, whether or not Fish
is recognised by evolutionary taxonomy, the real question is whether it is useful. We have already examined
this idea, and found that it is not useful. 5.1.4 Descent and modification, uniformity, homeostatic cohesion and selection So the group that has homeostatic cohesion as a result of
common ancestry will include the Tetrapods, and will not align with the Fish category. Again, the monophyletic category is better for the purposes of biology. Third, he notes
that higher taxa may have similar selection regimes, in virtue of which they should
be grouped together (1991, p. 92) but this doesn’t apply to Fish, because they have
very different selection regimes (including different predator/prey relationships). This
point was already touched upon when I pointed out that different fish have different
ecological niches.13 12 3 3 Synthese 5.1.5 Fish is not a biological category I have surveyed a range of possible ways for higher taxa to be biologically useful,
and found that in all cases, either Fish is not useful, or a rival grouping (given by
cladistics) is more useful. Thus, given that the category does not fulfil the two criteria
for a legitimate category, we can conclude that it should be rejected, at least when we’re
considering biological interests. Obviously, I have not shown that it is impossible for
Fish to fulfil some biological interest. But given that we have examined a range of
options, and none has worked out, it is reasonable to infer that it is unlikely that we
will find such a use. Now, if we accept that Fish is not a biological category, then we can begin to see the
whales/fish question in a new light. Contemporary biology does not rule that whales
are non-fish, because Fish is just not a category that biology deals with at all. This
point can be made clearer by distinguishing between two claims, which differ in scope: (i) A scientifically respectable taxonomy does not classify whales as fish. i) A scientifically respectable taxonomy does not classify whales as fish. (ii) A scientifically respectable taxonomy classifies whales as not fish. 14 Another suggestion could be made: that the modern folk use arose in order to more closely model a
different scientific discovery, the discovery that Fish is a bad scientific category. However, if the modern
folk usage were to accurately reflect this fact, then it should abandon the Fish category entirely, rather than
claim that whales are not fish. So either this is not the intended purpose of the modern folk usage, or it is
its intended purpose and modern folk usage fails to fulfil it adequately. (ii) A scientifically respectable taxonomy classifies whales as not fish. i) A scientifically respectable taxonomy classifies whales as not fish. (i) is true. Fish is not a biologically legitimate taxon, so biologists do not classify
whales as fish. Incidentally, of course, they do not classify tuna, herring, salmon etc. as fish, for the same reasons. I am not denying of course that biologists themselves
will still use the common word ‘fish’ to refer to herring, salmon etc. But this is just an
artefact of the fact that biologists’ informal nomenclature usually accords with folk
nomenclature. But from the point of view of legitimate biological taxonomy, nothing
is categorised either as fish or non-fish. But (ii) is false. For (ii) to be true, ‘fish’ would
have to be a biological category that biology can include and exclude organisms from. It is not plausible to claim that (ii) follows from (i). That would be to infer from
the fact that science does not recognise a particular category to the claim that science
classifies something as not falling into that category. This is clearly too extreme. The
category ‘beautiful painting’ is not a scientific category, but it would be absurd to infer
from this that science tells us that The Last Supper is not a beautiful painting. Of course, ‘mammal’ is a good scientific category: it is a perfectly respectable
monophyletic taxon, and whales do fall into that category. Ereshefsky and Reydon
take this to indicate that scientific taxonomy clashes with the claim that whales are
fish (2015, p. 984). But this is inaccurate. It doesn’t follow from the fact that they are
mammals to the claim that they are not fish. They could be both mammals and fish
(‘mammalian fish’ as Dupré calls them (1999)). There is nothing in biology that tells
us that mammals cannot be fish, because (to repeat a point already made) fish is not a
category that biology deals with. At this point, another taxonomy can easily be dealt with as well.: the non-modern
scientific use (lasting until at least the eighteenth century) on which whales are fish
(Burnett 2007, ch. 3). Should we accept this one? No. It should be uncontroversial 12 3 Synthese that modern biology has replaced this eighteenth century biological taxonomy with
one that is more useful for the purposes of biology. So, the interest-relative view tells
us that we should disregard the old scientific taxonomy in favour of the modern one. (ii) A scientifically respectable taxonomy classifies whales as not fish. The crucial point is this: once we realise that Fish is not a biological category, we
should drop the idea that there is a clash between contemporary biology and the claim
that whales are fish. For all of these reasons, we need to stop seeing the debate as a
clash between folk and scientific taxonomy, as suggested by Eresehefsky and Reydon
and Dupré in the quotations I gave above. There is no clash between the folk and
science, because science is silent on the question of whether whales are fish. 5.2 Modern folk use Dupré says: 5.3 Non-modern folk use Things are looking good for the non-modern folk use, on which whales are fish. However, this view does not win by default. It must be justified by serving a particular
interestbetterthantherivals.Whowouldhavesuchaninterest?Themostclearexample
is whalers. As we know (Burnett 2007) there was a time when whalers claimed that
whales are fish. This is an excellent case to examine because there is a clear folk
practice, that has a clear purpose, and which has a particular taxonomy. The practice is
whaling, the purpose is catching whales (skinning them, preparing their blubber, etc.),
and the taxonomic claim is that whales are fish. In order to vindicate this taxonomy,
then, we must ask whether this taxonomy is useful for this purpose. There are several factors that may lead us to the conclusion that it is useful for
whalers to group whales as fish. Many of the traits that are most important to a whaler
are its behavioural traits, which whales share with fish (at least at a coarse-grain). For example, how they move through the water, their threat avoidance behaviour, and
(most obviously) the fact that they live in the ocean, rather than on land. These are all
properties that whalers are interested in, and they are all properties that whales share
to a much greater extent with creatures like sharks, salmon, herring etc. than with any
other group of organisms. So (it may be thought) this gives us good reason to think
that the grouping of whales as fish is useful for whalers, and thus the claim that whales
are fish is vindicated. At this point, it could of course be pointed out that at a fine-grain of detail, the
properties shared by whales and fish are not very similar, and that there are many
dissimilarities between whales and things like salmon, etc. (they have lungs, they
surface for air, their skeletal structure is different, etc.) But even setting these issues
aside, there is a much deeper problem with the reasoning above that attempts to
vindicate the non-modern folk usage. Of course, it is true that whalers are primarily
interested in properties such as overall body shape, and marine environments, and it is
also true that these properties are shared more with herring, sharks etc. than with other
groups of organisms. However, to justify the preservation of the whalers’ taxonomy,
it must be shown that the category Fish is useful for the whaler. Dupré says: ‘I have already conceded the obvious fact that whales are not fish. Why not? We might ask both why this fact is obvious, and why it came to be a fact at
all. The first question is easy enough. Educated speakers will, I suspect, almost
unanimously refuse to apply the word ‘fish’ to, Blue whales, Killer whales…
Ultimately, I suppose that this is the only relevant evidence and that it is decisive.’
(1999, p. 467). Now, it could be that Dupré is here simply reporting a linguistic fact. On this
interpretation, when he says that whales are not fish, all he is saying is that this is not
how the word ‘fish’ is used in modern English. If that is all he means, then clearly he is
correct. I won’t debate this point. However, he could mean something more substantial. He may mean that the fact that people use the word ‘fish’ in this way is sufficient reason
for us to accept this taxonomy. This seems to be indicated by his claim that folk usage
is ‘decisive’ in determining that whales are not fish. But on the view proposed in this
paper, it is not enough to rely on linguistic usage. Rather, the categorisation has to be
useful for an interest, such that no alternative taxonomy is better for that interest. The modern folk category Fish fulfils a broad range of interests. I will restrict
myself to examining the interest which drove people to exclude whales from Fish. This is the most directly relevant to our purposes. Plausibly, this is because they
mistakenly believed that it was wrong to class whales as fish. They believed that
biology had proved that whales aren’t fish. As a result, the interest that drove people to
exclude whales from Fish is that they wanted their usage to accord with the scientific
discovery that whales are not fish. But the modern usage fails to do this, because
there is no scientific discovery that whales are not fish. As elaborated at length above,
this is because Fish is not a category that biology deals with, and so it is not one that
it can exclude organisms from. So the modern folk category of Fish fails to fulfil its
intended purpose: it fails to accord with a scientific discovery.14 For this reason, it
can be dismissed as failing on its own terms. Dupré says: 12 3 3 Synthese Of course, it may be the case that there is some other modern folk use that is
relevant to legitimating Fish, but as I said above, I do not have the space to examine
all such proposals. At the very least, we can reach the following conclusion: that for
the purpose that plausibly drove people to exclude whales from the Fish category, the
category fails to fulfil that interest. 15 Here there is a disanalogy with the taxonomy considered in Sect. 3, which groups polar bears, ABC
bears and grizzly bears together. Here it was explanatorily important that the taxonomy extended beyond
just the ABC bears themselves because interbreeding behaviours between all three groups is a core part of
the explanation in question. 5.3 Non-modern folk use The issue is that
this category is very broad, in a way that is not useful to the whaler. Whaling is not
concerned with creatures such as herring, salmon, mackerel etc. The whaler’s focus
is on whales, and only on whales. Given that they do not have interests that extend to
fish as a whole, the taxonomic practice that extends to the whole of fish does not have
a clear use for the practice of whaling. All the whaler needs to know is the properties
that whales have that impact on how to hunt them and prepare them. Exactly how
these relate to the properties of salmon, herring, sharks etc. is not important for the
whaler, because the whaler’s interests don’t extend that far. 123 12 Synthese Synthese Ultimately, the category of Fish (whether or not it includes whales) is not useful for
the interests of the whaler, because their interests do not extend so far that considering
the entire category of fish themselves is really useful. So, since the taxonomy is not
useful in this way, it again fails the test.15 Again, it could be the case that there is some
other non-modern folk usage for Fish. As before, I do not have space to examine all
such possible uses, but we can reach the conclusion that at least for the purposes of
the folk practice which is most relevant to the inclusion or exclusion of whales into
Fish, the category fails to fulfil the two criteria in question. The situation we are left with is this: we had four potential ways of settling the
question of whether whales are fish: the modern scientific one, the non-modern scien-
tific one, the modern folk one, and non-modern folk one. But using only the austere
resources of the interest-relative view, we have found reasons to be sceptical of the
utility of Fish for all of these practices. The result is that Fish should be rejected as
a category, for folk and scientific purposes. In a sense, then, the question of whether
whales are fish should be set to one side. The point is just that Fish is not itself a useful
(and therefore legitimate) category. category. Ultimately, this relatively liberal view may be just the view of taxonomy that
we need.16 category. Ultimately, this relatively liberal view may be just the view of taxonomy that
we need.16 Acknowledgements Funding was provided by Leverhulme Trust (Grant No. ECF-2015-088). Acknowledgements Funding was provided by Leverhulme Trust (Grant No. ECF-2015-088). Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 Interna-
tional License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution,
and reproduction in any medium, provided you give appropriate credit to the original author(s) and the
source, provide a link to the Creative Commons license, and indicate if changes were made. Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 Interna-
tional License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution,
and reproduction in any medium, provided you give appropriate credit to the original author(s) and the
source, provide a link to the Creative Commons license, and indicate if changes were made. 16 Thanks to Maja Spener, the editors and three anonymous referees for very helpful comments on earlier
drafts. Thanks to Alison Fernandes, Scott Sturgeon, Ema Sullivan-Bissett and Alastair Wilson for valuable
discussion. References Boyd, R. (1999). Kinds, complexity and multiple realization: Comments on Millikan’s “historical kinds
and the special sciences”. Philosophical Studies, 95, 67–98. Boyd, R. (2010). Homeostasis, higher taxa and monophyly. Philos Boyd, R. (2010). Homeostasis, higher taxa and monophyly. Philosophy of Science, 77, 686–701. Brummitt, R. (2002). How to chop up a tree. Taxon, 51, 31–41. Brummitt, R. (2002). How to chop up a tree. Taxon, 51, 31–41. Brummitt, R. (2003). Further dogged defence of paraphyletic taxa. Taxon, 52, 803–804. Burnett, D. (2007). Trying Leviathan. Princeton: Princeton University Press. Cahill, J., et al. (2013). Genomic evidence for island population conversion resolves conflicting theories of
polar bear evolution. PLoS Genetics, 9, 1–8. Cahill, J., et al. (2013). Genomic evidence for island
polar bear evolution. PLoS Genetics, 9, 1–8. polar bear evolution. PLoS Genetics, 9, 1–8. Cortés, J., et al. (2014). New approach to cancer therapy based on a molecularly defined cancer classification. CA: A Cancer Journal for Clinicians, 64, 70–74. pp
py
CA: A Cancer Journal for Clinicians, 64, 70–74. f
Dasgupta, S. (2018). Realism and the absence of value. Philosophical Review., 127, 279–322. De Querioz, K. (2007). Species concepts and species delimitation. Systematic Biology, 56, 879–88 Dupré, J. (1981). Natural kinds and biological taxa. The Philosophical Review, 90, 66–90. Dupré, J. (1993). The disorder of things. Cambridge: Harvard. J. (1993). The disorder of things. Cambridge: Harva Dupré, J. (1999). Are whales fish? In D. Medin & S. Atran (Eds.), Folkbiology. Hong Kong: Oxfo
University Press. Eldredge, N., & Cracraft, J. (1980). Phylogenetic patterns and the evolutionary process. New Yor
Columbia University Press. Ereshefsky, M. (1991). Species, higher taxa, and the units of evolution. Philosophy of Science, 58, 84–10 Ereshefsky, M. (1992). Eliminative pluralism. Philosophy of Science, 59(4), 671–690. Ereshefsky, M. (1998). Species pluralism and anti-realism. Philosophy of Science, 65(1), 103–120. ecies pluralism and anti-realism. Philosophy of Scien Ereshefsky, M. (2001). The poverty of the Linnean hierarchy. New York: CUP. Ereshefsky, M. (2010). Microbiology and the species problem. Biology and Philosophy, 25, 553–568. Ereshefsky, M., & Reydon, T. (2015). Scientific kinds. Philosophical Studies, 172, 969–986. Franklin-Hall, L. (2015). Natural kinds and categorical bottlenecks. Philosophical Studies, 172, 925–94 Ghiselin, M. (1987). Species concepts, individuality and objectivity. Biology and Philosophy, 2, 127–14 Goodman, N. (1955). Fact, fiction and forecast. Cambridge, MA: Harvard University Press. Hennig, W. (1950). Grundzuge einer theorie der phylogenetischen systematik. Berlin: Deutscher zentralver-
lag. Hennig, W. (1966). Phylogenetic systematics. 6 Conclusion We can see the ABC island bear case as an example of the way that the interest-
relative view can act positively, to vindicate alternative taxonomies of kinds: it justifies
preservation of several alternative concepts of species because they are useful for
biology. Conversely, the question of whether whales are fish is one way that the
interest relative view acts negatively: to dismiss taxonomies as illegitimate. Recall
that the main objection to the interest-relative view is that it is too permissive. But in
fact, examination of the question of Fish reveals that it is extremely restrictive: it has
within it the power to reject a great many taxonomies as illegitimate. To be clear: the problem is not that Fish is too folksy, or that it is ill-defined, or
that it is not a sharp category. A legitimate biological category can be difficult (even
impossible) to define, and many (even most) of them do not have sharp boundaries
(Hull 1965; Sober 1980; Boyd 1999). Further, being ‘folksy’ is no problem if we are
looking for reasons to think that a folk category is legitimate. Rather, the problem with
Fish is just that it’s not really useful. The purpose of this paper was to see how far the interest-relative view could go,
even if we resist the temptation to privilege some interests over others. It turns out it
can go surprisingly far: it gives us a good picture of why there are alternative species
concepts, and it’s powerful enough to give us a principled reason to reject the Fish 12 3 3 Synthese category. Ultimately, this relatively liberal view may be just the view of taxonomy that
we need.16 References Annual Review of Entomology, 10, 97–116. Hull, D. (1965). The effect of essentialism on taxonomy: Two thousand years of stasis (I). The Briti
Journal for the Philosophy of Science, 15, 314–326. Kitcher, P. (1984). Species. Philosophy of Science, 51, 308–333. Lakatos, I. (1970). Falsification and the methodology of scientific research programmes. In I. Lakatos &
A. Musgrave (Eds.), Criticism and the growth of knowledge. Cambridge, MA: CUP. Lewens, T. (2012). Pheneticism reconsidered. Biology and Philosophy, 27, 159–177. Linnaeus, C. (1756). Systema Naturae (9th ed.). Leiden: Haak. Ludwig, D. (2017). Indigenous and scientific kinds. British Journal for the Philosophy of Science., 6
187–212. 123 Synthese Magnus, P. D. (2012). Scientific enquiry and natural kinds: From planets to mallards. New York: Palgrav
Macmillan. Matthen, M. (2009). Chickens, eggs and speciation. Noûs, 43, 94–115. Mayr, E. (1957). Species concepts and definitions. In E. Mayr (Ed.), The species problem. Washington:
American Association for the Advancement of Science. 970). Population, species, and evolution. Cambridge, Mayr, E., & Brock, W. (2002). Classifications and other ordering systems. Journal of Zoological Systematics
and Evolutionary Research., 40, 169–194. Padial, J., Miralles, A., De Ia Riva, I., & Vences, M. (2010). The integrative future of taxonomy. Frontiers
in Zoology, 7, 1–14. Paterson, H. (1985). The recognition concept of species. In E. Vrba (Ed.), Species and speciation. Pretoria:
Transvall. Phillips, I. (2014). Catacean semantics: A reply to Sainsbury. Analysis, 74, 379–382. Pleijel, F. (1999). Phylogenetic taxonomy, a farewell to species and a revision of Heteropodarke (Hesionidae,
Polychaeta, Annelida). Systematic Biology, 48, 755–789. Prat, A., et al. (2013). Genomic analyses across six cancer types identify basal-like breast cancer as a unique
molecular entity. Scientific Reports., 3, 1–12. Preciado, J., et al. (2002). Radiative properties of polar bear hair. American Society of Medical Engineers
International Mechanical Engineering Congress and Exposition: Advances in Bioengineering, 53,
57–58. Redfield, R. (2001). Do bacteria have sex? Nature Reviews Genetics, 2, 634–639. Ridley, M. (1986). Evolution and classification. London: Longman. Rieppel, O. (2005). Monophyly, paraphyly, and natural kinds. Biology and Philosophy, 20, 465–487. Rieppel, O. (2005). Monophyly, paraphyly, and natural kinds. Biolog Rorty, R. (2000). Philosophy and social hope. New York: Penguin. Rosselló-Mora, R., & Amann, R. (2001). The species concept for prokaryotes. FEMS Microbiological
Review, 25, 39–67. Sainsbury, M. (2014). Fishy business. Analysis, 74, 3–5. Schlick-Steiner, B., Steiner, F., Seifert, B., Stauffer, C., Christian, E., & Crozier, R. (2010). References Integrative tax-
onomy: a multisource approach to exploring biodiversity. Annual Review of Entomology, 55, 421–438. Sider, T. (2011). Writing the book of the world. New York: OUP. Schlick-Steiner, B., Steiner, F., Seifert, B., Stauffer, C., Christian, E., & Crozier, R. (2010). Integrative ta
onomy: a multisource approach to exploring biodiversity. Annual Review of Entomology, 55, 421–43 Sider, T. (2011). Writing the book of the world. New York: OUP. Sider, T. (2011). Writing the book of the world. New York: OUP. Slater, M. (2015). Natural kindness. The British Journal for the Philosophy of Science, 66, 375–411. Sneath, P. (1995). Thirty years of numerical taxonomy. Systematic Biology, 44, 281281–281298. Sneath, P., & Sokal, R. (1973). Numerical taxonomy: The principles and practices of numerical classifica
tion. San Francisco: W.H. Freeman. Sober, E. (1980). Evolution, population thinking and essentialism. Philosophy of Science, 47, 350–383. Stamos, D. (2003). The species problem. Lanham: Lexington Books. Stamos, D. (2003). The species problem. Lanham: Lexington Books. Stanford, P. (1995). For pluralism and against realism about species. Philosophy of Science, 62, 70–91. Taylor, H. (2017). Attention, psychology and pluralism. The British Journal for the Philosophy of Science. https://doi.org/10.1093/bjps/axx030. https://doi.org/10.1093/bjps/axx030. Taylor, H. (2018). Emotions, concepts and the indeterminacy of natural kinds. Synthese. https://doi.org/10. 1007/s11229-018-1783-y. Taylor, H. (2019). Fuzziness in the mind: Can perception be unconscious? Philosophy and Phenomenolog-
ical Research. https://doi.org/10.1111/phpr.12592. Taylor, H. (2019). Fuzziness in the mind: Can perception be unconscious? Philosophy and Phenomenolo
i
l R
h htt
//d i
/10 1111/ h
12592 Taylor, H. (2019). Fuzziness in the mind: Can perception
ical Research. https://doi.org/10.1111/phpr.12592. ical Research. https://doi.org/10.1111/phpr.12592 Taylor, H., & Vickers, P. (2017). Conceptual fragmentation and the rise of eliminativism. European Journal
for the Philosophy of Science., 7, 17–40. Wilkins, J. (2018). Species: The evolution of the idea. Boca Raton: CRC Press. Wilkins, J. (2018). Species: The evolution of the idea. Boca Raton: CRC Press. ). Species: The evolution of the idea. Boca Raton: CR Zachos, F. (2016). Species concepts in biology. Cham: Springer. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps
and institutional affiliations. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps
and institutional affiliations. 123 123 12
|
https://openalex.org/W2159861873
|
https://journals.plos.org/plosone/article/file?id=10.1371/journal.pone.0005521&type=printable
|
English
| null |
Incidence and Prediction of Falls in Dementia: A Prospective Study in Older People
|
PloS one
| 2,009
|
cc-by
| 6,872
|
Abstract Additional funding was provid
Council, UK. The authors’ work was independent of the funding bodies. ted by the Alzheimer’s Society, UK and ER by the Alzheimer’s Research Trust. Additional funding was provided by the Medical Research
work was independent of the funding bodies. Competing Interests: The authors have declared that no competing interests exist. Competing Interests: The authors have declared that no competing interests exist. * E-mail: louise.allan@ncl.ac.uk Louise M. Allan1*, Clive G. Ballard2, Elise N. Rowan1, Rose Anne Kenny3 1 Institute for Ageing and Health, Wolfson Research Centre, Newcastle General Hospital, Newcastle upon Tyne, United Kingdom, 2 Wolfson Centre for Age Related
Disorders, King’s College London, London, United Kingdom, 3 Trinity College Institute of Neuroscience, Trinity College Dublin, Dublin, Ireland 1 Institute for Ageing and Health, Wolfson Research Centre, Newcastle General Hospital, Newcastle upon Tyne, United Kingdom
Disorders, King’s College London, London, United Kingdom, 3 Trinity College Institute of Neuroscience, Trinity College Dublin, Du Abstract Background: Falls are a major cause of morbidity and mortality in dementia, but there have been no prospective studies of
risk factors for falling specific to this patient population, and no successful falls intervention/prevention trials. This
prospective study aimed to identify modifiable risk factors for falling in older people with mild to moderate dementia. Methods and Findings: 179 participants aged over 65 years were recruited from outpatient clinics in the UK (38 Alzheimer’s
disease (AD), 32 Vascular dementia (VAD), 30 Dementia with Lewy bodies (DLB), 40 Parkinson’s disease with dementia (PDD),
39 healthy controls). A multifactorial assessment of baseline risk factors was performed and fall diaries were completed
prospectively for 12 months. Dementia participants experienced nearly 8 times more incident falls (9118/1000 person-years)
than controls (1023/1000 person-years; incidence density ratio: 7.58, 3.11–18.5). In dementia, significant univariate
predictors of sustaining at least one fall included diagnosis of Lewy body disorder (proportional hazard ratio (HR) adjusted
for age and sex: 3.33, 2.11–5.26), and history of falls in the preceding 12 months (HR: 2.52, 1.52–4.17). In multivariate
analyses, significant potentially modifiable predictors were symptomatic orthostatic hypotension (HR: 2.13, 1.19–3.80),
autonomic symptom score (HR per point 0–36: 1.055, 1.012–1.099), and Cornell depression score (HR per point 0–40: 1.053,
1.01–1.099). Higher levels of physical activity were protective (HR per point 0–9: 0.827, 0.716–0.956). Conclusions: The management of symptomatic orthostatic hypotension, autonomic symptoms and depression, and the
encouragement of physical activity may provide the core elements for the most fruitful strategy to reduce falls in people
with dementia. Randomised controlled trials to assess such a strategy are a priority. Citation: Allan LM, Ballard CG, Rowan EN, Kenny RA (2009) Incidence and Prediction of Falls in Dementia: A Prospective Study in Older People. PLoS ONE 4(5):
e5521. doi:10.1371/journal.pone.0005521
Editor: Bernhard Baune James Cook University Australia rd CG, Rowan EN, Kenny RA (2009) Incidence and Prediction of Falls in Dementia: A Prospective Study in Older People. PLoS ONE 4(5)
al.pone.0005521 Editor: Bernhard Baune, James Cook University, Australia Received October 8, 2008; Accepted March 7, 2009; Published May 13, 2009 Copyright: 2009 Allan et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits
unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Funding: LA was supported by the Alzheimer’s Society, UK and ER by the Alzheimer’s Research Trust. PLoS ONE | www.plosone.org Introduction hazards, muscle weakness and incontinence.[9] In community
settings, impairment on various cognitive measures has been
shown to be a risk factor for falls, although dementia itself has only
been examined and shown to be a risk factor for falls in
community dwelling people with Parkinson’s disease,[10] and in
residents of extended care settings.[7,11] The identification of
potentially modifiable risk factors has been critical in the
development of effective multifactorial falls intervention pro-
grammes, particularly for older people at high risk of falling.[12] It
is likely that falls in dementia are also multi-factorial in origin,
possibly with risk factors similar to those identified in the general
older population, but there may be other potentially modifiable
factors specific to dementia. It is thought that those with Lewy
body (LB) dementias are at particular risk, and a history of
recurrent falls is accepted as a supporting feature for the diagnosis
of dementia with Lewy bodies (DLB).[13] The prevalence of neurodegenerative disorders is increasing due
to changes in population demographics. It is estimated that by
2020 there will be 42 million people with a diagnosis of dementia
worldwide,[1] in whom the most common causes of dementia will
be Alzheimer’s disease (AD), Vascular dementia (VAD) and the
Lewy body dementias.[2,3] Falls are a significant cause of injuries,
loss of confidence, increased morbidity, institutionalisation and
mortality in all older people,[4,5] but particularly those with
dementia.[6,7] People with dementia recover less well after a fall
than those without dementia.[8] In view of the suffering caused by
such falls, and the enormous cost of caring for people with
dementia who have fallen, there is an urgent need to optimise the
prevention of falls in this group. Importantly,
there
have
been
no
prospective
studies
of
multifactorial risk factors for falling specific to the population
with dementia. Previous studies have identified multiple risk
factors for falls in the older population as a whole, in a variety of
settings. In multivariate studies significant risk factors include fall
history, gait, balance and mobility impairments, visual impair-
ment,
cognitive
impairment,
fear
of
falling,
environmental Only
one
previous
multifactorial
intervention
study
has
exclusively recruited participants with dementia.[14] Interventions
were based upon those used in successful trials in people without
dementia, and the study did not show a significant reduction in
falls or number of fallers. Citation: Allan LM, Ballard CG, Rowan EN, Kenny RA (2009) Incidence and Prediction of Falls in Dementia: A Prospective Study in Older People. PLoS ONE 4(5):
e5521. doi:10.1371/journal.pone.0005521 Autonomic Assessment All assessments took place in the morning; participants refrained
from consuming caffeinated drinks or smoking on the morning of
the assessment. Assessments were carried out according to the
protocols described in our previous study.[31] Briefly, blood
pressure was monitored using a digital photoplethysmograph
(Portapres, TNO, Amsterdam), which enables non-invasive beat-
to-beat blood pressure measurement. Orthostatic hypotension
(OH) was defined as a fall in systolic blood pressure of greater than
20 mm Hg or diastolic blood pressure of greater than 10 mm Hg
that did not return to baseline within 30 seconds from the start of
the active stand. Return to baseline was defined as the start of a
series of 3 consecutive beats in which the blood pressure was
within one standard deviation of the baseline blood pressure. Participants were asked to report symptoms on standing; if
dizziness,
lightheadedness,
unsteadiness
or
presyncope
were
reported in the presence of OH this was defined as symptomatic
OH. Other clinical autonomic function tests included isometric
exercise, Valsalva manoeuvre and deep breathing. Ewing’s battery
was used to identify the presence of a clinical autonomic
neuropathy for each patient who had complied with sufficient
tests for the classification scheme to be applied.[32] Outcome Measures The primary outcome measures were prevalence and incidence
of falls occurring during the 12 month follow up period. Secondary outcome measures were proportional hazard ratios
for time to first fall in dementia, according to diagnosis and status
of putative clinical predictors. A fall was defined as an event
whereby a person comes to lie on the ground or another lower
level with or without loss of consciousness. Participants were given
diaries to record the occurrence of falls, to be returned every four
weeks in a postage paid envelope. If the participants did not return
the diaries they were reminded by telephone after 2 weeks. The
caregiver was asked to assist in the completion of diaries when
dementia was present. All participants were over 65 years of age. All diagnoses of
dementia were made according to DSM IV criteria. Significant
medical causes of dementia were excluded during diagnostic
investigations. Diagnoses were made by operationalised criteria for
AD,[16] VAD,[17] DLB[18] and PDD,[19] which have been
validated
against
neuropathological
diagnosis
within
our
group.[20,21] Participants were excluded if they declined participation, died
or withdrew from the study before commencing falls diaries, were
unable to perform the gait assessments due to other co-morbid
conditions, had an MMSE score[22] less than or equal to 8, or
were too visually impaired to complete cognitive assessments. Controls were excluded if they had any evidence of dementia or
Parkinson’s disease. Design Design
Longitudinal cohort study. Design
Longitudinal cohort study. Longitudinal cohort study. Statistics Differences in baseline characteristics across groups were
compared using Fisher’s Exact test for categorical data, ANOVA
for normally distributed data and Kruskal-Wallis for non-normally
distributed data. Differences between individual groups were
compared using the Chi squared test for categorical data, Introduction However, 70% of the participants in this
trial lived in institutional settings, most had moderate to severe PLoS ONE | www.plosone.org May 2009 | Volume 4 | Issue 5 | e5521 1 May 2009 | Volume 4 | Issue 5 | e5521 Falls in Dementia in more than one high quality study in older people, and/or their
relevance to clinical features of dementia postulated to be
causative of falls in dementia. cognitive impairment and all had already presented to hospital
with a fall. A recent systematic review of interventions to prevent
falls in hospitals and care homes suggested that effect size was
smaller in trials where the prevalence of dementia was high, but
meta-regressions were limited by incomplete reporting of dementia
prevalence.[15] It remains possible that multifactorial interven-
tions would be successful in patients with less severe dementia, but
we suggest that interventions in future trials should be tailored to
potentially modifiable factors which can be shown to predict falls
in community dwelling people with mild-moderate dementia. There is therefore an urgent need to understand risk factors for
falls in dementia to enable a more tailored and effective
intervention to be developed. Assessments included medical history (duration of dementia,
residence, history of previous falls and medications). Objective
assessments included the cognitive subsection of the CAMDEX
(CAMCOG),[23] a physical activity scale previously validated in
older people,[24] body mass index, performance-oriented assess-
ment of mobility[25] and the motor subsection of the Unified
Parkinson’s disease rating scale (UPDRS[26]) to evaluate extra-
pyramidal signs. Participants were assessed after taking their usual
dose of levodopa, if applicable, and were allowed to use their usual
walking aids. Dementia specific scales were used to assess activities of daily
living (Bristol scale[27]), depression (Cornell scale[28,29]) and
behavioural and psychological symptoms of dementia (Neuropsy-
chiatric
Inventory[30]),
in
addition
to
detailed
autonomic
assessments as described below. These factors have all been
postulated as potential causes of falls in dementia, particularly in
the LB dementias.[13] Participants in this study completed a multifactorial baseline
assessment of putative predictors of falls and then completed
prospective falls diaries for a period of 12 months, which is known
to be a robust method of ascertainment of falls. We aimed to
identify potentially modifiable predictors of falls in older people
with mild-moderate dementia, the majority of whom lived in the
community. Participant recruitment We recruited consecutive cases with dementia (AD, VAD, DLB
or Parkinson’s disease dementia (PDD)) from Neurology, Old Age
Psychiatry and Geriatric Medical secondary care outpatient clinics
within the Northern Region of the United Kingdom. Cases were
referred to these clinics for assessment, diagnosis and management
by their primary care physician, or in the case of those with VAD
following a stroke, they may have been referred to a Psychiatrist by
their stroke physician. We recruited a healthy control group of
comparable age by local advertisement. Ethics statement This study was approved by the Joint Ethics Committee of
Newcastle and North Tyneside Health Authority, the University of
Newcastle upon Tyne and the University of Northumbria at
Newcastle and participants gave written informed consent in
accordance with the declaration of Helsinki. PLoS ONE | www.plosone.org May 2009 | Volume 4 | Issue 5 | e5521 Incidence of falls The incidence of falls in dementia was 9118 per 1000 person years,
which was significantly higher than in controls (1023; incidence
density ratio (IDR) adjusted for age and sex: 7.58, 95% CI: 3.11–
18.5), (table 2). If participants living in care homes were excluded, the
incidence of falls in dementia was 8763 which remained significantly
higher than in controls (IDR: 9.56, 95% CI: 6.07–15.1). With respect to dementia subtypes, the incidence of falls,
adjusted for age and sex, was higher in all dementia subtypes than
in controls (table 2): AD: 2486 falls/1000 person years (IDR: 1.95;
95% CI: 1.01–3.78), VAD: 3135 (IDR: 1.77, 95% CI: 1.17–2.69),
DLB: 9087 (IDR: 6.06, 95% CI: 3.53–10.4) and PDD: 19000
(IDR: 20.5, 95% CI: 10.4–40.2). The incidence of falls in PDD
was higher than in AD (IDR: 10.5, 95% CI: 3.32–33.1), VAD
(IDR: 11.6, 95% CI: 5.73–23.3), and DLB (IDR: 3.38, 95% CI:
2.66–4.31). The incidence of falls in DLB was also higher in AD
(IDR: 3.10, 95% CI: 1.16–8.28) and VAD (IDR: 3.41, 95% CI:
1.96–5.95). The incidence of falls in VAD was not significantly
higher than in AD (IDR: 0.907, 95% CI: 0.504–1.63). Following identification in univariate models, significant and
potentially modifiable risk factors were entered into a multivariate
forward stepwise Cox regression model, p 0.05 for entry, p 0.1 for
removal. Age and sex were included even if not significant. Where
similar clinical features were described by more than one significant
risk factor the factor with the higher level of significance in
univariate analyses was entered into multivariate analyses, in order
to avoid co-aggregation of predictors; e.g. abnormal pull test rather
than the full Tinetti balance scale as both these tests assess balance. Prevalence of falls The incidence density of falls in each diagnostic group was
calculated by the total number of falls in each group per number of
person years of diaries returned, expressed as number of falls per
1000 person years. Loglinear Poisson regression models (adjusted
for age and sex) were used to obtain an incidence density ratio for
dementia and its subtypes in comparison with healthy controls. AD, VAD and DLB were also used as reference groups to examine
the effect of dementia subtype upon incidence of falls. 81.6% of diaries were returned (82.9% in those participants
with dementia). During the 12 month follow up period 65.7% of
participants with dementia had at least one fall, compared with
35.9% of controls (relative hazard ratio (HR) adjusted for age and
sex: 3.03, 95% confidence intervals (CI) 1.71–5.35). With respect
to dementia subtypes, the prevalence of falls in AD was 47%; VAD
47%; DLB 77% and PDD 90%. Figure 2 shows the survival curves
until the occurrence of a fall for each diagnosis. In order to examine the associations between exposure to putative
risk factors for falling and the occurrence of falls in those participants
with dementia, Cox regression was used to obtain univariate
proportional hazard ratios for each risk factor, adjusted for age and
sex, using time to the occurrence of at least one fall as the dependent
variable. Hazard ratios were given according to presence or absence
of the risk factor, or per point on quantitative scales as appropriate. Analyses were performed initially for all participants with dementia,
and then repeated stratified by diagnosis. Baseline Clinical Assessment All participants received a detailed baseline assessment to
quantify putative risk factors for falls. The factors included were
selected on the basis of their previous identification as risk factors PLoS ONE | www.plosone.org May 2009 | Volume 4 | Issue 5 | e5521 May 2009 | Volume 4 | Issue 5 | e5521 2 Falls in Dementia Student’s t test for normally distributed data and Mann-Whitney
U test for non-normally distributed data. to take part (39 controls, 38 AD, 32 VAD, 30 DLB and 40 PDD). 116 (83%) of 140 participants with dementia and all of the controls
were residing in the community. A summary of baseline
characteristics of the participants is shown in table 1. to take part (39 controls, 38 AD, 32 VAD, 30 DLB and 40 PDD). 116 (83%) of 140 participants with dementia and all of the controls
were residing in the community. A summary of baseline
characteristics of the participants is shown in table 1. Student’s t test for normally distributed data and Mann-Whitney
U test for non-normally distributed data. Due to censoring of falls data from some participants (as a result of
death or withdrawal from the study), Cox regression (adjusted for age
and sex) was used to obtain a proportional hazard ratio for time to
first fall in dementia, using healthy controls as the reference group. Participants 289 patients were considered for inclusion; 65 participants were
excluded (Figure 1). 179 (80%) of 224 eligible participants agreed Significant modifiable and non-modifiable predictors of falls in all
participants with dementia are shown in table 3. Significant predic- Figure 1. Flow chart to show recruitment of patients to study. doi:10.1371/journal.pone.0005521.g001
PLoS ONE | www.plosone.org
3
May 2009 | Volume 4 | Issue 5 | e5521 Figure 1. Flow chart to show recruitment of patients to study. doi:10.1371/journal.pone.0005521.g001 PLoS ONE | www.plosone.org May 2009 | Volume 4 | Issue 5 | e5521 PLoS ONE | www.plosone.org 3 Falls in Dementia Table 1. Baseline characteristics by diagnosis: all participants Diagnosis (n)
Control (39)
AD (38)
VAD (32)
DLB (30)
PDD (40)
Mean Age (SD)*
75 (6.4)
79 (5.8)
79 (6.2)
76 (7.1)
72 (6.0)
Gender: male (%)
21 (53.8)
18 (47.4)
23 (71.9)
18 (60.0)
26 (65.0)
Median duration of dementia: months (IQR)
-
36 (21–48)
18 (9–30)
24 (12–48)
24 (15–48)
Mean CAMCOG score (SD)**
94 (4.7)
59 (15)
64 (18)
59 (15)
64 (16)
Resident in care home n (%)
0 (0)
6 (15.8)
6 (18.8)
9 (30)
3 (7.5)
History of falls in previous 12 months n (%) ***
13 (33.3)
19 (51.4)
21 (65.6)
18 (69.2)
33 (86.8)
History of recurrent falls in previous 12 months n (%) {
2 (5.1)
10 (26.3)
17 (53.1)
12 (40.0)
29 (72.5)
Uses walking aid or requires assistance to walk n (%) {{
3 (7.7)
5 (13.2)
9 (28.1)
14 (46.7)
27 (67.5)
Cardiovascular medication n (%) {{{
19 (48.7)
27 (71.1)
28 (87.5)
20 (66.7)
39 (97.5)
Psychotropic medication n (%) {
5 (12.8)
11 (28.9)
19 (59.4)
15 (50.0)
20 (50.0)
Symptomatic orthostatic hypotension n (%) {{
0/39 (0)
2/37 (5.4)
3/31 (9.7)
4/26 (15.4)
12/38 (31.6)
SD: Standard deviation; IQR: Inter-quartile range; AD: Alzheimer’s disease; VAD: Vascular dementia; DLB: Dementia with Lewy bodies; PDD: Parkinson’s disease dementia;
PD: Parkinson’s disease. Denominators are given for prevalence (%) where data is incomplete. *Control vs. patients p = 0.457; Control vs. AD p = 0.036; control vs. VAD p = 0.027; Control vs. DLB p = 0.945; Control vs. PDD p = 0.031. **Control vs. all patient groups p,0.001. No significant differences between patient groups. ***Control vs. AD p = 0.087; control vs. VAD p = 0.009; Control vs. DLB p = 0.005; Control vs. PDD p,0.001
{Control vs. Participants AD group
1
0.907 (0.504–1.63)
3.10 (1.16–8.28)
10.5 (3.32–33.1)
Incidence density ratio (95% CI) c.f. VAD group
1
3.41 (1.96–5.95)
11.6 (5.73–23.3
Incidence density ratio (95% CI) c.f. DLB group
1
3.38 (2.66–4.31)
Fractures: number of fractures recorded during study
0
1
1
2
3
doi:10.1371/journal.pone.0005521.t002 doi:10.1371/journal.pone.0005521.t002 symptom score and symptomatic orthostatic hypotension. In the
second model, stratified by diagnosis, predictors retained were
symptomatic orthostatic hypotension, use of cardioactive medica-
tion and physical activity score, which was protective (Table 4). preceding 12 months, use of cardioactive medication, autonomic
symptom scale greater than 7 and time taken for blood pressure to
return to baseline on standing remained significant predictors of
falls. Increased physical activity remained protective. Multivariate predictors of falls and recurrent falls in
dementia In the largest prospective study of predictors of falls in dementia
to date, we have demonstrated that older people with dementia
experience 8 times more incident falls than those without
dementia. These figures are even more striking when only
community dwelling people with dementia are considered, with
incidence in people with dementia nearly 10 times higher than in
those without dementia. Patients with Lewy Body dementias (DLB
or PDD) were at the highest risk, with DLB patients sustaining 6
times the number of falls in the control group and PDD 20 times Significant potentially modifiable predictors were entered into
multivariate analyses in the order: age, gender, Tinetti gait score,
Cornell depression score, physical activity score, autonomic
symptom score, symptomatic orthostatic hypotension, use of
cardioactive medication and time for systolic blood pressure to
return to baseline on standing. In the first model including all participants with dementia,
predictors retained were Cornell depression score, total autonomic Table 3. Univariate predictors of falls in participants with dementia
Diagnosis
n
All participants with dementia
Stratified by diagnosis
Relative
Hazard ratio
95% confidence
intervals
Relative
Hazard ratio
95% confidence
intervals
Non modifiable risk factors
Age (years)
140
0.969
0.940–0.999
0.999
0.968–1.03
Male Gender
140
1.08
0.710–1.65
1.04
0.679–1.60
Diagnosis of Lewy Body disorder (PDD or DLB)
140
3.33
2.11–5.26
-
-
Duration of dementia (per month)
133
1.01
0.998–1.02
1.02
1.00–1.02
Resident in care home
140
1.57
0.814–3.02
1.76
0.878–3.55
History of falls in previous 12 months
140
2.52
1.52–4.17
2.04
1.21–3.46
History of recurrent falls in previous 12 months
140
2.79
1.82–4.29
2.28
1.44–3.63
CAMCOG score (0–105, per point)
131
0.999
0.984–1.01
1
0.984–1.02
Potentially modifiable risk factors
Cardioactive medication
140
2.08
1.15–3.75
1.91
1.03–3.54
Psychotropic medication
140
1.49
0.986–2.25
1.36
0.885–2.07
Tinetti gait score ,7 and/or Tinetti balance score ,22
139
2.12
1.39–3.24
1.14
0.678–1.90
Physical activity score (0–9, per point)
140
0.818
0.723–0.926
0.873
0.766–0.994
NPI aberrant motor behaviour sub-score (0–12, per point)
121
0.989
0.922–1.06
0.982
0.916–1.05
Cornell score $10
132
2.01
1.18–3.43
1.26
0.728–2.18
Total autonomic symptom score $7
139
2.27
1.49–3.45
1.64
1.02–2.63
Definite, atypical or severe autonomic neuropathy
93
2.01
1.16–3.47
1.24
0.688–2.24
Symptomatic orthostatic hypotension
132
2.07
1.19–3.61
1.47
0.823–2.62
Time taken for blood pressure to return to baseline on
standing (per second)
133
1
1.00–1.01
1
1.00–1.01
doi:10 1371/journal pone 0005521 t003 Table 3. Univariate predictors of falls in participants with dementia Table 3. Participants AD p = 0.013; control vs. VAD p,0.001; Control vs. DLB p = 0.001; Control vs. PDD p,0.001
{{Control vs. AD p = 0.481; control vs. VAD p,0.024; Control vs. DLB p,0.001; Control vs. PDD p,0.001
{{{Control vs. AD p = 0.063; control vs. VAD p = 0.001; Control vs. DLB p = 0.151; Control vs. PDD p,0.001
{Control vs. AD p = 0.098; control vs. VAD p,0.001; Control vs. DLB p = 0.001; Control vs. PDD p = 0.001
{{Control vs. AD p = 0.234; control vs. VAD p,0.082; Control vs. DLB p,0.022; Control vs. PDD p,0.001
doi:10.1371/journal.pone.0005521.t001 Control (39)
AD (38)
VAD (32)
DLB (30)
PDD (40) matic OH and time taken for blood pressure to return to baseline on
standing. Age and increased physical activity were protective. tors included diagnosis of Lewy body disorder, history of falls or
recurrent falls in the preceding 12 months, use of cardioactive
medication, abnormal gait or balance score, Cornell depression score
$10, autonomic symptom scale .7, autonomic neuropathy, sympto- However, when univariate analyses were stratified by diagnosis,
only duration of dementia, history of falls or recurrent falls in the Figure 2. Survival curve showing time to first fall by diagnosis. doi:10.1371/journal.pone.0005521.g002
PLoS ONE | www.plosone.org
4
May 2009 | Volume 4 | Issue 5 | e5521 Figure 2. Survival curve showing time to first fall by diagnosis. doi:10.1371/journal.pone.0005521.g002 PLoS ONE | www.plosone.org May 2009 | Volume 4 | Issue 5 | e5521 PLoS ONE | www.plosone.org 4 Falls in Dementia Table 2. Annual incidence of falls Control (39)
AD (38)
VAD (32)
DLB (30)
PDD (40)
Incidence: number of falls/1000 person years
1023
2486
3135
9087
19 000
Incidence density ratio (95% CI) c.f. control group
1
1.95 (1.01–3.78)
1.77 (1.17–2.69)
6.06 (3.53–10.4)
20.5 (10.4–40.2)
Incidence density ratio (95% CI) c.f. AD group
1
0.907 (0.504–1.63)
3.10 (1.16–8.28)
10.5 (3.32–33.1)
Incidence density ratio (95% CI) c.f. VAD group
1
3.41 (1.96–5.95)
11.6 (5.73–23.3
Incidence density ratio (95% CI) c.f. DLB group
1
3.38 (2.66–4.31)
Fractures: number of fractures recorded during study
0
1
1
2
3
doi:10.1371/journal.pone.0005521.t002 Control (39)
AD (38)
VAD (32)
DLB (30)
PDD (40)
Incidence: number of falls/1000 person years
1023
2486
3135
9087
19 000
Incidence density ratio (95% CI) c.f. control group
1
1.95 (1.01–3.78)
1.77 (1.17–2.69)
6.06 (3.53–10.4)
20.5 (10.4–40.2)
Incidence density ratio (95% CI) c.f. doi:10.1371/journal.pone.0005521.t003 Multivariate predictors of falls and recurrent falls in
dementia These factors will be useful in identifying
individuals at particular risk, who may benefit from further
assessment and intervention. Even more importantly, a number of
the predictors identified are potentially modifiable, and should be
included as key elements of a multifactorial intervention. These
factors included use of cardioactive medications, autonomic
symptoms, symptomatic orthostatic hypotension, depression and
limitation of physical activity. We suggest that interventions
targeted towards these predictors could reduce the burden of falls
related morbidity and mortality in community dwelling people
with mild-moderate dementia. Our study has a number of strengths which reinforce our
findings. This was a prospective study and the majority of
participants had mild-moderate dementia and were residing in the
community. The incidence of falls was determined by completion
of daily diaries, with the support of a caregiver where appropriate,
with an adequate follow up period. The use of this method is
particularly important because it has the highest sensitivity for
accurate recording of falls. The number of participants in our
study was high (179); 80% of those approached agreed to take part
and compliance with fall diary returns was also high. Statistical
analysis was strengthened by the use of Cox regression models and
loglinear analysis of incidence densities, thus taking account of
censoring of falls diaries in some participants. All analyses included
adjustment for age and gender and the inclusion of multivariate
analyses enabled efficient estimate of significant associations while
adjusting for a number of confounding factors simultaneously. Prior to commencement of our study we identified only two
studies of falls in dementia which included a fully multifactorial
assessment of risk factors, the first of which had an inadequate
follow up period,[34] and in the second the primary outcome was
fall related serious injuries rather than falls.[35] Both of those
studies included fewer than 100 participants, and neither included
a control group. Two studies published since we commenced our
study have included multifactorial baseline assessments, but the
first of these was small (only 42 participants with advanced
AD).[36] The second was a high quality study of 124 participants
in which neuroleptic drug use and grade 2 white matter lesions
predicted falls in multivariate analyses, but only participants with
AD were included, and there was no control group[37]. Multivariate predictors of falls and recurrent falls in
dementia Univariate predictors of falls in participants with dementia May 2009 | Volume 4 | Issue 5 | e5521 5 Falls in Dementia Table 4. Multivariate predictors of falls in participants with dementia p
p
p
Diagnosis
All participants with dementia
Stratified by diagnosis
Relative Hazard
ratio
95% confidence
intervals
Relative Hazard
ratio
95% confidence
intervals
Predictors of falls
Cornell depression score (0–40, per point)
1.05
1.01–1.10
Total Autonomic symptom score (0–36, per point)
1.05
1.01–1.10
Symptomatic orthostatic hypotension
2.13
1.19–3.80
2.2
1.19–4.06
Physical activity score (0–9, per point)
0.827
0.716–0.956
Use of cardioactive medication
1.98
0.994–3.96
doi:10.1371/journal.pone.0005521.t004 All participants with dementia doi:10.1371/journal.pone.0005521.t004 biased the sample towards those with more progressive disease. However, in the northern region of the UK, suspected cases of
Parkinson’s disease are routinely referred to a secondary care
physician specialising in movement disorders (either a Neurologist
or Geriatrician). These patients should, therefore, be reasonably
representative of all patients presenting to primary care physicians
with symptoms suggestive of PD. There is a greater likelihood of
patients with AD or VAD not being referred to secondary care as
dementia is often undetected, or not managed by a specialist in
memory disorders. The reader should therefore bear in mind that
participants in this study may have had more severe or progressive
disease than those generally seen in primary care. In interpreting
our results, consideration should be given to the need to stratify for
dementia subtypes in univariate and multivariate analyses. Some
of the predictors were not significant when analyses were stratified
by dementia subtype, probably because these predictors were
actually surrogate markers of a diagnosis of a Lewy body disorder,
which itself predicted falls. In comparison with the proportions in
the general population the LB dementias were over represented in
our study, in order to make valid comparisons between dementia
subtypes. more falls. The annual incidence of falls was higher in the LB
dementias than in all other groups studied and much higher than
any previous reports in older people[33]. Incidence of falls was
higher in PDD than in DLB. To our knowledge, this is the first study which has identified
predictors specific to dementia, including the identification of non-
modifiable predictors such as a diagnosis of Lewy body disorder,
longer duration of dementia and previous history of falls or
recurrent falls. PLoS ONE | www.plosone.org Multivariate predictors of falls and recurrent falls in
dementia The
design of our study addressed these issues; to our knowledge, it is
the first to examine falls in both AD and non-AD dementias over
an adequate follow up period. There are some potential limitations to this study. Ideally, we
would have recruited both groups using random sampling from a
community population. Unfortunately the scale of such an exercise
would have been well beyond the resources available for this study. Healthy older people volunteering for participation in research in
response to advertisements are often fitter than the older
population at large, but interestingly the incidence of falls in our
control group was very similar to that found in most community
based studies, suggesting that the control group was reasonably
representative, despite the limitations of using this recruitment
method. The recruitment of the patient participants from
secondary care clinics rather than from primary care may have Exclusive to our study was the phasic measurement of blood
pressure using beat-to-beat recordings, thus enabling detection of
both magnitude and timing of blood pressure changes. Symptom-
atic OH and the length of time blood pressure fell below baseline
rather than the magnitude of the drop were predictive of falls. Our
group has previously shown an association between intermittent
hypotension and white matter lesions,[38] and this resonates with
the findings of Horikawa et al. that grade 2 white matter lesions
predicted falls in multivariate analyses[37]. White matter lesions
can be associated with abnormal gait,[39] and it is possible that PLoS ONE | www.plosone.org May 2009 | Volume 4 | Issue 5 | e5521 May 2009 | Volume 4 | Issue 5 | e5521 6 Falls in Dementia approach may increase opportunities for falling in individuals at
risk; similarly, aggressive treatment of motor features in LB
dementias might increase activity related opportunities to fall, and
also exacerbate OH. There is also a possibility that changes in
psychotropic medication might result in side effects such as
hypotension or somnolence, which could paradoxically increase
the risk of falls. This emphasises the importance of the conduction
of randomised controlled trials to ensure that modification of the
risk factors identified is the correct strategy. intermittent symptomatic hypotension against a background of
gait and balance instability exacerbates the likelihood that an older
person will fall. A higher level of physical activity was a protective
factor in our study. References van Doorn C, Gruber-Baldini AL, Zimmerman S, Hebel JR, Port CL, et al. (2003) Dementia as a risk factor for falls and fall injuries among nursing home
residents. Journal of the American Geriatrics Society 51: 1213–1218. 24. Brierley EJ, Johnson MA, James OF, Turnbull DM (1996) Effects of physical activity
and age on mitochondrial function. Quarterly Journal of Medicine 89: 251–
258. 8. Shaw FE (2002) Falls in cognitive impairment and dementia. Clinics in Geriatric
Medicine 18: 159–173. 9. National Institute for Health and Clinical Excellence (2004) Clinical practice
guideline for the assessment and prevention of falls in older people. London:
Royal College of Nursing. 25. Tinetti ME (1986) Performance-oriented assessment of mobility problems in
elderly patients. Journal of the American Geriatrics Society 34: 119–126. 10. Wood B, Walker R (2001) Parkinson’s disease: characteristics of fallers and non-
fallers. Age & Ageing 30: 423–424. 26. Fahn S, Elton R, Unified Parkinson’s disease rating scale committee (1987) In:
Fahn S, Marsden C, Calne D, eds. Recent developments in Parkinson’s disease. New York: Macmillan. pp 153–163. 11. Bueno-Cavanillas A, Padilla-Ruiz F, Jimenez-Moleon JJ, Peinado-Alonso CA,
Galvez-Vargas R (2000) Risk factors in falls among the elderly according to
extrinsic and intrinsic precipitating causes. European Journal of Epidemiology
16: 849–859. 27. Bucks RS, Ashworth DL, Wilcock GK, Siegfried K (1996) Assessment of
activities of daily living in dementia: development of the Bristol Activities of
Daily Living Scale. Age & Ageing 25: 113–120. 28. Alexopoulos GS, Abrams RC, Young RC, Shamoian CA (1988) Use of the
Cornell scale in non demented patients. Journal of the American Geriatrics
Society 36: 230–236. 12. Chang JT, Morton SC, Rubenstein LZ, Mojica WA, Maglione M, et al. (2004)
Interventions for the prevention of falls in older adults: systematic review and
meta-analysis of randomised clinical trials. BMJ 328: 680. y
29. Alexopoulos GS, Abrams RC, Young RC, Shamoian CA (1988) Cornell Scale
for Depression in Dementia. Biological Psychiatry 23: 271–284. 13. McKeith IG, Dickson DW, Lowe J, Emre M, O’Brien JT, et al. (2005) Diagnosis
and management of dementia with Lewy bodies: third report of the DLB
Consortium. Neurology 65: 1863–1872. 30. Cummings JL (1997) The Neuropsychiatric Inventory: assessing psychopathol-
ogy in dementia patients. Neurology 48: S10–16. 14. Shaw FE, Bond J, Richardson DA, Dawson P, Steen IN, et al. References 1. Ferri CP, Prince M, Brayne C, Brodaty H, Fratiglioni L, et al. (2005) Global
prevalence of dementia: a Delphi consensus study. Lancet 366: 2112–2117. 17. Roman GC, Tatemichi TK, Erkinjutti T (1993) Vascular dementia - diagnostic
criteria for research studies; report of the NINDS AIREN International
Workshop. Neurology 43: 250–260. 2. McKeith IG, Galasko D, Kosaka K, Perry EK, Dickson DW, et al. (1996)
Consensus guidelines for the clinical and pathologic diagnosis of dementia with
Lewy bodies (DLB): report of the consortium on DLB international workshop. Neurology 47: 1113–1124. 18. McKeith IG, Perry EK, Perry RH (1999) Report of the second dementia with
Lewy body international workshop: diagnosis and treatment. Consortium on
Dementia with Lewy Bodies. Neurology 53: 902–905. y
gy
19. Gibb WRG, Esiri MM, Lees AJ (1987) Clinical and pathological features of
diffuse Lewy body disease (lewy body dementia). Brain 110: 1131–1153. 3. Holmes C, Cairns N, Lantos P, Mann A (1999) Validity of current clinical
criteria for Alzheimer’s disease, vascular dementia and dementia with Lewy
bodies. British Journal of Psychiatry 174: 45–50. 20. McKeith IG, Ballard CG, Perry RH (1998) Predictive accuracy of clinical
diagnostic criteria for dementia with Lewy bodies - a prospective neuropath-
ological validation study. Neurology 50: 181. 4. Tinetti ME, Inouye SK, Gill TM, Doucette JT (1995) Shared risk factors for
falls, incontinence, and functional dependence. Unifying the approach to
geriatric syndromes. Journal of the American Medical Association 273:
1348–1353. 21. McKeith IG, Ballard CG, Perry RH, Ince PG, O’Brien JT, et al. (2000)
Prospective validation of consensus criteria for the diagnosis of dementia with
Lewy bodies. Neurology 54: 1050–1058. 5. Tinetti ME, Williams CS (1997) Falls, injuries due to falls, and the risk of
admission to a nursing home. New England Journal of Medicine 337: 1279–1284. y
gy
22. Folstein MF, Folstein SE, McHugh PR (1975) ‘‘Mini-mental state’’. A practical
method for grading the cognitive state of patients for the clinician. Journal of
Psychiatric Research 12: 189–198. 6. Morris JC, Rubin EH, Morris EJ, Mandel SA (1987) Senile dementia of the
Alzheimer’s type: an important risk factor for serious falls. Journal of
Gerontology 42: 412–417. 23. Roth M, Tym E, Mountjoy CQ, Huppert FA, Hendrie H, et al. (1986)
CAMDEX. A standardised instrument for the diagnosis of mental disorder in
the elderly with special reference to the early detection of dementia. British
Journal of Psychiatry 149: 698–709. 7. Multivariate predictors of falls and recurrent falls in
dementia The physical activity scale measured all
physical activity rather than wandering, which has been shown in
one study to be a predictor of falls in dementia.[40] It is likely that
the more active patients were those with fewer problems with
mobility,
whereas
the
participants
with
disturbed
mobility
restricted their activity because of fear of falling, thus having
fewer opportunities to fall. We conclude that whilst the outcome of future trials are awaited
best clinical practice should focus on identification and manage-
ment of orthostatic hypotension, depression and maintenance of
physical activity in individuals who do not have severely impaired
gait and balance, whilst bearing in mind the need to monitor
patients carefully because of the potential side effects of these
changes. We believe that randomised multifactorial intervention trials to
prevent falls in mild-moderate dementia should now be made a
priority. Possible management strategies could include manage-
ment of the potentially modifiable factors identified in this study;
for example, the use of selective serotonin reuptake inhibitors for
depression, manipulation of cardiovascular medications, adequate
hydration and targeted drug therapies such as fludrocortisone and
midodrine for OH. Such a focus would differ from multifactorial
interventions in older people without dementia, which prioritise
strength and balance exercises that are more difficult for those
with impairment of recall to continue following initial intervention. It is possible that encouragement of overall physical activity may
be successful in prevention of falls in dementia. However, such an Author Contributions Conceived and designed the experiments: LMA CB RAK. Performed the
experiments: LMA ENR. Analyzed the data: LMA CB. Contributed
reagents/materials/analysis tools: ENR. Wrote the paper: LMA. Reviewed
the analysis: CB RAK. Reviewed the paper: CB ENR RAK. Conceived and designed the experiments: LMA CB RAK. Performed the
experiments: LMA ENR. Analyzed the data: LMA CB. Contributed
reagents/materials/analysis tools: ENR. Wrote the paper: LMA. Reviewed
the analysis: CB RAK. Reviewed the paper: CB ENR RAK. 35. Asada T, Kariya T, Kinoshita T, Asaka A, Morikawa S, et al. (1996) Predictors
of fall-related injuries among community-dwelling elderly people with dementia.
Age & Ageing 25: 22–28. p
p
y
38. Kenny RA, Shaw FE, O’Brien JT, Scheltens PH, Kalaria R, et al. (2004)
Carotid sinus syndrome is common in dementia with Lewy bodies and correlates g
g
g
36. Camicioli R, Licis L (2004) Motor impairment predicts falls in specialized
Alzheimer care units. Alzheimer Disease & Associated Disorders 18: 214–218. 37. Horikawa E, Matsui T, Arai H, Seki T, Iwasaki K, et al. (2005) Risk of falls in
Alzheimer’s disease: a prospective study. Internal Medicine 44: 717–721. with deep white matter lesions. Journal of Neurology, Neurosurgery &
Psychiatry 75: 966–971. 39. Baloh RW, Ying SH, Jacobson KM (2003) A longitudinal study of gait and
balance dysfunction in normal older people. Archives of Neurology 60: 835–839. with deep white matter lesions. Journal of Neurology, Neurosurgery &
Psychiatry 75: 966–971.
39. Baloh RW, Ying SH, Jacobson KM (2003) A longitudinal study of gait and
balance dysfunction in normal older people. Archives of Neurology 60: 835–839.
40. Katz IR, Rupnow M, Kozma C, Schneider L (2004) Risperidone and falls in
ambulatory nursing home residents with dementia and psychosis or agitation:
secondary analysis of a double-blind, placebo-controlled trial. American Journal
of Geriatric Psychiatry 12: 499–508. 40. Katz IR, Rupnow M, Kozma C, Schneider L (2004) Risperidone and falls in
ambulatory nursing home residents with dementia and psychosis or agitation:
secondary analysis of a double-blind, placebo-controlled trial. American Journal
of Geriatric Psychiatry 12: 499–508. with deep white matter lesions. Journal of Neurology, Neurosurgery &
Psychiatry 75: 966–971. References (2003)
Multifactorial intervention after a fall in older people with cognitive impairment
and dementia presenting to the accident and emergency department:
randomised controlled trial. BMJ 326: 73. 31. Allan LM, Ballard CG, Allen J, Murray A, Davidson AW, et al. (2007)
Autonomic dysfunction in dementia. J Neurol Neurosurg Psychiatry 78:
671–677. 32. Ewing DJ, Martyn CN, Young RJ, Clarke BF (1985) The value of cardiovascular
autonomic function tests: 10 years experience in diabetes. Diabetes Care 8:
491–498. 15. Oliver D, Connelly J, Victor C, Shaw F, Whitehead A, et al. (2006) Strategies to
prevent falls and fractures in hospitals and care homes and effect of cognitive
impairment: systematic review and meta-analyses BMJ; doi:10.1136/
bmj.39049.706493.55. 33. Rubenstein LZ, Josephson KR (2002) The epidemiology of falls and syncope. Clinics in Geriatric Medicine 18: 141–158. 16. McKhann G, Drachman D, Folstein M (1984) Clinical diagnosis of Alzheimer’s
disease. Report of the NINCDS ADRDA work group under the auspices of the
Department of Health and Human Services Task forces on Alzheimer’s disease. Neurology 34: 939–944. 34. Ballard CG, Shaw F, Lowery K, McKeith I, Kenny R (1999) The prevalence,
assessment and associations of falls in dementia with Lewy bodies and
Alzheimer’s disease. Dementia & Geriatric Cognitive Disorders 10: 97–103. PLoS ONE | www.plosone.org 7 May 2009 | Volume 4 | Issue 5 | e5521 May 2009 | Volume 4 | Issue 5 | e5521 Falls in Dementia PLoS ONE | www.plosone.org May 2009 | Volume 4 | Issue 5 | e5521 8
|
https://openalex.org/W4292449066
|
https://www.frontiersin.org/articles/10.3389/fbioe.2022.857975/pdf
|
English
| null |
The reliability and validity of gait analysis system using 3D markerless pose estimation algorithms
|
Frontiers in bioengineering and biotechnology
| 2,022
|
cc-by
| 6,929
|
The reliability and validity of gait
analysis system using 3D
markerless pose estimation
algorithms OPEN ACCESS
EDITED BY
Rezaul Begg,
Victoria University, Australia
REVIEWED BY
Susanna Summa,
Bambino Gesù Children’s Hospital
(IRCCS), Italy
João Paulo Morais Ferreira,
Superior Institute of Engineering of
Coimbra (ISEC), Portugal
*CORRESPONDENCE
Guoru Zhao,
gr.zhao@siat.ac.cn
SPECIALTY SECTION
This article was submitted
to Biomechanics,
a section of the journal
Frontiers in Bioengineering
and Biotechnology
RECEIVED 19 January 2022
ACCEPTED 13 July 2022
PUBLISHED 10 August 2022
CITATION Shengyun Liang1,2,3, Yu Zhang1,2,3, Yanan Diao1,2,3, Guanglin Li 1,3
and Guoru Zhao1,3* 1CAS Key Laboratory of Human-Machine Intelligence-Synergy Systems, Research Center for Neural
Engineering, Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences, Shenzhen,
China, 2Shenzhen College of Advanced Technology, University of Chinese Academy of Sciences,
Shenzhen, China, 3Guangdong-Hong Kong-Macao Joint Laboratory of Human-Machine Intelligence-
Synergy Systems, Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences,
Shenzhen, China Quantifying kinematic gait for elderly people is a key factor for consideration in
evaluating their overall health. However, gait analysis is often performed in the
laboratory using optical sensors combined with reflective markers, which may
delay the detection of health problems. This study aims to develop a 3D
markerless
pose
estimation
system
using
OpenPose
and
3DPoseNet algorithms. Moreover, 30 participants performed a walking task. Sample entropy was adopted to study dynamic signal irregularity degree for gait
parameters. Paired-sample t-test and intra-class correlation coefficients were
used to assess validity and reliability. Furthermore, the agreement between the
data obtained by markerless and marker-based measurements was assessed by
Bland–Altman analysis. ICC (C, 1) indicated the test–retest reliability within
systems was in almost complete agreement. There were no significant
differences between the sample entropy of knee angle and joint angles of
the sagittal plane by the comparisons of joint angle results extracted from
different systems (p > 0.05). ICC (A, 1) indicated the validity was substantial. This
is supported by the Bland–Altman plot of the joint angles at maximum flexion. Optical
motion
capture
and
single-camera
sensors
were
collected
simultaneously, making it feasible to capture stride-to-stride variability. In
addition, the sample entropy of angles was close to the ground_truth in the
sagittal plane, indicating that our video analysis could be used as a quantitative
assessment of gait, making outdoor applications feasible. CITATION
Liang S, Zhang Y, Diao Y, Li G and Zhao G
(2022), The reliability and validity of gait
analysis system using 3D markerless
pose estimation algorithms. Front. Bioeng. Biotechnol. 10:857975. TYPE Original Research
PUBLISHED 10 August 2022
DOI 10.3389/fbioe.2022.857975 TYPE Original Research
PUBLISHED 10 August 2022
DOI 10.3389/fbioe.2022.857975 TYPE Original Research
PUBLISHED 10 August 2022
DOI 10.3389/fbioe.2022.857975 single-camera video, 3D markerless pose estimates, 3D marker-based motion analysis,
validity, reliability KEYWORDS
single-camera video, 3D markerless pose estimates, 3D marker-based motion analysis,
validity, reliability The reliability and validity of gait
analysis system using 3D
markerless pose estimation
algorithms doi: 10.3389/fbioe.2022.857975 COPYRIGHT
© 2022 Liang, Zhang, Diao, Li and Zhao. This is an open-access article
distributed under the terms of the
Creative Commons Attribution License
(CC BY). The use, distribution or
reproduction in other forums is
permitted, provided the original
author(s) and the copyright owner(s) are
credited and that the original
publication in this journal is cited, in
accordance with accepted academic
practice. No use, distribution or
reproduction is permitted which does
not comply with these terms. 01 Frontiers in Bioengineering and Biotechnology Frontiers in Bioengineering and Biotechnology frontiersin.org Liang et al. 10.3389/fbioe.2022.857975 10.3389/fbioe.2022.857975 Introduction human pose from video to measure gait parameters. Then, we
clarify associations and agreements of motion analysis using
markerless and marker-based systems and confirm the reliability
and validity of 3D human pose from video. 2) Dataset: we have
collected the synchronized motion capture cameras and a single-
camera video dataset of movement sequences for elderly and
young people. 3) Application: our video-based gait analysis
workflow is freely available, involves minimal user input, and
does not require prior gait analysis expertise. Gait parameters have been proposed as an index of overall
gait pathology for elderly people. It uses kinematic and kinetic
variables to be taken as a cleaner reflection of gait quality (Brach
et al., 2001). Kinematic analysis is often performed in laboratory
research using three-dimensional motion capture or wearable
sensors, which are expensive, immobile, data-limited, and
require expertise (Cronin, 2021). Recently, video-based pose
estimation suggests the potential for analyzing gait kinematic
parameters (Andriluka et al., 2014). Methods Video-based 2D body pose estimation is a well-studied
problem in computer vision, with state-of-the-art methods
being based on deep networks (Toshev and Szegedy, 2014;
Fang et al., 2017; Güler et al., 2018). With the advent of deep
neural networks, it is now possible to estimate joint angles
without the need for reflective markers. One of the more
popular approaches, CMU’s OpenPose enables key body
landmarks to be tracked from multiple humans in a video in
real-time (Cao et al., 2017). Yagi et al. (2020) used OpenPose to
detect multiple individuals and their joints in images to estimate
step positions, stride length, step width, walking speed, and
cadence, in comparison with multiple infrared camera motion
capture system OptiTrack (Lénárt et al., 2018). Kidziński et al. (2020) designed machine learning models (e.g., convolutional
neural networks, random forest, and ridge regression models) to
predict clinical gait metrics based on trajectories of 2D body
poses extracted from videos using OpenPose. Similarly, Stenum
et al. (2021) used OpenPose to compare spatiotemporal and
sagittal kinematic gait parameters of healthy adults against
recorded optical marker–based motion captured from walking
simultaneously. These previous studies have been performed in
comparison between markerless and marker-based methods;
however, they only learned to infer joint angles or joint
locations in the sagittal plane. In the laboratory workflow, participants were marked with a
total of 39 markers placed on bony landmarks. The Body39 joints
were labeled by the Plug-In Gait full body model in Nexus
software. Then, participants were required to perform the
clear step on the treadmill, keeping 3 s and repeating three
times. The positions of these markers were tracked by several
optical cameras, which were later reconstructed into 3D position
time series. Measures derived from the Vicon data served as
ground_truth labels. In our proposed workflow, we used a
camera to record the participant’s movement. The open-
source OpenPose algorithm was adopted to extract trajectories
of 2D key points from continuous images. The 2D joint positions
from OpenPose were used as input, and we adopted a 3D
keypoint detection algorithm (3DPoseNet) to estimate body
joint locations in 3-dimensional space. It was important to
note
that
these
two
workflows
were
synchronized
by
hardware. Finally, these signals from two workflows were
converted to joint angles as a function of time. Frontiers in Bioengineering and Biotechnology Experimental setting Figure1showedthelaboratoryenvironmentandsetup,allowing
ustocapturedatafromsevensensors(sixViconMXmotioncapture
camerasandoneVuevideocamera).Thedesignatedlaboratoryarea
wasabout5 m×8 m×3 m,whereparticipantswerefullyvisibleinall
cameras. The motion capturecameras were rigged on the wall shelf,
four on each left and right edge and two roughly mid-way on the
horizontal
edges,
for
recoding
three-dimensional
marker
trajectories at 60 Hz (Liang et al., 2021). The video camera was
also rigged on the wall shelf, for recording images of the walking
sequencesat60 Hz.ThecameraimageswereRGBfileswitha1920×
1080 pixel resolution. Multiple infrared cameras and digital video
camera recording were used for hardware synchronization so that
eachtimepointofthemotioncapturedatapointcorrespondedtothe
time of every video frame. Some researchers have detected 3D skeletons by existing 2D
human pose detectors from images/video (Martinez et al., 2017;
Moreno-Noguer, 2017; Nie et al., 2017) by directly using image
features (Zhou et al., 2016; Zhou et al., 2018). Martinez et al. (2017)
used a relatively simple deep feed-forward network to lift 2D pose
to 3D pose efficiently based on given high-quality 2D joint
information. Nakano et al. (2020) compared joint positions
estimated from the 3D markerless motion capture technique
based on OpenPose with multiple synchronized cameras against
recorded three-dimensional motion capture. Multicamera systems
are not easy to deploy in real-life environments. Instead, we try to
adopt the OpenPose to computer the 2D pose input to our
3DPoseNet based on a single-camera system and then estimate
gait parameters in sagittal, coronal, and transverse planes. Moreover, 15 healthy elders [mean (SD) age: 56.6 (2.53)
years] and 15 healthy young people [mean (SD) age: 27.27 (4.31)
years] wore minimal, close-fitting clothes and participated in
walking on the treadmill. The speed was set at 1.5 km/h for the The main contributions of this work include 1) novelty:
unlike previous studies, we did not directly extract the
trajectories of 2D body poses to predict gait metrics using
machine learning models. Instead, we try to estimate the 3D 02 frontiersin.org Liang et al. 10.3389/fbioe.2022.857975 younger and 0.8 km/h for the elderly. Motion capture was
recorded by tracking 39 markers, Figure 2A showed the
Body39 joints labeled by the Plug-In Gait full body model in
Extracting 2D keypoints with OpenPose
The OpenPose algorithm first estimates features from each
FIGURE 1
Overview of the experimental environment and setup. FIGURE 2
Body39 joints are based on the Plug-In Gait full body model (A) and OpenPose Body25 keypoint model (B). FIGURE 1
Overview of the experimental environment and setup. FIGURE 1
Overview of the experimental environment and setup. Experimental setting FIGURE 2
Body39 joints are based on the Plug-In Gait full body model (A) and OpenPose Body25 keypoint model (B). FIGURE 2
Body39 joints are based on the Plug-In Gait full body model (A) and OpenPose Body25 keypoint model (B). Frontiers in Bioengineering and Biotechnology Extracting 2D keypoints with OpenPose younger and 0.8 km/h for the elderly. Motion capture was
recorded by tracking 39 markers, Figure 2A showed the
Body39 joints labeled by the Plug-In Gait full body model in
Nexus software (Vicon®, 2002). The standard protocols of setting
markers were designed to match the skeletal configuration of the
Human3.6M dataset (Ionescu et al., 2014). The OpenPose algorithm first estimates features from each
image using a 10-layer VGG19 network (Cao et al., 2017). In
addition, the obtained feature map is put into two convolutional
neural networks for calculating the confidence and affinity vectors 03 frontiersin.org Liang et al. 10.3389/fbioe.2022.857975 FIGURE 3
Diagram with the basic building blocks of 3DPoseNet. FIGURE 3
Diagram with the basic building blocks of 3DPoseNet. FIGURE 3
Diagram with the basic building blocks of 3DPoseNet. rectified linear units (RELUs) (Nair and Hinton, 2010), and
residual connections (He et al., 2016). for each key point. The heat maps with confidence and with
affinity fields are obtained. Then, the 2D body joint locations are
clustered, according to dichotomy matching in graph theory. The
nonparametric representations called part affinity fields are used to
regress joint position and body segment connections between the
joints. Finally, the output has the confidence of prediction and X
and Y pixel coordinates. Figure 2B shows the Body25 joints labeled
by the OpenPose body model. Figure 3 shows a diagram with the basic building blocks of
3DPoseNet. The network is a multilayer convolutional neural
network which inputs an array of 2D joint positions and outputs
a series of joint positions in 3D. First, the linear layer applies
directly to the input, which increases its dimensionality to 1024. Then, there are two same residual blocks. Each block includes
two linear layers, Batch Norm, RELUs, and Dropout, with
residual connections. Before the final prediction, the linear
layer can be applied to produce outputs of size 3n. Initially,
the weights of our linear layers are set using Kaiming
initialization (He et al., 2015). The 3DPoseNet is trained for
200 epochs using Adam (Kingma and Ba, 2014), a starting
learning rate of 0.001 and exponential decay, using mini-
batches of size 64. In addition, we first train the network
using the Human3.6M dataset (Ionescu et al., 2014). Extracting 3D keypoints with 3DPoseNet After obtaining 2D detections using OpenPose, our goal is to
estimate 3D body joint locations. Formally, our input is a series of
2D points xϵR2n, and our output is a series of points in 3D space
yϵR3n. We aim to learn a function
fp: R2n →R3n that
minimizes the prediction error over a dataset of N poses: fp min
f
1
N
N
i1ℓf(xi) −yi,
(1) (1) Frontiers in Bioengineering and Biotechnology Gait parameter extraction where xi is 2D joint obtained from the output of OpenPose. fp is
a deep and multilayer neural network with batch normalization
(Ioffe and Christian, 2015), dropout (Srivastava et al., 2014), First of all, we filled gaps in keypoint trajectories using linear
interpolation and smoothed trajectories using a one-dimensional 04 frontiersin.org Liang et al. 10.3389/fbioe.2022.857975 FIGURE 4
Illustration of the human body coordinate system (A) and extracted joint angle features for the subject (B). FIGURE 4 FIGURE 4
Illustration of the human body coordinate system (A) and extracted joint angle features for the subject (B). zh xh × yh
xh × yh
. (4) unit-variance Gaussian filter. It should be noted that the
markerless-based system returns 3D coordinates resolved in a
local system around the middle of the hip joint. The 3D
coordinates provided by the marker-based system came from
a 3D global coordinate reference system fixed on the ground. These two different coordinate systems were moved to a new,
coincident human-based local coordinate system {O}, as shown
in Figure 4A. We centered each univariate time series by
subtracting the coordinates of the pelvis and scaling all values
by dividing them by the Euclidean distance between the right hip
and the right shoulder. (4) Commonly, the 3D limb skeleton can be represented by three
link segments: Upper Trunk, Thigh, and Shank. The 3D position
vector between the right and left shoulder was denoted with u. The 3D position of the right hip, knee, and ankle as joints was
presented as δ, α, and β, respectively. The upper trunk, right
thigh, and shank as three links were represented as ⃗U u −o,
⃗T α −δ, and ⃗S α −β, respectively, as shown in Figure 4B. The
knee angle was illustrated in the equation as follows: Commonly, the 3D limb skeleton can be represented by three
link segments: Upper Trunk, Thigh, and Shank. The 3D position
vector between the right and left shoulder was denoted with u. The 3D position of the right hip, knee, and ankle as joints was
presented as δ, α, and β, respectively. The upper trunk, right
thigh, and shank as three links were represented as ⃗U u −o,
⃗T α −δ, and ⃗S α −β, respectively, as shown in Figure 4B. Statistical analysis Reliability and validity are central characteristics that define
the quality of measurement methods and the test result potential
for application in research and clinical practice (Michelini et al.,
2020). First, we needed to assess the test–retest reliability within
markerless and marker-based motion analysis systems, using
intra-class correlation coefficients [ICC (C, 1)]. ICCs were
determined as follows: almost perfect, 0.81–1.0; substantial,
0.61–0.80;
moderate,
0.41–0.60;
fair,
0.21–0.40;
slight,
0.00–0.20 (Koo and Li, 2016). Moreover, we used sample entropy to measure the time series
of joint angles in a gait cycle. Sample entropy is a nonlinear
measurement way to analyze time series signals and is proposed
by Richman and Moorman (2000). A higher sample entropy
value indicates more randomness in time series, and lower value
shows more self-similarity. The computational procedures of
sample entropy (SampEn) are as follows (Pham, 2010): Next, to assess the potential difference in joint angles among
different measurement systems, an independent-sample t-test
was used, when the variable conformed to the normality and
homogeneity of variance simultaneously. If the variable did not
conform to the normality or homogeneity assumption, the
Wilcoxon nonparametric test was used for the difference
analysis (Liang et al., 2021). In the event of a statistically
significant
main
effect,
we
performed
post-hoc
pairwise
comparisons with Bonferroni corrections. Given a standardized (with zero mean and unit variance)
time series {x(j); 1 ≤j ≤N}, N is the total number of data points. Step
1. Construct
subsequences
of
length
m:
Xm(1), Xm(2), /, Xm(N −m),
where
Xm(i) {x(i + k);
0 ≤k ≤m −1} and m is called as embedding dimension. Step 2. Compute the distance between Xm(i) and Xm(j),
represented by d(Xm(i), Xm(j)), as: Furthermore, we calculated standard errors of measurements
(SEM) and intra-class correlation coefficients [ICC (A, 1)] with
95% confidence intervals (CI) of each joint angle to assess
correlations and consistency. The smallest detectable change
(SDC) was calculated as 1.96p
2
√
pSEM. SDC can be regarded
as the smallest change between any two steps that cannot be
attributed to measurement errors (Furlan and Sterr, 2018). In
addition, agreements between the joint angles at maximum
flexion
obtained
from
the
markerless
and
marker-based
systems
were
assessed
by
Bland–Altman
analysis,
which
permits the delineation of systematic analysis (Bland and
Altman,
1986). p-values
less
than
0.05
were
deemed
statistically significant. Data preprocessing, algorithms, and
statistical analyses were implemented using Python (version 3.5). Statistical analysis dXm(i), Xmj max
x(i + k) −xj + k
; 0 ≤k ≤m
−1, 1 ≤i, j ≤N −m, i ≠j
(6) (6) Step 3. Calculate the probability that any vector Xm(j) which is
similar to Xm(i) within T as follows: Ci(m, T) ni(m, T)
N −m + 1, i 1, /, N −m + 1,
(7) (7) where ni(m, T) is the number of vectors Xm(j) that are similar to
Xm(i) subject to the criterion of similarity: d(Xm(i), Xm(j)) ≤T. where ni(m, T) is the number of vectors Xm(j) that are similar to
Xm(i) subject to the criterion of similarity: d(Xm(i), Xm(j)) ≤T. Step 4. Calculate Step 4. Calculate ∅(m, T)
1
N −m + 1
N−m+1
i1
Ci(m, T). (8) (8) Results Step 5. Set m m + 1 and repeat steps 1–4. Step 5. Set m m + 1 and repeat steps 1–4. Gait parameter extraction to determine the appropriate values of the parameters m and T. Usually, the constant value of m is 1 or 2, T value ranging from
0.1 SD to 0.25 SD (SD is the standard deviation of time series)
(Lake et al., 2002). In this experiment, we selected m 2 and
T 0.2SD for each gait cycle data. to determine the appropriate values of the parameters m and T. Usually, the constant value of m is 1 or 2, T value ranging from
0.1 SD to 0.25 SD (SD is the standard deviation of time series)
(Lake et al., 2002). In this experiment, we selected m 2 and
T 0.2SD for each gait cycle data. Gait parameter extraction The
knee angle was illustrated in the equation as follows: RTS arccos
⃗T · ⃗S
⃗T
×
⃗S
. (5) The definition of the coordinate system was as follows: X was
anterior/posterior, Y was lateral/medial, and Z was inferior/
superior. The alignment procedure was taken from the study by
Kabsch,
(1976)
and
involved
the
initial
rotation
of
the
measurement systems, followed by the translation toward the
desired origin. The origin of the coordinate system {O} at the
time t was denoted as o(t) 1/2(PR(t) + PL(t)), where PR(t)
and PL(t) were the 3D position of the RASI and LASI joints at time
t, respectively. Two unit vectors can be determined by V1
(PR(t) −PL(t))/PR(t) −PL(t)
and
V2 (o(t) −o(t −1))/
o(t) −o(t −2). Accordingly, {O} [xh, yh, zh] can be denoted by (5) Furthermore, we calculated the joint angles between each link
segment { ⃗U, ⃗T, ⃗S} with respect to the normal vectors of the
sagittal,
coronal,
and
transverse
planes
{xh, yh, zh}
and
indicated as RUX, RTX,
RSX,
RUY, RTY,
RSY,
RUZ,
RTZ, and RSZ. It should be noted that we estimated aforementioned features
within each gait cycle, and the gait cycle was defined as the time
interval between two consecutive heel-strike events. A heel-strike
was the contact points between the heel and surface. We
calculated gait events of right heel-strikes in marker motion
capture and camera data by independently applying the different (2) xh V1,
(2)
yh V1 × V2
∥V1 × V2∥,
(3) xh V1,
yh V1 × V2
∥V1 × V2∥, (3) 05 Frontiers in Bioengineering and Biotechnology frontiersin.org Liang et al. 10.3389/fbioe.2022.857975 methods to each set of data. For marker data, we found one
derived time series helpful for improving the detection of the gait
cycle. The time series was the x-coordinates of the right ankles. Heel-strikes were defined by the time points of positive peaks in
the anterior–posterior ankle trajectories. For the markerless-
based system, heel-strike events were detected by visual
inspection. The process was greatly aided by the identification
of ankle key points obtained from the deep neural network
model. In this letter, the RUX, RTX, RSX, RUY, RTY, RSY,
RUZ, RTZ, and
RSZ were extracted at 2% increments
throughout
the
entire
cycle
Φ(t) ∈RK(k 30),
for
representing the gait pattern at each time stamp. Frontiers in Bioengineering and Biotechnology
06 Participant characteristics Elderly
Young
Normality
Homogeneity
Difference
Gender (men, %)
40%
67%
0.000*
0.478
0.217
Age (years)
56.60 ± 2.53
27.27 ± 4.31
0.000*
0.108
0.000*
Mass (kg)
61.28 ± 8.41
61.50 ± 8.59
0.163
0.662
0.947
Height (cm)
159.2 ± 8.41
168.73 ± 6.63
0.200
0.485
0.002*
BMI (kg/m2)
24.17 ± 2.81
21.53 ± 2.11
0.067
0.198
0.009*
Significant results are indicated with *. TABLE 2 Statistic property for marker and markerless motion analysis during test–retest. TABLE 2 Statistic property for marker and markerless motion analysis during test–retest. Markerless motion analysis
Marker motion analysis
Bias
ICC (95% CI; p-value)
Bias
ICC (95% CI; p-value)
RTS
0.032
0.734 [(0.592, 0.826); 0.000]
0.033
0.715 [(0.563, 0.814); 0.000]
RUX
0.015
0.714 [(0.561, 0.813); 0.000]
0.029
0.715 [(0.563, 0.814); 0.000]
RTX
0.075
0.617 ([0.405, 0.752]; 0.000)
0.078
0.611 [(0.391, 0.7500; 0.000]
RSX
0.044
0.649 [(0.380, 0.791); 0.000]
0.051
0.741 [(0.599, 0.833); 0.000]
RUY
−0.082
0.603 [(0.244, 0.775); 0.000]
−0.086
0.684 [(0.292, 0.837); 0.000]
RTY
0.066
0.610 [(0.394, 0.748); 0.000]
0.076
0.512 [(0.245, 0.684); 0.000]
RSY
0.104
0.519 [(0.253, 0.689); 0.000]
0.101
0.486 [(0.211, 0.665); 0.000]
RUZ
−0.040
0.658 [(0.476, 0.777); 0.000]
−0.40
0.699 [(0.537, 0.804); 0.000]
RTZ
0.169
0.506 [(0.148, 0.703); 0.000]
0.179
0.466 [(0.082, 0.678); 0.000]
RSZ
0.089
0.652 [(0.428, 0.784); 0.000]
0.095
0.581 [(0.317, 0.738); 0.000] Table 1, the elderly group demonstrated a larger BMI than the
young group (elderly: 24.17 ± 2.81, young: 21.53 ± 2.11, and p =
0.009). Moreover, age showed significant differences (elderly:
56.60 ± 2.53, young: 27.27 ± 4.31, and p = 0.000). It was indicated
that strategies used for the control of gait deviation related to age
between healthy young and elderly adult play an important role
(Mehdizadeh, 2018). comparisons of joint angles results extracted from video
recordings and results generated by the marker motion
capture system, is shown in Table 3 and illustrated by the
Bland–Altman plot (Figure 5). We calculated the sample entropy of joint angles across the
stride cycle that was averaged for the walking bout of each
individual participant, shown in Table 3. The SDC showed
small values, with ranges between 0.040 and 0.071. It was
indicated that small real differences between measurements
could be detected by our method. Participant characteristics Bias for sample entropy of
joint angles between measurements was also shown to be over
small, 0.034 for RTS, 0.018 for RUX, 0.081 for RTX, and 0.047 for
RSX, and they were not statistically different (p > 0.05). This is
supported by intra-class correlation coefficients that were
substantial and the ICC estimates exceeded 0.61, indicating
excellent reliability. Only reliability for RSY [95% CI = (0.291,
0.699)] and RTZ [95% CI = (0.378, 0.7360)] was lower, but still
on a level of good-to-excellent reliability (p < 0.05). The
Bland–Altman plots for each joint angle at maximum flexion
obtained from the markerless and marker-based systems are
provided in Figure 5. The x-axis and y-axis represent the average
and difference between the outputs of the two methods,
respectively. Thick and dotted lines denote the mean and zero
of difference, respectively. The green and red lines represent the
upper and lower limits of 95%, respectively, indicating most data The reliability of motion analysis system Mean values ±standard deviation and intra-class correlation
coefficients for sample entropy of each joint angle test-retested
using camera and Vicon are described in Table 2. There were
moderate ICCs of the data obtained by marker motion capture
(ICCs = 0.466–0.741 and p < 0.05). Similarly, ICCs of the data
obtained by markerless motion capture were moderate (ICCs =
0.506–0.734 and p < 0.05). The reliability was confirmed on the
joint angles measured using a single camera. Frontiers in Bioengineering and Biotechnology Participant characteristics Step 6. Calculate the sample entropy as Step 6. Calculate the sample entropy as The data for gender and age did not conform to the normality
(p < 0.05); therefore, the Wilcoxon nonparametric tests were
used for difference analysis. The data of mass, height, and BMI
were
consistent
with
the
normality
and
homogeneity
assumptions (p > 0.05); therefore, an independent-sample
t-test was used for the difference analysis. As shown in SampEn(N, m, T) −ln ∅(m + 1, T)
∅(m, T) . (9) (9) Hence, sample entropy is the negative natural logarithm of
condition
probability,
without
allowing
self-matches. To
calculate sample entropies of those time series, it is important frontiersin.org Frontiers in Bioengineering and Biotechnology
06 06 frontiersin.org Liang et al. 10.3389/fbioe.2022.857975 TABLE 1 Comparison of characteristics between elderly and young groups. TABLE 1 Comparison of characteristics between elderly and young groups. Elderly
Young
Normality
Homogeneity
Difference
Gender (men, %)
40%
67%
0.000*
0.478
0.217
Age (years)
56.60 ± 2.53
27.27 ± 4.31
0.000*
0.108
0.000*
Mass (kg)
61.28 ± 8.41
61.50 ± 8.59
0.163
0.662
0.947
Height (cm)
159.2 ± 8.41
168.73 ± 6.63
0.200
0.485
0.002*
BMI (kg/m2)
24.17 ± 2.81
21.53 ± 2.11
0.067
0.198
0.009*
Significant results are indicated with *. TABLE 2 Statistic property for marker and markerless motion analysis during test–retest. Markerless motion analysis
Marker motion analysis
Bias
ICC (95% CI; p-value)
Bias
ICC (95% CI; p-value)
RTS
0.032
0.734 [(0.592, 0.826); 0.000]
0.033
0.715 [(0.563, 0.814); 0.000]
RUX
0.015
0.714 [(0.561, 0.813); 0.000]
0.029
0.715 [(0.563, 0.814); 0.000]
RTX
0.075
0.617 ([0.405, 0.752]; 0.000)
0.078
0.611 [(0.391, 0.7500; 0.000]
RSX
0.044
0.649 [(0.380, 0.791); 0.000]
0.051
0.741 [(0.599, 0.833); 0.000]
RUY
−0.082
0.603 [(0.244, 0.775); 0.000]
−0.086
0.684 [(0.292, 0.837); 0.000]
RTY
0.066
0.610 [(0.394, 0.748); 0.000]
0.076
0.512 [(0.245, 0.684); 0.000]
RSY
0.104
0.519 [(0.253, 0.689); 0.000]
0.101
0.486 [(0.211, 0.665); 0.000]
RUZ
−0.040
0.658 [(0.476, 0.777); 0.000]
−0.40
0.699 [(0.537, 0.804); 0.000]
RTZ
0.169
0.506 [(0.148, 0.703); 0.000]
0.179
0.466 [(0.082, 0.678); 0.000]
RSZ
0.089
0.652 [(0.428, 0.784); 0.000]
0.095
0.581 [(0.317, 0.738); 0.000] TABLE 1 Comparison of characteristics between elderly and young groups. The validity of motion analysis system All gait parameters were normally distributed and showed
homogeneity of variance. Criterion validity, analyzed by the 07 frontiersin.org Liang et al. 10.3389/fbioe.2022.857975 TABLE 3 Comparison of characteristics between marker and markerless motion analysis. SDC
Bias
(95% CI; p-value)
ICC (95% CI; p-value)
RTS
0.053
0.034 [(0.000, 0.068); 0.315]
0.726 [(0.579, 0.821); 0.000]
RUX
0.071
0.018 [(−0.019, 0.056); 0.325]
0.716 [(0.565, 0.815); 0.000]
RTX
0.055
0.081 [(0.036, 0.127); 0.399]
0.644 [(0.454, 0.768); 0.000]
RSX
0.064
0.047 [(0.008, 0.087); 0.174]
0.760 [(0.632, 0.844); 0.000]
RUY
0.054
−0.098 [(−0.123, −0.074); 0.000]
0.765 [(0.639, 0.847); 0.000]
RTY
0.040
0.084 [(0.045, 0.124); 0.000]
0.613 [(0.406, 0.748); 0.000]
RSY
0.044
0.108 [(0.054, 0.162); 0.000]
0.538 [(0.291, 0.699); 0.000]
RUZ
0.060
−0.046 [(−0.080, −0.013); 0.008]
0.694 [(0.530, 0.800); 0.000]
RTZ
0.058
0.185 [(0.130, 0.240); 0.000]
0.595 [(0.378, 0.736); 0.000]
RSZ
0.061
0.102 [(0.056, 0.145); 0.000]
0.665 [(0.485, 0.782); 0.000] ICC (95% CI; p-value) ICC (95% CI; p-value) accurately measured. Another reason is that RUY, RTY, RSY,
RUZ, RTZ, and RSZ were defined as the coronal and transverse
plane angle of the normal vector relative to the upper trunk, right
thigh, and shank. In the experiment, we acquired a dataset from
stationary video camera recordings of healthy human gait, with
sagittal plane views. It is possible that camera angles would likely
affect results. are distributed within this range. The blue line shows that the
difference between the two methods increases/decreases as the
angle increases. The thick line was close to the dotted line on RTS,
RUX, RTX, RSX, RTY, and RTZ, indicating higher consensus
among the results of the two methods, otherwise on RUY, RSY,
RUZ, and RSZ. This was consistent with the results given in
Table 3. Some limitations of this study should be noted. First, some
sources of error may be intrinsic to 3D markerless pose
estimation. First and most obvious, it is difficult to track
human movements frame-by-frame perfectly from the video. For example, the left and right segments could interchange or
disappear
in
OpenPose. In
this
situation,
our
3DPoseNet algorithm cannot predict 3D pose from a failed
detector output. Second, the markerless-based identification
system is unlikely to be equivalent to the marker landmarks. While marker placement depends on manual palpation of bony
landmarks, a markerless-based system relies on visually labeled
generalized key points. The placement of motion capture
markers also owns some degree of error. The validity of motion analysis system These errors can
affect the validity of the precision evaluation of a markerless-
based system. Despite these limitations, the marker motion
capture has been recognized as the ground truth (Nakano
et al., 2020). As a result, the proposed method for evaluating
a markerless-based system can be considered to be reasonable. Frontiers in Bioengineering and Biotechnology Discussion In this study, the markerless pose estimation algorithm,
OpenPose combined with 3DPoseNet, and signal processing
techniques were used to evaluate a non-invasive method
capable of capturing gait parameters. The main findings show
that the proposed methods for extracting gait angles possess their
own good validity and reliability. Several previous studies have
used markerless-based analysis to study gait patterns of walking
or other human movements (Michelini et al., 2020; Nakano et al.,
2020; Ota et al., 2020). Our findings were consistent with these
reports in that 3D markerless pose estimation in providing
quantitative information about human movement is very
promising. First, ICC (C, 1) indicates the test–retest reliability within
markerless- and marker-based systems was in almost complete
agreement. At joint angles RTS, RUX, RTX, and RSX, there was
no significant difference between the markerless- and marker-
based motion analysis systems and the ICC (A, 1) was high
enough. However, RUY, RTY, RSY, RUZ, RTZ, and RSZ were fair. One of the reasons is that there are differences in measurement
methods of sample entropy of angles between markerless- and
marker-based systems. While a marker-based system consisted of
several cameras to form three-dimensional motion data, a
markerless-based motion analysis system combined with deep
algorithms to provide three-dimensional coordinates based on
one camera. Therefore, the rotation motion has not been We did not pre-estimate the sample size. However, it
exceeded what is commonly required for reliability studies
(Koo and Li, 2016), since 30 subjects were instructed to walk
on three times for the sake of collecting enough information to
perform the analysis. Our approach also includes manual
marking of the contact points between the heel and surface by
visual inspection, which we register in order to accumulate a
database to refine gait events. In the next work, it may be possible
to obtain more accurate video-based analyses by training gait-
specific networks from coronal and transverse views. It may be
beneficial to train networks that are specific to each population, Frontiers in Bioengineering and Biotechnology 08 frontiersin.org 10.3389/fbioe.2022.857975 Liang et al. GURE 5
and–Altman plot of RTS (A), RUX (B), RTX (C), RSX (D), RUY (E), RTY (F), RSY (G), RUZ (H), RTZ (I), and RSZ (J) at maximum flexion and
omparison between the markerless- and marker-based systems. Funding This study demonstrated the potential for combining
OpenPose
and
3DPoseNet
markerless
pose
estimation
algorithms to identify gait pathology. Given economic and
time constraint problems, we have gained several insights
from this exercise: 1) laboratory-based optical motion capture
is a reasonable baseline predictor, while 3D markerless pose
estimation networks were close to the ground_truth statistically
significantly; 2) quantitative evaluations indicate that our
proposed workflow trained on experimental movements can
be
generalized
to
non-experimental-specific
poses;
3)
correlation
between
the
quantified
results
of
network
convergence support our initial hypothesis that learning a
mapping from images to predict kinematics gait parameters is
feasible; 4) the test–retest reliability within the device was in
almost complete agreement. It was indicated that our video
analysis could be used as a quantitative assessment of gait
outside of a clinic. For predicting abnormal gait patterns or
fall risk, future work should also include elderly people who
experienced a fall. This work was supported in part by the National Key R&D
Program of China (2018YFC2001400/04, 2019YFB1311400/01), in
part by Shandong Key R&D Program (2019JZZY011112), in part by
the
Innovation
Talent
Fund
of
Guangdong
Tezhi
Plan
(2019TQ05Z735), in part by High Level-Hospital Program, Health
Commission of Guangdong Province (HKUSZH201901023), in part
by Guangdong-Hong Kong-Macao Joint Laboratory of Human-
Machine Intelligence-Synergy Systems (2019B121205007). Fang, H. S., Xie, S., Tai, Y. W., and Lu, C. (2017). “Rmpe: Regional multi-person
pose estimation,” in Proceedings of the IEEE international conference on computer
vision, 2334–2343. Author contributions such as elderly people who experience accidental falls or
abnormal gait. The markerless-based analysis described in the
current study is promising for future applications. Such a method
can classify different gait types and automatically extract
quantitative gait information from the video. GZ:
conceptualization,
methodology,
and
funding
acquisition; SL: data curation and writing—original draft
preparation;
YZ:
formal
analysis,
visualization,
and
investigation;
YD:
software
and
validation;
GL:
writing—reviewing and editing and funding acquisition. Acknowledgments The authors would like to thank all the volunteers for
participating in this study. Data availability statement The raw data supporting the conclusion of this article will be
made available by the authors, without undue reservation. Conflict of interest The authors declare that the research was conducted in the
absence of any commercial or financial relationships that could
be construed as a potential conflict of interest. Ethics statement All claims expressed in this article are solely those of the
authors and do not necessarily represent those of their affiliated
organizations, or those of the publisher, the editors, and the
reviewers. Any product that may be evaluated in this article, or
claim that may be made by its manufacturer, is not guaranteed or
endorsed by the publisher. The studies involving human participants were reviewed and
approved by the Shenzhen Institute of Advanced Technology
Institutional Review Board. The patients/participants provided
their written informed consent to participate in this study. Cronin, N. J. (2021). Using deep neural networks for kinematic analysis:
challenges
and
opportunities.
J.
Biomechanics
123,
110460.
doi:10.1016/j.
jbiomech.2021.110460 Discussion FIGURE 5
Bland–Altman plot of RTS (A), RUX (B), RTX (C), RSX (D), RUY (E), RTY (F), RSY (G), RUZ (H), RTZ (I), and RSZ (J) at maximum flexion and
comparison between the markerless- and marker-based systems. 09 Frontiers in Bioengineering and Biotechnology Frontiers in Bioengineering and Biotechnology 09 frontiersin.org Liang et al. 10.3389/fbioe.2022.857975 Cao, Z., Simon, T., Wei, S. E., and Yaser, S. (2017). “Realtime multi-person 2d
pose estimation using part affinity fields,” in Proceedings of the IEEE conference on
computer vision and pattern recognition, 7291–7299. Brach, J. S., Berthold, R., Craik, R., VanSwearingen, J. M., and Newman, A. B.
(2001). Gait variability in community-dwelling older adults. J. Am. Geriatrics Soc.
49 (12), 1646–1650. doi:10.1111/j.1532-5415.2001.49274.x Bland, J. M., and Altman, D. G. (1986). Statistical methods for assessing
agreement between two methods of clinical measurement. lancet 327 (8476),
307–310. doi:10.1016/s0140-6736(86)90837-8 Cao, Z., Simon, T., Wei, S. E., and Yaser, S. (2017). “Realtime multi-person 2d
pose estimation using part affinity fields,” in Proceedings of the IEEE conference on
computer vision and pattern recognition, 7291–7299.
Cronin, N. J. (2021). Using deep neural networks for kinematic analysis:
challenges
and
opportunities.
J.
Biomechanics
123,
110460.
doi:10.1016/j.
jbiomech.2021.110460
Fang, H. S., Xie, S., Tai, Y. W., and Lu, C. (2017). “Rmpe: Regional multi-person
pose estimation,” in Proceedings of the IEEE international conference on computer
vision, 2334–2343. Frontiers in Bioengineering and Biotechnology References Andriluka, M., Pishchulin, L., Gehler, P., and Schiele, B. (2014). “2d human pose
estimation: new benchmark and state of the art analysis,” in Proceedings of the IEEE
conference on computer vision and pattern recognition, 3686–3693. Bland, J. M., and Altman, D. G. (1986). Statistical methods for assessing
agreement between two methods of clinical measurement. lancet 327 (8476),
307–310. doi:10.1016/s0140-6736(86)90837-8 Brach, J. S., Berthold, R., Craik, R., VanSwearingen, J. M., and Newman, A. B. (2001). Gait variability in community-dwelling older adults. J. Am. Geriatrics Soc. 49 (12), 1646–1650. doi:10.1111/j.1532-5415.2001.49274.x 10 Frontiers in Bioengineering and Biotechnology frontiersin.org Liang et al. Liang et al. 10.3389/fbioe.2022.857975 Furlan, L., and Sterr, A. (2018). The applicability of standard error of
measurement and minimal detectable change to motor learning research—a
behavioral study. Front. Hum. Neurosci. 12, 95. doi:10.3389/fnhum.2018.00095 Michelini, A., Eshraghi, A., and Andrysek, J. (2020). Two-dimensional video gait
analysis: A systematic review of reliability, validity, and best practice considerations. Prosthet. Orthot. Int. 44 (4), 245–262. doi:10.1177/0309364620921290 Moreno-Noguer, F. (2017). “3d human pose estimation from a single image via
distance matrix regression,” in Proceedings of the IEEE conference on computer
vision and pattern recognition, 2823–2832. Güler, R. A., Neverova, N., and Kokkinos, I. (2018). “Densepose: dense human
pose estimation in the wild,” in Proceedings of the IEEE conference on computer
vision and pattern recognition, 7297–7306. Nair, V., and Hinton, G. E. (2010). “Rectified linear units improve restricted
boltzmann machines,” in International conference on machine learning, 807–814. He, K., Zhang, X., Ren, S., and Sun, J. (2016). “Deep residual learning for image
recognition,” in Proceedings of the IEEE conference on computer vision and pattern
recognition, 770–778. Nakano, N., Sakura, T., Ueda, K., Omura, L., Kimura, A., Iino, Y., et al. (2020). Evaluation of 3D markerless motion capture accuracy using OpenPose with multiple
video cameras. Front. Sports Act. Living 2, 50. doi:10.3389/fspor.2020.00050 He, K., Zhang, X., Ren, S., and Sun, J. (2015). “Delving deep into rectifiers:
Surpassing human-level performance on imagenet classification,” in Proceedings of
the IEEE international conference on computer vision, 1026–1034. Nie, B., Wei, P., and Zhu, S. C. (2017). “Monocular 3d human pose estimation by
predicting depth on joints,” in IEEE international conference on computer vision,
3447–3455. Ioffe, S., and Christian, S. (2015). “Batch normalization: Accelerating deep
network
training
by
reducing
internal
covariate
shift,”
in
International
conference on machine learning, 448–456. Ota, M., Tateuchi, H., Hashiguchi, T., Kato, T., Ogino, Y., Yamagata, M., et al. (2020). References Verification of reliability and validity of motion analysis systems during
bilateral squat using human pose tracking algorithm. Gait posture 80, 62–67. doi:10. 1016/j.gaitpost.2020.05.027 Ionescu, C., Papava, D., Olaru, V., and Sminchisescu, C. (2014). Human3. 6m:
large scale datasets and predictive methods for 3d human sensing in natural
environments. IEEE Trans. Pattern Anal. Mach. Intell. 36 (7), 1325–1339. doi:10.1109/tpami.2013.248 Pham, T. D. (2010). GeoEntropy: a measure of complexity and similarity. Pattern
Recognit. 43 (3), 887–896. doi:10.1016/j.patcog.2009.08.015 Kabsch, W. (1976). A solution for the best rotation to relate two sets of vectors. Acta Cryst. Sect. A 32 (5), 922–923. doi:10.1107/s0567739476001873 Richman, J. S., and Moorman, J. R. (2000). Physiological time-series analysis using
approximate entropy and sample entropy. Am. J. Physiology-Heart Circulatory
Physiology 278 (6), H2039–H2049. doi:10.1152/ajpheart.2000.278.6.h2039 Kidziński, Ł., Yang, B., Hicks, J. L., Rajagopal, A., Delp, S. L., Schwartz, M. H.,
et al. (2020). Deep neural networks enable quantitative movement analysis using
single-camera videos. Nat. Commun. 11 (1), 4054. doi:10.1038/s41467-020-17807-z Srivastava, N., Hinton, G., Krizhevsky, A., Sutskever, I., and Salakhutdinov, R. (2014). Dropout: a simple way to prevent neural networks from overfitting. J. Mach. Learn. Res. 15 (1), 1929–1958. Kingma, D. P., and Ba, J. (2014). Adam: a method for stochastic optimization. Comput. Sci. Stenum, J., Rossi, C., and Roemmich, R. T. (2021). Two-dimensional video-based
analysis of human gait using pose estimation. PLoS Comput. Biol. 17 (4), e1008935. doi:10.1371/journal.pcbi.1008935 Koo, T. K., and Li, M. Y. (2016). A guideline of selecting and reporting intraclass
correlation coefficients for reliability research. J. Chiropr. Med. 15 (2), 155–163. doi:10.1016/j.jcm.2016.02.012 Lake, D. E., Richman, J. S., Griffin, M. P., and Moorman, J. R. (2002). Sample
entropy analysis of neonatal heart rate variability. Am. J. Physiology-Regulatory,
Integr. Comp. Physiology 283 (3), R789–R797. doi:10.1152/ajpregu.00069.2002 Toshev, A., and Szegedy, C. (2014). “Deeppose: human pose estimation via deep
neural networks,” in Proceedings of the IEEE conference on computer vision and
pattern recognition, 1653–1660. Vicon® (2002). Vicon® manual, Vicon®612 motion systems. Oxford, UK: Oxford
Metrics Ltd. Plug-in-Gait modelling instructions Lénárt, Z., Nagymáté, G., and Szabó, A. (2018). Validation process of an upper
limb motion analyzer using optitrack motion capture system. Biomech. Hung. 11
(2), 93–99. doi:10.17489/2018/2/07 Yagi, K., Sugiura, Y., Hasegawa, K., and Saito, H. (2020). Gait measurement at
home using a single RGB camera. Gait posture 76, 136–140. doi:10.1016/j.gaitpost. 2019.10.006 Liang, S., Zhao, G., Zhang, Y., Diao, Y., and Li, G. (2021). References Stability region derived
by center of mass for older adults during trivial movements. Biomed. Signal Process. Control 69, 102952. doi:10.1016/j.bspc.2021.102952 Zhou, X., Zhu, M., and Leonardos, S. (2016). “Sparseness meets deepness: 3d
human pose estimation from monocular video,” in Proceedings of the IEEE
conference on computer vision and pattern recognition, 4966–4975. Martinez, J., Hossain, R., and Romero, J. (2017). “A simple yet effective baseline
for 3d human pose estimation,” in Proceedings of the IEEE international conference
on computer vision, 2640–2649. Zhou, X., Zhu, M., Pavlakos, G., Leonardos, S., Derpanis, K. G., Daniilidis, K.,
et al. (2018). Monocap: monocular human motion capture using a cnn coupled with
a geometric prior. IEEE Trans. Pattern Anal. Mach. Intell. 41 (4), 901–914. doi:10. 1109/tpami.2018.2816031 Mehdizadeh, S. (2018). The largest lyapunov exponent of gait in young and
elderly individuals: a systematic review. Gait Posture 60 (2), 241–250. doi:10.1016/j. gaitpost.2017.12.016 frontiersin.org Frontiers in Bioengineering and Biotechnology 11 Frontiers in Bioengineering and Biotechnology
|
https://openalex.org/W4311849176
|
http://repositorio.chlc.min-saude.pt/bitstream/10400.17/4461/1/Cureus%202022%20Periodontoid.pdf
|
English
| null |
Periodontoid Pseudotumor in Tuberous Sclerosis Associated With Neck Diffuse Lipomatosis
|
Curēus
| 2,022
|
cc-by
| 2,584
|
DOI: 10.7759/cureus.32663 Introduction Tuberous sclerosis (TS) is a genetic multisystem disorder, caused by a mutation in one of two different
genes, TSC1 and TSC2, that encode for different tumor suppressor proteins known as hamartin and tuberin,
respectively. In many cases, it occurs as a sporadic mutation, but it can also be inherited as autosomal
dominant. Alterations in the TSC1 and TSC2 genes lead to exaggerated activity of the mammalian target of
rapamycin (mTOR) pathway, causing the development of benign tumors in many organs including the skin,
brain, eyes, kidneys, lungs, and heart. This disorder is clinically variable from one person to another,
depending on the organs affected. Nearly all patients have skin abnormalities including hypomelanotic
macule and angiofibromas. The central nervous system (CNS) is frequently involved with cortical/subcortical
tubers, subependymal nodules, and subependymal giant astrocytoma (SEGA). Other major abnormalities
include kidney angiomyolipomas, lymphangioleiomyomatosis of the lung, heart rhabdomyoma, and retinal
hamartomas [1,2]. Six patients were reported to have diffuse overgrowth of fatty tissue in one part of the
body, defined as diffuse lipomatosis [3-6]. We describe the case of a patient with diffuse lipomatosis in the right scapular and cervical regions
associated with an incidental finding on MRI of a space-occupying lesion at the atlanto-axial joint similar to
a pseudotumor, a previously unreported association, to the best of our knowledge. Open Access Case
Report Open Access Case
Report Periodontoid Pseudotumor in Tuberous Sclerosis
Associated With Neck Diffuse Lipomatosis Eduarda Pinto , Marcos Veiga , Amets Sagarribay , Carla Conceição
1
2
3
2 Eduarda Pinto , Marcos Veiga , Amets Sagarribay , Carla Conceição
1
2
3
2 Review began 12/06/2022
Review ended 12/13/2022
Published 12/18/2022 1. Neuroradiology, Centro Hospitalar Universitário do Porto, Porto, PRT 2. Neuroradiology, Centro Hospitalar
Universitário de Lisboa Central, Lisbon, PRT 3. Neurosurgery, Centro Hospitalar Universitário de Lisboa Central,
Lisbon, PRT © Copyright 2022
Pinto et al. This is an open access article
distributed under the terms of the Creative
Commons Attribution License CC-BY 4.0.,
which permits unrestricted use, distribution,
and reproduction in any medium, provided
the original author and source are credited. © Copyright 2022
Pinto et al. This is an open access article
distributed under the terms of the Creative
Commons Attribution License CC-BY 4.0.,
which permits unrestricted use, distribution,
and reproduction in any medium, provided
the original author and source are credited. Corresponding author: Eduarda Pinto, eduardapinto92@gmail.com Abstract Tuberous sclerosis (TS) is a genetic multisystem disorder associated with the development of benign tumors
in many organs. Diffuse lipomatosis, which represents the overgrowth of fatty tissue in one part of the body,
is a very rare finding reported in patients with tuberous sclerosis. We describe the case of a patient with
diffuse lipomatosis in the right scapular, posterior cervical and perivertebral regions, associated with a
space-occupying lesion adjacent to the odontoid process of C2 that appeared to be a pseudotumor, and
discuss possible relation between these entities. Categories: Neurology, Pediatrics, Radiology
Keywords: neuroradiology, phakomatoses, periodontoid pseudotumor, diffuse lipomatosis, tuberous sclerosis Categories: Neurology, Pediatrics, Radiology How to cite this article
Pinto E, Veiga M, Sagarribay A, et al. (December 18, 2022) Periodontoid Pseudotumor in Tuberous Sclerosis Associated With Neck Diffuse
Lipomatosis. Cureus 14(12): e32663. DOI 10.7759/cureus.32663 To better characterize this lesion, a cervical computed tomography (CT) was later performed. It
demonstrated a diffuse lipomatous infiltration and enlargement of the right suboccipital, posterior
paravertebral, and prevertebral muscles, and of the ipsilateral posterior cervical, submandibular,
retropharyngeal, and carotid spaces. In addition, there was a periodontoid soft tissue mass, with bone
remodeling of the right lateral mass of C2, the odontoid and the anterior arch of C1 and less pronounced
remodeling of the clivus. These bone abnormalities were well delimited and exhibited cortical and
subchondral sclerosis (Figure 2) which also extended to the ipsilateral C3 and C4 body and transverse
process. Cervical spine MRI revealed a hypointense mass on T1 and T2-weighted images at the atlanto-axial
joint, with no contrast enhancement. It extended into the anterior epidural space, decreasing the
subarachnoid space, without compression of the brainstem or the spinal cord (Figure 3). The lesion remained
stable compared to the brain MRI performed a few months earlier. 2022 Pinto et al. Cureus 14(12): e32663. DOI 10.7759/cureus.32663 Case Presentation A 16-year-old patient was initially observed at a medical appointment because of a slowly growing lesion in
the right scapular region over 4 years, composed of fatty tissue, as assessed by soft tissue ultrasound. He was
diagnosed with tuberous sclerosis, with a mutation in the TSC1 gene. Brain magnetic resonance imaging (MRI) revealed cortical and subcortical tubers and radial migrating bands
in both cerebral hemispheres and small subependymal hamartomas, related to tuberous sclerosis. A Chiari I
malformation was also found, due to a short clivus, with a caudal descent of cerebellar tonsils, compressing
the posterior surface of the medulla. As an incidental finding, a space-occupying lesion at the atlanto-axial
joint was noted (Figure 1). Besides
diffuse lipomatous infiltration, axial T1-WI (a) and sagittal T2-WI (b)
show a hypointense mass at the atlanto-axial joint, extending into the
anterior epidural space, but without compression of the brainstem or
the spinal cord. No contrast enhancement is depicted on fat-suppressed
contrast-enhanced sagittal T1-WI (c). FIGURE 3: MRI of the cervical spine with axial T1-WI (a), sagittal T2-WI
(b) and fat supressed contrast-enhanced sagittal T1-WI (c). Besides
diffuse lipomatous infiltration, axial T1-WI (a) and sagittal T2-WI (b)
show a hypointense mass at the atlanto-axial joint, extending into the
anterior epidural space, but without compression of the brainstem or
the spinal cord. No contrast enhancement is depicted on fat-suppressed
contrast-enhanced sagittal T1-WI (c). These imaging findings, namely the bone remodeling and sclerosis, and the stability of the changes through
the different examinations suggested a process with slow growth and, therefore, a benign lesion. Its imaging
features resembled those of a periodontoid pseudotumor. The patient had no symptoms related to this space-occupying lesion at the craniocervical junction, so a
conservative expectant attitude was decided with clinical and imaging surveillance. FIGURE 1: Brain MRI with axial T2FLAIR-weighted image (a), axial T2*-
WI (b) and sagittal T1-WI (c). Cortical and subcortical tubers (short
arrow) and radial migrating bands (long arrow) in both cerebral
hemispheres are shown on axial T2FLAIR-WI (a). A small calcified
subependymal hamartoma in the frontal horn of the left lateral ventricle
is better depicted on axial T2*-WI (b). Sagittal T1-WI (c) shows a Chiari I
malformation, with a short clivus and caudal descent of cerebellar
tonsils, compressing the posterior surface of the medulla. It also
reveals a space-occupying lesion at the atlanto-axial joint. FIGURE 1: Brain MRI with axial T2FLAIR-weighted image (a), axial T2*-
WI (b) and sagittal T1-WI (c). Cortical and subcortical tubers (short
arrow) and radial migrating bands (long arrow) in both cerebral
hemispheres are shown on axial T2FLAIR-WI (a). A small calcified
subependymal hamartoma in the frontal horn of the left lateral ventricle
is better depicted on axial T2*-WI (b). Sagittal T1-WI (c) shows a Chiari I
malformation, with a short clivus and caudal descent of cerebellar
tonsils, compressing the posterior surface of the medulla. It also
reveals a space-occupying lesion at the atlanto-axial joint. 2 of 5 FIGURE 2: Cervical CT with axial images in soft window (a) (b) (d) and
coronal reformation in bone window (c) demonstrates a diffuse
lipomatous infiltration of the right suboccipital, posterior paravertebral,
and prevertebral muscles, as well as the ipsilateral posterior cervical,
retropharyngeal, and carotid spaces, with increased volume of these
regions, deforming the posterior wall of nasopharynx and oropharynx
(a). It reveals a periodontoid soft tissue mass (b), with remodeling of the
right lateral mass of C2, the odontoid process and the anterior arch of
C1 (c). The bone abnormalities were well delimited and exhibited
cortical, subchondral and medullary sclerosis (c). Diffuse lipomatosis of
the muscles and spaces of the right scapular region was also observed
(arrow) (d). FIGURE 2: Cervical CT with axial images in soft window (a) (b) (d) and
coronal reformation in bone window (c) demonstrates a diffuse
lipomatous infiltration of the right suboccipital, posterior paravertebral,
and prevertebral muscles, as well as the ipsilateral posterior cervical,
retropharyngeal, and carotid spaces, with increased volume of these
regions, deforming the posterior wall of nasopharynx and oropharynx
(a). It reveals a periodontoid soft tissue mass (b), with remodeling of the
right lateral mass of C2, the odontoid process and the anterior arch of
C1 (c). The bone abnormalities were well delimited and exhibited
cortical, subchondral and medullary sclerosis (c). Diffuse lipomatosis of
the muscles and spaces of the right scapular region was also observed
(arrow) (d). FIGURE 3: MRI of the cervical spine with axial T1-WI (a), sagittal T2-WI
(b) and fat supressed contrast-enhanced sagittal T1-WI (c). Besides
diffuse lipomatous infiltration, axial T1-WI (a) and sagittal T2-WI (b)
show a hypointense mass at the atlanto-axial joint, extending into the
anterior epidural space, but without compression of the brainstem or
the spinal cord. No contrast enhancement is depicted on fat-suppressed
contrast-enhanced sagittal T1-WI (c). These imaging findings, namely the bone remodeling and sclerosis, and the stability of the changes through
the different examinations suggested a process with slow growth and, therefore, a benign lesion. Its imaging
features resembled those of a periodontoid pseudotumor. Th
i
h d
l
d
hi
i
l
i
h
i
i
l j
i FIGURE 3: MRI of the cervical spine with axial T1-WI (a), sagittal T2-WI
(b) and fat supressed contrast-enhanced sagittal T1-WI (c). 2022 Pinto et al. Cureus 14(12): e32663. DOI 10.7759/cureus.32663 Conclusions Diffuse lipomatosis in one part of the body is a very rare finding in patients with tuberous sclerosis. Our
patient was initially observed because of a slowly growing tumefaction in the right scapular region over 4
years. The first follow-up brain MRI of tuberous sclerosis revealed an incidental space-occupying lesion at
the anterior atlanto-axial joint, the characteristics of which were of a periodontoid pseudotumor on both
cervical CT and MRI. These imaging studies also revealed diffuse lipomatosis in the right cervical spaces and
ipsilateral perivertebral muscles, leading to a probable association between these two entities. Disclosures Human subjects: Consent was obtained or waived by all participants in this study. Conflicts of interest: In
compliance with the ICMJE uniform disclosure form, all authors declare the following: Payment/services
info: All authors have declared that no financial support was received from any organization for the
submitted work. Financial relationships: All authors have declared that they have no financial
relationships at present or within the previous three years with any organizations that might have an
interest in the submitted work. Other relationships: All authors have declared that there are no other
relationships or activities that could appear to have influenced the submitted work. Publications
No. of patients
Age (years)
Diagnosis methods
Site
Klein et al. 1986 [6]
1
15
X-ray, biopsy/excision
Lower limb
Alcázar et al. 1998 [5]
1
37
CT scan, biopsy
Dorsal transthoracic
Mittal et al. 2017 [4]
1
24
MRI, fine needle aspiration
Lower limb
Ilyas et al. 2021 [3]
3
11
MRI, fine needle aspiration
Lower limb
14
Ultrasound, MRI
Lower limb
13
Ultrasound, MRI, biopsy/excision
Lower limb
TABLE 1: Previous reports of Diffuse Lipomatosis in TS. We hypothesized that this diffuse lipomatous infiltration of the muscles and neck spaces led to an atlanto-
axial joint hypermobility with the development of a presumed periodontoid pseudotumor. Publications
No. of patients
Age (years)
Diagnosis methods
Site
Klein et al. 1986 [6]
1
15
X-ray, biopsy/excision
Lower limb
Alcázar et al. 1998 [5]
1
37
CT scan, biopsy
Dorsal transthoracic
Mittal et al. 2017 [4]
1
24
MRI, fine needle aspiration
Lower limb
Ilyas et al. 2021 [3]
3
11
MRI, fine needle aspiration
Lower limb
14
Ultrasound, MRI
Lower limb
13
Ultrasound, MRI, biopsy/excision
Lower limb TABLE 1: Previous reports of Diffuse Lipomatosis in TS. We hypothesized that this diffuse lipomatous infiltration of the muscles and neck spaces led to an atlanto-
axial joint hypermobility with the development of a presumed periodontoid pseudotumor. Periodontoid pseudotumor reflects a non-neoplastic soft tissue mass adjacent to the odontoid process. It is
caused by various mechanisms and etiologies such as rheumatoid arthritis, trauma, atlantoaxial
hypermobility compensating for subaxial ankylosis, and deposition diseases (like calcium pyrophosphate). Some cases are idiopathic, but the underlying mechanism is believed to be a chronic subluxation of the
atlanto-axial joint. The histology of the mass mainly reveals connective tissue and cartilage [7]. Supporting
this hypothesis, previous surgery reports with posterior decompression and fusion led to a progressive
spontaneous regression of the pseudotumor [7]. Though we do not have histologic confirmation in our case, we postulate that a periodontoid pseudotumor
is the most plausible diagnosis, since it is a slowly growing extraosseous mass located around the odontoid
process causing bone remodeling and sclerosis. The diffuse lipomatosis of the right neck spaces and muscles
could explain an atlanto-axial subluxation and the development of an ipsilateral pseudotumor. 2022 Pinto et al. Cureus 14(12): e32663. DOI 10.7759/cureus.32663 Discussion Diffuse lipomatosis in one part of the body is a very rare finding in patients with tuberous sclerosis and has
been previously reported only in six patients. It is observed on the lower limbs in most cases, and in the
subcutaneous lumbar region and the thoracic region around the lungs [3-6]. These previous reports are
summarized in Table 1. Our patient had a diffuse lipomatous infiltration of the right neck spaces and the
right paravertebral and prevertebral muscles, causing the tumefaction observed on physical examination at
the right scapular region. 3 of 5 2022 Pinto et al. Cureus 14(12): e32663. DOI 10.7759/cureus.32663 References 1. DiMario FJ Jr, Sahin M, Ebrahimi-Fakhari D: Tuberous sclerosis complex. Pediatr Clin North Am. 2015,
62:633-48. 10.1016/j.pcl.2015.03.005
2. Curatolo P, Moavero R, Roberto D, Graziola F: Genotype/phenotype correlations in tuberous sclerosis
complex. Semin Pediatr Neurol. 2015, 22:259-73. 10.1016/j.spen.2015.10.002
3. Ilyas M, Quezada J, Opfer EK: Lipomatous infiltration in tuberous sclerosis complex - A case series and
literature review. Child Neurol Open. 2021, 8:2329048X211048065. 10.1177/2329048X211048065
4. Mittal A, Vinay K, De D, Handa S, Sinha A: Tuberous sclerosis complex and diffuse lipomatosis: Case report
of a rare association. Indian Dermatol Online J. 2018, 9:37-9. 10.4103/idoj.IDOJ_60_17
5. Alcázar JD, Ramos R, Verdugo J: Lipomatosis difusa dorsal transtorácica en un paciente con esclerosis
tuberosa familiar. Archivos de Bronconeumología. 1998, 34:468-9. 10.1016/s0300-2896(15)30377-x
6. Klein JA, Barr RJ: Diffuse lipomatosis and tuberous sclerosis. Arch Dermatol. 1986, 122:1298-302. 10.1001/archderm.1986.01660230090018
7. Yanni DS, Halim AY, Alexandru D: Odontoid pseudotumor and serial postfusion radiographic evaluation in a 1. DiMario FJ Jr, Sahin M, Ebrahimi-Fakhari D: Tuberous sclerosis complex. Pediatr Clin North Am. 2015,
62:633-48. 10.1016/j.pcl.2015.03.005 2. Curatolo P, Moavero R, Roberto D, Graziola F: Genotype/phenotype correlations in tuberous sclerosis
complex. Semin Pediatr Neurol. 2015, 22:259-73. 10.1016/j.spen.2015.10.002 3. Ilyas M, Quezada J, Opfer EK: Lipomatous infiltration in tuberous sclerosis complex - A case series and
literature review. Child Neurol Open. 2021, 8:2329048X211048065. 10.1177/2329048X211048065 4. Mittal A, Vinay K, De D, Handa S, Sinha A: Tuberous sclerosis complex and diffuse lipomatosis: Case report
of a rare association. Indian Dermatol Online J. 2018, 9:37-9. 10.4103/idoj.IDOJ_60_17 5. Alcázar JD, Ramos R, Verdugo J: Lipomatosis difusa dorsal transtorácica en un paciente con esclerosis
tuberosa familiar. Archivos de Bronconeumología. 1998, 34:468-9. 10.1016/s0300-2896(15)30377-x 6. Klein JA, Barr RJ: Diffuse lipomatosis and tuberous sclerosis. Arch Dermatol. 1986, 122:1298-302. 10.1001/archderm.1986.01660230090018 7. Yanni DS, Halim AY, Alexandru D: Odontoid pseudotumor and serial postfusion radiographic evaluation in a 2022 Pinto et al. Cureus 14(12): e32663. DOI 10.7759/cureus.32663 4 of 5 2022 Pinto et al. Cureus 14(12): e32663. DOI 10.7759/cureus.32663 patient with a C1-2 mass. J Neurosurg Spine. 2015, 22:605-10. 10.3171/2014.10.SPINE13987 5 of 5
|
https://openalex.org/W2519881422
|
http://www.ajas.info/upload/pdf/ajas-30-5-736.pdf
|
English
| null |
Cloning and sequence analysis of Wild Argali short palate, lung and nasal epithelium clone 1 cDNA
|
Asian-Australasian journal of animal sciences
| 2,016
|
cc-by
| 4,987
|
Cloning and sequence analysis of Wild Argali short palate, lung
and nasal epithelium clone 1 cDNA Wen Shen
1,2, Kaili Chen
3, Yanming Sun
1,*, Haiying Guo
1, Dongmei Chen
1, and Yang Cao
2 Objective: Experiments were conducted to clone the sequence of Wild Argali short palate, lung
and nasal epithelium clone 1 (SPLUNC1) cDNA, and to lay the foundation for further study
the biological function of Wild Argali SPLUNC1. Methods: The complete sequence of Wild Argali SPLUNC1 cDNA was generated by rapid Objective: Experiments were conducted to clone the sequence of Wild Argali short palate, lung
and nasal epithelium clone 1 (SPLUNC1) cDNA, and to lay the foundation for further study
the biological function of Wild Argali SPLUNC1. Methods: The complete sequence of Wild Argali SPLUNC1 cDNA was generated by rapid
amplification of cDNA ends. The entire coding sequence was inserted into the pPIC9K vector
and expressed in Pichia pastoris (P. pastoris) GS115. The recombinant SPLUNC1 protein was
detected by Western blot and purified by Ni
2+ chelate affinity chromatography. The test of effect
of the protein on Mycoplasma ovipneumoniae (MO) was performed with real-time polymerase
chain reaction. * Corresponding Author: Yanming Sun
Tel: +86-0993-2027809, Fax: +86-0993-2058079,
E-mail: sym@shzu.edu.cn * Corresponding Author: Yanming Sun
Tel: +86-0993-2027809, Fax: +86-0993-2058079,
E-mail: sym@shzu.edu.cn g
g
Methods: The complete sequence of Wild Argali SPLUNC1 cDNA was generated by rapid
amplification of cDNA ends. The entire coding sequence was inserted into the pPIC9K vector
and expressed in Pichia pastoris (P. pastoris) GS115. The recombinant SPLUNC1 protein was
detected by Western blot and purified by Ni
2+ chelate affinity chromatography. The test of effect
of the protein on Mycoplasma ovipneumoniae (MO) was performed with real-time polymerase
chain reaction.h Results: The Wild Argali SPLUNC1 cDNA was 1,076 bp with an open reading frame of 768 bp,
which encoded a 26.49 kDa protein composed of 255 amino acids. Its amino acid sequence
shared 98.4%, 96.9%, 94.5%, 90.2%, 80.8%, 78.4%, 78.3%, 72.5%, 72.3%, 68.8% identity with
those of SPLUNC1 cDNA from Ovis aries (accession no. NP_001288334.1), Capra hircus
(accession no. XP_005688516.1), Pantholops hodgsonii (accession no. XP_005979709.1), Bos
taurus (accession no. NP_776851.1), Felis catus (accession no. XP_006929910.1), Homo sapiens
(accession no. NP_001230122.1), Sus scrofa (accession no. NP_001005727.1), Chinchilla lanigera
(accession no. NP_001269294.1), Mus musculus (accession no. NP_035256.2), and Rattus
norvegicus (accession no. NP_742028.1), respectively. The recombinant protein corresponded to
the expected molecular mass of 25.47 kDa as judged by sodium dodecyl sulfate-polyacrylamide
gel electrophoresis, and it was detected in the supernatant of P. Open Access Asian-Australas J Anim Sci
Vol. 30, No. 5:736-742 May 2017
https://doi.org/10.5713/ajas.15.0557
pISSN 1011-2367 eISSN 1976-5517
Open Access 1 College of Animal Science and Technology, Shihezi
University, Shihezi, Xinjiang 832003, China
2 College of Life Science, South China Agricultural
University, Guangzhou, Guangdong 510000, China
3 College of Life Science, Shihezi University, Shihezi,
Xinjiang 832003, China Cloning and sequence analysis of Wild Argali short palate, lung
and nasal epithelium clone 1 cDNA pastoris, and it could be purified. The results from the test of inhibition effect of argali recombinant SPLUNC1 protein on MO
showed that the product could inhibit MO very well (p<0.01).hf p
y
(p
)
Conclusion: The amino acid sequence of Wild Argali SPLUNC1 was different from other
organisms. The recombinant SPLUNC1 protein has good biological activity. Keywords: Argali; Short Palate; Lung and Nasal Epithelium Clone 1 (SPLUNC1) cDNA; Clone;
Sequencing; Expression; Activity Detection Keywords: Argali; Short Palate; Lung and Nasal Epithelium Clone 1 (SPLUNC1) cDNA; Clone;
Sequencing; Expression; Activity Detection Keywords: Argali; Short Palate; Lung and Nasal Epithelium Clone 1 (SPLUNC1) cDNA; Clone;
Sequencing; Expression; Activity Detection Submitted Jul 4, 2015; Revised Nov 2, 2015;
Accepted Sept 10, 2016 Copyright © 2017 by Asian-Australasian Journal of Animal Sciences
This is an open-access article distributed under the terms of the Creative Commons Attribution Non-Commercial License
(http://creativecommons.org/licenses/by-nc/4.0/), which permits unrestricted non-commercial use, distribution, and repro
duction in any medium, provided the original work is properly cited. Copyright © 2017 by Asian-Australasian Journal of Animal Sciences Construction of expression vectorh The expression vector of recombinant SPLUNC1 protein was
constructed by inserting full length SPLUNC1 open reading
frame (ORF) into pPIC9K plasmid (Invitrogen, China). Briefly,
recombinant pMD-18T with SPLUNC1 ORF was digested with
SnaB I and Not I and the SPLUNC1 cDNA fragment was inserted
into pPIC9K vector previously linearized with similar enzymes
to obtain the pPIC9K/SPLUNC1 cDNA expression vector. The
recombinant vector was transformed into Escherichia coli and
eight positive clones were verified by enzyme digestion. The
recombinant pPIC9K/SPLUNC1 vector was linearized by using
Sac I and transformed into Pichia pastoris GS115 by using electro
poration-mediated method. His+ transformants were selected
on minimal dextrose (MD) plates and selected subsequently on
yeast extract peptone dextrose (YPD) plates containing genta
mycin 418 (G418; 0.25 to 3 mg/mL), and the genomic DNA of
the His+ and G418+ transformants were extracted and analyzed
by PCR using the 5′AOX1 primer and the 3′AOX1 primer. MATEREALS AND METHODS Oral palate mucosa materials and RNA preparation
Oral palate mucosa materials were collected by scrapping oral
palate from four healthy wild Argali and stored in Trizol (Invitro
gen, Beijing, China). Total RNA was extracted by using the Trizol
(Invitrogen, China) method and stored at –70°C until use. Rapid amplification of cDNA 5’ends (5’-RACE)h The 5’ ends were generated according to the instructions for
the 5’ Full RACE Core Set cDNA Kit (Clontech, Beijing, China). Briefly, cDNA was synthesized by reverse transcription (RT)
UPM primer (as provided in the kit). Based on the published
mRNA sequence of Bos taurus (accession no. NM_174426.3),
gene specific primers namely R2 (5′ATTGACCAGNGGGCAC
AC3′) and K2(5′CAGGCTGCCAGGGGAGTG3′) were employed
in the nested polymerase chain reaction (PCR) at concentration
of 20 pmol/μL each. The second amplification products were
recovered using ‘Takara Agarose gel DNA purification Kit
version 2.0’ according to the manufacturer’s instructions and
ligated into pMD-18T cloning vector (Takara Biotechnology,
Dalian, China) as described [6]. The sequences in the clones
were verified by the sanger DNA sequencing method using Taq
DNA Polymerase (Qiagen, Shanghai, China). Expression of recombinant SPLUNC1 protein A monoclonal strain that was picked from plate G418 was in
oculated into BMGY medium [BMGY: 20 g tryptone, 10 g yeast
extract, 100 mL 10×yeast nitrogen base (YNB W/O amino acids),
2 mL 500×D-Biotin (B), 100 mL 10× glycerol (Gl), 100 mL 1 M
phosphate buffer, 700 mL H2O], and cultured for 2 days (OD600
= 2-6) in the shaker. Pichia pastoris liquid was centrifuged. Then,
it was transferred into BMMY medium [BMMY: 20 g tryptone,
10 g yeast extract, 100 mL 10×YNB, 2 mL 500×B, 100 mL 10×
methanol (M), 100 mL 1 M phosphate buffer, 700 mL H2O],
which was used for inducing protein expression. Methanol was
added to a final concentration of 0.5% (v/v) every 24 h. After
3 days, yeast liquid was collected and centrifuged. Finally, the
supernatant was collected separately and analyzed by sodium
dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). Synthesis of complete SPLUNC1 cDNAh Synthesis of complete SPLUNC1 cDNAh The complete coding sequence was generated by RT-PCR as
described [6]. The primers used were as follows: SPB1(5′TAC
GTAATGCACCACCACCACCACCACCTGCTAGAAGCCCT
GCCCG3′); SPB2(5′GCGGCCGCTCAGACTTTGATGACA
AATTCTAGCCC3′) at concentrations of 20 pmol/μL each. The purified PCR products were ligated into pMD-18T cloning
vector (Takara Biotechnology, China). The sequences in the
clones were verified by the Sanger DNA sequencing method
using Taq DNA Polymerase (Qiagen, China). Shen et al (2017) Asian-Australas J Anim Sci 30:736-742 pecially expressed in the oral cavity, nasal cavity, nose, pharynx,
respiratory tract, nasopharyngeal secretions, saliva in secreted
form and may become the body's first line of defense against
bacteria and viruses [3]. SPLUNC1 is the decisive element in in
nate immune defense against pathogenic organisms in the
respiratory tract infected pulmonary mucosa [4]. SPLUNC1
plays an important role in the antibacterial and anti-inflam
matory response to respiratory tract infection caused by Gram-
negative bacteria [3]. Besides that, SPLUNC1 plays a very
important role in maintaining the steady-state aspects of up
per respiratory tract [5]. However, there is no report about the
SPLUNC1 sequence of Xinjiang Wild Argali. To study the dif
ference in nucleotide sequences and amino acid sequences of
SPLUNC1 gene between wild and domestic animals, the pres
ent study used the sequences from cattle and sheep as the
references to clone and analyze the cDNA sequence by RACE
technique and in vitro expression in eukaryotic cells, which
provides a basis for further studies on the differences in gene
functions and disease resistance. pMD-18T cloning vector (Takara Biotechnology, China). The
sequences in the clones were verified by the Sanger DNA se
quencing method using Taq DNA Polymerase (Qiagen, China). INTRODUCTION Argali (Ovis ammon), also known as the big sheep in Artiodactyla, Ruminantia, Boridae, Caprinae,
genus Ovis, belongs to wild animals under the second class state protection [1]. Wild Argali are
distributed in the mountains of Eurasia, and also found in Inner Mongolia, Ningxia, Gansu,
Xinjiang, Qinghai, and Tibet in China. Wild Argali’s weight is the highest among Caprinae, and
the rams weigh up to 200 kg. Wild Argali are up to 1.27 meters tall at the shoulder, have more
lean and less fat content, enabling fast growth, less fat, strong adaptability, high disease resistance
as the main characteristics [2], and are an excellent genetic resource. It is important to exploit
and make full use of the superior characteristics of Wild Argali in Xinjiang. Short palate, lung and nasal epithelium clone 1 (SPLUNCl) is a secreted protein molecule,
which is expressed highly on respiratory epithelium and has a signal peptide. SPLUNC1 is es 736 www.ajas.info www.ajas.info Determination of the activity of the expression product Determination of the activity of the expression product
Determination of the activity of the expression product of
SPLUNC1 was tested by incubating MO, 10
5 colony-forming
units (CFUs)/mL, with recombinant SPLUNC1 (make the final
concentration of recombinant SPLUNC1 protein up to 1 μg/mL,
3 μg/mL, 10 μg/mL in every well), phosphate-buffered saline
(PBS) (as control) in a 24-well plate for 48 hours at 37°C as de
scribed [8,9]. MO RNA was extracted, and real-time PCR was
performed to quantify MO 16S rRNA levels. Finally, statistical
analysis was carried out for the data by using one-way analysis
of variance by SPSS17.0 (www.xpgod.com). The initial ATG and stop TAG codons are shaded. The distal
polyadenylation signal sequence AATAAA are shaded. The
nucleotide and its amino acids are numbered along the left
margins respectively. RESULTS RACE and complete SPLUNC1 cDNA PCR products
Nucleotide sequencing combined with a search of the GenBank
database revealed eight clones generated by 5’RACE-PCR (≈635)
(Figure 1A), having no initiator element in their open reading
frames. The 3’RACE-PCR products (Figure 1B) had complete
3’ends containing polyadenylation tail of 12A’s at 224 bp down
stream of the stop codon TAA. All the transformation clones
had inserts of the expected SPLUNC1 cDNA-ORF as identified
by SnaB I and Not I enzyme digestion (Figure 2A). The sequences
were confirmed in all the nine clones. The 3’, 5’RACE and full
length sequences were 635 bp (3’ends), 620 bp (5’ends), and 1,076
bp (SPLUNC1 cDNA) respectively, each having 99% identity
to Ovis.aries mRNA. Western blot analysis of recombinant SPLUNC1 proteint Western blot analysis of recombinant SPLUNC1 proteint After the process of SDS-PAGE electrophoresis, the proteins were
electroblotted to polyvinylidene fluoride membrane. Then the
membrane was placed in blocking solution about 1 h; incubated
at 4°C overnight with Murine monoclonal antibody with anti-His
tag (1:2,000), next, the membrane was washed 3 times with Tris-
buffered saline and Tween 20 (TBST) (TBST: 50 mL 1 M pH
7.5 Tris HCl, 8 g NaCl, 0.2 g KCl, 0.5 mL Tween 20, 1,000 mL
H2O). Incubated at 25°C with the secondary antibody, horse
radish peroxidase labeled goat anti-mouse IgG (1:5,000), for about
90 min, then the membrane was washed three times with TBST,
and finally colored with diaminobenzidine coloring solution. Rapid amplification of cDNA 3′ ends (3’-RACE)h i
The 3’ends were synthesized according to the instructions in
3’Full RACE Core Set cDNA Kit (Takara Biotechnology, China). The 3-sites adapter Primer (provided in the Kit), the P1 primer
(5′CATCGTCTCTATGTCACC3′) and P2 primer (5′ATTGAC
CAGNGGGCACAC3′) were used at concentration of 20 pmol/μL
each. The 3’ end PCR products were purified and ligated into Purification of recombinant SPLUNC1 proteins
Following induction, the supernatant were collected were isolated www.ajas.info 737 Shen et al (2017) Asian-Australas J Anim Sci 30:736-742 as described by the manufacturer (Qiagen, China). Briefly, yeast
liquid which contained the expression product was collected by
centrifugation and the supernatant collected. The supernatant was
filtered through a 0.22 μm membrane. The filtrate was passed
through a Ni-NTA agarose affinity chromatography column with
10 to 15 volumes of binding buffer (pH 8.0 50 mM phosphate
buffer containing 0.5 M NaCl and 20 mM imidazole) until the
optical density (OD) value reached the 280 nm absorbance, and
was collected respectively. Then, it was eluted with elution buffer
(pH 8.0 50 mM phosphate buffer containing 0.5 M NaCl and
300 mM imidazole). Finally, the liquid effluent was collected
separately and analyzed by SDS- PAGE [7]. Characteristics of SPLUNC1 cDNA sequenceh The sequence of Wild Argali SPLUNC1 cDNA and the inferred
amino acids was shown in Figure 2B. The accuracy of the se
quence was determined on both strands and found to be in order. The cDNA insert was 1,076 bp long. SPLUNC1 cDNA-ORF
contains one long open reading frame of 768 bp from 64 nt to
831 nt. The first methionine codon ATG begins at 64 to 66 nt
and ends at presumed stop codon TAA at 829 to 831 nt. The
bases AATAAA at 1,034 to 1,039 nt downstream are the distal
polyadenylation stop signals for SPLUNC1 cDNA. However, the
TATA box was not found in the upstream of SPLUNC1 cDNA-
ORF. Figure 2. Restriction enzyme analysis of expression vector of pPIC9K-SPLUNC1 and the sequence of SPLUNC1 cDNA. (A) The recombinant plasmids of eight clones containing
inserts of the full-length SPLUNC1 cDNA-ORF identified by SnaB I and Not I enzyme digestion. Key: Lane 1: ≈744 bp SPLUNC1-ORF; Lane 2: empty vector DNA; Lane 3: DNA Maker I
molecular marker. (B) Nucleotide and predicted amino acid sequence of SPLUNC1cDNA. The complete SPLUNC1cDNA sequence consisting of 1,076 nucleotides and open reading
frame of 768 bp. The initiation ATG and stop TAA codons are shaded. The distal polyadenylation signal sequence AATAAA are shaded. The nucleotide and its amino acids are
numbered along the left margins respectively. The putative signal peptide sequence is shown as capital italicized letters. Predicted amino acid sequence of recombinant SPLUNC1 The open reading frame encoded a 255 amino acid residue
protein with an isoelectric point (pI) of 5.07 and a predicted Figure 1. The 5’-RACE and 3’-RACE PCR products of SPLUNC1. (A) The 5’-RACE-PCR products, Key: Lane 1: Positive 3’ends PCR products ≈635 bp; Lane 2: DNA Maker I molecular
marker. (B) The 3’-RACE-PCR products, Key: Lane 1: positive 5’ends PCR product ≈620 bp; Lane 2: DNA Maker I molecular marker. PCR, polymerase chain reaction. 301
302
303
304
305
306
(A) (B)
307
308
309
2,000 bp
100 bp
635 bp
500 bp
1 2
750 bp
2,000 bp
100 bp
750 bp
500 bp
1 2
620 bp (A)
2,000 bp
100 bp
635 bp
500 bp
1 2
750 bp (A)
2,000 bp
100 bp
635 bp
500 bp
1 2
750 bp (B)
2,000 bp
100 bp
750 bp
500 bp
1 2
620 bp (B)
2,000 bp
100 bp
750 bp
500 bp
1 2
620 bp (B) (A) Figure 1. The 5’-RACE and 3’-RACE PCR products of SPLUNC1. (A) The 5’-RACE-PCR products, Key: Lane 1: Positive 3’ends PCR products ≈635 bp; Lane 2: DNA Maker I molecular
marker. (B) The 3’-RACE-PCR products, Key: Lane 1: positive 5’ends PCR product ≈620 bp; Lane 2: DNA Maker I molecular marker. PCR, polymerase chain reaction. 308
309 Figure 1. The 5’-RACE and 3’-RACE PCR products of SPLUNC1. (A) The 5’-RACE-PCR products, Key: Lane 1: Positive 3’ends PCR products ≈635 bp; Lane 2: DNA Maker I molecular
marker. (B) The 3’-RACE-PCR products, Key: Lane 1: positive 5’ends PCR product ≈620 bp; Lane 2: DNA Maker I molecular marker. PCR, polymerase chain reaction. 308
309 738 www.ajas.info310
311 www.ajas.info310
311 Shen et al (2017) Asian-Australas J Anim Sci 30:736-742 2
3
1
744 bp
10,000 bp
1,000 bp
750 bp
250 bp
(A) 2
3
1
744 bp
10,000 bp
1,000 bp
750 bp
250 bp
(A) 2
3
1
744 bp
10,000 bp
1,000 bp
750 bp
250 bp
(A) (A) (B) (B) (B) Figure 2. Restriction enzyme analysis of expression vector of pPIC9K-SPLUNC1 and the sequence of SPLUNC1 cDNA. (A) The recombinant plasmids of eight clones containing
inserts of the full-length SPLUNC1 cDNA-ORF identified by SnaB I and Not I enzyme digestion. Key: Lane 1: ≈744 bp SPLUNC1-ORF; Lane 2: empty vector DNA; Lane 3: DNA Maker I
molecular marker. Predicted amino acid sequence of recombinant SPLUNC1 (B) Nucleotide and predicted amino acid sequence of SPLUNC1cDNA. The complete SPLUNC1cDNA sequence consisting of 1,076 nucleotides and open reading
frame of 768 bp. The initiation ATG and stop TAA codons are shaded. The distal polyadenylation signal sequence AATAAA are shaded. The nucleotide and its amino acids are
numbered along the left margins respectively. The putative signal peptide sequence is shown as capital italicized letters. molecular mass of 26,486.43 Da (≈26.49 kDa). The recombinant
protein corresponded to the expected molecular mass of 25.47
kDa as judged by SDS-PAGE and the expression product was
detected in the supernatant of the Pichia pastoris (Figures 3A,
B). The protein consists of 118 hydrophobic amino acids (Ala,
Ile, Leu, Phe, Trp, and Val), 57 Polar amino acids (Asn, Cys, Gln, Ser, Thr, and Tyr), 12 strongly basic amino acids (Lys and Arg)
and 19 strongly acidic amino acids (aspartic acid [Asp] and
glutamic acid [Glu]). Alignment of amino acid sequences of
SPLUNC1 cDNA against amino acid sequences from other
species revealed SPLUNC1 protein shared 98.4%, 96.9%, 94.5%,
90.2%, 80.8%, 78.4%, 78.3%, 72.5%, 72.3%, 68.8% identity with 739 www.ajas.info Shen et al (2017) Asian-Australas J Anim Sci 30:736-742 (A)
350
351
352
353
354
355
356
357
(B)
358
359
360
361
362
(C)
363
364
365
366
367
368
369
Figure 3. Expression, purification and biological activity analysis of SPLUNC1. (A) Expression of
370
SPLUNC1 protein. Gels were stained with Coomassie blue and 20 μL samples were loaded perlane. 371
Key: Lane 1: standard protein molecular weight marker; Lane 2: pPIC9K empty expression vector
372
(negative control); Lane 3-7: transformed Pichia pastoris supernatant after 0, 24, 48, 72, 96 h of
373
induction with methyl alcohol, respectively. (B) Purification of SPLUNC1 protein. Gels were stained
374
0
0.2
0.4
0.6
0.8
1
1.2
0μg/ml
1μg/ml
3μg/ml 10μg/ml
16S rRNA relative level
groups
**
** **
**
**
25.47 kDa
25 kDa
35 kDa
1 2 3 4 5 6
7
116 kDa
14.4 kDa
1 2
35 kDa
116 kDa
25 kDa
14.4 kDa
25.47 kDa
0 μg/mL
1 μg/mL
3 μg/mL 10 μg/mL
Groups
3. Expression, purification and biological activity analysis of SPLUNC1. (A) Expression of SPLUNC1 protein. Gels were stained with Coomassie blue and 20 μL samples were
erlane. Predicted amino acid sequence of recombinant SPLUNC1 NP_001005727.1), Chinchilla lanigera
(accession no. NP_001269294.1), Mus musculus (accession no. NP_035256.2), and Rattus norvegicus (accession no. NP_742028.1),
respectively (Figure 4). The phylogenetic tree generated from
with Coomassie blue and 10 μL samp
375
25.47 kDa SPLUNC1 protein; Lanes
376
analysis of Mycoplasma ovipneumoniae
377 those of SPLUNC1 cDNA from Ovis aries (accession no. NP_
001288334.1), Capra hircus (accession no. XP_005688516.1),
Pantholops hodgsonii (accession no. XP_005979709.1), Bos taurus
(accession no. NP_776851.1), Felis catus (accession no. XP_0069
29910.1), Homo sapiens (accession no. NP_001230122.1), Sus
scrofa (accession no. NP_001005727.1), Chinchilla lanigera
(accession no. NP_001269294.1), Mus musculus (accession no. NP_035256.2), and Rattus norvegicus (accession no. NP_742028.1),
respectively (Figure 4). The phylogenetic tree generated from
with Coomassie blue and 10 μL samp
375
25.47 kDa SPLUNC1 protein; Lanes
376
analysis of Mycoplasma ovipneumoniae
377 Predicted amino acid sequence of recombinant SPLUNC1 Key: Lane 1: standard protein molecular weight marker; Lane 2: pPIC9K empty expression vector (negative control); Lane 3-7: transformed Pichia pastoris supernatant
4, 48, 72, 96 h of induction with methyl alcohol, respectively. (B) Purification of SPLUNC1 protein. Gels were stained with Coomassie blue and 10 μL samples were loaded
Key: Lane 1: purified recombinant 25.47 kDa SPLUNC1 protein; Lanes 2: tandard protein molecular weight marker. (C) The effect analysis of Mycoplasma ovipneumoniae
n recombinant SPLUNC1 protein of Argali. (A) (A)
25.47 kDa
25 kDa
35 kDa
1 2 3 4 5 6
7
116 kDa
14.4 kDa (B)
1 2
35 kDa
116 kDa
25 kDa
14.4 kDa
25.47 kDa (B) (C)
0
0.2
0.4
0.6
0.8
1
1.2
0μg/ml
1μg/ml
3μg/ml 10μg/ml
16S rRNA relative level
groups
**
** **
**
**
0 μg/mL
1 μg/mL
3 μg/mL 10 μg/mL
Groups (C) 369
Figure 3. Expression, purification and biological activity analysis of SPLUNC1. (A) Expression of
370
SPLUNC1 protein. Gels were stained with Coomassie blue and 20 μL samples were loaded perlane. 371
Key: Lane 1: standard protein molecular weight marker; Lane 2: pPIC9K empty expression vector
372
(negative control); Lane 3-7: transformed Pichia pastoris supernatant after 0, 24, 48, 72, 96 h of
373
Figure 3. Expression, purification and biological activity analysis of SPLUNC1. (A) Expression of SPLUNC1 protein. Gels were stained with Coomassie blue and 20 μL samples were
loaded perlane. Key: Lane 1: standard protein molecular weight marker; Lane 2: pPIC9K empty expression vector (negative control); Lane 3-7: transformed Pichia pastoris supernatant
after 0, 24, 48, 72, 96 h of induction with methyl alcohol, respectively. (B) Purification of SPLUNC1 protein. Gels were stained with Coomassie blue and 10 μL samples were loaded
per lane. Key: Lane 1: purified recombinant 25.47 kDa SPLUNC1 protein; Lanes 2: tandard protein molecular weight marker. (C) The effect analysis of Mycoplasma ovipneumoniae
growth in recombinant SPLUNC1 protein of Argali. nucleotide sequences (Figure 4) revealed a close relationship
between Argali and Ovis aries. were loaded per lane. Key: Lane 1: purified recombinant
ndard protein molecular weight marker. (C) The effect
wth in recombinant SPLUNC1 protein of Argali. those of SPLUNC1 cDNA from Ovis aries (accession no. NP_
001288334.1), Capra hircus (accession no. XP_005688516.1),
Pantholops hodgsonii (accession no. XP_005979709.1), Bos taurus
(accession no. NP_776851.1), Felis catus (accession no. XP_0069
29910.1), Homo sapiens (accession no. NP_001230122.1), Sus
scrofa (accession no. Effects of recombination protein on MOh frame of 768 bp, which encoded a 26.49 kDa protein composed
of 255 amino acids, containing one initiation codon ATG and
one termination codon TAA. The 5’UTR ends at 63 nt and the
3’UTR ends at 245 bp in Argali, while the 5’UTR ends at 58 nt,
the 3’UTR ends at 265 bp in sheep. The amino acid sequence
homology of Argali and sheep is 98.4% by basic local alignment
search tool (BLAST). When we designed primer of expression
of Argali SPLUNC1 protein, removed the signal peptide, and
added His tag sequence, thus the target protein’s molecular weight
becomes 25.47 kDa. The results from the test of recombination SPLUNC1 protein
on MO showed that the expression product could inhibit MO
proliferation very well (p<0.01), which proved the expressed
protein had biological activity (Figure 3C). Western blot analysis of recombinant SPLUNC1 proteinh Western blot analysis of recombinant SPLUNC1 proteinh Western blot analysis of recombinant SPLUNC1 protein
The results showed that a distinct protein bands appeared at
25.47 KDa place (Figure 5) and the recombinant proteins can
bind specifically His tag antibody by Western blot analysis so Western blot analysis of recombinant SPLUNC1 protein
The results showed that a distinct protein bands appeared at The results showed that a distinct protein bands appeared at
25.47 KDa place (Figure 5) and the recombinant proteins can
bind specifically His tag antibody by Western blot analysis, so
the expression of recombinant SPLUNC1 protein was confirmed. 740 www.ajas.info Shen et al (2017) Asian-Australas J Anim Sci 30:736-742 Figure 4. Computer genetic distance and generated phylogenetic tree from SPLUNC1cDNA sequences. Sequences were derived from GenBank databases. Figure 4. Computer genetic distance and generated phylogenetic tree from SPLUNC1cDNA sequences. Sequences were derived from GenBank databases. DISCUSSION The SPLUNC1 cDNA sequence of the Xinjiang Wild Argali for
test was obtained by RT-PCR, 5'RACE, and 3'RACE methods. The SPLUNC1 cDNA was 1,076 bp long with an open reading Bacterial infection in the lung is a major cause of mortality
and morbidity, especially in high-risk groups such as immu
nocompromised mammalian species, the young and those with
other underlying pulmonary diseases such as pneumonias. The
antimicrobial function of SPLUNC1 probably plays a critical
role in host defense against pathogens not only in maintaining
homoeostasis of healthy individuals but also in protecting mam
mal whose lungs are compromised by a chronic lung disease. Results from both in vitro and in vivo studies indicate the an
timicrobial function of SPLUNC1 and suggest a defensive role
of SPLUNC1 in airways exposed to bacterial infection [10]. 1
M
kDa
100
30
25
14
25.47 kDa
Figure 5. Western blotting of purified SPLUNC1 protein. Key: M: Protein marker; 1:
recombinant SPLUNC1 purification of the sample. SPLUNC1 has been shown to be a key role in promoting host
innate immunity against Mycoplasma pneumoniae (Mp), and
previous studies have verified in mouse. Recombinant mSPLUNC1
protein inhibits Mp growth in a dose-dependent manner [11]. Using SPLUNC1 knockout and transgenic mice to test this
model against Mp in vivo, found that SPLUNC1 was highly
expressed in the trachea and intrapulmonary, and undetectable SPLUNC1 has been shown to be a key role in promoting host
innate immunity against Mycoplasma pneumoniae (Mp), and
previous studies have verified in mouse. Recombinant mSPLUNC1
protein inhibits Mp growth in a dose-dependent manner [11]. Using SPLUNC1 knockout and transgenic mice to test this
model against Mp in vivo, found that SPLUNC1 was highly
expressed in the trachea and intrapulmonary, and undetectable Figure 5. Western blotting of purified SPLUNC1 protein. Key: M: Protein marker; 1:
recombinant SPLUNC1 purification of the sample. 741 www.ajas.info Shen et al (2017) Asian-Australas J Anim Sci 30:736-742 and vicissitude of argali, Ovis ammon, in China. Biodivers Sci 2008;
16:197-204. in any tissues or bronchoalveolar lavage (BAL) fluid harvested
from SPLUNC1-/- mice, Mp levels in SPLUNC1-/- mouse lungs
were threefold higher than those in the SPLUNC1+/+ littermate
control mice after Mp infection 24 h. DISCUSSION Meanwhile, hSPLUNC1 over
expression in mouse airways resulted in doubling of SPLUNC1
protein in BAL fluid, lung Mp load in hSPLUNC1 transgenic
mice was about threefold lower than that in the wild-type litter
mate control mice after Mp infection. A critical in vivo role of
SPLUNC1 was demonstrated in host defense against Mp [12]. So we tested the activity of Argali recombinant SPLUNC1 protein
by using real-time PCR in this research, found that Argali re
combinant SPLUNC1 protein could remarkably inhibit the
growth of MO, have biological activity, and lay the foundation
of subsequent function study for this gene. 2. Sun YM, Chen KL, Shen W, Cui RP, Lu HF. Cloning and sequence
analysis of wild argali ISG15 cDNA. Asian-Australas J Anim Sci
2014;27:561-6. 3. Di YP, Harper R, Zhao Y, et al. Molecular cloning and characterization
of spurt, a human novel gene that is retinoic acid-inducible and
encodes a secretory protein specific in upper respiratory tracts. J
Biol Chem 2003;278:1165-73. 4. Liu Y, Bartlett JA, Di ME, et al. SPLUNC1/BPIFA1 contributes to
pulmonary host defense against Klebsiella pneumoniae respiratory
infection. Am J Pathol 2013;182:1519-31. 5. McGillivary G, Bakaletz LO. The Multifunctional host defense peptide
SPLUNC1 is critical for homeostasis of the mammalian upper airway. PloS ONE 2010;5:e13224. Mycoplasma pneumonia infections (also known as infections
of pleuropneunonia of sheep and goats), have clinical features of a
high fever, cough, chest and pleural serous and fibrinous inflam
mation, can be acute and chronic, and has a high mortality. The
main method for the treatment of the disease is with antibiotics
[5,13]. There is a difference in sensitivity of MO to antimicrobial
agents, resulting in a persistent disease in the sheep after the drug
was administered. However, Argali recombinant SPLUNC1
protein could remarkably inhibit the growth of MO (data not
shown), so there are good application prospects for development
and utilization of Argali SPLUNC1 gene and treatment of infec
tious diseases of lungs in sheep. 6. Sambrook J, Russell WD. Molecular cloning: A laboratory manual. 3rd edition. Cold Spring Harbor: NY : Cold Spring Harbor Laboratory
Press; 2001. 7. Marion MB. A rapid and sensitive method for the quantitation of
microgram quantities of protein utilizing the principle of protein-
dye binding. Anal Biochem 1976;72:248-54. 8. Chu HW, Gally F, Thaikoottathil J, et al. SPLUNC1 regulation in
airway epithelial cells: role of toll-like receptor 2 signaling. Respir
Res 2010;5:155-62. 9. ACKNOWLEDGMENTS 12. Gally F, Di YP, Smith SK, et al. SPLUNC1 promotes lung innate
defense against Mycoplasma pneumoniae infection in mice. Am J
Pathol 2011;178:2159-67. The research work was funded by a grant from the National
Natural Science Foundation of P. R. China (NO: 31460686). 13. Hao RX, Hao YQ, Xu CG, Liu B, Zhang C. Isolation of Mycoplasma
ovipneumoniae a sheep identification and culture characteristics. J Inner Mongolia Agric Univ 2013;34:10-3. DISCUSSION Xiao DL, Zhong LL, Liang M, et al. Expression of SPLUNC1 in
human bronchial epithelial cells induced by Mycoplasma peumoniae
lipid-associated membrane proteins in human bronchial epithelial
cells. J Clin Res 2013;30:1041-5. CONFLICT OF INTEREST 10. Di YP. Functional roles of SPLUNC1 in the innate immune response
against Gram-negative bacteria. Biochem Soc Trans 2011;39:1051-5. We certify that there is no conflict of interest with any financial
organization regarding the material discussed in the manuscript. 11. Chu HW, Thaikoottathil J, Rino JG, et al. Function and regulation
of SPLUNC1 protein in Mycoplasma infection and allergic infla
mmation. J Immunol 2007;179:3995-4002. REFERENCES J Inner Mongolia Agric Univ 2013;34:10-3. 1. Yu YQ , Ji MZ, Liu CG, Li KC, Guo SG. Geographical distribution 1. Yu YQ , Ji MZ, Liu CG, Li KC, Guo SG. Geographical distribution 742 www.ajas.info
|
https://openalex.org/W4391556902
|
https://hal.science/hal-04342085/document
|
French
| null |
Two Decades Of Industrial Policy In Morocco : What Has Been Achieved ?
| null | 2,023
|
public-domain
| 13,936
|
To cite this version: A L Assimi. Deux Decennies De Politique Industrielle Au Maroc : Quel Bilan ?. African Scientific
Journal, A paraître, 10.5281/zenodo.10256718. hal-04342085 Deux Decennies De Politique Industrielle Au Maroc :
Quel Bilan ?
A L Assimi A L Assimi Auteur 1 : AL ASSIMI Mohammed AL ASSIMI Mohammed (Enseignant-chercheur)
Ecole Supérieure de Technologie / Université Ibn Tofail-Kenitra (Maroc)
Laboratoire des Sciences Economiques et Politiques Publiques (LSEPP) AL ASSIMI Mohammed (Enseignant-chercheur) Déclaration de divulgation : L’auteur n’a pas connaissance de quelconque financement qui
pourrait affecter l’objectivité de cette étude. Conflit d’intérêts : L’auteur ne signale aucun conflit d’intérêts. Conflit d’intérêts : L’auteur ne signale aucun conflit d’intérêts. Pour citer cet article : AL ASSIMI .M (2023) « Deux Decennies De Politique Industrielle Au Maroc : Quel
Bilan ? », African Scientific Journal « Volume 03, Numéro 21 » pp: 0055 – 0079. Pour citer cet article : AL ASSIMI .M (2023) « Deux Decennies De Politique Industrielle Au Maroc : Quel
Bilan ? », African Scientific Journal « Volume 03, Numéro 21 » pp: 0055 – 0079. Date de soumission : Novembre 2023 Date de soumission : Novembre 2023 Date de publication : Décembre 2023 Date de publication : Décembre 2023 www.africanscientificjournal.com Page a
DOI : 10.5281/zenodo.10256718
Copyright © 2023 – ASJ DOI : 10.5281/zenodo.10256718
Copyright © 2023 – ASJ African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre 2023
Deux Decennies De Politique Industrielle Au Maroc : Quel Bilan ?
Two Decades Of Industrial Policy In Morocco : What Has Been Achieved ?
Auteur 1 : AL ASSIMI Mohammed
AL ASSIMI Mohammed (Enseignant-chercheur)
Ecole Supérieure de Technologie / Université Ibn Tofail-Kenitra (Maroc)
Laboratoire des Sciences Economiques et Politiques Publiques (LSEPP) African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre 2023
Deux Decennies De Politique Industrielle Au Maroc : Quel Bilan ? Two Decades Of Industrial Policy In Morocco : What Has Been Achieved ? Auteur 1 : AL ASSIMI Mohammed
AL ASSIMI Mohammed (Enseignant-chercheur)
Ecole Supérieure de Technologie / Université Ibn Tofail-Kenitra (Maroc)
Laboratoire des Sciences Economiques et Politiques Publiques (LSEPP) African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre 2023 Deux Decennies De Politique Industrielle Au Maroc : Quel Bilan ? Two Decades Of Industrial Policy In Morocco : What Has Been Achieved ? Two Decades Of Industrial Policy In Morocco : What Has Been Achieved HAL Id: hal-04342085
https://hal.science/hal-04342085v1
Submitted on 13 Dec 2023 L’archive ouverte pluridisciplinaire HAL, est
destinée au dépôt et à la diffusion de documents
scientifiques de niveau recherche, publiés ou non,
émanant des établissements d’enseignement et de
recherche français ou étrangers, des laboratoires
publics ou privés. HAL is a multi-disciplinary open access
archive for the deposit and dissemination of sci-
entific research documents, whether they are pub-
lished or not. The documents may come from
teaching and research institutions in France or
abroad, or from public or private research centers. Public Domain www.africanscientificjournal.com www.africanscientificjournal.com Page a African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre 2023 www.africanscientificjournal.com Résumé Cet article s’interroge sur l’impact de la politique industrielle adoptée au Maroc ces deux
dernières décennies sur la compétitivité de l’industrie marocaine, principalement les secteurs
clés désignés comme une locomotive de développement de l’économie marocaine. A travers la
mise en place de trois plans : Plan Emérence, le Pacte National pour l’Emergence Industrielle
et le Plan d’Accélération Industrielle. Dans un contexte national, régional et international
spécial. L’objectif principal consiste à faire une évaluation et une analyse de l’approche
industrielle marocaine par la spécialisation dans les Métiers Mondiaux du Maroc. Moyennant
un ensemble d’indicateurs de compétitivité durant la période 2005-2020. Aussi, de déterminer
les facteurs qui encouragent et/ou ceux qui entravent la compétitivité de l’industrie marocaine. Cependant, les résultats obtenus, démontrent clairement que malgré les ressources mobilisées
dans le cadre de la nouvelle politique industrielle. Le Maroc n’a pas encore consolidé une
véritable politique industrielle capable de surmonter les contraintes structurelles à son
développement. Exception faite pour les secteurs de l’Automobile et d’une façon moins
importante le secteur de l’Aéronautique. Car, la compétitivité de l’industrie marocaine est
compromise par des contraintes d’ordre institutionnel (climat d’affaire, limites de bonne
gouvernance, compétitivité des PME…) et des contraintes d’ordre transversal (foncier
industriel, la problématique du financement et d’accès au crédit, un système de formation non
adapté au besoin de la compétitivité…etc). Mots-clés : Politique industrielle, compétitivité, spécialisation, Métiers Mondiaux du Maroc. www.africanscientificjournal.com Page 55 African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre 2023 African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre 2023 African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre 2023 www.africanscientificjournal.com Introduction Les efforts ayant été accomplis au Maroc dans le cadre des Plans Émergences, durant ces
dernières années, méritent d’être relevés compte tenu de la dynamique qui a été déclenchée, et
des mesures apportées par ces plans à l’industrie nationale et particulièrement, les Métiers
Mondiaux du Maroc, dans un contexte international, régional et national spécial. Aussi de
s’interroger sur l’impact de la nouvelle politique industrielle sur la compétitivité du secteur
industriel. Cependant, l’évaluation de la compétitivité de l’industrie marocaine et la pertinence
de la nouvelle politique industrielle, durant la période 2005-2020. Cette dernière, qu’a été
marquée par la mise en place du Plan Émergence(PE) en 2005. Puis le Pacte National pour
l’Emergence Industrielle (PNEI) en 2009 et le Plan d’Accélération Industrielle en 2014-2020
d’une part et l’entrée en vigueurs de plusieurs accords de libre-échange, notamment, l’accord
Maroc et l’Union Européenne, Maroc-USA, Maroc-Turquie et l’accord d’Agadir (Maroc,
Tunisie, Egypte et Jordanie)… d’autre part et un contexte économique national marqué par un
creusement du déficit de la balance commerciale, une corrélation étroite entre ce creusement et
l’entrée en vigueur des ALE conclus par le Maroc. Ce déficit soulève la problématique de la
compétitivité de l’industrie marocaine (l’offre exportable) en termes de produits à forte valeur
technologique. D’où l’importance de faire une évaluation de la politique industrielle marocaine
adopter au Maroc durant ces deux dernières décennies. À ce niveau, nous partons de la question
fondamentale suivante : Est-ce que les réformes adoptées au Maroc ces dix dernières années
dans le cadre de la nouvelle politique industrielle, l’ont suffisamment préparée pour
devenir compétitive et en mesure de se spécialiser dans des secteurs à rendement élevé et
à forte valeur ajoutée ? L’objectif principal consiste donc à faire une évaluation de la nouvelle approche de la politique
industrielle par la spécialisation dans les Métiers Mondiaux du Maroc moyennant un ensemble
d’indicateurs de compétitivité. Par ailleurs, l’évaluation de la compétitivité de l’industrie dans
le cadre de la nouvelle politique industrielle au Maroc, revient à mesurer la contribution de cette
industrie au développement de l’économie nationale. À cet effet, notre positionnement
épistémologique s’appuie sur une analyse interprétative purement qualitative, basée sur une
étude de cas portant sur des secteurs industriels et sur certains indicateurs et grandeurs
économiques. Abstract This article examines the impact of the industrial policy adopted in Morocco over the last two
decades on the competitiveness of Moroccan industry, and mainly in the key sectors designated
as the locomotive for development of the Moroccan economy. Through the implementation of
three plans : Emergence Plan, National Pact for Industrial Emergence and the Industrial
Acceleration Plan. In a special national, regional and international context. The main objective
is to assess and analyze of the Moroccan industrial approach through specialization in
Morroco's Global Trades. Using a set of competitiveness indicators. To analyze Morroco's
industrial choices during this period, the results obtained and to highlight the strengths and
weaknesses facing the industrial sectors analyzed during the period 2005-2020, we would also
try to determine the factors which encourage and/or those which hinder the competitiveness of
Moroccan industry. However, the results obtained clearly demonstrate that despite the resources
mobilized under the new industrial policy. Morocco has yet to consolidate a genuine industrial
policy capable of overcoming the structural constraints to its development. The exceptions are
the automotive sector and, to a lesser extent, the aerospace industry. The competitiveness of
Moroccan industry is compromised by institutional constraints (business climate, limits to good
governance, competitiveness of SMEs, etc.) and cross-cutting constraints (industrial land, the
problem of financing and access to credit, a training system not adapted to the needs of
competitiveness, etc.). Keywords : Industrial policy, competitiveness, specialization, World Trades of Morocco. www.africanscientificjournal.com Page 56 African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre 2023 African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre 2023 www.africanscientificjournal.com Introduction Car, le raisonnement que nous allons adopter est basé sur une méthode déductive
,qui va nous permettre de déduire un certain nombre de conclusions sur le niveau de
compétitivité de notre industrie en égard des réformes qu’ont été adoptées ces deux dernières
décennies, les facteurs déterminants de cette compétitivité, particulièrement, en termes de
spécialisation sectorielle et géographique. www.africanscientificjournal.com Page 57 African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre 2023 African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre 2023 Au niveau de l’étude empirique, nous avons utilisé des données secondaires (des statistiques et
des données établies par certaines institutions marocaines publiques et /ou privés) ainsi, que les
rapports et études de quelques organismes internationaux spécialisés dans la question
industrielle. Ainsi, notre analyse est basée sur les indicateurs les plus utilisés dans les rapports
d’analyse économique pour évaluer la compétitivité industrielle. Certains sont traditionnels et
autorisent une lecture directe, comme la production industrielle, la valeur industrielle,
l’investissement industriel, les exportations industrielles, l’emploi industriel, le chiffre d’affaire
à l’export et le taux de croissance annuel moyen (TCAM). Par contre d’autre permettent une
décomposition qui nécessite des interprétations plus complètes qui doivent recouvrir plusieurs
variables, comme le calcul et la décomposition de la variation de la part de marché au niveau
mondial qui est considérée comme l’un des indicateurs les plus importants pour mesurer le
niveau de compétitivité des exportations marocaines. En effet, l’analyse de la compétitivité de
l’industrie marocaine nécessite l’étude de plusieurs paramètres, à savoir : le taux d’ouverture,
le taux de pénétration, la structure des exportations, les produits dynamiques, les avantages
comparatifs, la demande étrangère adressée au Maroc…etc. Le présent article sera décliné en trois axes : le premier présentera le cadre théorique de la
politique industrielle et un rappel du choix de politique industrielle au Maroc. Le deuxième axe
sera consacré à l’évaluation de l’industrie marocaine à travers les principales grandeurs
économiques. Tandis que le dernier axe analysera les résultats et la sous performance de
l’industrie marocaine. www.africanscientificjournal.com I.
LA POLITIQUE INDUSTRIELLE : FONDEMENTS THEORIQUES Dans la littérature, la politique industrielle se définit de plusieurs manières (Humbert, 1995),
avec un accent particulier sur les différents aspects de l’intervention de l’État en appui à
l’industrialisation. Il y a autant de définitions et d’explications de la politique industrielle que
de points de vue relatifs à son opportunité. Le point central de ces définitions est la
reconnaissance du fait que la politique industrielle comprend un effort délibéré de promotion
de l’industrie en général ou des secteurs industriels spécifiques jugés importants pour la
croissance économique et le développement en particulier. En effet, la politique industrielle
selon Robinson .J. A, (2010) a été couronnée de succès dans le cas où les décideurs politiques,
qui l’ont mise en œuvre, soit ont souhaité un décollage industriel, soit ont été forcés de
s’engager dans cette voie en raison des incitations engendrées par les institutions politiques. www.africanscientificjournal.com Page 58 African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre 2023 African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre 2023
Ainsi, Reich. R, (1997) a définit la politique industrielle comme une série d’actions
gouvernementales conçues pour appuyer les industries qui ont des potentialités majeures en
termes d’exportations, de création d’emplois et d’appui à la production des infrastructures. African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre 2023 African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre 2023 Ainsi, Reich. R, (1997) a définit la politique industrielle comme une série d’actions
gouvernementales conçues pour appuyer les industries qui ont des potentialités majeures en
termes d’exportations, de création d’emplois et d’appui à la production des infrastructures. Ainsi, Reich. R, (1997) a définit la politique industrielle comme une série d’actions
gouvernementales conçues pour appuyer les industries qui ont des potentialités majeures en
termes d’exportations, de création d’emplois et d’appui à la production des infrastructures. Une autre définition de la politique industrielle a été donnée par Ha-Joon Chang (2006) qui
considère les politiques industrielles comme « des actions gouvernementales visant à appuyer
la stimulation de la production et les capacités technologiques dans les industries jugées
stratégiques pour le développement national ». Cette définition implique que les gouvernements
essaient de faire une distinction entre les activités, les secteurs et les agents. www.africanscientificjournal.com I.
LA POLITIQUE INDUSTRIELLE : FONDEMENTS THEORIQUES Ces
actions peuvent être plus ou moins explicites ou sélectives, avec ou sans accompagnement
financier, mais elles sont d’abord déterminées par leur capacité à édicter les règles
d’organisation et de comportement permettant des formes de coopération stratégique explicite
ou implicite (Van Ha Nguyen ,2011). Elles sont appliquées à des processus précis, au profit des
cibles précises. Autrement dit, « l’indifférenciation » ou la « neutralité » éventuelle des
politiques industrielles s’exerce toujours à l’intérieur d’un ensemble de choix politiques qui
sont par nature, ni indifférenciées ni neutres » (Niosi. J et Bellon. B, 1995). C’est cela qui est
censé assurer les conditions les plus favorables à l’industrie nationales sans que pour autant les
autorités publiques définissent un projet d’ensemble sur ce que devraient être les structures et Quant à l’approche offensive elle qualifie la politique industrielle comme active et agressive
(Bellon et Niosi.1995) donnent une nouvelle dimension à la politique industrielle, cette dernière
est matérialisée par l’ensemble des actions publique ayant pour objectif d’agir directement ou
indirectement sur la création, le développement et la diffusion de la production industrielle et
de générer à long terme des avantages construits dans le cadre des mécanismes de marché. Ces
actions peuvent être plus ou moins explicites ou sélectives, avec ou sans accompagnement
financier, mais elles sont d’abord déterminées par leur capacité à édicter les règles
d’organisation et de comportement permettant des formes de coopération stratégique explicite
ou implicite (Van Ha Nguyen ,2011). Elles sont appliquées à des processus précis, au profit des
cibles précises. Autrement dit, « l’indifférenciation » ou la « neutralité » éventuelle des
politiques industrielles s’exerce toujours à l’intérieur d’un ensemble de choix politiques qui
sont par nature, ni indifférenciées ni neutres » (Niosi. J et Bellon. B, 1995). C’est cela qui est
censé assurer les conditions les plus favorables à l’industrie nationales sans que pour autant les
autorités publiques définissent un projet d’ensemble sur ce que devraient être les structures et
l’évolution de l’industrie et une stratégie comportant des interventions coordonnées pour
réaliser ce projet. Cet ensemble de politiques distinctes constitue des stratégies de suppression
des bons fonctionnements naturels des marchés. I.
LA POLITIQUE INDUSTRIELLE : FONDEMENTS THEORIQUES Elle implique
également que la distinction est basée sur l’idée selon laquelle les activités, les secteurs et les
agents privilégiés sont plus susceptibles que d’autres de favoriser le développement
économique. Car, le territoire est considéré comme un moteur de développement d’après
Courlet, (2007). En effet, si la proximité occupe une place importante dans l’industrie
marocaine, le territoire n’a qu’une place modeste dans notre politique (Piermay, 2011). Cependant, une analyse au niveau du paysage économique mondial indique que tous les pays
qui se sont industrialisés ont réussi à s’industrialiser en adoptant des politiques et des
programmes généralement financés par l’état pour les aider à passer de la production des
matières premières et des activités à rendements décroissants (concurrence parfaite) à des
activités à rendements croissants (concurrence imparfaite). Par conséquent, un aspect important
de la politique industrielle est qu’elle est sélective en matière d’activités économiques. Elle fait
de la discrimination et sélectionne de certaines industries et secteurs, et elle est spécifiquement
conçue pour chaque industrie ou secteur choisi sur un territoire national donné (Kwabena
Nyarko Otoo ,2013). C’est pour cette raison que les politiques industrielles ont été perçues
comme des efforts de l’état visant à créer des gagnants et des perdants. Le consensus qui se
dégage autour de la politique industrielle est que, en elle-même, elle est susceptible d’être moins
efficace dans la construction d’une base industrielle durable pour une économie. Elle doit être
considérée comme composante d’une stratégie plus globale de développement qui se focalise
sur l’amélioration de l’accumulation de capital et de savoirs (Cheval, 1983). Les politiques
industrielles doivent alors être perçues comme prenant en compte toutes les politiques conçues
pour appuyer l’industrie, y compris des mesures fiscales et monétaires incitatives pour des
investissements, des investissements publics et des programmes publics, des mesures
incitatives pour l’investissement dans la recherche et développement, des programmes clés de
création de « géants » dans des secteurs stratégiques, et des politiques d’appui aux petites et
moyennes entreprises. Les politiques industrielles du passé avaient du mal à trouver leur
justification dans le cadre de la théorie standard qui rejette systématiquement l’intervention de www.africanscientificjournal.com Page 59 African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre 2023 African Scientific Journal
ISSN : 2658-9311
V l
3 N
é
21 Dé
b l’État. En effet, cette théorie estime l’intervention de l’État ne se justifie qu’en cas
« d’imparfaite allocation des ressources par le marché ». www.africanscientificjournal.com I.
LA POLITIQUE INDUSTRIELLE : FONDEMENTS THEORIQUES Certain auteur (Morvan, 1991) estime
dans ce sens qu’il y a absence de fondement logique de toute politique industrielle. Selon lui la
politique industrielle ne devrait pas exister. Conçue comme un ensemble de relation entre l’État
et l’industrie, elle ne devrait pas avoir sa place dans un système économique réglé par les
mécanismes de marché, organisé selon les lois de la concurrence et dominé par le dogme de la
propriété privée ». Ainsi, la pensée défensive soutient l’idée de l’intervention en tant que
mesure correctrice des échecs de marché. La définition de l’économiste De Bandt J (1993)
résume cette approche. « Les échecs de marché tendent à se généraliser. Les solutions cherchant
à les réduire dans le cadre de la logique du marché ne sont pas satisfaisantes. Il en résulte des
exigences accrues de politique industrielle ». l’État. En effet, cette théorie estime l’intervention de l’État ne se justifie qu’en cas
« d’imparfaite allocation des ressources par le marché ». Certain auteur (Morvan, 1991) estime
dans ce sens qu’il y a absence de fondement logique de toute politique industrielle. Selon lui la
politique industrielle ne devrait pas exister. Conçue comme un ensemble de relation entre l’État
et l’industrie, elle ne devrait pas avoir sa place dans un système économique réglé par les
mécanismes de marché, organisé selon les lois de la concurrence et dominé par le dogme de la
propriété privée ». Ainsi, la pensée défensive soutient l’idée de l’intervention en tant que
mesure correctrice des échecs de marché. La définition de l’économiste De Bandt J (1993)
résume cette approche. « Les échecs de marché tendent à se généraliser. Les solutions cherchant
à les réduire dans le cadre de la logique du marché ne sont pas satisfaisantes. Il en résulte des
exigences accrues de politique industrielle ». Quant à l’approche offensive elle qualifie la politique industrielle comme active et agressive
(Bellon et Niosi.1995) donnent une nouvelle dimension à la politique industrielle, cette dernière
est matérialisée par l’ensemble des actions publique ayant pour objectif d’agir directement ou
indirectement sur la création, le développement et la diffusion de la production industrielle et
de générer à long terme des avantages construits dans le cadre des mécanismes de marché. I.
LA POLITIQUE INDUSTRIELLE : FONDEMENTS THEORIQUES À chacune correspond en fait une catégorie
d’entraves et c’est de la distinction de ces catégories d’entraves que résulte la distinction entre
des politiques qui sont cependant toutes destinées à assurer l’efficience des marchés et par voie
de conséquence assurer la performance de l’industrie, même si seule la dernière reçoit le nom
de politique industrielle . Billote, (1997) considère dans ce sens que «les développements
récents de la théorie des échanges intègrent les développements de l’économie industrielle et
notamment l’impact du changement technique et sa performance. www.africanscientificjournal.com Page 60 African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre 2023 African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre 2023 African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre 202 En effet, une autre approche (l’économie géographique) montre que les pouvoirs publics
peuvent modifier les choix des implantations, car la localisation géographique des entreprises
dépend de plusieurs facteurs, parmi lesquels les relations interindustrielles et l’importance des
externalités. Dans le cas des activités intensives en connaissances, polarisées dans quelques
régions, la mondialisation de la production des connaissances par les grands groupes est
motivée par la recherche de l’excellence technologique. Au-delà des activités de recherche-
développement, les entreprise recherchent des territoires riches en infrastructures de transport
et de télécommunications et fiscalement compétitifs et des territoires avantagés en terme de
capital humain et de compétences (Demazière, 2006). Notre analyse dans cet article va nous permettre de prendre connaissance des initiatives prises
par les pouvoirs publics pour soutenir la compétitivité au sein de leurs territoires. Toutefois,
notre objectif n’est pas d’établir un historique de l’action publique en la matière mais de
présenter, de manière synthétique, le cadre d’intervention des acteurs publics. Concernant l’initiative publique au Maroc, l’État rencontre un certain nombre de difficultés
dans la prise de décision et la mise en œuvre de ses actions dans l’économie. En effet,
l’expansion continue du commerce mondial, la globalisation financière, l’extension de la sphère
du marché, la montée de l’incertitude constituent autant de contraintes qui ont modifié les
politiques monétaires ou budgétaires nationales. Si l’objectif de compétitivité recherché par
notre pays peut prendre la forme de mesures nationales (fiscalité, création des plates-formes,
des organismes nationaux, etc.) il débouche également sur une mise en valeur des territoires,
car ces politiques ne visent pas seulement l’équité territoriale mais la valorisation des
spécificités et des potentialités économiques existantes. II.
RAPPEL DES CHOIX DE LA POLITIQUE INDUSTRIELLE MAROCAINE Au Maroc deux périodes peuvent être distinguées : de l’indépendance à l’application du
Programme d’Ajustement Structurel et de l’après PAS à nos jours. Car, le Maroc a choisi au
lendemain de son indépendance, une politique industrielle basée sur la substitution aux
importations qui lui a permis de développer une industrie légère de biens de consommation. Au Maroc deux périodes peuvent être distinguées : de l indépendance à l application du
Programme d’Ajustement Structurel et de l’après PAS à nos jours. Car, le Maroc a choisi au
lendemain de son indépendance, une politique industrielle basée sur la substitution aux
importations qui lui a permis de développer une industrie légère de biens de consommation. Ainsi, l’interventionnisme des années soixante, qui avait pour objectif de doter le Maroc d’un
système industriel ancré sur le marché intérieur, a montré ses limites et notre économie a été
marquée par d’importants déséquilibres économiques et financiers. Toutefois, il a été amené, à
partir des années quatre-vingt, dans le cadre de la politique d’ajustement structurel, à
abandonner ce choix, au profit d’une stratégie, axée sur la promotion des exportations, la
libéralisation du commerce et l’encouragement du secteur privé. Cette stratégie de promotion
des exportations traduit la volonté des pouvoirs publics d’amorcer une nouvelle dynamique
industrielle au Maroc. Or, le choix de la promotion des exportations n’allait-il trouver que des
applications limitées, liées à la valorisation des ressources agricoles ou minières suivis, grâce à
des mesures de soutien, par des produits d’activités exportatrices de sous-traitance, notamment,
l’industrie de Textile et Cuir. Ainsi, et suite au ralentissement de la dynamique d’ouverture au
début des années quatre-vingt-dix et l’absence d’une politique industrielle après le PAS. Cependant, la deuxième phase (de l’après PAS à nos jours) a été marquée par un passage du
paradigme d’internalisation à celui de la mondialisation et de l’avantage comparatif à l’avantage
compétitif. En effet, durant les années quatre-vingt-dix et les années deux mille, la déprotection
va se révéler contraignante pour les industries bâties dans le cadre de l’import-substitution. Ce
qui va donner un coup de fouet aux importations, annuler les progrès réalisés par la politique
de promotion des exportations, et rendre endémique encore une fois le déficit commercial. La
question de la compétitivité de l’industrie marocaine est demeurée entière, comparativement à
des pays du même niveau de développent (Jordanie, Malaisie, Tunisie) mais plus performants
à l’export. Ainsi, l’économie marocaine est engagée dans une ouverture. I.
LA POLITIQUE INDUSTRIELLE : FONDEMENTS THEORIQUES En fin, pour résoudre le problème de
l’asymétrie de l’information, on peut évoquer l’importance de la théorie des incitations pour la
politique industrielle est qu’elle permet aux pouvoirs publics d’élaborer une politique
d’incitation à la coopération entre les différentes institutions des systèmes productifs. Aussi,
elle permet d’étudier les mécanismes par lesquels les pouvoirs publics peuvent influencer
l’efficacité des coopérations industrielles en général et celles dans le domaine de la recherche-
développement en particulier. En effet, avec le renchérissement des nouvelles technologies, le
niveau de coopération nécessaire ne cesse de croître et il devient de plus en plus nécessaire de
mettre en commun des moyens financiers et des compétences complémentaires des agents en
matière de recherche. Le marché n’étant pas capable d’assurer cette incitation du fait de
l’existence des externalités positives, l’État doit donc favoriser la coopération entre les firmes
(Baudry B ,1993). www.africanscientificjournal.com www.africanscientificjournal.com Page 61 Page 61 African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre 2023 African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre Vol : 3, Numéro 21, Décembre 2023 Après avoir présenté les principales approches théoriques de la politique industrielle, on va
présenter un rappel des choix de la politique industrielle marocaine de l’avant PAS à nos jours,
les réformes adoptées et les conséquences de ces stratégies sur le redéploiement de la politique
industrielle marocaine, en vue de les analyser en terme d’objectifs, d’instruments et de
conditions de réussite ultérieurement. www.africanscientificjournal.com II.
RAPPEL DES CHOIX DE LA POLITIQUE INDUSTRIELLE MAROCAINE Un choix stratégique
irrévocable, concrétisé dans la série d’accords de libre-échange que la Maroc a conclus avec
l’Union Européenne, les pays membres de l’Association Européenne de Libre-échange Page 62 African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre 2023 African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre 2023 African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre 2 (AELE), avec les Etats-Unis d’Amérique, la Turquie et les pays formant la quadrilatérale
d’Agadir, en l’occurrence la Tunisie, l’Egypte et la Jordanie. Cette ouverture de l’économie
marocaine et l’exacerbation de la concurrence internationale vont mettre l’industrie marocaine
face à des véritables défis et soulèvent la problématique de la compétitivité de l’industrie
marocaine, comme en témoigne l’aggravation du déficit de la balance commerciale et font
ressortir les limites de l’articulation insuffisante entre politique industrielle et politique
commerciale. En parallèle, le Maroc s’est engagé dans une politique promotionnelle qui cherche
à faire des investissements directs étrangers (IDE) dans le secteur industriel principalement un
support stratégique de croissance économique. Les différents accords de libre-échange signé
par le Maroc constituent une occasion pour drainer un flux des investissements étrangers. Pour
faire face aux retombées principalement économiques de ces accords, les pouvoirs publics ont
mis en place dès 1997 un programme de mise à niveau dans le but de sauvegarder et consolider
les acquis de l’industrie marocaine et de renforcer sa compétitivité globale. Ainsi, le Maroc a mis en place plusieurs réformes pour améliorer la compétitivité de l’économie
marocaine, elles concernent entre autres la réforme du climat des affaires, la charte
d’investissement et celle de la PME, réformes fiscales, réglementation de la concurrence,
réforme du système bancaire et financier, l’amélioration de la gouvernance publiques, basée
essentiellement sur la privatisation des entreprises publiques et la réforme de l’administration
des douanes, la modernisation de la législation du travail ….etc. Ces réformes structurelles ont
amélioré relativement la performance de l’économie marocaine, mais le déficit commercial a
continué sa dégradation. Face à l’aggravation du déficit commercial et pour maîtriser les effets
de l’ouverture de l’économie et relancer la politique de mise à niveau afin que la Maroc puisse
se positionner dans les échanges mondiaux et faire face à la concurrence mondiale. www.africanscientificjournal.com II.
RAPPEL DES CHOIX DE LA POLITIQUE INDUSTRIELLE MAROCAINE L’État va
lancer à partir de 2005 une nouvelle politique industrielle «Plan Émergence », en 2009 le «
Pacte National pour l’Émergence Industrielle » et en 2014 « Plan d’Accélération industrielle ». Cette nouvelle orientation de la politique industrielle a comme principaux objectifs : attirer de
nouveaux investissements industriels au Maroc, développer les secteurs désignés comme
moteurs de croissance et dans lesquels notre pays dispose d’un avantage compétitif, intitulés
« Métiers Mondiaux du Maroc » à savoir : l’automobile, l’électronique, l’aéronautique,
l’offshoring, l’agroalimentaires, le textile et cuir…etc. Et réorienter les exportations vers les
marchés à forte expansion. Cette nouvelle stratégie industrielle, formulée sur la base d’un
diagnostic éclairant sur la réalité du tissu industriel et sa place dans l’économie nationale, repose
également sur la mise en place des plates-formes intégrées (P2I), la relance des piliers
existants et l’accélération de la modernisation compétitive. L’objectif principal étant de Page 63 African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre 2023 African Scientific Journal
ISSN : 2658-9311
é Vol : 3, Numéro 21, Décembre 2023
transformer le positionnement naturel du Maroc, basé sur la proximité géographique et la
disponibilité d’une main d’œuvre bon marché, à un positionnement stratégique basé sur la
logistique et l’offre compétitive. À cet effet, l’amélioration de la compétitivité de l’industrie
marocaine est un enjeu prioritaire pour le développement économique de notre pays. Cependant, le Maroc a accompli beaucoup d’efforts et a pris plusieurs mesures importantes
dans le cadre la nouvelle politique industrielle, en vue d’enclencher une dynamique du secteur
industriel, afin de jouer un rôle de levier de croissance et de développement de notre économie. À cela s’ajoute une batterie de mesure en matière du commerce international et d’attraction des
investissements directs étrangers (IDE). Car, le nouvel impératif de la politique industrielle
aussi bien dans les pays développés que dans les pays en développement est le renforcement de
l’attractivité d’un pays vis-à-vis des investissements étrangers. Il vient s’ajouter à celui plus
ancien de la compétitivité. Ces deux concepts sont la mesure où une économie compétitive est
aussi attractive. Ainsi, l’efficacité d’une politique industrielle dépend d’une politique nationale
d’attraction des investissements directs étrangers conforme aux objectifs de développement du
pays et de l’amélioration de sa compétitivité. En effet, la nouvelle stratégie industrielle
considère dans cette perspective les IDE comme un facteur et un levier d’émergence et de
modernisation technologique. www.africanscientificjournal.com II.
RAPPEL DES CHOIX DE LA POLITIQUE INDUSTRIELLE MAROCAINE D’où l’importance accordée par le Plan Émergence à la nécessité
d’adapter l’organisation institutionnelle en matière d’attraction des Investissements Directs
Étrangers, d’amélioration du climat d’affaires et de renforcement de la compétitivité des
entreprises. Pour une nouvelle accélération du développement industriel en capitalisant sur les
réalisations et les acquis du Plan Émergence, le Gouvernement marocain a lancé en avril 2014
un programme d’accélération industrielle 2014-2020 qui vise à donner une nouvelle impulsion
à l’ensemble du secteur industriel en maintenant le cap des nouveaux Métiers Mondiaux du
Maroc et d’intégrer les autres filières traditionnelles de notre tissu industriel. Le nouveau plan
s’appuie sur un ensemble de mesures et d’actions portant, notamment, sur le développement de
la logique d’écosystèmes autour de grands donneurs d’ordre et de la compensation industrielle
qui vise à renforcer l’investissement, à monter en valeur dans les filières et à améliorer la
balance des paiements en favorisant les achats de produits ou services auprès du tissu local. Ce
plan vise, également, à renforcer l’accompagnement de l’informel vers le formel avec la mise
en place d’un dispositif complet d’intégration de la TPE, qui inclut la création d’un statut de
l’autoentrepreneur, un volet fiscal adapté, une couverture sociale, un accompagnement et un
financement dédiés. De même, la qualification des ressources humaines et l’amélioration de la
compétitivité des PME constituent des axes prioritaires à même de faire face à l’enjeu de
l’adéquation des compétences avec les besoins réels des entreprises et l’émergence de PME transformer le positionnement naturel du Maroc, basé sur la proximité géographique et la
disponibilité d’une main d’œuvre bon marché, à un positionnement stratégique basé sur la
logistique et l’offre compétitive. À cet effet, l’amélioration de la compétitivité de l’industrie
marocaine est un enjeu prioritaire pour le développement économique de notre pays. Cependant, le Maroc a accompli beaucoup d’efforts et a pris plusieurs mesures importantes
dans le cadre la nouvelle politique industrielle, en vue d’enclencher une dynamique du secteur
industriel, afin de jouer un rôle de levier de croissance et de développement de notre économie. À cela s’ajoute une batterie de mesure en matière du commerce international et d’attraction des
investissements directs étrangers (IDE). Car, le nouvel impératif de la politique industrielle
aussi bien dans les pays développés que dans les pays en développement est le renforcement de
l’attractivité d’un pays vis-à-vis des investissements étrangers. Il vient s’ajouter à celui plus
ancien de la compétitivité. III. ÉVALUATION DE L’INDUSTRIE MAROCAINE A TRAVERS CERTAINS
INDICATEURS Ce paragraphe présentera une évaluation des principales grandeurs économiques de manière
générale, avant d’analyser d’une manière détaillée la compétitivité des principaux secteurs de
l’industrie marocaine, à travers l’analyse des grandeurs économiques. Par ailleurs, l’étude de la
compétitivité de l’industrie marocaine nécessite l’analyse de plusieurs autres paramètres, à
savoir : le taux d’ouverture, le taux de pénétration, la structure des exportations, les produits
dynamiques, les avantages comparatifs, la part de marché, la demande étrangère adressée au
Maroc…etc. II.
RAPPEL DES CHOIX DE LA POLITIQUE INDUSTRIELLE MAROCAINE Page 64 African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre 2023 African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre 2023
novatrices et compétitives. En matière d’instruments d’appui, le plan sera doté d’un fonds
d’investissement industriel public (FDI) d’une enveloppe de 20 milliards de dirhams à l’horizon
2020 visant à permettre au tissu industriel de se consolider, de se moderniser et de développer
sa capacité de production locale. Autour du FDI, six dynamiques sont mises en œuvre pour
soutenir la réforme industrielle avec la mobilisation de 1000 hectares de foncier public locatif,
l’engagement du secteur bancaire pour soutenir activement l’industrie, la refonte du système de
garantie publique aux PME, la refonte de la Charte de l’investissement ainsi que le
regroupement des outils d’appui et de promotion. African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre 2023 African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre 2023 novatrices et compétitives. En matière d’instruments d’appui, le plan sera doté d’un fonds
d’investissement industriel public (FDI) d’une enveloppe de 20 milliards de dirhams à l’horizon
2020 visant à permettre au tissu industriel de se consolider, de se moderniser et de développer
sa capacité de production locale. Autour du FDI, six dynamiques sont mises en œuvre pour
soutenir la réforme industrielle avec la mobilisation de 1000 hectares de foncier public locatif,
l’engagement du secteur bancaire pour soutenir activement l’industrie, la refonte du système de
garantie publique aux PME, la refonte de la Charte de l’investissement ainsi que le
regroupement des outils d’appui et de promotion. II.
RAPPEL DES CHOIX DE LA POLITIQUE INDUSTRIELLE MAROCAINE Ces deux concepts sont la mesure où une économie compétitive est
aussi attractive. Ainsi, l’efficacité d’une politique industrielle dépend d’une politique nationale
d’attraction des investissements directs étrangers conforme aux objectifs de développement du
pays et de l’amélioration de sa compétitivité. En effet, la nouvelle stratégie industrielle
considère dans cette perspective les IDE comme un facteur et un levier d’émergence et de
modernisation technologique. D’où l’importance accordée par le Plan Émergence à la nécessité
d’adapter l’organisation institutionnelle en matière d’attraction des Investissements Directs
Étrangers, d’amélioration du climat d’affaires et de renforcement de la compétitivité des
entreprises. Pour une nouvelle accélération du développement industriel en capitalisant sur les
réalisations et les acquis du Plan Émergence, le Gouvernement marocain a lancé en avril 2014
un programme d’accélération industrielle 2014-2020 qui vise à donner une nouvelle impulsion
à l’ensemble du secteur industriel en maintenant le cap des nouveaux Métiers Mondiaux du
Maroc et d’intégrer les autres filières traditionnelles de notre tissu industriel. Le nouveau plan
s’appuie sur un ensemble de mesures et d’actions portant, notamment, sur le développement de
la logique d’écosystèmes autour de grands donneurs d’ordre et de la compensation industrielle
qui vise à renforcer l’investissement, à monter en valeur dans les filières et à améliorer la
balance des paiements en favorisant les achats de produits ou services auprès du tissu local. Ce
plan vise, également, à renforcer l’accompagnement de l’informel vers le formel avec la mise
en place d’un dispositif complet d’intégration de la TPE, qui inclut la création d’un statut de
l’autoentrepreneur, un volet fiscal adapté, une couverture sociale, un accompagnement et un
financement dédiés. De même, la qualification des ressources humaines et l’amélioration de la
compétitivité des PME constituent des axes prioritaires à même de faire face à l’enjeu de
l’adéquation des compétences avec les besoins réels des entreprises et l’émergence de PME transformer le positionnement naturel du Maroc, basé sur la proximité géographique et la
disponibilité d’une main d’œuvre bon marché, à un positionnement stratégique basé sur la
logistique et l’offre compétitive. À cet effet, l’amélioration de la compétitivité de l’industrie
marocaine est un enjeu prioritaire pour le développement économique de notre pays. Cependant, le Maroc a accompli beaucoup d’efforts et a pris plusieurs mesures importantes
dans le cadre la nouvelle politique industrielle, en vue d’enclencher une dynamique du secteur
industriel, afin de jouer un rôle de levier de croissance et de développement de notre économie. www.africanscientificjournal.com III.1 Aperçu au niveau global La figure N°1, montre que la situation industrielle au Maroc a connu, de manière générale, une
évolution relativement positive durant la période 2005-2020, exception faite de l’année 2009,
marquée par l’impact de la crise financière internationale. Cette situation est appréhendée à
travers l’analyse de l’évolution d’un certain nombre de grandeurs économiques relatives à
l’industrie marocaine, notamment, la production industrielle, les exportations industrielles,
l’investissement industriel, la valeur ajoutée et la création d’emplois. www.africanscientificjournal.com Page 65 African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre 2023
Fi
N° 1
É
l
i
d
i
i
l
d
é
i
l
i
i d
i African Scientific Journal
ISSN : 2658-9311 Vol : 3, Numéro 21, Décembre 2023 Figure N° 1 : Évolution des principales grandeurs économiques relatives aux industries
de transformation au Maroc entre 2004 et 2020 (En pourcentage)
Source : Données du Ministère de l’Industrie, du Commerce, de l’Investissement et de
l’Économie Numérique, élaboration de l’auteur. Figure N° 1 : Évolution des principales grandeurs économiques relatives aux industries de transformation au Maroc entre 2004 et 2020 (En pourcentage) Source : Données du Ministère de l’Industrie, du Commerce, de l’Investissement et de
l’Économie Numérique, élaboration de l’auteur. Ainsi, le taux de croissance annuel moyen (TCAM) affiché par les principales grandeurs
économiques durant la période 2005-2020, comme le montre le tableau ci- dessous (N°1),
donne le constat suivant : Tableau N°1 : Le taux de croissance annuel moyen (TCAM) affiché par les
principales grandeurs économiques durant la période 2005-2020
Grandeurs
TCAM
Exportations industrielles
5,4%
Production industrielle
5,9%
Investissement industriel
0,8%
Valeur ajoutée industrielle
5,4%
Emploi industriel
-3,2%
Source : Données du Ministère de l’Industrie, du Commerce, de l’Investissement et
de l’Économie Numérique, calculs de l’auteur. www.africanscientificjournal.com III.2 Évaluation des principales grandeurs économiques Après avoir donné un aperçu sur la situation globale de l’industrie marocaine à travers un certain
nombre de grandeurs économiques durant la période 2005-2020. On va procéder dans ce sous www.africanscientificjournal.com www.africanscientificjournal.com Page 66 Page 66 African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre 2023 African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre 2023
paragraphe, à une analyse sectorielle de la compétitivité des principales filières de l’industrie
marocaine. Ainsi, la présentation des secteurs industriels se fera selon une approche positive
permettant d’analyser les secteurs en question, notamment, l’agroalimentaire, le textile et cuir,
l’industrie chimique et para chimiques, l’industrie mécanique, l’électronique et électronique. En termes de leurs potentiels de croissance, d’emploi, d’exportations et de valeur ajoutée, tout
en se référant à des statistiques et des indicateurs du ministère de l’industrie, de l’office des
changes, du haut-commissariat au plan…etc. que nous avons collectés précédemment mesurant
l’évolution de la compétitivité intrinsèque de chaque secteur. African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre 2023 African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre 202 paragraphe, à une analyse sectorielle de la compétitivité des principales filières de l’industrie
marocaine. Ainsi, la présentation des secteurs industriels se fera selon une approche positive
permettant d’analyser les secteurs en question, notamment, l’agroalimentaire, le textile et cuir,
l’industrie chimique et para chimiques, l’industrie mécanique, l’électronique et électronique. En termes de leurs potentiels de croissance, d’emploi, d’exportations et de valeur ajoutée, tout
en se référant à des statistiques et des indicateurs du ministère de l’industrie, de l’office des
changes, du haut-commissariat au plan…etc. que nous avons collectés précédemment mesurant
l’évolution de la compétitivité intrinsèque de chaque secteur. 1 Une baisse de 5% et une perte de près de 4 milliards de dirhams d’après le rapport du centre marocain de
promotion des exportations, 2011. www.africanscientificjournal.com III.2.1 Des exportations tirées par la performance de l’industrie automobile et des
industries chimiques et parachimiques En effet, malgré la libéralisation des
échanges commerciaux du Maroc avec un ensemble des pays dans le cadre des Accords de
Libre Échange, les exportations marocaines n’arrivent pas toujours à se diversifier ou à leur
trouver de nouveaux débouchés géographiques, ce qui nous interroge sur le degré de
compétitivité des produits exportés (à faible technicité et valeur ajoutée). l’agroalimentaire+11%.Contrairement aux performances relativement faibles des secteurs de
l’électronique et du Textile durant cette période qui ne dépassent pas respectivement 4% et 2%. Il est a signalé que la concentration des exportations nationales par produit et par marché met à
mal la position concurrentielle et compétitive du Maroc. En effet, malgré la libéralisation des
échanges commerciaux du Maroc avec un ensemble des pays dans le cadre des Accords de
Libre Échange, les exportations marocaines n’arrivent pas toujours à se diversifier ou à leur
trouver de nouveaux débouchés géographiques, ce qui nous interroge sur le degré de
compétitivité des produits exportés (à faible technicité et valeur ajoutée). Figure 2 : Évolution des exportations industrielles au Maroc
Pour la période 2004-2010 (En%)
Source : Données du Ministère de l’Industrie, du Commerce, de l’Investissement et de
l’Économie Numérique, calcul et élaboration de l’auteur. 6,0
5,0
10,1
19,4
12,6
-16,1
31,2
24,6
0,7
4,6
8,6
4,6
2,0
13,2
48,5
2,1
-0,3
-20,0
-10,0
0,0
10,0
20,0
30,0
40,0
50,0
60,0
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 Figure 2 : Évolution des exportations industrielles au Maroc
Pour la période 2004-2010 (En%) Source : Données du Ministère de l’Industrie, du Commerce, de l’Investissement et de
l’Économie Numérique, calcul et élaboration de l’auteur. Source : Données du Ministère de l’Industrie, du Commerce, de l’Investissement et de
l’Économie Numérique, calcul et élaboration de l’auteur. www.africanscientificjournal.com III.2.1 Des exportations tirées par la performance de l’industrie automobile et des
industries chimiques et parachimiques Les exportations industrielles nationales ont plus que triplées depuis le lancement du plan
émergence, enregistrant presque 228 milliards de dirhams en 2020 contre 125 milliards de
dirhams en 2014 et 53,8 milliards de dirhams seulement en 2005, soit une hausse de presque de
351 %. Cette croissance importante s’explique par le développement essentiellement des
Métiers Mondiaux du Maroc durant cette période et tirées principalement par la performance
des industries chimiques et para chimiques (OCP) et de l’industrie automobile. Une analyse de la structure des exportations industrielles nous a permis de constater une
inversion des parts des exportations qu’a été dominée principalement par le secteur du textile
(32%) et de la chimie et parachimie (28%) en 2005 au profit de l’industrie mécanique et
métallurgique qui a enregistré une part de 45,3% en 2020 tirée par l’automobile, suivi du secteur
de la chimie et parachimie avec 22%. Quant au secteur des industries agro-alimentaires il a
enregistré une baisse de sa part, passant de 19,9 % à 14 % durant la même période. Puis le
secteur du textile et cuir qui a enregistré une très faible variation ne dépasse pas 1% et a chuté,
passant d’une part des exportations de 31,2 % en 2005 à 13 % seulement en 2020. Ainsi, le taux de croissance annuel moyen des exportations industrielles durant cette période
(2005-2020) est de 5,4 %. Exception de l’année 2009 qu’a été marquée par l’impact de la crise
financière internationale1. Car le meilleur taux 10,2% a été réalisé durant la période du PNEI
(2010-2014) contre 5,1% entre 2005-2009 (PI) et seulement 0,8% (2015-2020) durant la
période du Plan d’Accélération Industrielle. Cette dynamique des exportations s’explique
principalement, par la performance remarquable enregistrée entre 2009 et 2014 des secteurs de
l’automobile avec plus que 98, suivi par l’aéronautique avec +76%, et de l’Offshoring +70,2% , www.africanscientificjournal.com www.africanscientificjournal.com Page 67 African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre 2023 African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre 2023
l’agroalimentaire+11%.Contrairement aux performances relativement faibles des secteurs de
l’électronique et du Textile durant cette période qui ne dépassent pas respectivement 4% et 2%. Il est a signalé que la concentration des exportations nationales par produit et par marché met à
mal la position concurrentielle et compétitive du Maroc. III.2.2 Une évolution de la production industrielle Il faut noter que le profil de spécialisation du Maroc marqué par la prédominance des secteurs
de la chimie parachimie est dû au rôle pionnier que joue l’OCP et de l’industrie automobile. Figure N°2 : Évolution de la production industrielle au Maroc
Entre 2005 et 2020 (En %)
Source : Données du Ministère de l’Industrie, du Commerce, de l’Investissement et
de l’Économie Numérique, calcul et élaboration de l’auteur. 12
12
12
16
-12
21
18
1
0
0
-5
-1
4
7
56
-7
-20
-10
0
10
20
30
40
50
60
2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 Figure N°2 : Évolution de la production industrielle au Maroc
Entre 2005 et 2020 (En %) Figure N°2 : Évolution de la production industrielle au Maroc Source : Données du Ministère de l’Industrie, du Commerce, de l’Investissement et
de l’Économie Numérique, calcul et élaboration de l’auteur. III.2.2 Une évolution de la production industrielle La production industrielle nationale a connu une progression soutenue durant la période 2005-
2020, qu’à passer de 210 Milliards en 2005 à 593 Milliards de dirhams en 2020. Cette évolution
est stimulée essentiellement par les Métiers Mondiaux du Maroc depuis le lancement de la
première stratégie industrielle (Plan Émergence I) est en moyenne de 5% entre 2005-2009, mis
à part la baisse enregistrée en 2009 (-12%) qui s’explique par l’effet de la crise financière
internationale. Une tendance confirmée après la mise en œuvre du Pacte National pour
l’Émergence Industrielle (PNEI) de 6,5 % entre 2010-2014 et une régression sensible durant
la période 2014-2020 (PAI) avec un taux de croissance annuel moyen de 5,9%. Par ailleurs, la variation de la production industrielle durant la période 2005-2020 a fait révéler
que le secteur des industries chimiques et para chimiques domine le secteur industriel marocain www.africanscientificjournal.com Page 68 African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre 2023 et celui qu’a connu la croissance de la production la plus importante avec presque 38 % de la
production industrielle totale en 2005 contre 28 % en 2020. Suivi par les industries
agroalimentaires qui représentent 27 % du total de la production industrielle en 2020 et les
industries mécaniques et métallurgiques viennent en troisième position avec un pourcentage
d’environ 26 % en 2020 contre 14,5 en 2005, ces trois secteurs seuls représentent presque 81
% de la production industrielle Totale au Maroc en 2020. et celui qu’a connu la croissance de la production la plus importante avec presque 38 % de la
production industrielle totale en 2005 contre 28 % en 2020. Suivi par les industries
agroalimentaires qui représentent 27 % du total de la production industrielle en 2020 et les
industries mécaniques et métallurgiques viennent en troisième position avec un pourcentage
d’environ 26 % en 2020 contre 14,5 en 2005, ces trois secteurs seuls représentent presque 81
% de la production industrielle Totale au Maroc en 2020. Quant au secteur de l’électrique et électronique qui a enregistré une amélioration sensible de
presque 8 % en 2020 au lieu de 6,6 % en 2005. Puis le secteur de textile et cuir qu’a enregistré
une chute de la production industrielle, soit de 12,3 % en 2005 à 6,1 % seulement en 2014 puis
une reprise fluctuée durant la période 2015-2020 avec une moyenne de 13% . www.africanscientificjournal.com III.2.3 Une progression insuffisante de la valeur ajoutée industrielle La valeur ajoutée industrielle dégagée par l’industrie marocaine a été de 171 223 milliards de
dirhams en 2020 contre presque 113 milliards de dirhams en 2014 et 64 milliards de dirhams
seulement en 2005, soit une augmentation annuelle moyenne de 5,41 % par an durant la période
2005-2020. La structure de la valeur ajoutée industrielle au Maroc par grands secteurs,
représentant l’ossature de l’industrie nationale, n’a pas évoluée de manière significative durant
les trois ans après le lancement du Plan Émergence (2005-2008) ni après le lancement du PNEI
en 2009 et reste relativement stable durant la période du lancement du Plan d’Accélération www.africanscientificjournal.com www.africanscientificjournal.com Page 69 African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre 2023 Industrielle (2014-2020). À part les secteurs de de la Chimie-Parachimie et celui de
l’agroalimentaire qui ont passées respectivement de 23 milliards à plus que 50 milliards de
dirhams et de 19 à 37 milliards de dirhams entre 2005 et 2020. Ainsi que le secteur des industries
mécaniques et métallurgiques qu’a enregistré une progression notable ces dernières années,
passant de 7,9 milliards de dirhams en 2005 à 34,8 milliards de dirhams en 2020. Ainsi, l’analyse de la variation de la contribution des secteurs à la valeur ajoutée totale a fait
montrer que le secteur de la chimie et parachimie occupe la première place avec 32 % en 2020
contre 30% en 2005. Suivi par l’agroalimentaire avec 24 % en 2020 contre 30% en 2005. Quant
à l’industrie mécanique et métallurgique, elle a enregistrée un dédoublement de sa part durant
cette période avec 22,4% en 2020 contre 12,4% en 2005. Suivi par l’industrie électrique et
électronique avec environ 7% et en fin, le secteur de textile et cuir qu’a chuté et voir sa part de
la valeur ajoutée a passée de 15,4% en 2005 à 7,1 en 2014 avant de progresser ces dernières
années. De point de vue création de richesse, on note une progression insuffisante de la valeur
ajoutée industrielle durant cette période (5,41 % par an), dont l’amélioration est attribuable à la
dynamique des MMM principalement les deux secteurs clés l’automobile et l’aéronautique. Figure N° 3 : Évolution de la valeur ajoutée industrielle entre 2005 et 2020 (En %)
Source : Données du Ministère de l’Industrie, du Commerce, de l’Investissement et
de l’Économie Numérique, calcul et élaboration de l’auteur. Figure N° 3 : Évolution de la valeur ajoutée industrielle entre 2005 et 2020 (En %) Figure N° 3 : Évolution de la valeur ajoutée industrielle entre 2005 et 2020 (En % Source : Données du Ministère de l’Industrie, du Commerce, de l’Investissement et
de l’Économie Numérique, calcul et élaboration de l’auteur. III.2.3 Une progression insuffisante de la valeur ajoutée industrielle 8
4
13
3
3
18
6
0
4
9
-5
5
3
5
31
7
-10
-5
0
5
10
15
20
25
30
35
2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 III.2.4 Une amélioration limitée de la création de l’emploi industriel Notre analyse concerne l’emploi industriel au Maroc depuis le lancement de la nouvelle
politique industrielle en 2005 jusqu’à 2020, nous a permis de constater une amélioration
généralement trop limitée de la création de l’emploi industriel durant cette période (Figure N°4). En effet, le nombre total se situe à près de 643578 personnes en 2020 contre 597 278 personnes
en 2014 et 481 551 personnes en 2005. L’augmentation de l’emploi total moyen sur 15 ans n’a www.africanscientificjournal.com www.africanscientificjournal.com Page 70 African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre 2023 African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre 2023
été que de 162027 postes soit un taux de croissance annuel moyen de 2 % par an entre 2005 et
2009 (PE) et de 4% durant la période 2010-2014 (PNEI) avant que ce taux enregistre une chute
libre négative de – 16% entre 2015 et 2020 (PAI) , cette régression s’explique principalement
par l’impact de COVID 19 en 2020 (-24). Ainsi, les emplois crées dans les secteurs dynamiques
sont perdus dans les secteurs en repli, notamment, dans le textile et cuir et l’agroalimentaire. Par ailleurs, le secteur des industries mécanique et métallurgique (secteur de l’automobile) reste
le seul secteur qu’a enregistré une progression notable récemment, leur part à plus que tripler,
passant de 54 318 personnes en 2005 à 252019 en 2020. Cependant, une comparaison avec la
période avant le lancement du programme émergence, en se basant sur l’analyse des résultats
d’une enquête menée par le Ministère de l’industrie en partenariat avec la Banque mondiale sur
l’évaluation du climat de l’investissement au Maroc en 2005. La part moyenne de l’emploi
permanent dans l’effectif total des employés dans l’industrie est passée de 85,3% durant la
période 1996-2000 à 91,1% durant la période 2001-2006. Cet indicateur se situe à 72,5% pour
le secteur agroalimentaire contre 97,2% pour l’industrie textile/cuir entre 2001 et 2006.Les
secteurs de l’industrie du électrique/électronique (47,9%), agroalimentaire (38,8%) et chimique
para chimiques (32,1%) ont contribué positivement à la création nette d’emplois permanents
durant la période 2001-2006. Le secteur du textile a vu sa part dans la création d’emplois
permanents diminuer pour enregistrer un taux négatif durant la période 2001-2006 (-32%) alors
qu’il était le premier secteur générateur d’emplois au Maroc durant la décennie 1990 (66,1%). III.2.4 Une amélioration limitée de la création de l’emploi industriel Généralement, le comportement de l’emploi industriel au Maroc ces dix dernières années se
caractérise par une évolution en dent de scie, ce qui pose de façon récurrente la problématique
de création/destruction de l’emploi dans le secteur industriel, et invite à distinguer entre les
secteurs industriels créateurs de l’emploi et les secteurs qui en perdent. African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre 2023 African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre 2023 été que de 162027 postes soit un taux de croissance annuel moyen de 2 % par an entre 2005 et
2009 (PE) et de 4% durant la période 2010-2014 (PNEI) avant que ce taux enregistre une chute
libre négative de – 16% entre 2015 et 2020 (PAI) , cette régression s’explique principalement
par l’impact de COVID 19 en 2020 (-24). Ainsi, les emplois crées dans les secteurs dynamiques
sont perdus dans les secteurs en repli, notamment, dans le textile et cuir et l’agroalimentaire. Par ailleurs, le secteur des industries mécanique et métallurgique (secteur de l’automobile) reste
le seul secteur qu’a enregistré une progression notable récemment, leur part à plus que tripler,
passant de 54 318 personnes en 2005 à 252019 en 2020. Cependant, une comparaison avec la
période avant le lancement du programme émergence, en se basant sur l’analyse des résultats
d’une enquête menée par le Ministère de l’industrie en partenariat avec la Banque mondiale sur
l’évaluation du climat de l’investissement au Maroc en 2005. La part moyenne de l’emploi
permanent dans l’effectif total des employés dans l’industrie est passée de 85,3% durant la
période 1996-2000 à 91,1% durant la période 2001-2006. Cet indicateur se situe à 72,5% pour
le secteur agroalimentaire contre 97,2% pour l’industrie textile/cuir entre 2001 et 2006.Les
secteurs de l’industrie du électrique/électronique (47,9%), agroalimentaire (38,8%) et chimique
para chimiques (32,1%) ont contribué positivement à la création nette d’emplois permanents
durant la période 2001-2006. Le secteur du textile a vu sa part dans la création d’emplois
permanents diminuer pour enregistrer un taux négatif durant la période 2001-2006 (-32%) alors
qu’il était le premier secteur générateur d’emplois au Maroc durant la décennie 1990 (66,1%). III.2.4 Une amélioration limitée de la création de l’emploi industriel Généralement, le comportement de l’emploi industriel au Maroc ces dix dernières années se
caractérise par une évolution en dent de scie, ce qui pose de façon récurrente la problématique
de création/destruction de l’emploi dans le secteur industriel, et invite à distinguer entre les
secteurs industriels créateurs de l’emploi et les secteurs qui en perdent. igure N°4 : Évolution de l’emploi industriel au Maroc entre 2005-2020 (En %) Figure N°4 : Évolution de l’emploi industriel au Maroc entre 2005-2020 (En %) Figure N°4 : Évolution de l’emploi industriel au Maroc entre 2005-2020 (En %) Figure N°4 : Évolution de l’emploi industriel au Maroc entre 2005-2020 (En %)
Source : Données du Ministère de l’Industrie, du Commerce, de l’Investissement et
de l’Économie Numérique, calcul et élaboration de l’auteur. -4,0
4,0
2,4
-0,7
-1,8
2,2
14,8
-1,4
-1,8
5,3
0,0
0,7
5,6
17,8
13,6
-24,3
-30,0
-25,0
-20,0
-15,0
-10,0
-5,0
0,0
5,0
10,0
15,0
20,0
2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 Source : Données du Ministère de l’Industrie, du Commerce, de l’Investissement et
de l’Économie Numérique, calcul et élaboration de l’auteur. Page 71 African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre 2023 African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre 2023 www.africanscientificjournal.com III.2.5
Evolution moyenne des investissements industriels www.africanscientificjournal.com Page 72 Page 72 African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre 2023 African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre 2023 Figure N°5 : Évolution de l’investissement industriel entre 2005 et 2020 (En%) igure N°5 : Évolution de l’investissement industriel entre 2005 et 2020 (En%) Figure N 5 : Évolution de l investissement industriel entre 2005 et 2020 (En%)
Source : Données du Ministère de l’Industrie, du Commerce, de l’Investissement et
de l’Économie Numérique, calcul et élaboration de l’auteur. 20
19
16
38
-3
-15
28
-11
6
29
0
-43
-4
9
24
70
-60
-40
-20
0
20
40
60
80
2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 Source : Données du Ministère de l’Industrie, du Commerce, de l’Investissement et
de l’Économie Numérique, calcul et élaboration de l’auteur. 20
19
16
38
-3
-15
28
-11
6
29
0
-43
-4
9
24
70
-60
-40
-20
0
20
40
60
80
2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 Source : Données du Ministère de l’Industrie, du Commerce, de l’Investissement et
de l’Économie Numérique, calcul et élaboration de l’auteur. III.2.5
Evolution moyenne des investissements industriels L’investissement industriel total est passé de 13 milliards de dirhams en 2005 à
presque 31,7 milliards de dirhams en 2014, soit une augmentation de presque 144 %
durant cette période (2005-2014) et continue sa progression jusqu’au 2020 pour
atteindre 40 Milliards de dirhams. Cette évolution notable a été marquée par le
dynamisme de l’industrie chimique et para chimique, enregistrant presque 20,5
milliards de dirhams en 2014 (63,1% du total investissement) contre 5,3 milliards de
dirhams en 2005 (malgré la conjoncture internationale difficile en 2009). Suivi par le
secteur des industries agroalimentaires, passant de 2,9 milliards de dirhams en 2005 à
7,7 milliards de dirhams en 2014 et 10 milliards en 2020 (25 % du total
investissement). De même, les industries métalliques et mécaniques ont enregistrés
une progression des investissements durant cette même période, respectivement de
presque 2,6 milliards de dirhams en 2005 à presque 9 milliards de dirhams en 2020
soit 22% du total investissement. Ces IDE stimulés essentiellement par les MMM
(automobile, aéronautique, etc.). En revanche, les investissements dans le secteur des
industries de textiles et cuir ont enregistrés une contreperformance durant la même
période, passant de 1,6 milliards de dirhams en 2005 à 0,73 milliards de dirhams
seulement en 2014, soit un taux de croissance (régression) annuel moyen négatif de -
4,1% par an avant de reprendre une augmentation sensible avec l’implémentation du
plan d’accélération industrielle. Il est signalé que la part de l’industrie dans les
investissements directs étrangers, continue à attirer les investissements directs
étrangers avec 26% du total du montant investi en 2012, 24% en 2011 et 18% en 2010. L’année 2012 a été marquée, essentiellement, par la forte performance du secteur
automobile en matière d’attractivité des investissements directs étrangers (Renault-
Nissan) dans la mesure où ce secteur a pour la première fois devancé les autres secteurs
avec 26% du total des investissements directs étrangers implantés au Maroc. Ainsi,
que l’installation en 2019 du constructeur automobile français Peugeot Citroën (PSA). IV.
DISCUTIONS DES RESULTATS OBTENUS Les résultats obtenus au niveau de notre analyse des principales grandeurs et indicateurs de
compétitivité de l’industrie marocaine, durant la période 2005-2020, démontre clairement que
le Maroc n’a pas encore consolidé une véritable politique industrielle capable de surmonter les
contraintes structurelles à son développement. Exception faite pour les industries mécaniques
tirées, principalement, par les bonnes performances du secteur de l’Automobile et d’une façon
moins importante le secteur de l’Aéronautique. Les résultats globaux tirés à travers l’analyse des principales grandeurs économique entre 2005
et 2020 restent mitigés et loin des résultats escomptés. Au niveau des exportations industrielles,
une croissance importante a été constatée durant cette période, et qui s’explique par le
développement essentiellement des Métiers Mondiaux du Maroc, tirées principalement par la
performance des industries chimiques et para chimiques (OCP), l’industrie automobile et
l’industrie aéronautique. En contrepartie, une très forte croissance des importations (presque le
double des exportations), ce qu’a entraîné une aggravation du déficit de la balance commerciale
(non seulement en valeur absolue mais aussi en poids relatif) 3% du PIB en 2005 contre 17%
du PIB en 2020. Ce déficit est plus prononcé que dans d’autres pays émergents benchmarkés. En 2020, il représente 13% en Tunisie, 8% en Inde, 5 % en Turquie et 1% au Brésil. Ainsi, le secteur industriel seul représente 60% du déficit global, traduisant une insuffisante
d’industrialisation du Maroc au regard de ses besoins. Cependant, le déficit commercial du
Maroc est synonyme de la difficulté de l’offre industrielle marocaine à se positionner sur le
marché mondial, ce qui pose le problème de la compétitivité de l’offre marocaine. Au niveau du marché domestique la demande intérieure de biens et services est couvertede Les résultats globaux tirés à travers l’analyse des principales grandeurs économique entre 2005
et 2020 restent mitigés et loin des résultats escomptés. Au niveau des exportations industrielles,
une croissance importante a été constatée durant cette période, et qui s’explique par le
développement essentiellement des Métiers Mondiaux du Maroc, tirées principalement par la
performance des industries chimiques et para chimiques (OCP), l’industrie automobile et
l’industrie aéronautique. En contrepartie, une très forte croissance des importations (presque le
double des exportations), ce qu’a entraîné une aggravation du déficit de la balance commerciale
(non seulement en valeur absolue mais aussi en poids relatif) 3% du PIB en 2005 contre 17%
du PIB en 2020. Ce déficit est plus prononcé que dans d’autres pays émergents benchmarkés. www.africanscientificjournal.com IV.
DISCUTIONS DES RESULTATS OBTENUS En 2020, il représente 13% en Tunisie, 8% en Inde, 5 % en Turquie et 1% au Brésil. Ainsi, le secteur industriel seul représente 60% du déficit global, traduisant une insuffisante
d’industrialisation du Maroc au regard de ses besoins. Cependant, le déficit commercial du
Maroc est synonyme de la difficulté de l’offre industrielle marocaine à se positionner sur le
marché mondial, ce qui pose le problème de la compétitivité de l’offre marocaine. Au niveau du marché domestique, la demande intérieure de biens et services est couverte de Page 73 African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre 2023 African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre 2023 manière croissante par les importations, au détriment de la production locale qui croît moins
rapidement. Ainsi, le taux de couverture de la demande intérieure par les importations est passé
de 32% en 2005 à 50% en 2020. Ce fort besoin en importations peut s’expliquer d’une part par
le décalage entre le profil de la production nationale, concentrée sur des produits à faible valeur
ajoutée et les besoins nationaux, portant de plus en plus sur des produits sophistiqués à forte
valeur ajoutée. D’autre part, la qualité des produits « made in Morocco » perçue comme
moyenne par les consommateurs par rapport à la concurrence étrangère, puisque plus de 66%
des consommateurs marocains jugent le produit marocain de qualité moyenne. Au niveau du marché international, en plus de la faible part du marché mondiale, on constate
une stagnation de croissance sur la période 2005-2020. Cela démontre que l’offre marocaine
n’a pas su tirer profit de la croissance des échanges mondiaux, à l’instar de pays tels que la
Turquie (+0,25%) ou le Brésil (+0,38%). Cet état de fait est en partie dû à la faiblesse des
échanges intra régionaux et à une diversification relativement faible des partenaires
commerciaux (en 2020, presque les 2/3 des exportations marocaines sont à destination de
l’Union Européenne,) zone ayant connu une croissance économique limitée ces dernières
années. Mais la stagnation des parts de marché internationales du Maroc s’explique surtout par
la fragilité du positionnement de l’offre marocaine à l’export. www.africanscientificjournal.com IV.
DISCUTIONS DES RESULTATS OBTENUS Ces résultats réalisés restent loin des objectifs escomptés par les trois plans
Emergences, et qui peuvent être s’expliquées par des contraintes d’ordre structurel qui
continuent à handicaper son développement à l’image du foncier industriel qui constitue un
véritable obstacle à l’investissement au point qu’il pousse des groupes étrangers à renoncer à
leurs projets au Maroc. Par ailleurs, la part de l’industrie dans les investissements directs
étrangers a enregistrée une progression notable et a continuée à attirer les investissements directs
étrangers. Principalement dans les Métiers Mondiaux du Maroc. Cela confirme l’intérêt
manifesté par les investisseurs étrangers quant aux projets d’investissements industriels,
principalement, ceux prévus dans le cadre du Plan Émergence. est d’environ 5,4 % par an. Cependant, la part de la valeur ajoutée industrielle du Maroc au
niveau mondial reste faible entre 2005 et 2020 (entre 0,9% et 0,11%), ce qui confirme la relative
industrialisation de notre pays. Cela découle d’une part aux contraintes pesant toujours sur la
productivité et qui sont liées, entre autres, à l’environnement économique des entreprises
(ressources humaines, problématique de financement, matières premières). À l’instar de la
composition de la production industrielle, la valeur ajoutée industrielle est dominée par les
mêmes secteurs (la chimie et parachimie, l’agroalimentaire et l’industrie métallique et
mécanique) avec plus que 84% de la valeur ajoutée industrielle totale en 2020. Quant à
l’investissement industriel, une progression insuffisante a été constatée durant cette période
avec un TCAM de 1,17% entre 2005-2009, -4,91% entre 2010-2014 et 6% durant la période
2015-2020. Ces résultats réalisés restent loin des objectifs escomptés par les trois plans
Emergences, et qui peuvent être s’expliquées par des contraintes d’ordre structurel qui
continuent à handicaper son développement à l’image du foncier industriel qui constitue un
véritable obstacle à l’investissement au point qu’il pousse des groupes étrangers à renoncer à
leurs projets au Maroc. Par ailleurs, la part de l’industrie dans les investissements directs
étrangers a enregistrée une progression notable et a continuée à attirer les investissements directs
étrangers. Principalement dans les Métiers Mondiaux du Maroc. Cela confirme l’intérêt
manifesté par les investisseurs étrangers quant aux projets d’investissements industriels,
principalement, ceux prévus dans le cadre du Plan Émergence. Concernant l’emploi industriel, le nombre total des emplois crées par le secteur pendant 15
ans est d’environ 162027 emploi seulement. www.africanscientificjournal.com IV.
DISCUTIONS DES RESULTATS OBTENUS Or, l’analyse comparative de la compétitivité-prix menée par l’institut royal des études
stratégiques de l’offre Maroc sur quelques catégories représentatives de produits montre que
le Maroc est généralement moins compétitif sur les prix que les pays comparables les plus
performants : le différentiel de prix à l’export entre le produit marocain et le produit du pays
concurrent le plus compétitif peut aller de 10% à 98% selon les produits et les secteurs
considérés. En ce qui concerne l’évolution de la production industrielle durant la période 2005-
2020, notre analyse a fait apparaitre une progression relativement faible depuis le lancement
du Plan Émergence en 2005. Cette progression s’est confirmée avec le lancement de pacte
national pour l’émergence industrielle en 2009, malgré la régression enregistrée en cette année
même (2009), qui s’explique par l’impact de la crise financière internationale. Ainsi, l’analyse
de la composition de la production industrielle par grands secteur durant la même période, fait
ressortir une domination des secteurs de la chimie te parachimie (OCP), l’agroalimentaire
et l’industrie métallique et mécanique (principalement le secteur de l’automobile et de
l’aéronautique) avec 78% de la production total. Parallèlement, la valeur ajoutée industrielle
n’a pas évoluée de manière significative, ni après le lancement du Plan Émergence 2005 ni
après le lancement du PNEI en 2009 et même durant la période 2015-2020 dans le cadre du
plan d’accélération industrielle. Le taux de croissance annuel moyen durant toute cette période Page 74 African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre 2023 est d’environ 5,4 % par an. Cependant, la part de la valeur ajoutée industrielle du Maroc au
niveau mondial reste faible entre 2005 et 2020 (entre 0,9% et 0,11%), ce qui confirme la relative
industrialisation de notre pays. Cela découle d’une part aux contraintes pesant toujours sur la
productivité et qui sont liées, entre autres, à l’environnement économique des entreprises
(ressources humaines, problématique de financement, matières premières). À l’instar de la
composition de la production industrielle, la valeur ajoutée industrielle est dominée par les
mêmes secteurs (la chimie et parachimie, l’agroalimentaire et l’industrie métallique et
mécanique) avec plus que 84% de la valeur ajoutée industrielle totale en 2020. Quant à
l’investissement industriel, une progression insuffisante a été constatée durant cette période
avec un TCAM de 1,17% entre 2005-2009, -4,91% entre 2010-2014 et 6% durant la période
2015-2020. IV.
DISCUTIONS DES RESULTATS OBTENUS Loin de l’objectif escompté par le plan émergence
I qu’a prévoyait la création de 220 000 emplois sans comparaison avec les objectifs fixés par
le PAI. Ainsi, son évolution durant cette période (2005-2020) se caractérise par une création
d’emploi dans des secteurs et une perte d’emploi dans d’autres. En effet, un certain
dynamisme d’emploi a été constaté au niveau des secteurs de l’automobile, de l’électronique
et de l’aéronautique, contre une perte d’emploi dans le secteur de textile. Cependant, l’emploi
industriel total qui a été gagné en création d’emplois dans les secteurs dynamiques a été perdu
dans les secteurs en repli. Par ailleurs, notre évaluation de la compétitivité de l’industrie marocaine au niveau
macroéconomique durant la période 2005-2020, a révélé une hausse du taux d’ouverture. Cette
tendance vers l’ouverture est plus accentuée à l’importation qu’à l’exportation, ce qui se
traduit par un déficit de la balance commerciale du pays sur toute la période. En parallèle, une augmentation du taux de pénétration, passant de 30,3% en 2005 à 44% en
2014 et à 40% en 2020, ce qui se répercute sur l’industrie marocaine. Cette situation peut être Page 75 African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre 2023 African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre 2023
s’expliquer par l’impact des accords de libre-échange que le Maroc a conclus, principalement,
celui avec l’union européenne, vu le poids et l’importance de cet accord pour nos échanges
extérieurs. Aussi, une croissance de la demande étrangère adressée au Maroc a été constatée
durant cette période. Ce qui nous interroge sur le degré de compétitivité extérieure de nos
exportations et l’impact de l’orientation géographique du commerce extérieur. Cela est
confirmé par la stagnation de la part du marché marocaine au niveau mondial (0,10% en 2005 à
0,11% en 2020). African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre 2023 African Scientific Journal
ISSN : 2658-9311
Vol : 3 Numéro 21 Décembr s’expliquer par l’impact des accords de libre-échange que le Maroc a conclus, principalement,
celui avec l’union européenne, vu le poids et l’importance de cet accord pour nos échanges
extérieurs. Aussi, une croissance de la demande étrangère adressée au Maroc a été constatée
durant cette période. Ce qui nous interroge sur le degré de compétitivité extérieure de nos
exportations et l’impact de l’orientation géographique du commerce extérieur. www.africanscientificjournal.com Conclusion « Aucun pays ne peut imaginer rester sur la trajectoire de croissance s’il perd son industrie et
que le développement de l’activité industrielle, productrice de biens matériels, est essentiel
quand on sait que plus de 70% des exportations et plus de 80% des dépenses de recherche et
développement sont liées à l’industrie ». Cette citation de Gaffard, J- L. (2012) reflète des
explications beaucoup plus approfondies de ce que nous croyons déjà sur l’importance de
l’industrialisation. Il est vrai que le développement industriel est indispensable pour la
modernisation du Maroc. Ce développement est intervenu suite à une série de réforme et
mesures importantes qu’ont été déployées ces dernières années, en vue d’enclencher une
dynamique de son secteur industriel appelé à jouer un rôle de levier de croissance et de
développement de l’économie nationale. Nous avons vu que le Maroc a mis en place une
diversité des stratégies et politiques industrielles adoptées au Maroc depuis l’indépendance
jusqu’à nos jours. Malgré cette diversification, l’industrie marocaine n’arrive pas aujourd’hui à
répondre aux exigences de la compétitivité internationale et semble encore loin de développer
une capacité d’anticipation et d’adaptation. Cependant, plus qu’une stratégie industrielle, le
Maroc a besoin d’une politique de développement territorial intégré (El Kadiri, N. 2008) à
même d’atténuer l’ampleur des déséquilibres entre les différentes régions du pays. Par ailleurs,
notre analyse des principales grandeurs et indicateurs de compétitivité de l’industrie marocaine,
principalement, des Métiers Mondiaux du Maroc durant la période 2005-2020, et leur
comparaison avec un échantillon des pays concurrents, à montrer que la compétitivité de
l’industrie marocaine s’est améliorée d’une manière relative dans la durée, mais cette
amélioration demeure faible. Notre politique industrielle n’a pas connu de spécialisation, mais
parfois même une déspécialisation. C’est pourquoi nos échanges internationaux s’est heurtés à
plusieurs handicapes (une économie mal adaptée à la demande mondiale, une spécialisation à
faible intensité et très fragile du point de vue de constitution). Ainsi, les liens entre les activités
industrielles ne sont pas bien structurés. Car, notre tissu industriel est constitué des unités
isolées les aunes des autres, et quand il y a une possibilité d’une présence forte dans certains
secteurs, la coopération industrielle fait défaut. Aussi, la politique de développement
industrielle a-t-elle induit des relations distendues entre entreprises et territoire national. Conclusion À cet effet, le progrès technique, l’innovation, la recherche-développement, la valorisation des
ressources humaines et l’utilisation rationnelle des technologies de l’information et de la « Aucun pays ne peut imaginer rester sur la trajectoire de croissance s’il perd son industrie et
que le développement de l’activité industrielle, productrice de biens matériels, est essentiel
quand on sait que plus de 70% des exportations et plus de 80% des dépenses de recherche et
développement sont liées à l’industrie ». Cette citation de Gaffard, J- L. (2012) reflète des
explications beaucoup plus approfondies de ce que nous croyons déjà sur l’importance de
l’industrialisation. Il est vrai que le développement industriel est indispensable pour la
modernisation du Maroc. Ce développement est intervenu suite à une série de réforme et
mesures importantes qu’ont été déployées ces dernières années, en vue d’enclencher une
dynamique de son secteur industriel appelé à jouer un rôle de levier de croissance et de
développement de l’économie nationale. Nous avons vu que le Maroc a mis en place une
diversité des stratégies et politiques industrielles adoptées au Maroc depuis l’indépendance
jusqu’à nos jours. Malgré cette diversification, l’industrie marocaine n’arrive pas aujourd’hui à
répondre aux exigences de la compétitivité internationale et semble encore loin de développer
une capacité d’anticipation et d’adaptation. Cependant, plus qu’une stratégie industrielle, le
Maroc a besoin d’une politique de développement territorial intégré (El Kadiri, N. 2008) à
même d’atténuer l’ampleur des déséquilibres entre les différentes régions du pays. Par ailleurs,
notre analyse des principales grandeurs et indicateurs de compétitivité de l’industrie marocaine,
principalement, des Métiers Mondiaux du Maroc durant la période 2005-2020, et leur
comparaison avec un échantillon des pays concurrents, à montrer que la compétitivité de
l’industrie marocaine s’est améliorée d’une manière relative dans la durée, mais cette
amélioration demeure faible. Notre politique industrielle n’a pas connu de spécialisation, mais
parfois même une déspécialisation. C’est pourquoi nos échanges internationaux s’est heurtés à
plusieurs handicapes (une économie mal adaptée à la demande mondiale, une spécialisation à
faible intensité et très fragile du point de vue de constitution). Ainsi, les liens entre les activités
industrielles ne sont pas bien structurés. Car, notre tissu industriel est constitué des unités
isolées les aunes des autres, et quand il y a une possibilité d’une présence forte dans certains
secteurs, la coopération industrielle fait défaut. Aussi, la politique de développement
industrielle a-t-elle induit des relations distendues entre entreprises et territoire national. IV.
DISCUTIONS DES RESULTATS OBTENUS Cela est
confirmé par la stagnation de la part du marché marocaine au niveau mondial (0,10% en 2005 à
0,11% en 2020). En clos, malgré l’apport des trois plans (Plan Emergence 2005-2009, Pacte National pour
l’Emergence Industrielle 2009-2014 et le Plan d’Accélération Industrielle 2014-2020) le
modèle industriel marocain demeure vulnérable. Car, il se caractérise par une persistance des
déficiences en termes des prérequis de base facilitant l’intégration des entreprises marocaines
dans les chaînes de valeurs mondiales. Plusieurs contraintes limitent la contribution des
entreprises industrielles dans la création de la valeur ajoutée industrielle à la croissance
économique. D’où l’interrogation sur l’impact encore trop limité du plan émergence, du Pacte
national pour l’Emergence Industrielle et du plan d’accélération industrielle sur
l’industrialisation et la compétitivité globale du Maroc ? www.africanscientificjournal.com Page 76 Page 76 African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre 2023 African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre 2023 www.africanscientificjournal.com www.africanscientificjournal.com Bibliographie Bandt, J. D. (1993). Politiques industrielles et de la concurrence : quelles priorités, quelles
combinaisons ? Revue d'économie industrielle, 63(1), 207-221. Baudry, B. (1993). Partenariat et sous-traitance : une approche par la théorie des incitations. R
d’é
i i d
t i ll
V l 66 4
t i
t
51 Bandt, J. D. (1993). Politiques industrielles et de la concurrence : quelles priorités, quelles
combinaisons ? Revue d'économie industrielle, 63(1), 207-221. Bandt, J. D. (1993). Politiques industrielles et de la concurrence : quelles priorités, quelles
combinaisons ? Revue d'économie industrielle, 63(1), 207-221. Baudry, B. (1993). Partenariat et sous-traitance : une approche par la théorie des incitations. Revue d’économie industrielle, Vol. 66, 4eme trimestre, pp. 51. Billote, P J. (1997). Concurrence technologique et normalisation : enjeux publics et stratégies
industrielles. Edition AFNOR. Revue d’économie industrielle, Vol. 66, 4eme trimestre, pp. 51. Billote, P J. (1997). Concurrence technologique et normalisation : enjeux publics et stratégies
industrielles. Edition AFNOR. Billote, P J. (1997). Concurrence technologique et normalisation : enjeux publics et stratégies
industrielles. Edition AFNOR. Cheval,J. (1983). Un examen des politiques industrielles des pays de l’OCDE. Analyse de la
SEDÉIS. Cheval,J. (1983). Un examen des politiques industrielles des pays de l’OCDE. Analyse de la
SEDÉIS. Courlet, C. (2007). Territoire et développement économique au Maroc : Le cas des systèmes
productifs localisés. Edition l’Harmattan. Courlet, C. (2007). Territoire et développement économique au Maroc : Le cas des systèmes
productifs localisés. Edition l’Harmattan. Demazière, C. (2006). Décisions et mise en œuvre des politiques économiques : de l’État
souverain au polycentrisme institutionnel. Cahiers français N°335. Demazière, C. (2006). Décisions et mise en œuvre des politiques économiques : de l’État
souverain au polycentrisme institutionnel. Cahiers français N°335. El Kadiri, N. (2008). Développement territorial et politique industrielle au Maroc : Entre
impératifs économiques et enjeux sociaux et politiques. Gaffard, J- L. (2012). Compétitivité et développement industriel : un défi Européen. OFCE et
SKEMA Business Scool. Ha-Joon, Chang. (2006). Sur la relation entre les institutions et le développement économique. Revue d’économie politique, N° 30,53-67. Humbert, M. (1995). Élargissement des fondements théoriques standards des politiques
industrielles. Revue d’économie industrielle, vol 71, pp.143-162. Kwabena , N.O. (2013). Politiques industrielles en Afrique de l’ouest. Publication de Friedrich-
Ebert-Stiftung. Morvan, Y. (1991). Fondements d’économie industrielle. 2éme édition, Economica. Niosi, J et Bellon, B. (1995). « Une interprétation évolutionniste des politiques industrielles »
Revue d’économie industrielle, vol. 71(1), 213-226. Piermay, J-L (2011). Conclusion www.africanscientificjournal.com Page 78 Page 78 African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre 2023 Conclusion À cet effet, le progrès technique, l’innovation, la recherche-développement, la valorisation des
ressources humaines et l’utilisation rationnelle des technologies de l’information et de la
communication, deviennent des options stratégiques incontournables de compétitivité. Malgré À cet effet, le progrès technique, l’innovation, la recherche-développement, la valorisation des
ressources humaines et l’utilisation rationnelle des technologies de l’information et de la
communication, deviennent des options stratégiques incontournables de compétitivité. Malgré
les efforts déployés en vue de booster l’émergence industrielle du Maroc et les résultats
encourageants de certains métiers mondiaux du Maroc (l’automobile et l’aéronautique), la Page 77 African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre 2023 African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre 2023
transformation structurelle de l’économie nationale et sa sophistication restent handicapées
notamment par une spécialisation axée sur la sous-traitance, une formation défaillante et une
gouvernance déficiente. La réussite de toute stratégie industrielle au Maroc dépendra aussi
bien des facteurs internes sources de la croissance endogène que des facteurs externes
favorisant la dynamique exogène. Les facteurs internes concernent entre autres l’adéquation
de la formation avec le marché d’emploi, l’effort d’innovation et de recherche, l’amélioration
de gouvernance institutionnelle et de l’infrastructure industrielle, la bonne articulation entre
les politiques publiques et la cohérence des stratégies sectorielles. Les facteurs externes
concernent notamment l’évolution de l’environnement international, le comportement et la
pérennité des flux des Investissements Directs Étrangers orientés vers les secteurs stratégiques
ainsi que la dynamique des chaines de valeurs mondiales. African Scientific Journal
ISSN : 2658-9311
Vol : 3, Numéro 21, Décembre 2023 transformation structurelle de l’économie nationale et sa sophistication restent handicapées
notamment par une spécialisation axée sur la sous-traitance, une formation défaillante et une
gouvernance déficiente. La réussite de toute stratégie industrielle au Maroc dépendra aussi
bien des facteurs internes sources de la croissance endogène que des facteurs externes
favorisant la dynamique exogène. Les facteurs internes concernent entre autres l’adéquation
de la formation avec le marché d’emploi, l’effort d’innovation et de recherche, l’amélioration
de gouvernance institutionnelle et de l’infrastructure industrielle, la bonne articulation entre
les politiques publiques et la cohérence des stratégies sectorielles. Les facteurs externes
concernent notamment l’évolution de l’environnement international, le comportement et la
pérennité des flux des Investissements Directs Étrangers orientés vers les secteurs stratégiques
ainsi que la dynamique des chaines de valeurs mondiales. www.africanscientificjournal.com Bibliographie L’attractivité sans le territoire ? L’industrie marocaine a-t-elle besoin du
territoire ? Annales de géographie n°678. Reich, R. (1997). L’économie mondialisée. Edition Dunod. pp. 23-67. Robinson, J. (2010). Politique industrielle et développement : analyse en termes d'économie
politique. Revue d'économie du développement, 18, 21-45. Van Ha Nguyen. (2011). Stratégie d’industrialisation et compétitivité de l’économie
vietnamienne. Economies et finances. www.africanscientificjournal.com www.africanscientificjournal.com Page 79 Page 79
|
https://openalex.org/W3090938295
|
https://ieeexplore.ieee.org/ielx7/6287639/8948470/09210068.pdf
|
English
| null |
LOAF: Load and Resource Aware Federation of Multiple Sensor Sub-Networks
|
IEEE access
| 2,020
|
cc-by
| 19,041
|
2020, accepted September 27, 2020, date of publication September 30, 2020, date of current version October 12, 2020. Received August 25, 2020, accepted September 27, 2020, date of publication September 30, 2020, date of current version October 12, 2020. eceived August 25, 2020, accepted September 27, 2020, date of p Digital Object Identifier 10.1109/ACCESS.2020.3027821 LOAF: Load and Resource Aware Federation
of Multiple Sensor Sub-Networks SOOKYOUNG LEE
1, MOHAMED YOUNIS
2, (Senior Member, IEEE),
MOHAMMED ALSOLAMI
2, AND MEEJEONG LEE
1, (Member, IEEE)
1Department of Computer Science and Engineering, Ewha Womans University, Seoul 03760, South Korea
2Department of Computer Science and Electrical Engineering, University of Maryland, Baltimore County, Baltimore, MD 21250, USA
Corresponding author: Meejeong Lee (lmj@ewha.ac.kr) This work was supported by the National Research Foundation of Korea (NRF) funded by the Korean Government (MSIT) under
Grant 2018R1D1A1B07043671. ABSTRACT Wireless sensor networks (WSNs) serving in hostile environments are susceptible to multiple
collocated failures due to explosives, and natural calamities such as avalanches, landslides etc., which could
partition the network into disjoint segments. A similar scenario is encountered when autonomous WSNs need
to collaborate for achieving a common task; therefore, federating the segments or individual WSNs would
be essential for sharing data between them. The federation may be achieved by populating relay nodes and
providing perpetual inter-segment paths. In this paper, we tackle the federation problem while considering
constrained relay availability that makes the problem more challenging. We exploit the use of a limited
number of mobile relays to provide intermittent inter-segment connectivity. We propose LOAF, a novel
algorithm for LOad and resource Aware Federation of multiple sensor sub-networks. LOAF strives to group
the segments into multiple clusters considering the amount of energy consumed to support the inter-segment
traffic as well as proximity between segments. The formed inter-cluster topology is star-shaped where the
segments in each cluster are served by a distinct mobile relay and a central cluster provides inter-cluster data
delivery. During forming a central cluster, LOAF opts to balance the load on the individual mobile relays
in terms of energy consumed in travel and in wireless communication. We analyze the properties of LOAF
mathematically and validate its performance through extensive simulation experiments. INDEX TERMS Intermittent connectivity, federation, mobile data carrier, network partitioning, topology
repair, wireless sensor networks. This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0 r a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
VOLUME 8, 2020 A. RELAY NODE PLACEMENT where the relay count is not sufficient to form a perpetual
topology; instead multiple mobile data carriers (MDCs) are
employed to form intermittent communication links among
the segments. Particularly, we solve the federation problem
when the MDC count is so constrained that it is not feasible
to assign an MDC to each link in a minimum spanning tree
of the segments. In general, deterministic placement of relay nodes that are
richer in computation, energy and communication resources
than sensor nodes, has been pursued as a means for shaping
the WSN topology in order to satisfy some desired per-
formance goals. While minimizing the required relay count
has been the prime optimization objective, many variants of
relay placement have been proposed to achieve additional
goals, e.g., degree of connectivity, path length, etc. Achieving
connectivity with the least relay count is equivalent to iden-
tifying a Steiner minimum tree with minimal Steiner points
and bounded edge-length which is shown to be NP-hard by
Lin and Xue [4]. Therefore, heuristics have been pursued. Published heuristics can be categorized into three groups. Given the MDC availability constraint, each MDC has to
serve more than two segments. We thus ought to divide the
set of segments S into multiple groups or clusters. This is
mapped into a set cover problem which is NP-hard. There-
fore, we pursue heuristics. While grouping segments in mul-
tiple clusters, we opt to balance the energy consumed by the
MDCs for touring the segments of the individual clusters and
uploading/downloading data over wireless links. We propose
a novel algorithm for LOad and resource Aware Federa-
tion of multiple sensor sub-networks (LOAF) by forming
a k-clustered topology where the energy consumption over-
head in touring and data transporting is balanced among
the k MDCs. LOAF consists of two phases. In the first group, unconstrained setups are considered
where the nodes can be virtually placed anywhere in the
area of interest. The placement algorithms in this group
strive to minimize the relay count needed for restoring con-
nectivity in a partitioned network [5], [8]. In addition to
reducing the number of relays, the second group considers
more objectives, such as high energy efficiency to prolong
network lifetime [9], [10], or high degree of connectivity by
providing multiple disjoint inter-node paths in the formed
network [11], [14]. A. RELAY NODE PLACEMENT Solutions in the third group tackle the fed-
eration problem under node position constraints [15], [16] or
various quality of services [17], [18]. Overall, the approaches
in this category provide perpetual connectivity and do not
consider resource-availability constraints. LOAF solves the
federation problem when the relay count is insufficient for
forming stable data paths and establishes intermittent connec-
tivity instead. p
In the 1st phase, LOAF first tries to find the energy-based
center of mass of S while considering the inter-segment
data volume, hereafter denoted as eG. Basically, to calcu-
late eG, LOAF considers the in/out data volume for each
segment Si and calculates the corresponding energy con-
sumption for up/downloading data at Si via wireless com-
munication. Based on the found eG, LOAF opts to form
(k −1) non-central-clusters Ci’s around eG while minimiz-
ing the overall energy consumed in visiting the segments
of Ci, i < k by an MDCi. During the 2nd phase, LOAF
opts to form a central-cluster Ck around eG while balancing
energy consumption among MDCi’s, ∀i in terms of travelling
a cluster and transporting data. The clustering results in a
star-shaped inter-cluster topology via which every pair of
segments in different clusters can exchange data at most
within two cluster-hops. LOAF is validated through simu-
lation experiments and is shown to outperform competing
schemes. B. MOBILITY-ASSISTED SOLUTIONS Mobile agents have been employed to transport data in sparse
networks by playing a role of a base-station or a data carrier. A mobile base-station moves in the area of interest and gath-
ers data from sensor nodes over multi-hop paths. Meanwhile,
a data carrier tours sensors and downloads their readings in
one hop to relay or carry them to a sink or another node in
the network. Therefore, a mobile data carrier (MDC) even-
tually forms an intermittently connected topology. Published
MDC-based federation schemes fall into one of two cate-
gories depending on whether more objectives are considered
in addition to establishing connectivity. The rest of the paper is organized as follows. LOAF is com-
pared to related work in Section II. In Section III, the prob-
lem is formally defined and the considered system model is
described. The details of LOAF are provided in Section IV. The validation results are presented in Section V. The paper
is finally concluded in Section VI. The approaches in the first category do not consider a con-
straint on the availability of the mobile nodes [20]–[23]. For
instance, in [20] Almasaeid et al. model the mobile-assisted
federation problem as a closed queuing network and focus on
the effect of the network parameters related to mobile relays
on the end-to-end delay. They expanded their studies in [21]
by including two other roles of mobile agents, i.e., mobile
sinks and mobile collectors. On the other hand, Li and Hua
deploy mobile nodes to form a second-tier mesh network to
transfer data from sensor nodes to a sink such that the data
delivery latency due to buffer space limitation of the mobile
nodes is minimized. Unlike LOAF, the approaches introduced I. INTRODUCTION get destroyed. Restoring the connectivity among segments is
essential for enabling full network operation. Another sce-
nario is when autonomous networks are to be federated in
order to aggregate their capabilities and achieve an emerging
task such as search-and-rescue, military situation awareness,
criminal hunting, etc. WSNs have attracted increased interest in recent years due
to their numerous civil, scientific and military applications. In a WSN, a large set of sensors are deployed to form
a mesh topology and coordinate their actions to carry out
a common task [1]. Thus, the inter-sensor connectivity has a
significant influence on the effectiveness of WSNs and should
be sustained all the time. Moreover, a WSN often operates
in harsh environments and may suffer from a major damage
which results in simultaneous failure of multiple collocated
nodes and causes the network to get partitioned into disjoint
segments. For example, in a battlefield, parts of the deploy-
ment area may be attacked by explosives and some nodes Federating a set of disjoint segments or standalone net-
works has recently received growing attention from the
research community. Most published solutions exploit the
deployment of stationary relay nodes and formulate the fed-
eration problem as finding the locations of the least relay
count to form a stable inter-segment topology [2]. In other
words, a connected topology is formed by populating suf-
ficient relays to provide stable inter-segment data paths. However, resource scarcity makes the federation problem
more challenging. In this paper, we consider the situation The associate editor coordinating the review of this manuscript and
approving it for publication was Kashif Saleem
. 179466 179466 VOLUME 8, 2020 VOLUME 8, 2020 S. Lee et al.: LOAF: Load and Resource Aware Federation of Multiple Sensor Sub-Networks II. RELATED WORK D (Ss, Sd) = DT
Trips,d
+ σDC(Datas,d) + DR
wts,d
=
1
VM
· Trips,d
+
σ
BM
· Datas,d
+
wts,d
(1)
E (Mi) = EM(TR(Mi)) + EC
DataMi
= (µ · TR(Mi)) +
DataMi · PC(R)
(2) D (Ss, Sd) = DT
Trips,d
+ σDC(Datas,d) + DR
wts,d
=
1
VM
· Trips,d
+
σ
BM
· Datas,d
+
wts,d
(1) (1) Meanwhile, ToCS pays more attention to reducing the
average and maximum delay for delivering data between seg-
ments by finding the balanced tour paths among MDCs [32]. Like LOAF, ToCS groups a set of segments into k clusters
that form a star topology in order to reduce the inter-segment
latency. Then it opts to equalize the tour length of clusters by
adjusting the size of the cluster at the center. However, ToCS
does not consider balancing the energy consumption among
MDCs during operation. Unlike ToCS, LOAF strives to even
the energy overhead for each MDC to tour and transport a
distinct volume of data between the terminals which it serves. The consumed energy and the communication delay of LOAF
are compared to that of MINDS, LEEF, ToCS, and FOCUS
in Section V. E (Mi) = EM(TR(Mi)) + EC
DataMi
= (µ · TR(Mi)) +
DataMi · PC(R)
(2) In (1), the delay incurred while carrying data from
Ss to Sd is primarily determined by a trip latency DT ,
a data transmission latency DC, and a relaying latency DR. DT mainly depends on the length of the travel path between
two terminals Ss and Sd, denoted as Trips,d and the speed of
the involved MDCs. We assume a constant speed VM for all
MDCs and DT is thus a function of only the tour length of
the individual MDC. In addition, DC is determined by the
volume of data transfer between Ss and Sd, and the capability
of the radio transceiver (bits per second), represented as
Datas,d and BM in (1), respectively. Recall that LOAF forms
a star-like inter-cluster topology; thus, inter-cluster relaying
may take place. The constant σ captures that fact that the
communication delay is incurred more than once since two
MDCs will be involved in exchanging the data. II. RELATED WORK Published
techniques
for
federating
the
distinct
seg-
ments or autonomous WSNs can be categorized into two
groups [2], [3]: (i) approaches that employ relay nodes and
form stable data paths among segments. The main objective
of these approaches is usually to minimize the number of
populated relays to achieve full connectivity; (ii) approaches
that exploit mobility for solving the federation problem. These mobility-assisted techniques opt to create intermit-
tent links. This section discusses related work in these two
categories. 179467 179467 179467 VOLUME 8, 2020 VOLUME 8, 2020 S. Lee et al.: LOAF: Load and Resource Aware Federation of Multiple Sensor Sub-Networks in this group do not consider the limited supply of mobile
nodes and thus careful assignment of MDCs is not studied. In addition, Wang et al. [22] and Ma [23] have focused on a
single MDC setup and how to limit the length of data paths. However, such work does not deal with segments of nodes
and address inter-segments connectivity. among multiple standalone WSNs in order to achieve a
common mission such as search-and-rescue, or environment/
creature monitoring. In both scenarios, meeting some inter-
segment requirements, such as link capacity, may be neces-
sary to achieve the mission; in this paper, we consider the
data volume, represented as bits per data collection round,
to be transported between every pair of segments. Such an
inter-segment requirement can be: (i) just a byproduct of the
damage depending on the size of the individual segments,
or (ii) required to provide the application service though the
federated segments. Like LOAF, the second category solves the federation
problem with a limited MDC count. IDM-kMDC [24],
FeSMoR [25], MINDS [26] and LEEF [27] first find a
stable inter-segment topology based on which MDC tours
are formed. Basically, an MDC may be dedicated to tour
the link between a pair of segments or tour a subset of the
segments. In addition, MINDS and LEEF strive to even tour
length among MDCs and is compared to LOAF in Section V. Moreover, MiMSI and RCR strive to solve a more constrained
version of the problem where only few relays are movable,
and the others are stationary [28], [29]. However, unlike
LOAF these approaches focus more on tour lengths rather
than energy consumed by the MDCs and do not factor in the
inter-segment communication latency. II. RELATED WORK The value of
σ is set to 1 if Ss and Sd are in the same cluster, to 2 if either Ss
or Sd belong to the central cluster, and to 3 if Ss and Sd are part
of two distinct outer clusters. The data volume transported II. RELATED WORK The federation is to be achieved using k MDCs that provide
intermittent data paths among the segments subject to the
data volume requirement. The main optimization objective
for such a federation is to determine the MDC tours that
reduce inter-segment communication delay and balance the
energy overhead experienced by the MDCs while travelling
and transporting data. In other words, we employ a set M of
k MDCs {M1, M2, ..., Mk} to connect a set S of n distinct
terminals {S1, S2, ... Sn}, n > 2 which play the role of a
gateway node that serves as an interface for the segment. In the rest of the paper the terms terminal and segment are
used interchangeably. In LOAF, S is to be clustered into Ci’s,
i = 1,.., k, i.e., Ci ⊂S and S
∀i Ci = S. Each cluster Ci
is served by Mi along a travel route TR(Mi) that contains a
list of coordinates at which Mi stops to upload or download
data from/to the terminals in Ci. The delay incurred while
delivering the inter-segment data between Ss to Sd denoted as
D(Ss, Sd) and the energy consumed by Mi, E(Mi), could be
calculated by (1) and (2), respectively. g
y
On the other hand, there are some approaches that focus on
reducing data delivery latency. I. F. Senturk et al. propose a
delay-aware clustering algorithm which groups segments for
k MDCs to reduce the maximum tour length among MDCs
and data delivery delay between segments to a sink [30]. In addition, FOCUS factors in the data delivery delay in the
federation by carefully scheduling the motion of MDCs [31]. It first forms overlapped k clusters served by k MDCs. The
intersection segment (IS) of two overlapping clusters plays
the role of a rendezvous point. In order to avoid storing data
and holding MDCs at the ISs, FOCUS tries to synchronize the
arrival of two MDCs by carefully setting their motion speed. However, FOCUS does not care for the data delivery delay
and energy concerns. III. PROBLEM STATEMENT The mobility-assisted federation problem tackled in this
paper may arise in two scenarios; (i) restoring lost
inter-segment connectivity after a major node damage, e.g.,
inflicted by explosives in a battlefield or natural calamities
such as landslides or avalanches, and (ii) linking individual
batches of sensor nodes, for example to enable collaboration VOLUME 8, 2020 179468 S. Lee et al.: LOAF: Load and Resource Aware Federation of Multiple Sensor Sub-Networks between every pair of terminals is assumed to vary based on
the size of the corresponding segments. data delivery delay is reduced and energy overhead for the
individual MDCs is balanced. In addition, we suppose that all
MDCs have the same capabilities with enough data storage to
handle the inter-segment data traffic in each cluster, i.e., we
do not consider any buffering constraints in our solution. Without considering any additional objectives, the formulated
problem can be mapped into solving a k-means clustering
problem that is known NP-hard [4], and thus LOAF pursues
heuristics. Lastly, DR means the loading time for inter-cluster data
to be relayed from an MDC to another if Ss and Sd do not
belong to the same cluster. In the proposed solution MDCs
meet directly to exchange data at a computed position and
the inter-cluster data relaying may thus involve the waiting
time for rendezvous between MDCs depending on their trip
schedule. Therefore, like DC, DR is computed depending on
the locations of the communicating segment pair Ss and Sd. In case Ss and Sd belong to the same cluster Ci, i.e., both
segments are served by the same MDC Mi, DR becomes zero. Otherwise, the extra rendezvous time may be encountered
once or twice depending on which clusters Ss and Sd belong
to as explained above when discussing σ. A sensor does not need to have a global map of the area. The network is modeled at the level of segments rather than
the level of sensors. Discovering segments can be done by
land-based robots or UAVs. Once the segment locations are
known, the area is mapped into a grid and the segment cells
are identified, as explained in detail in the next section. LOAF could be executed at a centralized command center
or by one of the MDCs and then the tours are communi-
cated to the individual MDCs. An MDC is assumed to have
sufficient buffer space for the data transported in one tour. IV. THE LOAF APPROACH To provide energy balanced federation of the n terminals
using k MDCs, LOAF groups the terminals into a set of
k clusters in a star inter-cluster topology, where a central-
cluster Ck serves for data relaying between pairs of clusters. In the first phase, LOAF computes eG, i.e., an energy-based
center of mass, of a set of segments S based on the
inter-segment related communication energy which is pro-
portional to traffic volume exchanged between segments. Then it strives to form non-central-clusters Ci’s i<k around
eG while minimizing the overall energy consumption in the
clusters in terms of motion. In the second phase, LOAF opts
to form a central cluster around eG while balancing energy
among Ci’s, ∀i. The details of LOAF are provided in the
balance of this section. EM (Mi) = µ · TR(Mi)
(3)
EC (Mi) = DataMi · PC (R) , where PC (R)=α+β · R∂
(4) (3)
(4) (3) In (3), µ ranges from 0.1 to 1 J/m [33]. In addition,
the energy required to transmit 1 bit is 2 · 10−6 Joule, where
α = 100nJ, β = 0.1nJ/m∂, ∂= 2 in (4) [34]. During the
federation, LOAF strives to balance E(Mi)’s, ∀i by factoring
in the motion and communication related energy overhead,
i.e., EM(Mi) and EC(Mi) respectively. Overall, the problem
that we tackle in this paper is captured mathematically by the
following formula: Find a set of Ci,
i = 1, 2, . . . k,
where
[
∀i Ci = S such thatD(Sp, Sq) and
X∀Ci,Cj
i̸=j
E (Mi) −E
Mj
are minimized
(5) (5) III. PROBLEM STATEMENT Additionally, the paper focuses on the algorithmic aspect of
the inter-networking problem without considering diversity
of the physical, link and network layers. It is also assumed
that all MDCs have the same communication range R which
is equal to that of a sensor r, i.e., R = r. This is a sim-
plifying assumption to ease the presentation. In addition,
coverage is not the focus of LOAF although MDCs may
have sensing capabilities that can mitigate coverage loss
caused by the damage during their tours. Finally, we assume
that the data volume generated by sensors, i.e. sampling
rate, is fixed and determined by various application-specific
missions [38], [39]; thus the communication load between
segments stays constant during network operation. Nonethe-
less, we study the effect of variation in the data volume
in Section V. Meanwhile, the consumed energy by Mi in serving Ci, i.e.,
E(Mi) includes the energy for motion, i.e., EM(Mi), and the
ancillary energy for communication, i.e., EC(Mi). In detail,
EM(Mi) represents the consumed energy during Mi motion
while it tours the segments of Ci once. In addition, EC(Mi)
includes the required energy for Mi to upload/download
two types of data traffic: the intra-cluster data in Ci and
inter-cluster data, i.e., data imported from or exported to
another cluster Cj. Therefore, EM predominantly scales with
a trip distance of Mi serving Ci, i.e., TR(Mi) in meters
while EC is dependent on the volume of data, DataMi,
uploaded/downloaded over a wireless link, i.e., DataMi =
X
∀Ss,Sd∈Ci
Data (Ss, Sd)
+
X
Sx∈Ci,Sy∈Cj∀j̸=i,
Data
Sx, Sy
DataMi =
X
∀Ss,Sd∈Ci
Data (Ss, Sd) DataMi =
X
∀Ss,Sd∈Ci
Data (Ss, Sd) +
X
Sx∈Ci,Sy∈Cj∀j̸=i,
Data
Sx, Sy
Sx∈Ci,Sy∈Cj∀j̸=i, We assume using an energy cost model for EM that is
proportional to the distance, as seen in (3), while EC is
primarily proportional to the power for transmitting a single
bit over a wireless link seen in (4). VOLUME 8, 2020 1) COMPUTING AN ENERGY-CONSIDERED CORE OF MASS 1) COMPUTING AN ENERGY-CONSIDERED CORE OF MASS
Prior to grouping segments into clusters, LOAF first tries
to find the appropriate location for a central-cluster Ck via
which the inter-segment data traffic between clusters is for-
warded in a star-shaped cluster topology. Since the objective
of LOAF is to balance energy consumed by MDCs that
serve clusters, the ideal location of the central cluster ought
to be at the center of groups of segments with respect to
energy consumption of the MDCs, i.e., a sum of EM (Mi)
and EC (Mi) each of which is mainly affected by data traffic
volume between segments and the inter-segment proximity,
respectively. In other words, LOAF finds the center of mass
of segments i.e., eG, considering data volume as well as
proximity. For federating the segments Si∀i ∈S with k MDCs in an
energy balanced manner, LOAF tries to group the Si’s in a
star inter-cluster topology, where each cluster is served by
a distinct MDC and the inter-segment data delivery can be
operated within at most two cluster hops. Then, LOAF tries
to equalize the energy overhead of MDCs in adjacent clusters. In order to reduce the overall energy consumed in the clusters
by the MDCs, LOAF opts to group segments to minimize
energy overhead experienced by each MDC serving Ci, i < k. Therefore, the segments are grouped into (k −1) non-central
clusters in rounds during which LOAF strives to reduce the
sum of E (Mi) ∀i, i.e., P
∀i E(Mi) in each round. Initially
we assume that a central cluster Ck is formed by placing
an MDC Mk at eG as a stationary relay and includes the
segments within a radio range of Mk that is R, i.e., Ck = {Si|
EuclideanDist(Si, eG) ≤R }. Therefore,
for
computing
eG(cx,
cy)
LOAF
first
calculates
the
total
data
volume
coupled
with
each
segment Si, i.e., a sum of traffic originated from/destined
to Si which is hereafter denoted as DataSi that is equal to
P
∀Sx∈ST ,Si̸=Sx Data (Si, Sx). Then the x- and y- coordinates,
cx and cy are calculated in the same way to find a center of
mass of the weighted vertices. In other words, cx and cy equal
P
∀i x(Si)·DataSi
P
∀i DataSi
, and
P
∀i y(Si)·DataSi
P
∀i DataSi
, respectively, where x(Si)
and y(Si) represents x- and y- coordinate of a segment Si. A. 1st PHASE: FORMING NON-CENTRAL CLUSTERS
CONSIDERING HETEROGENEOUS INTER-SEGMENT
DATA VOLUME The problem presented in formula (5) is to find a set of
clusters Ci’s that covers S, each of which includes a set of
terminals that an MDC Mi visits such that the inter-terminal LOAF first tries to find the center of mass eG of S con-
sidering energy consumption required for the inter-segment 179469 S. Lee et al.: LOAF: Load and Resource Aware Federation of Multiple Sensor Sub-Networks FIGURE 1. In (a) the number on an edge represents data volume exchanged between two segments in megabits. In order to intermittently connect
12 segments by k MDCs, LOAF first computes the weighted core of mass eG of 12 segments with regards to the energy consumption considering
inter-segment data volume. The numbers written in each circle shown in (b) represent the aggregated amount of inter-segment data traffic involved
to each segment and eG is thus determined towards the circles which have larger numbers. FIGURE 1. In (a) the number on an edge represents data volume exchanged between two segments in megabits. In order to intermittently connect
12 segments by k MDCs, LOAF first computes the weighted core of mass eG of 12 segments with regards to the energy consumption considering
inter-segment data volume. The numbers written in each circle shown in (b) represent the aggregated amount of inter-segment data traffic involved
to each segment and eG is thus determined towards the circles which have larger numbers. data transportation. Then non-central clusters, Ci’s i < k are
formed around the found eG while minimizing energy expe-
rienced while MDCs travel the formed non-central clusters. The details of the first phase are provided below. the volume of data that is exchanged between Si and Sji ̸= j
in Figure 1-(a). In consideration of DataSi’s, the location
of eG moves towards the position where the heavy data
exchange overhead is involved like around S3. B. 2nd PHASE: FORMING A CENTRAL-CLUSTER Ck WHILE
BALANCING ENERGY AMONG CLUSTERS While in the 1st phase LOAF focuses on reducing the overall
energy overhead experienced by Mi, i < k in touring non-
central clustering, the major objective of the 2nd phase is
to equalize energy overhead for all MDCs. Since a central
cluster Ck is initially formed by placing Mk at eG as a
stationary relay in the 1st phase, we adjust the size of Ck for
energy balance during the 2nd phase. In other words, the star
inter-cluster topology centered at Ck is re-formed by moving
Pi outwards or inwards depending on the relative energy
consumption of Mi. This phase consists of two steps. In the
first step, Ck is grown by shirking the non-central clusters
whose MDCs consume more energy than the average among
all MDCs, denoted as EAVG. In other words, the rendezvous
point, Pj is moved outwards from eG for a non-central
cluster Cj whose designated MDC, Mj, consumes more
than EAVG. The step is iterative since the tour length of
re-formed clusters will change and subsequently EAVG. The
1st step terminates as Mk consumes energy more than EAVG. Then, the second step is geared to adjust the size of
non-central clusters whose MDCs consume less than EAVG, Figure 2 shows the formation of non-central clusters using
the example seen in Figure 1 as k = 5. We assume that
R is 30m and the required energy to move along TR(Mi)
is computed according to equation (3), where µ = 1 J/m. In addition, 2 Joules are required to transmit 1 megabit over
wireless communication according to (4). For these settings,
Figure 2 shows that the twelve segments are grouped into
Ci, i= 1, 2, 3, 4, in 8 rounds. In the 1st rounds (r = 0),
M5 is to be placed at eG where S3 happens to locate and
forms C5= {S3} and eleven individual clusters are formed,
e.g., Ci = {Si, eG} for i = 0, 1, 2 and Ci−1 = {Si, eG} for
i = 4, . . . , 11. Then during r = 1, 2, and 3, each pair of two
clusters, i.e., (C4, C5), (C7, C8), and (C9, C10) is combined
and becomes one cluster, {S4, S5, eG}, {S7, S8, eG}, and
{S9, S10, eG}, respectively since merging two clusters reduces
the most energy consumed for travelling in each round. Algorithm 1 Pseudocode of the 1st Phase of LOAF Algorithm 1 Pseudocode of the 1st Phase of LOAF
1. S ←a set of given segments Si, ∀i;
2. DataSi ←a set of Data(Si, Sj), Si ̸= Sj ∈S; // Data
volume exchanged between Si and Sj
// Computing a weighted core of mass of S, eG(cx, cy)
3. cx =
P
∀i x_(Si)·DataSi
P
∀i DataSi
, cy =
P
∀i y_(Si)·DataSi
P
∀i DataSi
// forming (k −1) non-central clusters minimizing the
overall energy consumption of MDCs
4. Ck = {Si| EuclideanDist(Si, eG) ≤R };
5. S−=Ck;
6. i = 0
7. for each Sj∈(S−Ck){
8. Ci = {Sj, eG }; i + +;
9. } end for
10. r = 1; N 0
cluster = |S| −|Ck|; Ncluster = N 0
cluster;
11. do {
12. Cx, Cy =
13. min
x,y
PN 0
cluster−r
i=1,i̸=x,y E (Ci) + ME
Cx ∪Cy
,
where ME
Cx, Cy
= EM
Cx ∪Cy
+EC
Cx ∪Cy
14. Cx∪= Cy; Cy = CNcluster ;
15. r = r + 1; Ncluster−=1;
16. } while (r < N 0
cluster−(k−1)) min
x,y
XN 0
cluster−r
i=1,i̸=x,y E (Ci) + ME
Cx ∪Cy
,
where ME
Cx ∪Cy
= EM
Cx ∪Cy
+ EC
Cx ∪Cy
(6) 1. S ←a set of given segments Si, ∀i; g
g
i,
;
2. DataSi ←a set of Data(Si, Sj), Si ̸= Sj ∈S; // Data
volume exchanged between Si and Sj
// Computing a weighted core of mass of S, eG(cx, cy) (6) p
g
g
,
( x,
y)
3. cx =
P
∀i x_(Si)·DataSi
P
∀i DataSi
, cy =
P
∀i y_(Si)·DataSi
P
∀i DataSi
// forming (k −1) non-central clusters minimizing the
overall energy consumption of MDCs p
g
g
,
( x,
y)
3. cx =
P
∀i x_(Si)·DataSi
P
∀i DataSi
, cy =
P
∀i y_(Si)·DataSi
P
∀i DataSi
// forming (k −1) non-central clusters minimizing the
overall energy consumption of MDCs In (6), ME(Cx ∪Cy) represents energy required for an
MDC to serve the merged cluster that includes Cx and Cy
for transporting data and travelling. EC
Cx ∪Cy
is mainly
affected by the aggregated data volume, i.e., DataSx +DataSy
and EM
Cx ∪Cy
, the updated trip that is the shortest path
in the merged cluster Cz. Algorithm 1 Pseudocode of the 1st Phase of LOAF Since the goal of this step is to
group segments in the form which minimizes P
∀i EM(Mi)
and P
∀i EC(Mi) and P
∀i EC(Mi) is already determined by
given data transport requirements between segments, LOAF
focuses on reducing P
∀i EM(Mi) in the clustering process. Therefore, a pair of Cx and Cy whose merged tour path
decreases P
∀i EM(Mi), will be selected. overall energy consumption of MDCs
4. Ck = {Si| EuclideanDist(Si, eG) ≤R };
5. S−=Ck;
6. i = 0
7. for each Sj∈(S−Ck){
8. Ci = {Sj, eG }; i + +;
9. } end for
10. r = 1; N 0
cluster = |S| −|Ck|; Ncluster = N 0
cluster;
11. do {
12. Cx, Cy =
13. min
x,y
PN 0
cluster−r
i=1,i̸=x,y E (Ci) + ME
Cx ∪Cy
,
where ME
Cx, Cy
= EM
Cx ∪Cy
+EC
Cx ∪Cy
14. Cx∪= Cy; Cy = CNcluster ;
15. r = r + 1; Ncluster−=1;
16. } while (r < N 0
cluster−(k−1)) ∀i
Computing such a merged path of Cx and Cy is equivalent
to the problem which finds the shortest Hamiltonian cycle
to visit Si′s∀i ∈(Cx∪Cy) considering R that is NP-hard. Therefore, LOAF uses the heuristic solution in [24], [35], [36]
to compute a tour path of Cz. The clustering procedure is
repeated until Ncluster become (k −1) and the 2nd step of
the first phase thus terminates as r=N 0
cluster −(k −1). We refer to the point Pi on the circle of radius R and centered
at eG, where Mk is placed, as the rendezvous points between
Mii < k and Mk. When reaching Pi, Mi will be able to
establish communication with Mk and exchange relevant data
payload. The position of rendezvous points will be adjusted in
the 2nd phase of LOAF as we explain in the next subsection. 1) COMPUTING AN ENERGY-CONSIDERED CORE OF MASS Figure 1-(b) shows the found eG assuming the area of interest
is 270 × 360m2 and the cell width is corresponding to
R
√
2,
R = 30m. The eG is located in a different position from the
centroid G computed only considering locations of segments. In the figure, the number in a circle of Si represents DataSi in
megabits based on the numbers on an edge e(Si, Sj) that means In the first round (r = 0), for each Si∈(S−Ck) an indi-
vidual cluster Ci = {Si,eCoM} , i= 0, . . . ,N 0
cluster is initially
formed, where N 0
cluster equals Nseg −|Ck|, where Nseg is
the number of segments. Then, each E (Mi) ∀i is computed
as a sum of EM(Mi) and EC(Mi). Since every Ci includes
only two elements, eG and Si, EC(Mi) is mainly determined
by DataSi and EM(Mi), the length of TR(Mi) which follows
the straight path between eG and Si. Thus |TR(Mi)| is equal
to (EuclideanDist(eG, Si) – 2R), where EuclideanDist(a, b)
is an Euclidean distance between a and b. Obviously, it is
sufficient for an MDC to be at a distance R from the seg-
ments and from Mk in order to establish communication links,
assuming free space signal propagation model. In the second 179470 179470 VOLUME 8, 2020 VOLUME 8, 2020 S. Lee et al.: LOAF: Load and Resource Aware Federation of Multiple Sensor Sub-Networks four non-central clusters C0 = {S0, S1, eG }, C1 = {S2, S9,
S10, S11, eG}, C2 = {S6, S7, S8, eG }, and C3 = {S4, S5,
eG} are formed with the least amount of energy P4
i=1 E(Mi),
where E(M1) = 320, E(M2) = 526, E(M3) = 470, and
E(M4) = 330. The pseudo code of the1st phase is described
in Algorithm I. round (r
= 1), LOAF opts to combine two clusters Cx
and Cy whose merging leads to the greatest decrease in the
overall tour length. This step is performed repetitively. Thus,
the number of clusters in round r denoted as Ncluster decreases
by one in each round. In other words, a pair of Cx and Cy
which satisfies the equation (6) is selected in each
round r ≥1. Algorithm 1 Pseudocode of the 1st Phase of LOAF B. 2nd PHASE: FORMING A CENTRAL-CLUSTER Ck WHILE
BALANCING ENERGY AMONG CLUSTERS The
selection of a pair of two clusters to merge is repeated in
the subsequent rounds. Finally in the last round (r = 7), 179471 VOLUME 8, 2020 S. Lee et al.: LOAF: Load and Resource Aware Federation of Multiple Sensor Sub-Networks FIGURE 2. In the 2nd step of the 1st phase, LOAF opts to form (k −1) non-central clusters, C0, ..., Ck−1, (k = 5) during which the overall energy
consumption by MDCs is minimized. The step operates in rounds. In the first round, each segment and eG form an individual cluster. Since S3 is within R
from eG, (Nseg−1) clusters are formed, where Nseg is the number of segments. After that two clusters are combined in each round while minimizing the
overall energy experienced by MDCs. The process is repeated until the number of clusters equals (k −1) = 4. FIGURE 2. In the 2nd step of the 1st phase, LOAF opts to form (k −1) non-central clusters, C0, ..., Ck−1, (k = 5) during which the overall energy
consumption by MDCs is minimized. The step operates in rounds. In the first round, each segment and eG form an individual cluster. Since S3 is within R
from eG, (Nseg−1) clusters are formed, where Nseg is the number of segments. After that two clusters are combined in each round while minimizing the
overall energy experienced by MDCs. The process is repeated until the number of clusters equals (k −1) = 4. where the rendezvous points for these clusters are moved
inwards toward eG. Both steps are iterative in nature and
are performed in rounds. The following notation used in the
discussion. tour of the segments of Ci i.e., EM(Mi), based on the
membership of Ci in round r. EC(Mi) and EM(Mi) are
computed based on equation (3) and (4), respectively. r
r • Cr
i
: reflects the segments grouped into a cluster Ci
in round r. C0
k includes segments that are reachable from
Mk which is placed at eG i.e., {Si|EuclideanDist(Si,
eCoM) ≤R} and C0
i ’s, i < k are the same as the
1st phase. With the provided notations, the 2nd phase of LOAF opts
to balance the energy overhead by adjusting the MDC tours. Basically, LOAF tries to adjust rendezvous points Pi’s of Ci’s
outwards or inwards based on the average energy overhead. B. 2nd PHASE: FORMING A CENTRAL-CLUSTER Ck WHILE
BALANCING ENERGY AMONG CLUSTERS In other words, LOAF relocates Pm outwards towards Cm if
Em > EAVG in order to reduce the heavy energy overhead of
Cm or reposition Pl inwards towards eG, where Ck is initially
formed if El < EAVG for expanding Cl. Since the computed
eG during the 1st phase lies inside the convex hull of the
segments, as will be proven in Lemma 1, and the initial energy
overhead E0
k of forming Ck at the eG, is less than EAVG, as
will be proven in Lemma 2, (i) the rendezvous points Pm’s for
Cm’s whose Er
m > Er
AVG are moved outward while Er
k < Er
AVG
and then (ii) Pl’s for Cl’s whose Er
l < Er
AVG are also handled
while Er
k > Er
AVG. Each of the steps (i) and (ii) is repeated
in rounds until there is no more improvement of the standard
deviation (SD) of Ei∀i as explained below. • Pr
i : It represents a rendezvous point where Mii < k
meets Mk in round r. Pr
i ’s ∀i are always on the straight
line between eG and CoMr−1
i
which will be explained
later. • TRr
i : The travel path that Mi takes in round r while visit-
ing the segments in Ci is denoted as TRr
i . It is equivalent
to the shortest path where Mi departs from Pr
i and visits
every segment Si ∈Ci once and returns to Pr
i . Since
finding TRr
i is mapped into the Travelling Salesman
Problem, the heuristic solution of [24], [35], [12] will
be used to find TRr
i . i
• CoMr
i
: It represents a core of mass of the poly-
gon formed by the segments of Ci and Pr
i in round r. In round r of the 2nd phase of LOAF, Ck may be
expanded by moving Pr
i towards CoMr−1
i
of Ci or Cj
may be extended by relocating its Pr
j towards CoMr−1
k
. (i) moving P′
is outwards: During this step, LOAF iter-
atively shrinks the boundary of clusters, Cm’s, of heavily
loaded MDCs, and accordingly expands Ck. In the first round
(r = 0), for each cluster Ci, i = 1, . . . B. 2nd PHASE: FORMING A CENTRAL-CLUSTER Ck WHILE
BALANCING ENERGY AMONG CLUSTERS , k −1, P0
i , TR0
i , and
CoM0
i are set up based on the intra-cluster topology formed
in the 1st phase. Then the initial energy consumption, E0
i is
computed as a sum of EC(Mi) and EM(Mi), e.g., E0
1 = 320, j
• Er
i : denotes the energy experienced by Mi in round r for
up/downloading intra-cluster data in Ci and inter-cluster
data at Pr
i i.e., EC(Mi) and also during completing one 179472 VOLUME 8, 2020 VOLUME 8, 2020 VOLUME 8, 2020 S. Lee et al.: LOAF: Load and Resource Aware Federation of Multiple Sensor Sub-Networks E0
2
= 526, E0
3
= 470, E0
4
= 330, and E0
5
= 194 in
the example seen in Figure 2-(f), and which corresponds to
the initial star-shaped cluster arrangement. In particular, E0
5
reflects only EC(M5) which includes energy consumption for
transporting the inter-cluster data traffic volume represented
using double dotted lines in Figure 3. Therefore, EC(Mk)
increases when high volume of data traffic is delivered
between clusters. during this process as will be proven in Theorem 1. The same
process of Ck expansion is repeated in successive rounds and
terminates in round f where Mk consumes energy more than
average, i.e., Ek ≥EAVG. Therefore, the process is converged
in O( Dist(eG,L)
ε
), where L is the smallest polygon that includes
all segments, and Dist(eG, L) represents a distance from eG
to the furthest point on L. (ii) moving Pl inwards: This step adjusts Pl’s of Cl’s,
whose serving Ml requires less energy than average, i.e.,
El < EAVG. Thus LOAF tries to extend the size of Cl’s by
moving the Pl’s towards Ck, exactly towards eG or a core of
mass of Ck, i.e., CoMk in case Ck includes more than one
segment. Accordingly, TRk is shortened. For repositioning
of Pl’s, LOAF also computes ϕ′l which reflects the relative
energy shortage of Ml to the least energy consuming MDC
and is found in the opposite way of computing ϕi, i.e., ϕ′l =
Eleast
El
and then Pl moves inwards (ϕ′l ×ε) meters along a line
from Pl to CoMk. Using ϕ′l, the less energy overhead Ml
incurs, the more distance Pl moves inwards towards CoMk. Thus Cl is accordingly extended and El also increases. After
that, TRl, CoMk, and El, Ek and EAVG are updated. B. 2nd PHASE: FORMING A CENTRAL-CLUSTER Ck WHILE
BALANCING ENERGY AMONG CLUSTERS The same
procedure is repeated until Ek ≤EAVG. FIGURE 3. In the 1st phase of LOAF, the non-central clusters C1, C2, C3
and C4 are formed centered at eG where M5 is placed. C5 that includes
segments reachable from M5 is formed. Thus E0
5 contains only energy for
up/downloading inter-cluster data traffic represented using double
dotted lines. procedure is repeated until Ek ≤EAVG. Figure 4 shows the procedure of the 2nd phase of LOAF
as ε is equal to R. In the first round (r = 0) of the Ck
extension step, the initial rendezvous points P0
i ’s ∀i < k
are located at eG. Then each Ci relocates ( E0
i
E0
5 × R) meters,
i
=
2,3,4 respectively towards CoM0
i i= 2, 3, 4 as seen
in Figure 4-(b). This is because M0
2 , M0
3 , and M0
4 con-
sume more energy than E0
AVG. Accordingly, C1
5 is extended
towards C0
2, C0
3, and C0
4 as seen in Figure 4-(c). In addi-
tion, the same process is applied to C0
2, and C0
3 in a sub-
sequent round as shown in Figure 4-(d). After that S10
becomes unreachable to M2, and accessible to M5; there-
fore, S10 changes association to C5. Figure 4-(e) to (h) show
the 2nd step. The first round of that step, C0
1, C0
3, and
C0
4 grow towards C5 by adjusting their rendezvous points. Similarly, C0
1, and C0
4 are extended in the second round. Finally, the resulting intermittent topology of k (= 5)
MDCs formed in an energy-balanced manner for federat-
ing twelve segments after the first round of the 2nd phase
is presented in Figure 4-(i). The pseudo code of the 2nd
phase of LOAF is described in Algorithm II. Lines 9-12
describe the first step and lines 23-32 explain the second step. As a result, LOAF returns k sets of segments each of which
is served by an individual MDC and their tour paths, i.e.,
{Ci, TRi∀i } in line 33. Figure 4 shows the procedure of the 2nd phase of LOAF
as ε is equal to R. In the first round (r = 0) of the Ck
extension step, the initial rendezvous points P0
i ’s ∀i < k
are located at eG. B. 2nd PHASE: FORMING A CENTRAL-CLUSTER Ck WHILE
BALANCING ENERGY AMONG CLUSTERS Then each Ci relocates ( E0
i
E0
5 × R) meters, FIGURE 3. In the 1st phase of LOAF, the non-central clusters C1, C2, C3
and C4 are formed centered at eG where M5 is placed. C5 that includes
segments reachable from M5 is formed. Thus E0
5 contains only energy for
up/downloading inter-cluster data traffic represented using double
dotted lines. LOAF relocates Pi’s away from eG starting with clusters
whose MDCs consume energy more than average and adjusts
their tour accordingly. For the relocation of Pi’s, LOAF first
determines the relative movement rate ϕi of each Pi based
on the energy imbalance among the MDCs, i.e., ϕi =
Ei
Eleast ,
where Eleast = min
∀j E0
j . In other words, ϕi factors in the
relative excessive energy overhead of each Mi to the least
energy consuming MDC. In order to adjust the actual moving
distance in meters, a constant parameter, ε (≥1) is multiplied
to ϕi. With excessively large values of ε, the convergence to
form the energy-balanced clusters may be fluctuated while
small ε slows the convergence of energy balancing. Thus,
an appropriate value of ε, should be selected. Since the
formed clusters are served by MDCs, the communication
range of an MDC, i.e., R is the recommended setting for ε
in practice. Based on ϕi and ε, LOAF expands a size of Ck and
accordingly reduces that of Ci by moving Pi away (ϕi × ε)
meters along a line from Pi to a core of mass of a polygon
formed by segments grouped to Ci in round r = 0 denoted
as CoM0
i . After that tour paths of Mk and Mi, i.e., TRk and
TRi, are recomputed using an updated rendezvous point Pi. Accordingly, Ek, Ei and CoMi are also updated. In addition,
re-grouping of segments may occur, e.g., Sx ∈Ci may be
included to Ck if Sx becomes reachable from Mk after expand-
ing Ck. Grouping all segments to k clusters is maintained C. ALGORITHM ANALYSIS LOAF is analyzed in this subsection. We mainly focus on
proving that the energy imbalance among MDCs is mini-
mized by LOAF while forming a star inter-cluster topology. In addition, the complexity of LOAF is analyzed to show
that the resulting cluster topology is formed within an exe-
cution time bound. We introduce the following theorems and
lemmas: 179473 VOLUME 8, 2020 S. Lee et al.: LOAF: Load and Resource Aware Federation of Multiple Sensor Sub-Networks S. Lee et al.: LOAF: Load and Resource Aware Federation of Multiple Sensor Sub Networks
FIGURE 4. The 2nd phase of LOAF is dedicated to balance energy among MDCs by adjusting the rendezvous points Pi , i < k. Balancing is performed in an
iterative manner during which LOAF expands Ck by moving Pi outwards towards Ci if Ei > EAVG in order to reduce more-than-energy consuming
clusters Ci ’s,(a)-(e) and then Pl ’s of Cl ’s whose El <EAVG are moved inwards towards Ck, (f)-(h). (i) shows the result of the two steps. FIGURE 4. The 2nd phase of LOAF is dedicated to balance energy among MDCs by adjusting the rendezvous points Pi , i < k. Balancing is performed in an
iterative manner during which LOAF expands Ck by moving Pi outwards towards Ci if Ei > EAVG in order to reduce more-than-energy consuming
clusters Ci ’s,(a)-(e) and then Pl ’s of Cl ’s whose El <EAVG are moved inwards towards Ck, (f)-(h). (i) shows the result of the two steps. a solid polygon in Figure 5-(b). In addition, Figure 5-(c)
shows the resulting topology formed by LOAF. Lemma 1: In case that Nseg is more than two, the energy-
based center of mass among segments during the 1st phase of
LOAF, denoted as eG, always lies inside the convex hull of all
segments. shows the resulting topology formed by LOAF. Therefore, the computed eG in the 1st phase of LOAF always
lies inside the convex hull of the segments. Proof: The computed energy-based center of mass
among segments, denoted as eG can be one of three
cases [40], [41]. Lemma 2: The energy consumption of Mk serving for the
central cluster formed in the 1st phase of LOAF is larger than
the average energy of all MDCs, i.e., Eo
k > E0
AVG. C. ALGORITHM ANALYSIS The maximum number of
non-central clusters corresponds to the first round (r = 0) and
equals (N 0
cluster −k + 1), where N 0
cluster
≤Nseg
. The main
time complexity of each round comes from finding a tour
path of each cluster, i.e., TR(Mi), which is computed using
the heuristic whose execution time bounds to O(n4logn) [24],
where n is the number of segments in a cluster in our case,
i.e., 1, 2, 3, ..., (N 0
cluster −k + 1) each round. Thus the 1st
phase’s time complexity equals P(N 0
cluster−k+1)
n=1
n4logn which
is O( (Nseg −k) · (Nseg −k)4log(Nseg −k)). □ i
TR0
i ;
7. } end for
8. E0
AVG ←Average(E0
i , ∀i); // Eo
k < E0
AVG
9. SD0 ←Standard Deviation of Ef
i ∀i; rr ←0;
10. do {
11. r = 1; //1st step during which P′
ms move out-
ward
12. do{
13. Er−1
least = Er−1
i
;
14. for Er−1
i
≥Er−1
AVG, ∀i {
15. Lr
i = Line from Pr−1
i
towards CoMr−1
i
;
16. Pr
i moves
ε × Er−1
i
Er−1
least
meters along Lr
i ;
17. }end for
18. Update TRr
k based on Pr
i ;
19. if ∃Sx ∈Cr
i in a polygon formed by TRr
k then {
20. Cr
k ∪= {Sx}; Cr
i −= {Sx};
21. } end if
22. Compute TRr
i and CoMr
i , i < k and Er
i , ∀i;
23. Er
AVG = Average(Er
i , ∀i); r+ = 1;
24. } while ( Er
k < Er
AVG)
d 11. r = 1; //1st step during which P′
ms move out-
ward is O( (Nseg −k) · (Nseg −k)4log(Nseg −k)). g
g
g
Theorem 1: LOAF guarantees the convergence to grouping
Nseg segments into k-clusters in a star topology. Theorem 1: LOAF guarantees the convergence to grouping
Nseg segments into k-clusters in a star topology. i
AVG
15. Lr
i = Line from Pr−1
i
towards CoMr−1
i
;
16. Pr
i moves
ε × Er−1
i
Er−1
least
meters along Lr
i ;
17
}end for g
Proof: As proven in Lemma 3, LOAF groups Nseg seg-
ments into k-clusters in its first phase. C. ALGORITHM ANALYSIS Pr
i moves
ε × Er−1
i
Er−1
least
meters along Lr
i ;
17. }end for
18. Update TRr
k based on Pr
i ;
19. if ∃Sx ∈Cr
i in a polygon formed by TRr
k then {
20. Cr
k ∪= {Sx}; Cr
i −= {Sx};
21. } end if
22. Compute TRr
i and CoMr
i , i < k and Er
i , ∀i;
23. Er
AVG = Average(Er
i , ∀i); r+ = 1;
24. } while ( Er
k < Er
AVG)
25. f = r = r −1; // 2nd step during which P′
ls move
inward
26. do {
27. CoMr
k = CoMf
k;
28. for each Cr
i whose Er
l < Er
AVG {
29. Lr
i = Line from Pr
i towards CoMr
k;
30. Pr
i moves
ε × Er−1
least
Er−1
i
meters along Lr
i ;
31. Update TRr
i , Er
i , and CoMr
i ;
32. }end for
33. r+ = 1;
34. } while ( Er
k > Er
AVG)
35. rr+ = 1; SDrr ←Standard Deviation of Er
i ∀i;
36. } while ( SDrr < SDrr−1)
37. return {Cr−1
i
, TRr−1
i
∀i } Eo
k cannot be less than E0
AVG because P
∀i,j,i̸=j Data
Si, Sj
≈
P
∀i
P
Sj∈Ci DataSj
k
and TR
L
(k−1)
> 0. □ Eo
k cannot be less than E0
AVG because P
∀i,j,i̸=j Data
Si, Sj
≈
P
∀i
P
Sj∈Ci DataSj
k
and TR
L
(k−1)
> 0. □ Algorithm 2 Pseudo Code of the 2nd Phase of LOAF
// Forming a central-cluster while balancing E(Mi), ∀i
(
)
Lemma 3: The energy consumption of Mi serving a non-
central cluster Ci, i < k is minimized in O((N seg −k) ·
(Nseg −k)4log(Nseg −k)).
(
)
Lemma 3: The energy consumption of Mi serving a non-
central cluster Ci, i < k is minimized in O((N seg −k) ·
(Nseg −k)4log(Nseg −k)). g
g
Proof: The 1st phase of LOAF is devoted to forming (k −1)
non-central clusters, during which at most Nseg of clusters
are grouped into (k −1) clusters via successive merging
two clusters in subsequent rounds. C. ALGORITHM ANALYSIS Thus, for the proof
of the convergence of Nseg segments into k-clusters, it is
sufficient to show that (i) a segment Si which is not visited
by the updated tour of Mi is necessarily visited by Mk during
the first step of the 2nd phase, and (ii) vice versa, for Si that is
not visited by TRr
k ought to be covered by TRr
i , ∃i < k during
the second step. 17. }end for 18. Update TRr
k based on Pr
i ; 19. if ∃Sx ∈Cr
i in a polygon formed by TRr
k then {
20. Cr
k ∪= {Sx}; Cr
i −= {Sx}; k
21. } end if 22. Compute TRr
i and CoMr
i , i < k and Er
i , ∀i; p
For proving (i), it is required to show that a candidate Si for
re-clustering is always placed between a convex hull CHr of
{Sj∀j ∈Ci, Pr
i } and TRr
i which is R away from CHr and will
be visited by Mi or Mk in the next round (r+1). It is proven by
contradiction. If Si resides outside CHr or inside R away from
TRr
i , it self-proves the found CHr or TRr
i is incorrect due to
the natural features of a convex hull and the way of computing
a tour path described in [24] respectively. Therefore, Si that
will be re-grouped into Ck is undoubtedly inside CHr and in
a range R from TRr
i . In addition, in the cases in which Si will
not be reachable from TRr+1
k
and thus not be re-clustered into
Ck as seen in Figure 6-(c) and (d), the Si will be necessarily
visited by the updated tour path TRr+1
i
as Figure 6-(a) and (b)
show. The same proof applies to (ii) and thus k-clustering
of the segments formed in a star topology is guaranteed
by LOAF. □ Theorem 2: The time complexity of LOAF is proportional
to the computation time for TRr
i which is determined by a
gap between Nseg and the number of the available MDCs (k),
as ((N seg + Nrk) · (Nseg −k)4 · log(Nseg −k)), where
Nr is the maximum iteration of the 2nd phase and bounds to
O( Dist(eG,L)
ε
), k is the number of available MDCs and Nseg is
the number of segments in the area L. C. ALGORITHM ANALYSIS (1) If there are only two segments, i.e., Nseg = 2, then
eG lies on the line connecting the two segments. Proof: Via the 1st phase of LOAF, the initial inter-cluster
topology is formed in a star shape, where a central cluster Ck
is formed by placing Mk as a stationary relay at the computed
eG, around which (k −1) non-central clusters are formed. Segments residing within R from Mk are made part of Ck and
the rest of segments are grouped into Ci, i < k. Therefore,
the energy consumed by Mk, Eo
k is completely determined
by EC(Mi), which is primarily affected by the sum of all
inter-segment data volume, i.e., Eo
k ≈P
∀i,j,i̸=j Data(Si, Sj). In addition, the average energy consumption of all MDCs, (2) If Nseg ≥3 and the segments form a convex polygon
as illustrated by the example in Figure 2, then eG lies
inside the polygon. (3) If Nseg ≥4 and the segments form a concave polygon as
an example seen in Figure 5, then eG may lie outside the
polygon depending on the inter-segment data volume
as seen in Figure 5-(a). However, the eG lies inside the
convex hull which includes all segments, represented as 179474 VOLUME 8, 2020 VOLUME 8, 2020 S. Lee et al.: LOAF: Load and Resource Aware Federation of Multiple Sensor Sub-Networks Algorithm 2 Pseudo Code of the 2nd Phase of LOAF
// Forming a central-cluster while balancing E(Mi), ∀i
1. for ∀i {// based on the initial inter-cluster topology
formed during the 1st phase
2. C0
i ←Ci, in the 1st phase;
3. TR0
i ←Tour path of Mi in Ci computed by
[24], [35], [12];
4. P0
i ←Rendezvous point where Mi meets Mk;
5. E0
i ←Energy consumed by Mi in Ci;
6. CoM0
i ←Core of mass of a polygon formed by
TR0
i ;
7. } end for
8. E0
AVG ←Average(E0
i , ∀i); // Eo
k < E0
AVG
9. SD0 ←Standard Deviation of Ef
i ∀i; rr ←0;
10. do {
11. r = 1; //1st step during which P′
ms move out-
ward
12. do{
13. Er−1
least = Er−1
i
;
14. for Er−1
i
≥Er−1
AVG, ∀i {
15. Lr
i = Line from Pr−1
i
towards CoMr−1
i
;
16. C. ALGORITHM ANALYSIS E0
AVG
is mainly determined by a total data volume
up/downloaded by each Mi i.e., P
∀i
P
Sj∈Ci DataSj and a
sum of TR(Mi), i
<
k. Since each TR(Mi) is bound
to the tour length of each Mi in the area of (
L
(k−1)),
where L is the smallest polygon that includes all segments,
E0
AVG
≈
1
k
P
∀i
P
Sj∈Ci DataSj + TR(
L
(k−1))
. Therefore, Proof: As proven in lemma 3 the 1st phase of LOAF is
bound to O((Nseg−k)·(Nseg −k)4log(Nseg −k)). In addition,
the 2nd phase of LOAF is iterative in two steps. The number 179475 VOLUME 8, 2020 S. Lee et al.: LOAF: Load and Resource Aware Federation of Multiple Sensor Sub-Networks FIGURE 5. (a) An example setup of 9 segments that form a concave polygon. The numbers on each line represent data volume exchanged between the
segments at the two ends in megabits; (b) Showing an energy-based center of mass, denoted as eG that lies outside the concave polygon due to the
heavy data communication load between S1 and S9. However, eG lies inside the convex hull of segments represented by a solid line; (c) Depicting the
formed data collection topology of five clusters, i.e., C1 = {S1}, C2 = {S2, S3, S4}, C3 = {S6, S7, S8,}, C4 = {S9}, C5 = {S5} and the MDCs’ routes, assuming
an area of interest = 270m×360m, R = 30m, and MDC count of five. FIGURE 5. (a) An example setup of 9 segments that form a concave polygon. The numbers on each line represent data volume exchanged between the
segments at the two ends in megabits; (b) Showing an energy-based center of mass, denoted as eG that lies outside the concave polygon due to the
heavy data communication load between S1 and S9. However, eG lies inside the convex hull of segments represented by a solid line; (c) Depicting the
formed data collection topology of five clusters, i.e., C1 = {S1}, C2 = {S2, S3, S4}, C3 = {S6, S7, S8,}, C4 = {S9}, C5 = {S5} and the MDCs’ routes, assuming
an area of interest = 270m×360m, R = 30m, and MDC count of five. FIGURE 6. During the first step of the 2nd phase, Sd covered by Mi in
round r is necessarily visited by Mi (a,b) or Mk(c,d) in the next round
(r + 1). V. PERFORMANCE EVALUATION In this section, the effectiveness of LOAF is validated through
simulation. The simulation experiments study the perfor-
mance with respect to the number of RNs populated by LOAF
in comparison with the best known and recently published
algorithms. The quality of the resulting topology of LOAF is
also discussed. FIGURE 6. During the first step of the 2nd phase, Sd covered by Mi in
round r is necessarily visited by Mi (a,b) or Mk(c,d) in the next round
(r + 1). The inner solid line and inner dotted line represent TRr
i and TRr+1
i
respectively. In (a) and (b), Sd remains in Ci since it is R away from TRr+1
k
and the convex hull used in round r is modified for computing TRr+1
i
considering Sd since Sd becomes a border segment in Cr+1
i
. C. ALGORITHM ANALYSIS The inner solid line and inner dotted line represent TRr
i and TRr+1
i
respectively. In (a) and (b), Sd remains in Ci since it is R away from TRr+1
k
and the convex hull used in round r is modified for computing TRr+1
i
considering Sd since Sd becomes a border segment in Cr+1
i
. of rounds Nr for balancing energy in the steps is bound to
O( Dist(eCoM,L)
ε
), where L is the smallest polygon that includes
all segments, and Dist(eG, L) represents a distance from eG
to the furthest point on L in meters. In addition, the execution
time of each round in both steps is determined by computing
TR(Mi) for at most (k −1) clusters and the average number of
elements in clusters tentatively formed during the 2nd phase
is Nseg
k . In addition, the two steps are iterative while the stan-
dard deviation of Ei, ∀i has been improved. Therefore, the 2nd
phase of LOAF requires the execution time complexity which
is bounded by (Nrk · ( Nseg
k )
4 · log( Nseg
k )). In conclusion,
the time complexity of LOAF for grouping Nseg segments into
k-clusters in an energy-balanced manner equals {(Nseg −k) ·
(Nseg −k)4log(Nseg −k) + Nrk · ( Nseg
k )
4 · log( Nseg
k )} which
bounds to ((N seg +Nrk)·(Nseg −k)4 ·log(Nseg −k), where
Nr = O( Dist(eCoM,L)
ε
). □ □ □ B. BASELINE APPROACHES Then each tour path is determined as done
in IDM-kMDC [24] and the average tour length TLavg
is found. ToCS then adjusts the size of Ck towards Ci in
order to balance the MDC tour lengths in the 2nd phase. During the 2nd phase, Ck expands towards Ci whose tour
length TLi is larger than TLavg by moving Pi along the
line between G and x of Ci until TLi > TLavg. In the case
of TLi < TLavgPi moves towards G. The same process
is repeated for each Ci∀i and the 2nd phase terminates
when TLavg ≈TLi, ∀i. In ToCS the waiting time at the
rendezvous point contributes to the inter-segment com-
munication delay due to the need for MDC synchroniza-
tion where data uploading/downloading is performed. • ToCS forms clusters around the center of the area G
and consists of two phases [32]. In the first phase, each
cluster segment initially becomes a cluster Ci, and then
combines two Ci’s whose merging cost based on G is the
least. The merging is repeated until (k −1) clusters are
formed. Then a center-cluster Ck is formed by including
(k −1) rendezvous points Pi, i = 1,..,(k −1), each of
which is computed as a mid-point between G and the
closest point ‘‘x′′ to G that is on the convex hull of seg-
ments of Ci. Then each tour path is determined as done
in IDM-kMDC [24] and the average tour length TLavg
is found. ToCS then adjusts the size of Ck towards Ci in
order to balance the MDC tour lengths in the 2nd phase. During the 2nd phase, Ck expands towards Ci whose tour
length TLi is larger than TLavg by moving Pi along the
line between G and x of Ci until TLi > TLavg. In the case
of TLi < TLavgPi moves towards G. The same process
is repeated for each Ci∀i and the 2nd phase terminates
when TLavg ≈TLi, ∀i. In ToCS the waiting time at the
rendezvous point contributes to the inter-segment com-
munication delay due to the need for MDC synchroniza-
tion where data uploading/downloading is performed. • Maximum inter-segment communication delay (Dmax):
Obviously, LOAF strives to minimize the maximum
data delivery latency between segments. Dmax is mainly
determined by rendezvous time along the path between
two furthest apart segments, which is affected by a count
of clusters and inter-segment data traffic volume. B. BASELINE APPROACHES • Energy balance among MDCs: This is also a main
objective of LOAF. Since LOAF tries to balance
energy considering motion and communication both,
energy balance among MDCs is computed as a stan-
dard deviation of total energy consumption of each
MDC for motion and communication, i.e., it equals
r PNMDC
i=1
|E(Mi)−
¯
E(M)|2
NMDC
, where
¯
E(M) =
PNMDC
i=1
E(Mi)
NMDC
and
E (Mi) = EM (Mi) + EC (Mi). • Average energy consumption of all MDCs (AEC):
It measures the average energy consumed by all MDCs
before the first MDC dies. It will represent the absolute
amount of energy consumption of MDCs for intermit-
tently connecting segments. p
g
g
p
• FOCUS operates in three phases [31]. In the 1st phase,
segments are grouped into k disjoint clusters based on
proximity. Each cluster is to be served by an MDC. In order to minimize the tour path of an MDC, clustering
is initiated by selecting k farthest away segments from
the centroid and then the remaining segments are joined
in a greedy way. The set of k disjoint clusters is repre-
sented as a directed graph G = (V, E), where V reflects
clusters and E contains a set of weighted directed edges
−−−→
Ca, Cb between every pair of clusters. The weight of
−−−→
Ca, Cb is equal to the increase in the tour length of
Ca when Ca is extended towards Cb by including a
segment from Cb. During the 2nd phase FOCUS opts to
overlap the k clusters in order to minimize tour lengths
of MDCs for the inter-cluster communication; it does
so by selecting an intersection segment (IS) between • FOCUS operates in three phases [31]. In the 1st phase,
segments are grouped into k disjoint clusters based on
proximity. Each cluster is to be served by an MDC. In order to minimize the tour path of an MDC, clustering
is initiated by selecting k farthest away segments from
the centroid and then the remaining segments are joined
in a greedy way. The set of k disjoint clusters is repre-
sented as a directed graph G = (V, E), where V reflects
clusters and E contains a set of weighted directed edges
−−−→
Ca, Cb between every pair of clusters. B. BASELINE APPROACHES may increase the traffic volume where more data is
exchanged among segments and thus inter-MDC energy
balancing becomes more complicated. The performance of LOAF is compared to four compet-
ing approaches. The first two solutions focus on reducing
inter-segment communication delay. One of them, namely
ToCS [32], forms clusters of segments in a star topology
like LOAF and balance tours of MDCs. The second is
FOCUS [31], which opts to reduce the tour lengths of MDCs
and adjusts their speed to minimize the travel distance and
time required for transferring inter-clustering data between
MDCs. Meanwhile, the focus of the third baseline approach,
namely, MINDS [26], is more on reducing and balancing
the MDC tours than reducing data delivery delay between
segments. Like LOAF, the fourth approach is LEEF [27], opts
to achieve MDCs’ energy balance and optimize inter-segment
delivery latency. These baseline approaches address the same
problem tackled by LOAF using different solution strategies,
as summarized below. • Number of MDCs (NMDC): The given number of MDCs
is assumed to be less than the least count of the required
relay nodes to form a perpetual topology of Nseg seg-
ments, denoted as NRN. We use the algorithm in [7]
to compute NRN which primarily depends on Nseg and
the layout of segments. In the simulation, NMDC is
determined based on NRN as explained in subsection C
below. As the value of NMDC gets reduced, MDCs may
consume more energy for visiting more segments. • Average Inter-Segment Data Volume (ISDV): It repre-
sents the data traffic requirements between each pair of
segments. The value of ISDV would have influence on
energy consumption of MDCs for wireless communica-
tion and the inter-MDC energy balancing. The performance of LOAF is assessed using the following
four metrics: • ToCS forms clusters around the center of the area G
and consists of two phases [32]. In the first phase, each
cluster segment initially becomes a cluster Ci, and then
combines two Ci’s whose merging cost based on G is the
least. The merging is repeated until (k −1) clusters are
formed. Then a center-cluster Ck is formed by including
(k −1) rendezvous points Pi, i = 1,..,(k −1), each of
which is computed as a mid-point between G and the
closest point ‘‘x′′ to G that is on the convex hull of seg-
ments of Ci. A. SIMULATION ENVIRONMENT AND PERFORMANCE
METRICS We have implemented a simulation environment in C. The
environment articulates the effect of damage on a single WSN
that originally covers an area or multiple autonomous WSNs. Basically, varying numbers of segments are randomly located
in an area of interest 1200m×1200m such that the segments
are evenly distributed in the quadrants. The following param-
eters are used to vary the network characteristics: segments visited by an MDC. The length of tour path is
also affected by R, specifically, MDCs need to travel a
shorter path with a larger R. • Number of segments (Nseg): Having high segment count
may increase the connectivity requirement and thus
more energy may be needed for MDCs because of
the longer tour. In addition, involving more segments • Number of segments (Nseg): Having high segment count
may increase the connectivity requirement and thus
more energy may be needed for MDCs because of
the longer tour. In addition, involving more segments • Communication range of relays (R): In general, R has
the influence on forming clusters which include a set of 179476 VOLUME 8, 2020 S. Lee et al.: LOAF: Load and Resource Aware Federation of Multiple Sensor Sub-Networks 1) MAXIMUM INTER-SEGMENT DATA DELIVERY
LATENCY (Dmax) • LEEF groups the segments into a set of k clusters in a star
topology where a hub-cluster Ck facilitates data relaying
between clusters. In the first phase, LEEF models an
area of interest as a grid of equal-size square-shaped
cells, based on which a set CST of the fewest cells that
cover all segments is identified by evaluating each cell’s
reachability to segments, and proximity to the center G
of the area. Then LEEF groups the cells in CST into
k VCs considering inter-cell proximity. In the second
phase, LEEF opts to use the VCs to guide the forma-
tion of energy balanced k-clusters in a greedy manner
through a two-step process: greedy-expansion and opti-
mization. During greedy-expansion, LEEF operates in
rounds. Starting with Ck = VCk and Ci = {G, the clos-
est cell to G ∈VCi}, in each round r energy overhead
of Ci, denoted as Er
i , is computed based on the intra-
and inter-cluster data up/download and the MDC tours
between the involved segments in Ci up to round r. Then,
the least energy consumed cluster Cleast is selected for
expansion by adding a segment. In the optimization step,
balancing inter-cluster energy is performed by adjusting
cluster membership to enable Cleast to grow. Figure 7 shows that LOAF experiences less worst-case inter-
segment data transfer latency than all baseline approaches in
terms of R, Nseg, NMDC and ISDV a. Dmax is mainly affected
by rendezvous time along the path between the furthest apart
segments; Therefore, Dmax is affected by a count of clusters
in the path and inter-segment data volume and not much
dependent on R. However, as seen in Figure 7-(a), MINDS,
FOCUS, LEEF and LOAF show less Dmax with larger R. This
is because R has influence on the inter-cluster path length and
traffic volume in their algorithms. In other words, in MINDS
and FOCUS the inter-cluster traffic is delivered via a segment
overlapped by multiple clusters and larger R makes a smaller
number of cluster hops between two furthest apart segments. g
In addition, both LEEF and LOAF ensure that there are
at most two hops between any pair of segments and unlike
ToCS which also forms a star topology, they consider energy
affected by R during clustering. C. SIMULATION RESULTS We have simulated multiple configurations, each has different
combinations of values R, Nseg, NMDC, and ISDV a. The value
of R is varied from 50 to 250 with increment of 25 and Nseg
takes the values between 21 and 39 with increment of 3. In addition, the data volume between segments is randomly
picked using a Gaussian distribution with mean, that takes
values from {1, 2, 4, 8, 16, 32, and 64} with a standard
deviation of 3. The other parameters for simulation setup are
summarized in Table 1. We note that the value for SpeedMDC
is based on nominal speeds of search-and-rescue robots,
e.g., iRobot PackBot 510 [36] whose maximum speed is
9.6 km/h, i.e., about 2.59 meter/s. The results of the individual
experiments are averaged over 30 runs and all results thus
stay within 10% of the sample mean due to 90% confidence
interval analysis. B. BASELINE APPROACHES Spitting is done by finding a center segment
Vc of L from which the path on mst to the furthest
segment is shortest. Then, the segments in L are divided
into two groups ga and gb and Vc is designated as a
rendezvous point between MDCa and MDCb. Forming
ga and gb is based on the node degree of Vc. If there
are more than two edges connected to Vc, then grouping
takes three steps. First ga is formed by the segments
connected through the edge along which the furthest seg-
ment is reached from Vc. Then the remaining segments
are grouped into gb in the same way. If there are still
ungrouped segments, each of them belongs to the closest
group between ga and gb. In the 3rd phase, the tour path
of MDCs is computed as done in [24]. • MINDS strives to balance tour lengths among k MDCs
and consists of three phases. In the 1st phase, it cal-
culates an mst of segments. If only one MDC is avail-
able, the MDC tours the segments along the mst edges. Otherwise, MINDS opts to split the longest tour L into
two groups until the number of groups equals k in the
2nd phase. Spitting is done by finding a center segment
Vc of L from which the path on mst to the furthest
segment is shortest. Then, the segments in L are divided
into two groups ga and gb and Vc is designated as a
rendezvous point between MDCa and MDCb. Forming
ga and gb is based on the node degree of Vc. If there
are more than two edges connected to Vc, then grouping
takes three steps. First ga is formed by the segments
connected through the edge along which the furthest seg-
ment is reached from Vc. Then the remaining segments
are grouped into gb in the same way. If there are still
ungrouped segments, each of them belongs to the closest
group between ga and gb. In the 3rd phase, the tour path
of MDCs is computed as done in [24]. B. BASELINE APPROACHES The weight of
−−−→
Ca, Cb is equal to the increase in the tour length of
Ca when Ca is extended towards Cb by including a
segment from Cb. During the 2nd phase FOCUS opts to
overlap the k clusters in order to minimize tour lengths
of MDCs for the inter-cluster communication; it does
so by selecting an intersection segment (IS) between • Network lifetime (NL): This indicates how long the
formed intermittent topology stays fully operational and
is measured as the duration until the first MDC fully
depletes its on-board energy supply. In other words,
NL represents the length of time during which every
MDC operates and serves their clusters. This metric
indicates the maximum energy consumed by an MDC. We measure NL as (initE−AEC)
BPC
, where initE is the initial
energy of an MDC, BPC is battery power consumption
of an MDC per second. • Buffer space required at gateway segments: This met-
ric is relevant to only the case in which some MDCs
share a segment in their tours and uploads/downloads
inter-cluster data from/at the buffer of the common
segment. The maximum buffer space required within a
certain time window is measured. 179477 179477 179477 VOLUME 8, 2020 S. Lee et al.: LOAF: Load and Resource Aware Federation of Multiple Sensor Sub-Networks two adjacent clusters Ca and Cb. The pair of Ca and
Cb corresponds to two end vertices in each edge of the
computed minimum spanning tree (mst) of G. In the 3rd
phase, FOCUS adjusts the motion speed of MDCs for
reducing the time required at IS for the data communi-
cation between clusters. However, ToCS does not schedule MDCs’ rendezvous time. FOCUS also strives to reduce buffering space and time con-
sumed at the common segment by adjusting the motion speed
of MDCs. However, ToCS, FOCUS, and MINDS do not
consider energy issues like LEEF and LOAF. • MINDS strives to balance tour lengths among k MDCs
and consists of three phases. In the 1st phase, it cal-
culates an mst of segments. If only one MDC is avail-
able, the MDC tours the segments along the mst edges. Otherwise, MINDS opts to split the longest tour L into
two groups until the number of groups equals k in the
2nd phase. 1) MAXIMUM INTER-SEGMENT DATA DELIVERY
LATENCY (Dmax) Nonetheless, LOAF yields
the least Dmax among all approaches, which is attributed to
the following four design features: (i) forming a star-shaped
inter-cluster topology where the number of clusters between
every pair of segments becomes at most three (like ToCS
and LEEF, and unlike FOCUS and MINDS); (ii) MDCs in
LOAF do not need to pause and wait for exchanging data
between clusters during their travel and thus no extra time
for rendezvous is needed in the data delivery (like LEEF,
unlike ToCS); (iii) LOAF tries to keep a pair of segments
which exchange large volumes of data in the same cluster
by considering the inter-segment data exchange overhead
during cluster formation. This reduces the inter-cluster traffic
volume. Overall, the four baseline approaches all try to minimize tour
lengths of MDCs. Like LOAF, LEEF and ToCS form a star
inter-cluster topology while FOCUS and MINDS establish
a cluster-based mst. Also, in FOCUS, MINDS, and LEEF,
the MDCs rendezvous via a common segment for transferring
inter-cluster traffic while in ToCS MDCs meet at a point
where no segment is located. In order to minimize the wait-
ing time for rendezvous ToCS strives to even MDC tours. 179478 179478 179478 VOLUME 8, 2020 S. Lee et al.: LOAF: Load and Resource Aware Federation of Multiple Sensor Sub-Networks TABLE 1. Simulation setup. TABLE 1. Simulation setup. lower latency than all baselines regardless of Nseg. This is
because they strive to include a pair of segments which
exchange data in the same cluster, and hence reduces inter-
cluster trip for data transfer and diminishes inter-segment
data delivery latency. In addition, LOAF outperforms LEEF,
which is again attributed to how LOAF places the central
cluster that plays the role of a hub for the inter-segment com-
munication. It is worth noting that the performance of ToCS
is worse than FOCUS and MIND due to the implicit waiting
time for MDC rendezvous to support data up/downloading. In addition, LOAF yields less Dmax than all baselines
for various numbers of MDCs as seen in Figure 7-(c). Such performance is due to the advantage of forming a star
inter-cluster topology, in which data delivery between any
pair of two segments requires at most two MDC rendezvous. Although both LEEF and ToCS also form a star topology,
ToCS involves waiting time during MDC rendezvous and
LEEF pursues inferior clustering procedure as mentioned
earlier. 1) MAXIMUM INTER-SEGMENT DATA DELIVERY
LATENCY (Dmax) It is worth to note that ToCS and LOAF experience
little changes in Dmax when varying NMDC, while Dmax
slightly decreases as NMDC grows. This is because ToCS,
LOAF and LEEF form a star topology where Dmax is mainly
affected by a tour length in a central cluster. Unlike ToCS
and LOAF which strive to equalize the tour length for all
MDCs, LEEF initially forms a central cluster by placing an
MDC at a geographical center of an area and (NMDC −1) of
virtual non-central clusters which are merged. Thus, a larger
value of NMDC tends to shorten a tour length in a central
cluster of LEEF. Meanwhile, increasing NMDC raises Dmax in
MINDS and FOCUS since they form a mst-based topology
of clusters where the more MDCs are involved, the higher
inter-cluster communication traffic along the path between
two the furthest apart segments becomes and consequently
Dmax increases. As seen in Figure 7-(d), growing the inter-segment data
volume boosts Dmax. This is fairly anticipated because
increasing the data exchange between segments implies that
the MDCs take more time in data up/downloading. However,
it is worth noting that Dmax for LEEF and LOAF is much less
than its value in ToCS, FOCUS and MINDS. For instance,
with ISDV of 64Mb LEEF and LOAF yield 9% of Dmax of
ToCS and 12% of Dmax of FOCUS. In addition, as ISDV
grows, Dmax increases at a much lower rate in LEEF and
LOAF than the other three approaches. These results are
due to the fact that LEEF and LOAF factor data volume
between segments in determining the cluster membership. Meanwhile, ToCS, FOCUS and MINDS do not consider data
traffic volume between segments during clustering. While the latter feature is also supported by LEEF, LOAF
distinguishes itself by forming a more efficient inter-segment
topology, particularly in designating the central cluster. LOAF first locates the central cluster at the energy-balanced
center of mass among all segments in terms of communica-
tion and motion; then non-central clusters are formed consid-
ering segments’ individual position. Meanwhile, LEEF iden-
tifies the central cluster after forming the non-central ones. Accordingly, the inter-cluster topology formed by LOAF
better factors in inter-segment data volume and proximity
and thus yields less inter-segment data exchange latency. It is
worth noting that ToCS which also forms a star inter-cluster
topology has the largest Dmax with all values of R. 1) MAXIMUM INTER-SEGMENT DATA DELIVERY
LATENCY (Dmax) This is
because ToCS does not reduce rendezvous time by careful
scheduling of the MDC tours. VOLUME 8, 2020 2) ENERGY BALANCE AMONG MDCs Moreover, as seen in Figure 7-(b) Dmax grows for all
approaches when the number of segments, Nseg increases. Such a growth is intuitive since adding more segments not
only requires longer trips of MDCs but also involves heavier
inter-segment data traffic, and subsequently more stops and
longer loading time. Nonetheless, LOAF and LEEF yield In Figure 8, the effectiveness of LOAF in terms of energy
balance among MDCs is validated. LOAF yields better per-
formance than all baselines regardless of R, Nseg, NMDC
and ISDV since it: (i) considers energy consumed by
MDCs in up/downloading inter-segment data traffic, and
in touring between clusters, and (ii) strives to equalize the 179479 VOLUME 8, 2020 S. Lee et al.: LOAF: Load and Resource Aware Federation of Multiple Sensor Sub-Networks FIGURE 7. The performance comparison of LOAF to LEEF, ToCS, FOCUS, and MINDS in terms of the inter-segment data delivery latency (Dmax ). (a), (b), (c) and (d) show the results with respective to various R, Nseg, NMDC , and ISDV respectively with a fixed value of R = 100, Nseg = 30,
NMDC = 9, and ISDV = 4. FIGURE 7. The performance comparison of LOAF to LEEF, ToCS, FOCUS, and MINDS in terms of the inter-segment data delivery latency (Dmax ). (a), (b), (c) and (d) show the results with respective to various R, Nseg, NMDC , and ISDV respectively with a fixed value of R = 100, Nseg = 30,
NMDC = 9, and ISDV = 4. Moreover, Figure 8-(c) demonstrates that the star inter-
cluster topology formed by LOAF, LEEF and ToCS has more
positive impact on load balancing as NMDC grows than that
of the mst-based topology formed by FOCUS and MINDS. Formation of clusters around a central cluster provides higher
flexibility in adjusting proximity and traffic volume between
clusters than a mst-based topology; such increased flexibility
becomes even more influential as NMDC increases. Finally,
Figure 8-(d) demonstrates how LOAF efficiently copes with
the increased inter-segment data volume and thus it outper-
forms all baselines in terms of energy consumption balance. This result comes from the fact that LOAF tries to keep a high
volume of data exchanging segments in the same clusters and
the positive impact of the design principle becomes clearer as
the inter-segment data volume, ISDV grows. energy overhead while forming the inter-cluster topology. 2) ENERGY BALANCE AMONG MDCs Moreover, LOAF’s optimized design enables it to outper-
form LEEF which also considers energy balancing dur-
ing cluster formation. In other words, unlike LEEF which
locates the central cluster geographically at the center of
the deployment area, LOAF forms the central cluster to
reduce the energy overhead. In addition, LOAF considers
locations of individual segments and thus optimizes the for-
mation of balanced clusters using more precise computation
than LEEF. As seen in Figure 8-(a), balancing the energy overhead
among MDCs has nothing to do with R for all approaches. Meanwhile, all approaches show more balanced clustering
topology with smaller Nseg as shown in Figure 8-(b). This
phenomenon is because an MDC consumes its power for
up/downloading data traffic volume exchanged between seg-
ments as well as touring the segments. Thus, equalizing
energy consumption among MDCs gets highly complicated
when more segments are involved and subsequently the com-
binations of inter-segment data volume vary. This is espe-
cially evident in FOCUS, MINDS and ToCS, all of which do
not consider energy used for the inter-segment data exchanges
during clustering and thus the inter-MDC energy balancing
rapidly drops with large Nseg. Meanwhile, LOAF outperforms
all baselines for various values of Nseg and its superiority
scales with Nseg. 3) AVERAGE ENERGY CONSUMPTION OF ALL MDCs Figure 9 reports the average energy consumed by all MDCs. LOAF outperforms all baselines for varying R, Nseg, NMDC
and ISDV. Figure 9-(a) shows that ToCS, LEEF and LOAF
require less energy as R increases, unlike FOCUS and
MINDS which pursue a mst-based inter-cluster topology. Enlarging the communication range of an MDC, R, helps
to shorten the tour length in a star inter-cluster topology
and in turn results in reducing the average energy con-
sumption of MDCs. Meanwhile, the tour length is hardly VOLUME 8, 2020 179480 S. Lee et al.: LOAF: Load and Resource Aware Federation of Multiple Sensor Sub-Networks FIGURE 8. (a), (b), (c) and (d) show the performance comparison of LOAF to LEEF, ToCS, FOCUS, and MINDS in terms of energy balance with respect to
various R, Nseg, NMDC , and ISDV respectively with a fixed value of R = 100, Nseg = 30, NMDC = 9, and ISDV = 4. FIGURE 8. (a), (b), (c) and (d) show the performance comparison of LOAF to LEEF, ToCS, FOCUS, and MINDS in terms of energy balance with respect to
various R, Nseg, NMDC , and ISDV respectively with a fixed value of R = 100, Nseg = 30, NMDC = 9, and ISDV = 4. and ISDV on NL, measured as (initE−AEC)
BPC
, where initE and
BPC are set as 54400 Joule and 0.5 Joule/sec respectively. The results shown in Figure 10 are very much expected. Basically, LOAF yields the longest network lifetime in all
cases and the results are consistent with Figure 9 where LOAF
outperforms all baselines in terms of the average energy
consumption. and ISDV on NL, measured as (initE−AEC)
BPC
, where initE and
BPC are set as 54400 Joule and 0.5 Joule/sec respectively. The results shown in Figure 10 are very much expected. Basically, LOAF yields the longest network lifetime in all
cases and the results are consistent with Figure 9 where LOAF
outperforms all baselines in terms of the average energy
consumption. affected under a mst-based inter-cluster topology. Thus,
Figures 9-(a) and 10-(a) evidently show that LOAF groups
segments in an energy-efficient and balanced way. For
instance, LOAF requires about 300% less energy and
achieves approximately 480% more energy-balanced cluster-
ing in comparison to ToCS. g
p
In addition, larger values of Nseg or ISDV
require
more
MDC
energy
for
all
approaches
as
seen
in
Figure 9-(b) and 9-(d), respectively. 5) BUFFER SPACE REQUIRED AT GATEWAY SEGMENTS Figure 11 presents the maximum buffer requirement of ren-
dezvous segments while delivering the inter-cluster data
traffic. Therefore, this performance metric is applicable to
only FOCUS, MINDS, LEEF and LOAF in which an MDC
uploads/downloads the inter-cluster data from/at a buffer of
a segment that is to be visited by another MDC serving a
different cluster. As seen in Figure 11, LOAF requires the
least buffer space regardless of R, Nseg, NMDC and ISDV. This
is because LOAF opts to not only reduce inter-cluster data
traffic by keeping a pair of segments that exchange large data
volume in the same cluster but also place a central cluster at
the energy-balanced location considering the inter-segment
traffic volume. Such design diminishes the buffer require-
ment of segments at rendezvous points. 3) AVERAGE ENERGY CONSUMPTION OF ALL MDCs This is expected because
serving more segments or delivering more inter-segment data
traffic would enlarge a total tour length and/or data loading
time for MDCs. Figure 9-(c) shows another advantage of a
star topology where the energy overhead diminishes dramati-
cally as NMDC grows. We note that for small values of NMDC,
LOAF achieves outstanding reduction in the energy overhead
as compared to ToCS and LEEF. Such superiority is attributed
to the energy-centric design principle of LOAF which strives
to place a central cluster at the energy balanced center of all
segments and based on which the formation of non-central
clusters is optimized. 4) NETWORK LIFETIME (NL) Figure 10 demonstrates the network lifetime during which
the formed topology by each algorithm normally operates. In other words, it shows how long the MDCs tour as planned
after federation and thus every pair of inter-segment commu-
nication is maintained. We study the effect of R, Nseg, NMDC, Figure 11-(a) shows how the buffer space requirement is
affected by the communication range of an MDC, i.e., R. For
FOCUS, MINDS and LOAF little change can be observed
because R does not have a direct influence on a count of 179481 VOLUME 8, 2020 S. Lee et al.: LOAF: Load and Resource Aware Federation of Multiple Sensor Sub-Networks FIGURE 9. The performance comparison of LOAF in terms of the average energy consumption of MDCs with a fixed value of R = 100, Nseg = 30,
NMDC = 9, and ISDV = 4. (a), (b), (c) and (d) show the results with respective to various R, Nseg, NMDC , and ISDV . FIGURE 9. The performance comparison of LOAF in terms of the average energy consumption of MDCs with a fixed value of R = 100, Nseg = 30,
NMDC = 9, and ISDV = 4. (a), (b), (c) and (d) show the results with respective to various R, Nseg, NMDC , and ISDV . FIGURE 10. The performance comparison of LOAF with respect to network lifetime with a fixed value of R = 100, Nseg = 30, NMDC = 9, and ISDV = 4. (a),
(b), (c) and (d) show the results with respective to various R, Nseg, NMDC , and ISDV . FIGURE 10. The performance comparison of LOAF with respect to network lifetime with a fixed value of R = 100, Nseg = 30, NMDC = 9, and ISDV = 4. (a),
(b), (c) and (d) show the results with respective to various R, Nseg, NMDC , and ISDV . FIGURE 11. The performance comparison of LOAF in terms of the required buffer space for MDCs with a fixed value of R = 100, Nseg = 30, NMDC = 9,
and ISDV = 4. (a), (b), (c) and (d) show the results with respective to various R, Nseg, NMDC , and ISDV . FIGURE 11. The performance comparison of LOAF in terms of the required buffer space for MDCs with a fixed value of R = 100, Nseg = 30, NMDC = 9,
and ISDV = 4. 4) NETWORK LIFETIME (NL) (a), (b), (c) and (d) show the results with respective to various R, Nseg, NMDC , and ISDV . segments belonging to a cluster or the inter-cluster data traffic
volume; particularly, the latter affects the maximum buffer
space of segments at rendezvous points. In addition, LOAF
yields better performance than FOCUS, MINDS and LEEF
since it places a central cluster at a mass of inter-segment traf-
fic demands and thus reduces burdens on the shared segment in the central cluster. It is worth noting that LEEF requires
more buffers as R grows. This is because LEEF exploits a cell-
based grid architecture where the size of a cell is determined
by R and identifies the smallest set CST of the selected cells
that covers all segments, based on which a star-shaped cluster
topology is formed. Thus, as R gets larger one cell in CST segments belonging to a cluster or the inter-cluster data traffic
volume; particularly, the latter affects the maximum buffer
space of segments at rendezvous points. In addition, LOAF
yields better performance than FOCUS, MINDS and LEEF
since it places a central cluster at a mass of inter-segment traf-
fic demands and thus reduces burdens on the shared segment VOLUME 8, 2020 179482 S. Lee et al.: LOAF: Load and Resource Aware Federation of Multiple Sensor Sub-Networks FIGURE 12. Comparison of the effect of various parameters R, Nseg, NMDC and ISDV on performance of LOAF in terms of (a) the inter-segment data
delivery latency (Dmax ), (b) the energy balance among mobile data collators, (c) the average energy consumption and (d) buffer size. In each chart,
the left and right bars represent the results of minus and plus 50% variations from a set of the base values for input parameters which is {R = 100,
Nseg = 30, NMDC = 9, ISDV = 4}. FIGURE 12. Comparison of the effect of various parameters R, Nseg, NMDC and ISDV on performance of LOAF in terms of (a) the inter-segment data
delivery latency (Dmax ), (b) the energy balance among mobile data collators, (c) the average energy consumption and (d) buffer size. In each chart,
the left and right bars represent the results of minus and plus 50% variations from a set of the base values for input parameters which is {R = 100,
Nseg = 30, NMDC = 9, ISDV = 4}. FIGURE 13. 4) NETWORK LIFETIME (NL) Comparison of the effect of change of the inter-segment data volume. ISDV is initially set at 8 Mbps and later changes to 2, 16, or 64 Mbps
and the other input parameters are set as R = 100, Nseg = 30, NMDC = 9. Each of (a), (b), (c), and (d) shows LOAF maintains a performance advantage in
term of inter-segment data delivery latency, energy balance among MDCs, energy consumption of MDCs and buffer requirement for MDC respectively. FIGURE 13. Comparison of the effect of change of the inter-segment data volume. ISDV is initially set at 8 Mbps and later changes to 2, 16, or 64 Mbps
and the other input parameters are set as R = 100, Nseg = 30, NMDC = 9. Each of (a), (b), (c), and (d) shows LOAF maintains a performance advantage in
term of inter-segment data delivery latency, energy balance among MDCs, energy consumption of MDCs and buffer requirement for MDC respectively. probably covers more segments which in turns requires more
buffer space for storing the inter-cluster data if the cell is
included in a central cluster. Meanwhile, Figure 12-(d) shows that Nseg and NMDC have
more effect on the maximum buffer size required at gateway
segments than ISDV. This is because such a requirement is
mostly determined by the formed topology, i.e., MDC travel
paths to serve a group of segments, rather than the inter-
segment data volume. Moreover, as much expected and seen in Figure 11-(b)
and 11-(d), larger buffer space is required for all approaches
as Nseg or ISDV rises due to the increased inter-segment
data volume. Figure 11-(c) captures the effect of the number
of MDCs. Growing NMDC implies more clusters, which in
turn introduces more inter-cluster data traffic. Therefore, the
shared segments need more buffer space as NMDC increases. 7) COMPARING THE EFFECT OF VARIATION IN THE DATA
VOLUME Figure 13 shows how LOAF and the baseline approaches
respond to changes in the inter-segment data volume (ISDV)
during operation. In the experiment, ISDV is initially set at
8 Mbps and later changes to 2, 16, or 64 Mbps. The other input
parameters are fixed as R = 100, Nseg = 30, NMDC = 9. With
this setup, the initial energy-balanced cluster is formed in the
first phase of LOAF considering 8 Mbps inter-segment data
volume and maintained during the second phase where data
volume between segments changes to 2, 16, or 64 Mbps. The
same steps are applied to LEEF which also works based on
clustering. As seen in Figure 13, LOAF maintains its supe-
riority and outperforms LEEF as well as FOCUS, MINDS
and ToCS even when the value of ISDV changes in terms of
inter-segment data delivery latency, energy balance between
MDCs and energy consumption of MDCs and required buffer
space for MDCs. 6) COMPARING THE IMPACT OF VARIOUS PARAMETERS ON
LOAF Figure 12 captures the sensitivity of the performance to
configuration parameters, R, Nseg, NMDC and ISDV. In each
chart, the left and right bars represent the results of minus and
plus 50% variations from the base values for input param-
eters, i.e., R = 100, Nseg = 30, NMDC = 9, ISDV = 4. As seen in Figure 12-(a), (b) and (c), an increment in the
ISDV value seems to influence LOAF more than that of R,
Nseg, and NMDC for Dmax, energy balance between MDCs
and the average energy consumption. This is because Dmax
is determined by rendezvous time between MDCs which
is primarily determined by the inter-segment data volume. The energy consumption and balance are also more affected
by the data communication load between segments than the
travel path. Thus, a 50% increase of ISDV from the base-
line setting causes larger performance variations than other
parameters. These results further confirm the importance of
factoring in the data upload/offload overhead in the solutions,
something that sets LOAF apart from competing approaches. REFERENCES Akkaya, ‘‘Mobile data collector assignment and
scheduling for minimizing data delay in partitioned wireless sensor net-
works,’’ in Proc. Int. Conf. Ad Hoc Netw., Oct. 2013, pp. 15–31. [7] F. Senel and M. Younis, ‘‘Relay node placement in structurally damaged
wireless sensor networks via triangular Steiner tree approximation,’’ Com-
put. Commun., vol. 34, pp. 1932–1941, May 2011. [31] B. Kalyanasundaram and M. Younis, ‘‘Using mobile data collec-
tors to federate clusters of disjoint sensor network segments,’’ in
Proc. IEEE Int. Conf. Commun. (ICC), Budapest, Hungary, Jun. 2013,
pp. 1496–1500. [8] S. Lee and M. Younis, ‘‘Optimized relay node placement for connect-
ing disjoint wireless sensor networks,’’ Comput. Netw., vol. 56, no. 12,
pp. 2788–2804, Aug. 2012. [32] J. L. V. M. Stanislaus and M. Younis, ‘‘Mobile relays based federation
of multiple wireless sensor network segments with reduced-latency,’’ in
Proc. IEEE Int. Conf. Commun. (ICC), Budapest, Hungary, Jun. 2013,
pp. 6407–6411. [9] Z. Cheng, M. Perillo, and W. B. Heinzelman, ‘‘General network lifetime
and cost models for evaluating sensor network deployment strategies,’’
IEEE Trans. Mobile Comput., vol. 7, no. 4, pp. 484–497, Apr. 2008. [10] S. Misra, N. E. Majd, and H. Huang, ‘‘Approximation algorithms for
constrained relay node placement in energy harvesting wireless sensor net-
works,’’ IEEE Trans. Comput., vol. 63, no. 12, pp. 2933–2947, Dec. 2014. [33] D. K. Goldenberg, J. Lin, and A. S. Morse, ‘‘Towards mobility as a network
control primitive,’’ in Proc. 5th ACM Int. Symp. Mobile Ad Hoc Netw. Comput. (MobiHoc), 2004, pp. 163–174. [11] F. Senel, M. F. Younis, and K. Akkaya, ‘‘Bio-inspired relay node placement
heuristics for repairing damaged wireless sensor networks,’’ IEEE Trans. Veh. Technol., vol. 60, no. 4, pp. 1835–1848, May 2011. [34] V. Rodoplu and T. H. Meng, ‘‘Minimum energy mobile wireless net-
works,’’ IEEE J. Sel. Areas Commun., vol. 17, no. 8, pp. 1333–1344, 1999. [35] A. Kashyap, S. Khuller, and M. Shayman, ‘‘Relay placement for higher
order connectivity in wireless sensor networks,’’ in Proc. IEEE 25TH IEEE
Int. Conf. Comput. Commun. (INFOCOM), Barcelona, Spain, Apr. 2006,
pp. 1–12. [12] S. Lee, M. Younis, and M. Lee, ‘‘Connectivity restoration in a partitioned
wireless sensor network with assured fault tolerance,’’ Ad Hoc Netw.,
vol. 24, pp. 1–19, Jan. 2015. [13] S. K. Gupta, P. Kuila, and P. K. Jana, ‘‘Genetic algorithm for k-connected
relay node placement in wireless sensor networks,’’ in Proc. 2nd Int. Conf. Comput. Commun. Technol. REFERENCES [1] I. F. Akyildiz, W. Su, Y. Sankarasubramaniam, and E. Cayirci, ‘‘Wireless
sensor networks: A survey,’’ Comput. Netw., vol. 38, no. 4, pp. 393–422,
2002. [24] F. Senel and M. Younis, ‘‘Optimized interconnection of disjoint wireless
sensor network segments using K mobile data collectors,’’ in Proc. IEEE
Int. Conf. Commun. (ICC), Ottawa, ON, Canada, Jun. 2012, pp. 492–496. [2] M. Younis, I. F. Senturk, K. Akkaya, S. Lee, and F. Senel, ‘‘Topology
management techniques for tolerating node failures in wireless sensor
networks: A survey,’’ Comput. Netw., vol. 58, pp. 254–283, Jan. 2014. [25] J. L. V. M. Stanislaus and M. Younis, ‘‘Delay-conscious federation of
multiple wireless sensor network segments using mobile relays,’’ in Proc. IEEE Veh. Technol. Conf. (VTC Fall), Québec City, QC, Canada, Sep. 2012,
pp. 1–5. [3] M. Younis, S. Lee, I. F. Senturk, and K. Akkaya, ‘‘Topology man-
agement techniques for tolerating node failure,’’ in The Art of Wire-
less Sensor Networks, H. Ammari, Ed. New York, NY, USA: Springer,
Oct. 2013. [Online]. Available: http://www.springer.com/engineering/
signals/book/978-3-642-40008-7 [26] Y. K. Joshi and M. Younis, ‘‘Mobility-based Internetworking of disjoint
segments,’’ in Proc. 27th Biennial Symp. Commun. (QBSC), Kingston, ON,
Canada, Jun. 2014, pp. 193–197. [27] S. Lee, M. Younis, B. Anglin, and M. Lee, ‘‘LEEF: Latency and energy
efficient federation of disjoint wireless sensor segments,’’ Ad Hoc Netw.,
vol. 71, pp. 88–103, Mar. 2018. [4] G.-H. Lin and G. Xue, ‘‘Steiner tree problem with minimum number of
Steiner points and bounded edge-length,’’ Inf. Process. Lett., vol. 69, no. 2,
pp. 53–57, Jan. 1999. [28] A. Abbas and M. Younis, ‘‘Establishing connectivity among disjoint ter-
minals using a mix of stationary and mobile relays,’’ Comput. Commun.,
vol. 36, no. 13, pp. 1411–1421, Jul. 2013. [5] F. M. Al-Turjman, H. S. Hassanein, and M. Ibnkahla, ‘‘Optimized relay
placement to federate wireless sensor networks in environmental applica-
tions,’’ in Proc. 7th Int. Wireless Commun. Mobile Comput. Conf., Istanbul,
Turkey, Jul. 2011, pp. 2040–2045. [29] Y. K. Joshi and M. Younis, ‘‘Restoring connectivity in a resource con-
strained WSN,’’ J. Netw. Comput. Appl., vol. 66, pp. 151–165, May 2016. [6] S. Lee and M. Younis, ‘‘Recovery from multiple simultaneous failures in
wireless sensor networks using minimum Steiner tree,’’ J. Parallel Distrib. Comput., vol. 70, no. 5, pp. 525–536, May 2010. [30] I. F. Senturk and K. VI. CONCLUSION In this paper, we have studied the federation problem of
disjoint WSN segments using a limited number (k) of mobile
relays that provide intermittent inter-segment connectivity. We have proposed LOAF, a novel approach that groups the 179483 VOLUME 8, 2020 S. Lee et al.: LOAF: Load and Resource Aware Federation of Multiple Sensor Sub-Networks segments into k clusters and defines energy-efficient and also
balanced tour paths for the mobile relays by factoring in
both inter-segment data volume and proximity to the formed
clusters. LOAF finally forms a star topology where a cen-
tral cluster is located at the center of the area considering
inter-segment data volumes as well as geographical distance
between segments. The simulation results demonstrate that
LOAF outperforms competing approaches in terms of the
maximum inter-segment data delivery latency, energy bal-
ance among MDCs, average energy consumption of MDCs
and network lifetime with a competitively small amount of
buffer space requirement. The effectiveness of LOAF grows
with larger MDC communication ranges, and with increased
segment and MDC counts. [17] S. Lee and M. Younis, ‘‘EQAR: Effective QoS-aware relay node placement
algorithm for connecting disjoint wireless sensor subnetworks,’’ IEEE
Trans. Comput., vol. 60, no. 12, pp. 1772–1787, Dec. 2011. [18] A. Nigam and Y. K. Agarwal, ‘‘Optimal relay node placement in delay
constrained wireless sensor network design,’’ Eur. J. Oper. Res., vol. 233,
no. 1, pp. 220–233, Feb. 2014. [19] H. M. Almasaeid and A. E. Kamal, ‘‘Data delivery in fragmented wireless
sensor networks using mobile agents,’’ in Proc. 10th ACM Symp. Modeling,
Anal., Simulation Wireless Mobile Syst. (MSWiM), 2007, p. 96. [20] H. M. Almasaeid and A. E. Kamal, ‘‘Modeling mobility-assisted data
collection in wireless sensor networks,’’ in Proc. IEEE GLOBECOM-IEEE
Global Telecommun. Conf., Dec. 2008, pp. 1–5. [21] K. Li and K. A. Hua, ‘‘Mobile data collection networks for wireless
sensors,’’ in Proc. 5th Int. Conf. Multimedia Commun., Services Secur. (MCSS), Krakow, Poland, May 2012, pp. 200–211. [22] W. Wang, V. Srinivasan, and K.-C. Chua, ‘‘Extending the lifetime of
wireless sensor networks through mobile relays,’’ IEEE/ACM Trans. Netw.,
vol. 16, no. 5, pp. 1108–1120, Oct. 2008. [23] M. Ma, Y. Yang, and M. Zhao, ‘‘Tour planning for mobile data-gathering
mechanisms in wireless sensor networks,’’ IEEE Trans. Veh. Technol.,
vol. 62, no. 4, pp. 1472–1483, May 2013. REFERENCES New Delhi, India: Springer, 2016,
pp. 721–729. [36] Robotyka,
Roboty,
Mechatronika
I
Automatyka. Accessed:
Dec. 23, 2019. [Online]. Available: http://www.asimo.pl/materialy/
download/iRobot_510_PackBot_for_HazMat_Technicians.pdf [37] Ah Rechargeable Lithium-Ion Battery. Accessed: Dec. 23, 2019. [Online]. Available: https://www.bren-tronics.com/bt-70791ck.html [14] S. Lee, M. Younis, and M. Lee, ‘‘Optimized bi-connected federation of
multiple sensor network segments,’’ Ad Hoc Netw., vol. 38, pp. 1–18,
Mar. 2016. [38] B.-S. Kim, K.-I. Kim, B. Shah, F. Chow, and K. Kim, ‘‘Wireless sen-
sor networks for big data systems,’’ Sensors, vol. 19, no. 7, p. 1565,
Apr. 2019. [15] S. Misra, S. Dong, G. Xue, and J. Tang, ‘‘Constrained relay node place-
ment in wireless sensor networks: Formulation and approximations,’’
IEEE/ACM Trans. Netw., vol. 18, no. 2, pp. 434–447, Apr. 2010. [39] M. Kuorilehto, M. Hännikäinen, and T. D. Hämäläinen, ‘‘A survey of
application distribution in wireless sensor networks,’’ EURASIP J. Wireless
Commun. Netw., vol. 2005, no. 5, Dec. 2005, Art. no. 859712. [16] D. Yang, S. Misra, X. Fang, G. Xue, and J. Zhang, ‘‘Two-tiered constrained
relay node placement in wireless sensor networks: Efficient approxima-
tions,’’ in Proc. 7th Annu. IEEE Commun. Soc. Conf. Sensor, Mesh Ad Hoc
Commun. Netw. (SECON), Boston, MA, USA, Jun. 2010, pp. 1–9. [40] W. Beiglböck, ‘‘The center-of-mass in Einsteins theory of gravitation,’’
Commun. Math. Phys., vol. 5, no. 2, pp. 106–130, Apr. 1967. 179484 VOLUME 8, 2020 S. Lee et al.: LOAF: Load and Resource Aware Federation of Multiple Sensor Sub-Networks MOHAMMED ALSOLAMI received the B.S. degree in education and computer science from
King Abdulaziz University, Jeddah, Saudi Arabia,
in 2005, the M.A. degree in career and techni-
cal education from Western Michigan University,
USA, in 2012, and the M.S. degree in computer
science from Western Michigan University, USA,
in 2013. He is currently pursuing the Ph.D. degree
in computer science with the University of Mary-
land, Baltimore County, Baltimore, MD, USA. Since 2005, he joined the Royal Commission, Yanbu Industrial, Saudi
Arabia, where he is also a Faculty Member and a Research Member. His
awards and honors include the SACM Fellowship, Monshaat’s Entrepreneur-
ship, AI Hackathon Award, Saudi Research Award, ISO 9001 Award, and
TAMAKON Award. He has been involved in research in the field of computer
science, and STEM education and evaluation. His current research interests
include robotics and artificial intelligent of networks, network architectures,
protocols, and future Internet of everything. [41] L. Costa, O. Filipe, and J. REFERENCES Natário, ‘‘Center of mass, spin supplemen-
tary conditions, and the momentum of spinning particles,’’ in Equations
of Motion in Relativistic Gravity. Cham, Switzerland: Springer, 2015,
pp. 215–258. SOOKYOUNG LEE received the B.S. and M.S. degrees in computer science from Ewha Wom-
ans University, South Korea, in 1995 and 1997,
respectively, and the Ph.D. degree in computer sci-
ence from the University of Maryland, Baltimore
County, USA, in 2010. She has been with LG
Electronics Inc., the Electronics and Telecommu-
nications Research Institute, the Korea Electron-
ics Technology Institute, and Samsung Electronics
Company Ltd., South Korea, from 1998 to 2004. While at LG, she has developed the IP data server over ATM switch and
implemented virtual private network service for multiprotocol label switch-
ing system. She was a volunteer of IPv6 forum Korea while at ETRI and
has developed the network address and protocol translation system between
IPv4 and IPv6. At Samsung, she was a broadband convergence network
designer especially focusing on requirements for QoS and IPv6. She is
currently a Research Professor with the Department of Computer Science
and Engineering, Ewha Womans University. Her research interests include
network architectures and protocols, topology restoration and fault tolerance
in wireless sensor networks, and network modeling and performance analysis
for dynamic and sparse ad-hoc networks. MOHAMED YOUNIS (Senior Member, IEEE)
received the Ph.D. degree in computer science
from the New Jersey Institute of Technology, USA. He is currently an Associate Professor with the
Department of Computer Science and Electrical
Engineering, University of Maryland, Baltimore
County (UMBC). Before joining UMBC, he was
with the Advanced Systems Technology Group,
an Aerospace Electronic Systems Research and
Development Organization of Honeywell Inter-
national Inc. While at Honeywell, he led multiple projects for building
integrated fault tolerant avionics and dependable computing infrastructure. He also participated in the development of the Redundancy Management
System, which is a key component of the Vehicle and Mission Computer
for NASA’s X-33 space launch vehicle. He has published over 260 technical
articles in refereed conferences and journals. He has seven granted and
three pending patents. In addition, he serves/served on the editorial board of
multiple journals and the organizing and technical program committees of
numerous conferences. His technical interests include network architectures
and protocols, wireless sensor networks, embedded systems, fault tolerant
computing, secure communication, and distributed real-time systems. He is
also a Senior Member of the IEEE Communications Society. REFERENCES MEEJEONG
LEE
(Member, IEEE) received
the
B.S. degree
in
computer
science
from
Ewha Womans University, Seoul, South Korea,
in 1987, the M.S. degree in computer sci-
ence from the University of North Carolina,
Chapel Hill, in 1989, and the Ph.D. degree
in
computer
engineering
from
North
Carolina State University, Raleigh, in 1994. In 1994, she joined the Department of Computer
Science and Engineering, Ewha Womans Univer-
ity, where she is currently a Professor. She has been involved in research
n the field of computer communications and networks and performance
modeling and evaluation. Her current research interests include protocols
and architectures of mobile and wireless networks and future Internet. MEEJEONG
LEE
(Member, IEEE) received
the
B.S. degree
in
computer
science
from
Ewha Womans University, Seoul, South Korea,
in 1987, the M.S. degree in computer sci-
ence from the University of North Carolina,
Chapel Hill, in 1989, and the Ph.D. degree
in
computer
engineering
from
North
Carolina State University, Raleigh, in 1994. In 1994, she joined the Department of Computer
Science and Engineering, Ewha Womans Univer-
sity, where she is currently a Professor. She has been involved in research
in the field of computer communications and networks and performance
modeling and evaluation. Her current research interests include protocols
and architectures of mobile and wireless networks and future Internet. 179485 VOLUME 8, 2020
|
https://openalex.org/W2068515030
|
https://bmcpsychiatry.biomedcentral.com/counter/pdf/10.1186/1471-244X-13-163
|
English
| null |
Prevalence and sociodemographic associations of common mental disorders in a nationally representative sample of the general population of Greece
|
BMC psychiatry
| 2,013
|
cc-by
| 10,807
|
RESEARCH ARTICLE Open Access Prevalence and sociodemographic associations of
common mental disorders in a nationally
representative sample of the general population
of Greece Petros Skapinakis*, Stefanos Bellos, Sotirios Koupidis, Ilias Grammatikopoulos, Pavlos N Theodorakis
and Venetsanos Mavreas © 2013 Skapinakis et al.; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the
Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use,
distribution, and reproduction in any medium, provided the original work is properly cited. Skapinakis et al. BMC Psychiatry 2013, 13:163
http://www.biomedcentral.com/1471-244X/13/163 Skapinakis et al. BMC Psychiatry 2013, 13:163
http://www.biomedcentral.com/1471-244X/13/163 Skapinakis et al. BMC Psychiatry 2013, 13:163 Abstract Background: No study in Greece has assessed so far the full range of common mental disorders using a
representative sample of the population from both mainland and insular regions of the country. The aim of the
present paper was to present the results of the first such study. Methods: The study was carried out between 2009–2010 in a nationally representative sample of 4894 individuals
living in private households in Greece. Common mental disorders in the past week were assessed with the revised
Clinical Interview Schedule (CIS-R). We also assessed alcohol use disorders (using AUDIT), smoking and cannabis use Results: 14% of the population (Male: 11%, Female: 17%) was found to have clinically significant psychiatric
morbidity according to the scores on the CIS-R. The prevalence (past seven days) of specific common mental
disorders was as follows: Generalized Anxiety Disorder: 4.10% (95% CI: 3.54, 4.65); Depression: 2.90% (2.43, 3.37);
Panic Disorder: 1.88% (1.50, 2.26); Obsessive-Compulsive Disorder: 1.69% (1.33, 2.05); All Phobias: 2.79% (2.33, 3.26);
Mixed anxiety-depression: 2.67% (2.22, 3.12). Harmful alcohol use was reported by 12.69% of the population (11.75,
13.62). Regular smoking was reported by 39.60% of the population (38.22, 40.97) while cannabis use (at least once
during the past month) by 2.06% (1.66, 2.46). Clinically significant psychiatric morbidity was positively associated
with the following variables: female gender divorced or widowed family status low educational status and g
g
,
y
,
unemployment. Use of all substances was more common in men compared to women. Common mental disorders
were often comorbid, undertreated, and associated with a lower quality of life. Conclusions: The findings of the present study can help in the better planning and development of mental health
services in Greece, especially in a time of mental health budget restrictions. * Correspondence: p.skapinakis@gmail.com
Department of Psychiatry, University of Ioannina, School of Medicine,
Ioannina 45110, Greece Sampling procedure use. Surveys of the general population are the first step in
this process and are valuable in providing unbiased and
representative information on the prevalence and asso-
ciations of mental disorders in a defined geographical area
[4]. Several countries in Europe and elsewhere in the
world
have
already
conducted
large
epidemiological
surveys for these reasons [5,6]. Sampling methodology was designed and implemented
by a research agency in Greece with substantial expe-
rience in conducting nationwide surveys of social or
political issues using representative samples of the
general adult population of Greece. According to the
latest Population Census (2001) the survey population
consisted of approximately 7,200,000 individuals. A
three-stage sampling design was used with enumerator
areas (one or more unified city blocks) based on the
2001 census survey selected at the first stage, households
within the selected areas at the second stage and indivi-
duals within the households at the third stage. In Greece, no study so far has assessed the full range
of common mental disorders using a representative
sample of the population from both mainland and insular
regions of the country. Two previous nationwide studies
were conducted in 1978 and 1984 [7] but these focused
mainly in the assessment of general psychiatric morbidity,
they used an instrument that predated modern diagnostic
criteria (Langner’s scale) and, regarding specific diagnoses,
they only covered the diagnosis of depression. In addition,
only households from mainland Greece were sampled. More recently, a series of telephone surveys were con-
ducted, but these also focused on depression only [8,9]. Therefore, there is a need for a large scale, population-
based, survey of psychiatric disorders in the whole
country, using a detailed diagnostic interview to assess the
prevalence and associations of common mental disorders
and harmful substance use. The primary sampling units (enumerator areas) were
first
stratified
by
allocating
the
Municipalities
and
Communes included in each Region according to the
degree of urbanization (stratum 1: urban areas; stratum
2: semi-urban; stratum 3: rural areas). The stratification
for the two major cities, Athens and Thessaloniki, were
different (Athens was divided into 31 strata of equal size
and Thessaloniki into 9 strata). The projected sample
size for the whole survey was 9,800 individuals with a
sampling fraction 1/λ for each stratum considered
constant and equal to 0.085%. Sampling procedure At the first stage of the sampling procedure primary
sampling units (enumerator areas) had a probability of
being selected proportional to their size (number of
households according to the 2001 census). At the second
stage from each selected area (primary sampling unit)
the sample of secondary units (households) was selected. Actually, in the second stage a random systematic
sample of households was drawn. Systematic sampling is
functionally similar to random sampling because each
element (household) had a known and equal probability
to be selected. Systematic sampling starts by selecting a
random starting point (using maps of the enumerator
areas) and then every kth element in the sampling frame
is selected, where k is the sampling interval. In any
selected primary unit, the sample size was determined
from the sampling interval which was calculated using
data from the 1st stage. At the third stage one eligible
member (aged 18–70 years) of the household was
selected using simple random sampling. The present study was designed to fulfill this aim. The current paper reports the prevalence, comorbidity,
use of services and basic sociodemographic associations
of the common mental disorders in Greece. We also
report basic data on smoking, harmful alcohol use and
cannabis use. A companion paper will focus more
specifically in the association between socioeconomic
status, subjective financial difficulties, unemployment
and common mental disorders (Skapinakis et al., in
preparation). Background the general population [1,2]. According to projections in
2030 depression will be the leading cause of disability
globally accounting for approximately 6% of the total
disability [3]. Regarding anxiety disorders, two of the most
disabling disorders, panic disorder and obsessive compul-
sive disorder are associated with almost the same amount
of disability as schizophrenia. Alcohol use disorders are
also in the list of the top twenty leading causes of disability
worldwide [1,3]. During the past 30 years several epidemiological surveys
around the world have shown that mental disorders are
a major source of disability accounting for approximately
12% of the global burden of disease [1]. Among these
disorders, depression and anxiety disorders (often referred
to as “common mental disorders”) are perhaps the most
important from a public health perspective due to their
prevalence and persistence even in unselected samples of Given these figures it is important that national health
systems are better organized for the optimal management
of the common mental disorders or harmful substance * Correspondence: p.skapinakis@gmail.com
Department of Psychiatry, University of Ioannina, School of Medicine,
Ioannina 45110, Greece Skapinakis et al. BMC Psychiatry 2013, 13:163
http://www.biomedcentral.com/1471-244X/13/163 Skapinakis et al. BMC Psychiatry 2013, 13:163
http://www.biomedcentral.com/1471-244X/13/163 Page 2 of 14 Page 2 of 14 General description of the study
Th
i
l
d The cross-sectional study reported here is part of the
2009–2010 Psychiatric Morbidity Survey carried out in
Greece using a nationally representative sample of the
adult population (18 – 70 years). The study was orga-
nized by the Ministry of Health and carried out by the
University of Ioannina. Data collection was conducted
between September 2009 and February 2010. Regarding
provision of healthcare in Greece, there are 7 “Regional
Health
Authorities”
(RHA)
covering
all
geographic
regions of the country. Eligible for participation were all
adults living in households in each of the seven RHAs. Due to the high costs incurred by sampling all islands of
the Aegean, Crete was excluded from the sampling
(as Crete has been covered in the past by other smaller
scale surveys), therefore the sample included participants
from all areas of Greece excluding Crete. Measurement of psychiatric morbidity
h
b d
d Psychiatric morbidity was assessed with the revised
clinical interview schedule (CIS-R), a fully structured
psychiatric interview designed to be used by trained lay
interviewers [10]. The CIS-R was the main instrument
used in the national psychiatric morbidity surveys in the
UK [11] and has been used in several other similar sur-
veys around the world [12-15]. A computerized version
has also been developed and found to be comparable
with the regular interview [16]. The CIS-R assesses the
presence and severity of 14 different common psychi-
atric symptoms during the past 7 days (psychosomatic
symptoms,
fatigue,
concentration/memory
problems,
sleep problems, irritability, worry about physical health,
depressive mood, depressive ideas, general worry, free-
floating anxiety, phobias, panic, compulsions and obses-
sions). Two screening questions in each section ask about
the presence of the symptom during the past month and
then there is a more detailed assessment of the presence,
frequency, duration, and severity of the symptom during
the past seven days. Additional questions, including ques-
tions assessing the impairment of functioning, enable the
diagnosis of six common mental disorders (depressive
episode, generalized anxiety disorder, all phobias com-
bined, panic disorder, obsessive compulsive disorder,
mixed anxiety and depression disorder) according to the
ICD-10 research diagnostic criteria using specially deve-
loped computerized algorithms. g
Current smoking status and current (past-month)
cannabis use was obtained from the participants by
direct questioning. Regarding smoking, participants were
asked to report the average number of cigarettes they
smoked per day during the past month. A second
question asked the participants to classify themselves
into one of the following categories: never-smoker,
ex-smoker, occasional/light smoker, moderate smoker,
heavy smoker. We combined those two questions to
define a binary variable of “regular smoker” in the past
30 days (all those who were at least moderate smokers
OR reported more than 2 cigarettes per day on average
during the past month). Regarding cannabis, we asked
two questions, the first for lifetime use (five categories:
never, 1–2 times, 3–10 times, >10 times/regular use, do
not wish to reply) and the second for past 30 days use
(“have you used cannabis during the past 30 days?” with
three possible answers: Yes, No, do not wish to reply). We classified participants as users of cannabis during
the past 30 days if they replied yes to the second
question OR reported regular use to the first. Measurement of psychiatric morbidity
h
b d
d The Greek version of the CIS-R has been validated
and its psychometric properties have been published
elsewhere [17]. Each symptom section is scored from 0
to 4 (except depressive ideas from 0 to 5) and a score of
2 or more denotes a clinically significant symptom [10]. Using the CIS-R psychiatric morbidity can be assessed
either in a dimensional way, using the total score on the
CIS-R (by adding-up all 14 symptom dimensions), or in
a categorical form using the six diagnostic categories. For the purposes of the present study we have selected to
use both in our analyses in order to be able to investigate
potential differences between general psychopathology
and specific diagnostic categories. For the dimensional Data collection and response rate p
In each RHA a pool of 20 trained researchers and 2
supervisors were employed. All instruments used were
computerized and responses to the questions/interview
were entered directly to a laptop computer. The average
time for completion of the instrument was from 30 to
45 minutes depending on the psychopathology. Appro-
ximately 35% of the participants entered their data into
the laptop without any further assistance from the Skapinakis et al. BMC Psychiatry 2013, 13:163
http://www.biomedcentral.com/1471-244X/13/163 Page 3 of 14 interviewer after the first guidance. The remaining 65%
required some help from the interviewer. assessment, we have defined four groups of severity based
on previous work with the UK and Greek samples
[10,11,17]: “no/minimal symptoms” (CIS-R score = 0-5),
“subthreshold symptoms” (CIS-R
score = 6-11), “mild
symptoms” (CIS-R score = 12-17) and “severe symptoms”
(CIS-R score > =18). A score on the CIS-R ≥12 (by com-
bining the last two groups into one) is usually considered
as the cut-off for “clinically significant” psychiatric mor-
bidity [10,11,17]. Overall response rate was 54% with a range between
51% and 60% between regions. Refusals were more
common from women and the middle aged participants
(40–55). Differences between the sample and the 2001
census population data were small. A full detail of the
study design, sampling procedures, sampling distribution
within each regional health authority and data collection
are available from the technical reports submitted to the
Ministry of Health and are available by the authors on
request. Assessment of substance use Alcohol- related disorders were assessed with AUDIT
[18]. In the present study we used the first three ques-
tions in AUDIT (consumption, frequency, binge drink-
ing) to calculate the AUDIT-C subscale with a range of
scores from 0 to 12. The AUDIT-C is considered a
reliable alcohol screen for use in general population
surveys to identify people with hazardous drinking or
active alcohol abuse and dependence [19]. To define
harmful alcohol use we used the cut points suggested by
Aalto et al. [20] who used data from a general popula-
tion survey. These were a score of ≥6 for men and ≥4
for women. The cut points are different for the two
genders as this has been supported by recent research
findings on AUDIT [21]. Assessment of health status We assessed current health status with the EuroQoL
EQ-5D, a generic, preference-based, measure of health-
related quality of life [22]. This has been validated in
Greece by Kontodimopoulos et al. [23]. For the purposes
of the present paper we calculated the EQ-5D index
scores based on responses to the 5-item questionnaire. Page 4 of 14 Page 4 of 14 Page 4 of 14 Skapinakis et al. BMC Psychiatry 2013, 13:163
http://www.biomedcentral.com/1471-244X/13/163 Skapinakis et al. BMC Psychiatry 2013, 13:163
http://www.biomedcentral.com/1471-244X/13/163 Assessment of socio-demographic and other variables Assessment of socio-demographic and other variables
Information about sex, age, marital status, employment
status and educational qualifications were obtained from
the participants. Regarding employment status, we distin-
guished between unemployment (i.e. the participant did
not do any paid work but looked for any kind of paid work
in the past 4 weeks) and economic inactivity (the partici-
pant did not do any paid work but did not look for any
paid work in the past 4 weeks; additional questions
clarified the reason for not seeking any work: a) looking
after the house, b) retired, and c) a residual category of
“other economically inactive” (including students, persons
doing their mandatory military service, those living with
parents, those unable to work, living on other income
such as rents or shares and other non-specific reasons). Participants were also presented with a list of chronic and
severe medical conditions (cardiovascular diseases in-
cluding coronary heart disease and stroke, chronic lung
diseases, diabetes, any malignancy, chronic kidney disease)
and asked to report whether they suffered from them. Prevalence of common mental disorders The prevalence of common mental disorders by gender
is shown in Table 1 (for a graphic representation by
gender and age please see the Additional file 1: Figures
S1 and S2 respectively). It can be seen that the most
common disorder was generalized anxiety disorder (GAD)
with a prevalence of 4.10% (95% CI: 3.54 – 4.65) followed
by depressive episode at 2.90% (95% CI: 2.43 – 3.37). All psychiatric diagnoses were significantly more common
in women except for the non-specific “mixed anxiety-
depression” which was marginally non-significant (p =
0.08 for the latter; p < 0.001 for the remaining disorders
except p = 0.03 for OCD). Criteria for at least one depres-
sive or anxiety disorder were met by 7.67% of the po-
pulation (95% CI: 6.92 – 8.41). Prevalence of depression
showed an increase in the age groups 50–59 and 60–70,
while that of anxiety disorders in the age group 60–70. Further analyses showed that this association was con-
founded by the presence of severe chronic medical Description of the sample Overall 4,894 adults were interviewed (overall response
rate 54.2%, see methods for details). Women represented
50.4% of the final sample. Mean age was 42 years with a
standard deviation of 15. The majority of the par-
ticipants were married (61.2%), employed (59.6%), had
graduated from senior high school (48.0%) and were
living in an urban environment (54.8%). A detailed table
of the distribution of the various sociodemographic
variables in the whole sample is given in the appendix
(Additional file 1: Table S1). Results The scoring algorithm for the EQ-5D index descriptive
system used in this paper is based on UK community
preferences as analogous preferences are lacking in
Greece [23]. The mean (SD) value of the EQ index in
the present sample was 0.82 (0.23) and was very similar
to the value of 0.80 (0.27) reported by Kontodimopoulos
et al. [23] in their validation study. Assessment of course and mental health service use Persistence of illness was assessed by asking the partici-
pants to retrospectively assess the duration of their
symptoms. Use of mental health services was assessed
by asking participants whether they had visited a mental
health professional (either a psychiatrist or psychologist)
during the past 12 months for any reason concerning
their mental health. Prevalence of psychiatric symptoms
h
h p y
y
p
The CIS-R estimates the presence of 14 common psy-
chiatric symptoms during the past week. Figure 1
presents the prevalence of symptoms by gender using a
score of 2 or more as the cut-off for clinical significance
(see methods). Fatigue, irritability and worry were the
most common symptoms with a prevalence in the whole
sample of >15%, while obsessions, phobias and panic
were the least common with a prevalence of <5%. All
symptoms were significantly more common in women
(p < 0.001) except irritability (p = 0.26). Figure 2 shows the
prevalence of general psychopathology by gender using
the total CIS-R score to denote mild (score ≥12 & ≤17)
or severe (score ≥18) psychopathology. In total, ap-
proximately 14.06% (95% CI: 13.08 – 15.03) of the
population had either mild or severe psychopathology
with a significant gender effect (p < 0.001, Figure 2). A
detailed table of the distribution of the 14 symptoms
assessed by the CIS-R in men, women and the total
sample is given in the appendix (Additional file 1:
Table S2). Statistical analysis
h Data were weighted to account for the complex sam-
pling design and non-response. We used the survey
commands in Stata version 10.0 to calculate prevalence
estimates and 95% confidence intervals [24]. These
commands take into account the complex sampling
design and compute robust standard errors. Associations
between the common mental disorders and socio-
demographic associations were examined using odds
ratios. These and their 95% confidence intervals were
calculated with a series of adjusted logistic regression
models using the survey commands in Stata 10.0. All
evaluations of statistical significance are based on two-
sided tests using the 5% level of significance. Skapinakis et al. BMC Psychiatry 2013, 13:163
http://www.biomedcentral.com/1471-244X/13/163 Skapinakis et al. BMC Psychiatry 2013, 13:163 Page 5 of 14 y
y
http://www.biomedcentral.com/1471-244X/13/163 Figure 1 Prevalence of psychiatric symptoms, N = 4894. Figure 1 Prevalence of psychiatric symptoms, N = 4894. conditions (see below on the sociodemographic asso-
ciations section and in the Additional file 1: Figures S2,
S2a and S2b). 1.66 – 2.46) and it was also more common in men. Harmful alcohol use and smoking were less common in
the age group 60–70 (p < 0.001 compared to 18–29), while
cannabis use was higher in the age group 18–29 (p < 0.001
compared to all other age groups). Prevalence of substance use disorders Harmful alcohol use, defined as a score of ≥6 for men
and ≥4 for women in audit-c (see methods), was reported
by 12.69% of the population (95% CI: 11.75 – 13.62) and
was more common in men compared to women (16.95%
vs. 8.51% respectively, p < 0.001) [Table 1]. Approximately
40% of the population reported regular smoking during
the past month with a significant gender effect (50% for
men vs. 30% for women, p < 0.001). The prevalence of
Cannabis use during the past month was 2.06% (95% CI: Comorbidity, health status and health services use
Table 2 (left columns) shows comorbidity rates for five
mental disorders. Panic disorder was rare in “pure” (i.e. non-comorbid) form, while depression and GAD were
more common in pure forms (41% and 47% respectively). OCD and phobic disorders were in the middle with
approximately one fifth of the cases being “pure”. Figure 3
shows a graphical representation of comorbidity rates. Figure 2 Prevalence of general psychopathology, N = 4894. Figure 2 Prevalence of general psychopathology, N = 4894. Figure 2 Prevalence of general psychopathology, N = 4894. Skapinakis et al. BMC Psychiatry 2013, 13:163
http://www.biomedcentral.com/1471-244X/13/163 Page 6 of 14 Table 1 Prevalence (1-week) of common mental disorders and substance use by gender in a representative sample of
the general population of Greece 18–70, (weighted data, N = 4894) Table 1 Prevalence (1-week) of common mental disorders and substance use by gender in a representative sample of
the general population of Greece 18–70, (weighted data, N = 4894)
Male
Female
p-value
Total
% (95% CI1)
% (95% CI1)
(male vs. Prevalence of substance use disorders female)
% (95% CI1)
Depressive Episode
2.10% (1.53 – 2.67)
3.68% (2.94 – 4.42)
0.001
2.90% (2.43 – 3.37)
Any Anxiety Disorder
4.49% (3.66 – 5.31)
8.45% (7.36 – 9.55)
<0.001
6.49% (5.80 – 7.18)
Generalized Anxiety Disorder
2.55% (1.92 – 3.18)
5.62% (4.71 – 6.53)
<0.001
4.10% (3.54 – 4.65)
Panic Disorder
1.19% (0.76 – 1.62)
2.55% (1.93 – 3.17)
<0.001
1.88% (1.50 – 2.26)
Obsessive Compulsive Disorder
1.27% (0.83 – 1.72)
2.10% (1.54 – 2.67)
0.025
1.69% (1.33 – 2.05)
All Phobias
1.93% (1.39 – 2.43)
3.64% (2.90 – 4.38)
<0.001
2.79% (2.33 – 3.26)
Any Anxiety or Depressive Disorder
5.43% (4.53 – 6.33)
9.87% (8.69 – 11.05)
<0.001
7.67% (6.92 – 8.41)
Other Non Specific Psychiatric Morbidity
(“mixed anxiety-depression”)
2.26% (1.67 – 2.85)
3.07% (2.39 – 3.75)
0.078
2.67% (2.22 -3.12)
Harmful Alcohol Use
16.95% (15.45 – 18.44)
8.51% (7.40 – 9.60)
<0.001
12.69% (11.75 – 13.62)
Smoking Use (regular user - past month)
49.56% (47.57 – 51.56)
29.80% (28.0 – 31.61)
<0.001
39.60% (38.22 – 40.97)
Cannabis Use (at least once - past month)
3.54% (2.80 – 4.27)
0.61% (0.30 – 0.91)
<0.001
2.06% (1.66 – 2.46)
Score on the CIS-R ≥122
10.84% (9.60 – 12.08)
17.21% (15.72 – 18.70)
<0.001
14.06% (13.08 – 15.03)
1 CI Confidence intervals, 2 A score on the CIS-R ≥12 represents a dimensional view of clinically significant psychiatric morbidity. Table 1 Prevalence (1-week) of common mental disorders and substance use by gender in a representative sample of
the general population of Greece 18–70, (weighted data, N = 4894) grouped into one category (data for specific disorders are
available upon request by authors). Table 2 (right columns) also shows the health status
(assessed with EQ-5D) and mental health service use
(during the past 12 months) associated with the pres-
ence of common mental disorders in pure and comorbid
forms. Health status was lower in comorbid forms of
panic disorder, OCD and phobic disorders, while there
were no significant differences for depression and GAD. Compared to participants without psychiatric morbidity,
scores on the health status index were significantly lower
for those with psychiatric morbidity (0.52 [sd: 0.30] vs
0.84
[sd:
0.20]
for
all
common
mental
disorders
combined, p < 0.001). Regarding mental health service
use, roughly one third (~32%) of those with a current
common mental disorder reported that they had visited
a mental health professional during the past year. Duration of common mental disorders We retrospectively assessed the duration of the core
symptoms of common mental disorders and in Figure 4
we present the percentage of cases with a duration of
symptoms for at least one year (or more). It can be seen
that chronicity (duration of illness for one year or more)
was reported by approximately half of those with a
common mental disorder. OCD and depression had the
longest duration with 65% and 56% persistence over one
year respectively (Figure 4). Prevalence of substance use disorders Service
use
was more
common
in panic disorder
(approximately 45% of those with current panic had seen
a mental health professional in the past 12 months). With the exception of panic, participants with comorbid
forms of disorders had visited more often a mental
health professional but this reached statistical signifi-
cance only for OCD where only 10% with “pure” form
had visited a professional compared to 36.5% of the
comorbid form (p = 0.024). For depression, female gender, being divorced/sepa-
rated or widowed and being unemployed were associated
with an increased prevalence, while having a better
education with a lower prevalence (p = 0.024 for the
linear trend in education). For anxiety disorders, female
gender and being divorced/separated or widowed was
associated with an increased prevalence. Harmful alco-
hol consumption was less likely in women, the older age
group, the better educated (p < 0.001 for the linear trend
in education) and for those looking after their house or
were economically inactive. It was higher in single per-
sons and the divorced/separated. Presence of a chronic
severe medical condition was also strongly associated
with both depression or anxiety disorders but not with
harmful alcohol use (Table 3). Sociodemographic associations Table 3 presents adjusted odds ratios and their 95% confi-
dence intervals for the association between sociodemo-
graphic variables and common mental disorders/harmful
alcohol use (for crude odds ratios please see Additional
file 1: Table S3). We also present the associations for our
dimensional definition of psychiatric morbidity (CIS-R
≥12)
for
comparison. Anxiety
disorders
have
been Main findings In this nationally representative sample of the Greek
general population the one-week prevalence of general Table 2 Co-morbidity, health status and use of mental health services of participants with common mental disorders in a representative sample of the general
population of Greece (18–70), N = 4894
Percentage (95% CI1) with
Health status (Score on the EQ-5D index) &
Use of services (% of cases who visited a Mental Health Professional past
12 months)
Pure disorder
1 additional
diagnosis
2 additional
diagnoses
≥3 additional
diagnoses
Total
Pure disorder
≥1 additional diagnoses
Depressive Episode
40.8%
22.5%
19.0%
17.6%
EQ-5D index (SD2)
0.46 (0.30)
0.46 (0.28)
0.46 (0.32)
(33.7 – 49.0)
(15.6 – 29.5)
(12.5 – 25.5)
(11.3 – 23.9)
Visited MH Prof. 3
34.5%
27.6%
39.3%
GAD4
46.8%
27.4%
12.9%
12.9%
EQ-5D2 index (SD2)
0.51 (0.28)
0.53 (0.28)
0.49 (0.28)
(39.8 -53.7)
(21.1 – 33.6)
(8.2 – 17.6)
(8.2 – 17.6)
Visited MH Prof. 3
29.3%
23.4%
34.6%
Panic Disorder
4.3%
41.3%
22.8%
31.5%
EQ-5D2 index (SD2)
0.49 (0.31)
0.22 (0.13)7
0.50 (0.31)7
(0.1 – 8.6)
(31.0 – 51.6)
(14.1 – 31.6)
(21.8 – 41.2)
Visited MH Prof. 3
44.6%
50.0%
44.3%
OCD5
24.1%
22.9%
21.7%
31.3%
EQ-5D2 index (SD2)
0.55 (0.33)
0.68 (0.32)8
0.51 (0.32)8
(14.7 – 33.5)
(13.7 – 32.1)
(12.6 – 30.7)
(21.1 – 41.5)
Visited MH Prof. 3
30.1%
10.0%9
36.5%9
All Phobias
21.9%
36.5%
20.4%
21.2%
EQ-5D2 index (SD2)
0.55 (0.31)
0.69 (0.26)8
0.51 (0.31)8
(14.9 – 28.9)
(28.3 – 44.7)
(13.6 – 27.3)
(14.2 – 28.1)
Visited MH Prof. 3
40.9%
43.3%
40.2%
Score on the CIS-R ≥125
NA6
EQ-5D2 index (SD2)
0.54 (0.30)
NA6
Visited MH Prof. 3
23.69%
No current Psychiatric morbidity
(score on the CIS-R <12)
NA6
EQ-5D2 index (SD2)
0.86 (0.18)
NA6
Visited MH Prof. 3
3.28%
1 CI Confidence intervals, 2 SD Standard deviation, 3 MH Prof. Mental health professional (either psychiatrist or psychologist), 4 GAD Generalized anxiety disorder, 5 OCD Obsessive compulsive disorder, 6 A score on the
CIS-R ≥12 represents a dimensional view of clinically significant psychiatric morbidity; 6 NA: not applicable; 7p = 0.08 (EQ-5D - pure vs. comorbid); 8p < 0.05 (EQ-5D - pure vs. comorbid); 9p < 0.05 (Visited MH Prof - pure
vs. comorbid). Comparison with previous studies in Greece Comparison with previous studies in Greece
We have previously carried out a number of smaller
scale studies in Greece using the CIS-R to assess psychi-
atric morbidity. Two of them were carried out in islands
of the Aegean sea, the first in small Northern Aegean
communities [28,29] and the second in Paros [30]. In all
studies, fatigue irritability and worry were the three
more prevalent symptoms, although in the small island
communities general psychiatric morbidity was reported
higher, especially in Paros where 30% of the women and
13% of men were found to have a CIS-R score of ≥12
compared to 17% and 11% respectively in the current
study. These differences are more likely explained by the
sampling procedure that resulted in an overrepresenta-
tion of women in the Paros study. Differences in the
Northern Aegean study were smaller. A third study was
carried out in older adolescents (16–18 years old)
attending senior high schools in Greece and since this
was carried out in the school context and in this specific
age group results are more difficult to compare [17]. Main findings Table 2 Co-morbidity, health status and use of mental health services of participants with common mental disorders in a representative sample of the general
population of Greece (18–70), N = 4894
Percentage (95% CI1) with
Health status (Score on the EQ-5D index) & bidity, health status and use of mental health services of participants with common mental disorders in a representative sample of the general
G
(
) services of participants with common mental disorders in a representative sample of the general Health status (Score on the EQ-5D index) & Skapinakis et al. BMC Psychiatry 2013, 13:163
http://www.biomedcentral.com/1471-244X/13/163 Skapinakis et al. BMC Psychiatry 2013, 13:163 Page 8 of 14 p
y
y
http://www.biomedcentral.com/1471-244X/13/163 Figure 3 Comorbidity of common mental disorders. igure 3 Comorbidity of common mental disorders. Figure 3 Comorbidity of common mental disorders. common mental disorders and other variables do not
have any causal implications. Reverse causality may be
equally true for some of the reported associations. psychiatric morbidity of at least mild severity was
14.06% (95% CI: 13.08 – 15.03). Generalized anxiety
disorder and depressive episode were the most common
specific disorders. A pattern of harmful alcohol use was
noted in 12.69% (11.75 – 13.62) of the population. Women were more likely to report a common mental
disorder but less likely to report use or abuse of all
substances studied. All disorders were associated with a
considerable reduction in health related quality of life
and a duration of symptoms longer than one year was
common. Despite these characteristics, only one third of
those with a disorder reported use of mental health
services within the past 12 months. Limitations
h
f
d These findings should be interpreted in the context of
the following limitations: a) The response rate of the
survey was relatively low at 54.2%. This is typical of what
can be achieved in general population surveys in coun-
tries without considerable tradition in medical or psychi-
atric survey research, such as Greece. The possibility of
selection bias cannot be ruled out, although the age and
sex distribution of the participants was comparable to
the national data. In any case, we have weighted our
results to account for non-response; b) The “clinical”
validity of the common mental disorders we have
defined in this paper has not been studied in the context
of the present study. A few studies in the past have tried
to test the validity of psychiatric diagnoses obtained by
structured diagnostic instruments administered by non-
clinicians,
and
generally
report
low
to
moderate
concordance with all diagnostic interviews [25-27]; c)
Chronic physical illnesses were assessed crudely by self-
report and this will be inaccurate. In addition, the dur-
ation
of
psychiatric
symptoms
was
retrospectively
assessed and there is a risk for recall bias which could
lead to overestimation depending on the severity of the
current episode of illness; d) Finally, as this is a cross-
sectional study, any reported associations between the There are two previous general population surveys of
psychiatric morbidity in Greece, both carried out by
Madianos and his colleagues in 1978 and 1984 [7]. Sam-
ples (of 4083 and 3706 participants respectively) were
drawn from mainland Greece only, excluding house-
holds from insular Greece. Psychiatric morbidity was
assessed with Langner’s scale (an instrument which
predated modern diagnostic criteria) and the Center for
Epidemiologic Studies of Depression Scale (CES-D). Due
to their design, these studies have provided data on gen-
eral psychiatric morbidity. Regarding specific diagnostic
categories only major depression was covered. These
studies reported clinical significant psychiatric morbidity
in 24% of women and 11% of men in 1978 and in 34%
vs 19% respectively in 1984. Limitations
h
f
d Prevalence of depression Table 3 Sociodemographic associations of common mental disorders and harmful alcohol use in a representative sample of the general population of Greece
(18–70), N = 4894
Depressive episode
Any anxiety disorder
Harmful alcohol use
Score on the CIS-R ≥121
%
Adjusted OR2
95% CI3
%
Adjusted OR2
95% CI3
%
Adjusted OR2
95% CI3
%
Adjusted OR2
95% CI3
Gender
Men
2.10%
1.00
Ref
4.49%
1.00
Ref
16.95%
1.00
Ref
10.85%
1.00
Ref
Women
3.68%
1.66
1.12 – 2.47
8.45%
2.12
1.61 – 2.78
8.51%
0.53
0.43 – 0.64
17.21%
1.69
1.39 – 2.05
Age
18-29
1.38%
1.00
Ref
4.64%
1.00
Ref
14.85%
1.00
Ref
8.16%
1.00
Ref
30-39
2.03%
1.58
0.75 – 3.34
5.80%
1.19
0.76 – 1.87
13.86%
0.99
0.75 – 1.31
10.95%
1.29
0.92 – 1.80
40-49
2.46%
1.65
0.74 – 3.65
6.52%
1.16
0.72 – 1.88
13.70%
1.00
0.73 – 1.37
14.78%
1.58
1.11 – 2.24
50-59
3.86%
2.11
0.94 – 4.73
6.84%
0.97
0.58 – 1.63
12.97%
0.87
0.61 – 1.24
16.83%
1.41
0.97 – 2.05
60-70
5.56%
2.05
0.85 – 4.94
9.44%
0.92
0.51 – 1.65
7.11%
0.40
0.25 – 0.65
22.44%
1.29
0.84 – 1.96
Marital Status
Married
2.74%
1.00
Ref
6.24%
1.00
Ref
11.15%
1.00
Ref
14.76%
1.00
Ref
Single
1.52%
1.11
0.57 – 2.13
4.08%
0.79
0.52 – 1.21
16.11%
1.41
1.09 – 1.84
8.23%
0.88
0.65 – 1.19
Divorced/Separated
5.83%
2.10
1.13 – 3.90
12.08%
1.89
1.22 – 2.93
18.33%
1.77
1.23 - 2.54
22.50%
1.67
1.19 – 2.34
Widowed
11.27%
2.44
1.39 – 4.30
16.43%
1.66
1.05 – 2.63
4.69%
0.57
0.29 – 1.13
34.27%
1.72
1.21 – 2.44
Educational Qual. None/Primary Educ. Limitations
h
f
d 5.40%
1.00
Ref
9.72%
1.00
Ref
12.63%
1.00
Ref
23.65%
1.00
Ref
Lower Secondary
4.27%
1.18
0.72 – 1.93
7.28%
0.99
0.68 – 1.45
15.18%
0.79
0.59 – 1.07
19.32%
1.05
0.81 – 1.36
Upper Secondary
1.92%
0.67
0.40 – 1.14
5.15%
0.81
0.55 – 1.17
12.65%
0.55
0.41 – 0.73
10.14%
0.60
0.46 – 0.78
Technical Vocational
1.82%
0.61
0.26 – 1.42
4.56%
0.69
0.39 – 1.22
10.02%
0.41
0.27 – 0.62
8.66%
0.49
0.33 – 0.74
Tertiary Education
1.30%
0.40
0.15 – 1.09
5.47%
0.82
0.47 – 1.44
10.94%
0.50
0.32 – 0.75
10.16%
0.56
0.37 – 0.84
Employment Status
Full-time/part-time
2.19%
1.00
Ref
5.42%
1.00
Ref
14.91%
1.00
Ref
11.62%
1.00
Ref
Looking after house
3.47%
0.73
0.41 – 1.30
6.80%
0.64
0.43 – 0.95
6.37%
0.60
0.42 – 0.87
17.95%
0.80
0.60 – 1.06
Unemployed
5.43%
2.46
1.20 – 5.04
9.24%
1.58
0.92 – 2.73
19.57%
1.37
0.93 – 2.04
18.48%
1.68
1.12 – 2.54
Retired
5.37%
0.72
0.39 – 1.31
9.88%
0.88
0.56 – 1.39
8.32%
0.89
0.58 – 1.37
22.36%
0.85
0.61 – 1.17
Other Econ. Inactive
2.49%
1.04
0.54 – 1.99
5.93%
1.06
0.69 – 1.64
11.09%
0.69
0.50 – 0.94
11.85%
1.06
0.77 – 1.47 l disorders and harmful alcohol use in a representative sample of the general population of Greece Skapinakis et al. BMC Psychiatry 2013, 13:163
http://www.biomedcentral.com/1471-244X/13/163 Table 3 Sociodemographic associations of common mental disorders and harmful alcohol use in a representative sample of the general population of Greece
(18–70), N = 4894 (Continued)
Type of locality
Urban
3.02%
1.00
Ref
6.07%
1.00
Ref
10.07%
1.00
Ref
13.50%
1.00
Ref
Semi-rural
1.97%
0.58
0.31 – 1.08
7.39%
1.18
0.82 – 1.69
15.16%
1.53
1.18 - 1.99
14.66%
1.02
0.78 – 1.33
Rural
3.05%
0.93
0.64 – 1.36
6.85%
1.13
0.87 – 1.48
16.14%
1.61
1.33 – 1.94
14.77%
1.01
0.84 – 1.23
Presence of Chronic Physical Diseases
No
1.01%
1.00
Ref
2.49%
1.00
Ref
12.92%
1.00
Ref
5.70%
1.00
Ref
Yes
5.80%
3.5
2.34 – 5.24
12.64%
3.96
2.96 – 5.31
12.33%
1.22
0.92 – 1.62
26.89%
3.62
2.92 – 4.50
1 A score on the CIS-R ≥12 represents a dimensional view of clinically significant psychiatric morbidity; 2 OR Odds Ratio adjusted for all other variables of the table; 3 CI Confidence intervals. Figures in bold statistically
significant at p < 0.05. Limitations
h
f
d s and harmful alcohol use in a representative sample of the general population of Greece Skapinakis et al. BMC Psychiatry 2013, 13:163
http://www.biomedcentral.com/1471-244X/13/163 Skapinakis et al. BMC Psychiatry 2013, 13:163 Page 11 of 14 p
y
y
http://www.biomedcentral.com/1471-244X/13/163 Figure 4 Duration of common mental disorders. Figure 4 Duration of common mental disorders. was reported by 4.5% of women and 2.4% of men in
1978 and 6.6% vs 3.8% respectively in 1984 [7]. These
figures and especially those from the 1984 study are con-
siderably higher than the ones reported in the present
study, but the different instruments and sampling frame
are probably the major source of discrepancy. During
the same period, Mavreas et al. [31] carried out a smaller
scale study in Athens (N = 489) using a structured psy-
chiatric interview, the Present State Examination (PSE)
to assess psychiatric disorders. It is worth noting that in
that study general psychiatric morbidity was reported by
16% of the population, very similar to the one reported
here, given the differences in methods and the length of
time interval (24 years) between the two studies. More
recently, a number of telephone cross-sectional surveys
were organized in Greece with the aim to assess depres-
sion [8,9]. Three such studies have documented an in-
creasing trend in the prevalence of depression (assessed
with the depression module of the Structured Clinical
Interview – SCID-I), from 3.3% in 2008, to 6.8% in 2009
and 8.2% in 2011. The present study, carried out in
2009–2010, has reported a lower figure for the preva-
lence of depression than the ones reported in these tele-
phone surveys. It would be interesting to investigate
whether this is due to the application of diagnostic cri-
teria or there is a higher reporting of the core depressive
symptoms in the telephone surveys compared to the
structured clinical interview we used in the present
study. It is likely that the use of structured psychiatric
interviews yields more conservative estimates of clinic-
ally significant disorders, while the methods used in the
telephone surveys are more likely to result in over-
reporting of milder forms of psychopathology especially
in those above 60 [32]. Limitations
h
f
d It is worth noting that our esti-
mates of depression and anxiety disorders are associated
with low health-related quality of life (Table 2) and sub-
stantial chronicity, arguing in favor of a higher clinical
significance threshold in our study compared to these telephone surveys. A recent study in four European sites
has also confirmed that the CIS-R is less sensitive but
more specific in its depression diagnosis compared to
other structured interviews such as the composite inter-
national diagnostic interview [15]. was reported by 4.5% of women and 2.4% of men in
1978 and 6.6% vs 3.8% respectively in 1984 [7]. These
figures and especially those from the 1984 study are con-
siderably higher than the ones reported in the present
study, but the different instruments and sampling frame
are probably the major source of discrepancy. During
the same period, Mavreas et al. [31] carried out a smaller
scale study in Athens (N = 489) using a structured psy-
chiatric interview, the Present State Examination (PSE)
to assess psychiatric disorders. It is worth noting that in
that study general psychiatric morbidity was reported by
16% of the population, very similar to the one reported
here, given the differences in methods and the length of
time interval (24 years) between the two studies. More
recently, a number of telephone cross-sectional surveys
were organized in Greece with the aim to assess depres-
sion [8,9]. Three such studies have documented an in-
creasing trend in the prevalence of depression (assessed
with the depression module of the Structured Clinical
Interview – SCID-I), from 3.3% in 2008, to 6.8% in 2009
and 8.2% in 2011. The present study, carried out in
2009–2010, has reported a lower figure for the preva-
lence of depression than the ones reported in these tele-
phone surveys. It would be interesting to investigate
whether this is due to the application of diagnostic cri-
teria or there is a higher reporting of the core depressive
symptoms in the telephone surveys compared to the
structured clinical interview we used in the present
study. It is likely that the use of structured psychiatric
interviews yields more conservative estimates of clinic-
ally significant disorders, while the methods used in the
telephone surveys are more likely to result in over-
reporting of milder forms of psychopathology especially
in those above 60 [32]. Limitations
h
f
d It is worth noting that our esti-
mates of depression and anxiety disorders are associated
with low health-related quality of life (Table 2) and sub-
stantial chronicity, arguing in favor of a higher clinical
significance threshold in our study compared to these Regarding alcohol and other substances our results are
compatible with previous surveys in Greece that show
high rates for smoking and low for cannabis compared
to other European countries [33-35]. Most previous
studies of alcohol in Greece have taken a more clinical
approach and therefore it is difficult to compare their
results with our study where we have used a more
dimensional approach. It should be noted however that
all associations with harmful alcohol use were in the
expected
direction
supporting
the
validity
of
our
approach. Comparison with other studies in Europe and elsewhere
Wittchen & Jacobi in 2005 [36] reviewed 27 general
adult population studies conducted in Europe and have
recently (2011) updated this review extending their
coverage to children and adolescent studies [5]. It should
be noted that our results refer to 1-week prevalence
while the review used data on 12-month prevalence. Based on previous studies (see for example [37] and
[38]) that have reported both one-month and 12-month
prevalence of common mental disorders, a conservative
estimate for the ratio of one year vs. one month preva-
lence is 1.5 for depression and 1.3 for anxiety disorders
which translates to projected estimates of 12-month
prevalence for our study of 4.3% for depression and 8.4%
for anxiety disorders. Comparing these estimates to the
other European studies one can see that depression is on
the lower side (median prevalence in European studies
of 5.7% vs 4.3% projected in the present) while genera-
lized anxiety (median 2% vs. 5.3% projected) and other
anxiety disorders are on the higher (with the exception
of phobias). Page 12 of 14 Skapinakis et al. BMC Psychiatry 2013, 13:163
http://www.biomedcentral.com/1471-244X/13/163 Skapinakis et al. BMC Psychiatry 2013, 13:163
http://www.biomedcentral.com/1471-244X/13/163 Page 12 of 14 Another interesting study that can be used for compari-
son is the World Mental Health Survey that was carried
out in 14 countries from all continents [6]. Sociodemographic associations Most of the associations reported for common mental
disorders (e.g. female gender, marital status, educational
status) are in the expected direction and confirm the
findings of previous studies both in Greece and else-
where. These have been discussed thoroughly in previ-
ous reviews [5]. We would like to comment here on the
association between common mental disorders and age. In the unadjusted analysis there was a significant asso-
ciation between older age and depression or anxiety
disorders. The effect was stronger for depression (for ex-
ample prevalence of depression in the older age group
was 5.56% vs 1.38% in the younger group, p < 0.001,
while for anxiety disorders was 9.44% vs 4.64% respect-
ively,
p < 0.001). This
same
finding
has
been
also
reported by other surveys in Greece for depression (see
for example [8]) or for general psychiatric morbidity
assessed in health-related quality of life instruments (see
for example validation studies of EQ-5D [23] and SF-36
[41]. Traditionally, in studies carried out in developed
countries, older age is usually associated with a reduced
prevalence [11,42,43]. It is worth noting that in our ad-
justed analysis the age effect was greatly reduced and be-
came non-significant. This was mainly due to the
inclusion into the model of the presence of a comorbid
chronic medical condition. There is a strong association
between chronic physical disorders and common mental
disorders [44,45] and this was also confirmed in our
study. As the prevalence of chronic medical conditions
increase sharply with age, this creates a spurious associ-
ation between older age and depression or anxiety that
is due to the confounding effect of the comorbid chronic
physical disorders. It is difficult however to explain why
this is shown in the Greek studies and not in other Limitations
h
f
d According to
this study, our prevalence estimates would place Greece in
the middle regarding prevalence of depression along with
Germany, Italy, Mexico and Spain (with Asian countries
in the low and US and other Northern European countries
such as France, Netherlands and Ukraine in the high). Regarding anxiety disorders, this would be middle to high. A similar pattern of lower prevalence of depression
compared to anxiety disorders and especially generalized
anxiety disorder had also been reported from Athens in a
WHO cross-cultural study of common mental disorders
in primary care [39]. It is also reminded that rates of
suicide, often associated with severe depression, in Greece
are among the lowest in the world [40]. It is possible that
socio-cultural factors, including strong family cohesion,
high levels of social support and higher religiosity, may
play an important role in explaining these differences. epidemiological surveys elsewhere, where the association
between chronic physical disorders and mental disorders
is also strong [45]. One possibility is the less optimal
management of chronic physical disorders in the Greek
health system which is mostly focused on the medical
dimension and not on the overall quality of life including
mental health issues. The Greek health system is still
physician-centred
rather
than
multidisciplinary
[46]. This issue merits further investigation in the future as
the older age group appears quite affected by both men-
tal and physical disorders and their combined harmful
effect in quality of life has not been recognized and
appreciated in Greece. Unemployment was also associated with common
mental disorders and especially with depression, a com-
mon finding in epidemiological research [47]. However,
in Greece there has been a sharp increase in unemploy-
ment recently and the health effects of unemployment
will become a serious public health issue, taking into
account the difficulty in establishing an effective safety
net for the unemployed in an economy which is in deep
recession [48]. In a companion paper we examine more
in detail the association between various socioeconomic
indicators and unemployment and common mental
disorders and suicidal behaviour (Skapinakis et al. in
preparation). Use of services
U
f
l h Use of mental health services during the past 12 months
was reported by approximately one third of those who
met criteria for a psychiatric diagnosis, a common
finding in several surveys around the world [49]. It was
higher for panic disorder and for comorbid vs. pure
disorders with the exception of panic and phobic
anxiety. An analysis of the correlates of service use is
beyond the scope of the current paper and it will be
explored in more detail in future reports. Authors’ contributions
PS
ibl ( ith 15. Pez O, Gilbert F, Bitfoi A, Carta MG, Jordanova V, Garcia-Mahia C, Mateos-
Alvarez R, Prince M, Tudorache B, Blatier C, Kovess-Masfety V: Validity across
translations of short survey psychiatric diagnostic instruments: CIDI-SF
and CIS-R versus SCID-I/NP in four European countries. Soc Psychiatry
Psychiatr Epidemiol 2010, 45:1149–1159. PS was responsible (with VM) for the conception and design of the study,
helped in data collection, contributed to the statistical analysis and drafted
the manuscript. SB helped in data collection, in the statistical analysis, in the
writing of the manuscript and interpretation of the results. SK helped in
obtaining funding for the study and contributed to the writing of the paper
and interpretation of the results. IG helped in data collection, writing of the
manuscript and interpretation of the results. PNT helped in data collection,
writing of the manuscript and interpretation of the results. VM was
responsible (with PS) for the conception and design of the study and helped
in obtaining funding for the study, in the writing of the paper and
interpretation of the results. All authors read and approved the final
manuscript. PS was responsible (with VM) for the conception and design of the study,
helped in data collection, contributed to the statistical analysis and drafted
the manuscript. SB helped in data collection, in the statistical analysis, in the
writing of the manuscript and interpretation of the results. SK helped in
obtaining funding for the study and contributed to the writing of the paper
and interpretation of the results. IG helped in data collection, writing of the
manuscript and interpretation of the results. PNT helped in data collection,
writing of the manuscript and interpretation of the results. VM was 16. Lewis G, Pelosi AJ, Glover E, Wilkinson G, Stansfeld SA, Williams P, Shepherd M:
The development of a computerized assessment for minor psychiatric
disorder. Psychol Med 1988, 1988(18):737–745. 17. Skapinakis P, Anagnostopoulos F, Bellos S, Magklara K, Lewis G, Mavreas V:
An empirical investigation of the structure of anxiety and depressive
symptoms in late adolescence: cross-sectional study using the Greek
version of the revised clinical interview schedule. Psychiatry Res 2011,
186:419–426. responsible (with PS) for the conception and design of the study and helped
in obtaining funding for the study, in the writing of the paper and
interpretation of the results. All authors read and approved the final
manuscript. 18. References . Vos T, Flaxman AD, Naghavi M, Lozano R, Michaud C, Ezzati M, et al: Years
lived with disability (YLDs) for 1160 sequelae of 289 diseases and
injuries 1990–2010: a systematic analysis for the Global Burden of
Disease Study 2010. Lancet 2012, 380:2163–2196. 2. World Health Organization: The Global Burden of disease: 2004 update. Geneva: WHO; 2008. 3. Eaton WW, Martins SS, Nestadt G, Bienvenu OJ, Clarke D, Alexandre P:
The burden of mental disorders. Epidemiol Rev 2008, 30:1–14. 4. Skapinakis P, Lewis G: Epidemiology in community psychiatric research:
common uses and methodological issues. Epidemiol Psichiatr Soc 2001,
10:18–26. 5. Wittchen HU, Jacobi F, Rehm J, Gustavsson A, Svensson M, Jfnsson B,
Olesen J, Allgulander C, Alonso J, Faravelli C, Fratiglioni L, Jennum P, Lieb R,
Maercker A, van Os J, Preisig M, Salvador-Carulla L, Simon R, Steinhausen
HC: The size and burden of mental disorders and other disorders of the
brain in Europe 2010. Eur Neuropsychopharmacol 2011, 21:655–679. 6. Kessler RC, Aguilar-Gaxiola S, Alonso J, Chatterji S, Lee S, Ormel J, Ustun TB,
Wang PS: The global burden of mental disorders: an update from the
WHO World Mental Health (WMH) surveys. Epidemiol Psichiatr Soc 2009,
18:23–33. 7. Madianos MG, Stefanis CN: Changes in the prevalence of symptoms of
depression and depression across Greece. Soc Psychiatry Psychiatr
Epidemiol 1992, 27:211–219. 8. Madianos M, Economou M, Alexiou T, Stefanis C: Depression and
economic hardship across Greece in 2008 and 2009: two cross-sectional
surveys nationwide. Soc Psychiatry Psychiatr Epidemiol 2011, 46:943–952. Received: 6 January 2013 Accepted: 25 May 2013
Published: 4 June 2013 p
g
The authors declare that they have no competing interests. The authors declare that they have no competing interests. 14. Liu SI, Mann A, Cheng A, Tjung JJ, Hwang LC: Identification of common
mental disorders by general medical doctors in Taiwan. Gen Hosp
Psychiatry 2004, 26:282–288. Authors’ contributions
PS
ibl ( ith Saunders JB, Aasland OG, Babor TF, de la Fuente JR, Grant M: Development
of the alcohol Use disorders identification test (AUDIT): WHO
collaborative project on early detection of persons with harmful alcohol
consumption–II. Addiction 1993, 88:791–804. Conclusions We presented the results of the first general population
survey of psychiatric morbidity that used a nationally
representative sample from both mainland and insular
Greece. These results show that a considerable propor-
tion of the population suffer from common mental dis-
orders. In this figure we should also add those suffering
from serious mental disorders not covered in the present
survey, such as bipolar disorder or schizophrenia. These
results highlight the need for better organization of
services taking into account that the majority of those
with a mental illness will not visit a mental health
professional either due to stigmatization [50] or low
recognition [39]. This is even more important today in
Greece because of the restrictions imposed on the health
budget that could disproportionately affect funds available Skapinakis et al. BMC Psychiatry 2013, 13:163
http://www.biomedcentral.com/1471-244X/13/163 Page 13 of 14 Skapinakis et al. BMC Psychiatry 2013, 13:163
http://www.biomedcentral.com/1471-244X/13/163 Page 13 of 14 for mental health. Our study also highlights the close
association between mental and physical disorders sup-
porting the view that there is no health without mental
health [51]. Our finding that depression and anxiety are
greatly increased in older people with chronic physical
disorders also calls for a more careful assessment of
common mental disorders in this group. In times of
economic crises, the observed increases in morbidity and
mortality is the result of common diseases that are less
optimally managed. The Russian example during the 90’s
is quite informative: the mortality fluctuations observed
there during the 90’s were mainly due to changes in
mortality from vascular diseases and harmful alcohol use
[52]. The additional burden of mental illness in this group
of patients may also be a contributing factor that needs to
be taken into account if we want to reduce the impact of
the economic crisis in the health of the population. Additional file 9. Economou M, Madianos M, Peppou LE, Patelakis A, Stefanis CN: Major
depression in the Era of economic crisis: A replication of a cross-
sectional study across Greece. J Affect Disord 2012. http://dx.doi.org/
10.1016/j.jad.2012.08.008. Additional file 1: Table S1. Basic Description of the Sample in a
representative sample of the general population of Greece (18-70,
N = 4894). Table S2. Prevalence of Psychiatric Symptoms by gender in a
representative sample of the general population of Greece (18-70,
N = 4894). Table S3. Crude odds ratios of the association between
sociodemographic associations and common mental disorders/harmful
alcohol use in a representative sample of the general population of
Greece (18-70, N = 4894). Figure S1. Prevalence of common mental
disorders by gender in a representative sample of the general population
of Greece (18-70), N = 4894. Figure S2. Age distribution of common
mental disorders in a representative sample of the general population of
Greece (18-70), N = 4894. 10. Lewis G, Pelosi AJ, Araya R, Dunn G: Measuring psychiatric disorder in the
community: a standardised assessment for use by lay interviewers. Psychol Med 1992, 22:465–486. 11. Singleton N, Bumpstead R, O’Brien M, Lee A, Meltzer H: Psychiatric Morbidity
Among Adults Living in Private Households. London: The Stationery Office; 2001. 12. Botega NJ, Pereira WA, Bio MR, Garcia CJR, Zomignani MA: Psychiatric
morbidity among medical in-patients: a standardized assessment (GHQ-
12 and CIS-R) using ‘lay’ interviewers in a Brazilian hospital. Soc Psychiatry
Psychiatr Epidemiol 1995, 30:127–131. 13. Araya R, Rojas G, Fritsch R, Acu AJ, Lewis G: Common mental disorders in
Santiago, Chile: prevalence and socio-demographic correlates. Br J
Psychiatry 2001, 178:228–233. Additional file 1: Table S1. Basic Description of the Sample in a
representative sample of the general population of Greece (18-70,
N = 4894). Table S2. Prevalence of Psychiatric Symptoms by gender in a
representative sample of the general population of Greece (18-70,
N = 4894). Table S3. Crude odds ratios of the association between
sociodemographic associations and common mental disorders/harmful
alcohol use in a representative sample of the general population of
Greece (18-70, N = 4894). Figure S1. Prevalence of common mental
disorders by gender in a representative sample of the general population
of Greece (18-70), N = 4894. Figure S2. Age distribution of common
mental disorders in a representative sample of the general population of
Greece (18-70), N = 4894. Acknowledgements This study was co-funded by the European Social Fund and National
Resources. The authors would like to thank Mrs Maria Trochani and Mrs Eleni
Petsi from the Ministry of Health, Mrs Maria Triantafillou and Mr Makis
Ragousis from MRB Hellas, and the employees of the seven regional health
authorities who were involved in the study. SK would like to acknowledge
his current affiliation with Evangelismos General Hospital, and PNT with the
Open University of Cyprus. 19. Bush K, Kivlahan DR, McDonell MB, Fihn SD, Bradley KA: The AUDIT alcohol
consumption questions (AUDIT-C): an effective brief screening test for
problem drinking. Ambulatory care quality improvement project
(ACQUIP). alcohol use disorders identification test. Arch Intern Med 1998,
158:1789–1795. 20. Aalto M, Alho H, Halme JT, Seppd K: AUDIT and its abbreviated versions in
detecting heavy and binge drinking in a general population survey. Drug Alcohol Depend 2009, 103:25–29. Page 14 of 14 Skapinakis et al. BMC Psychiatry 2013, 13:163
http://www.biomedcentral.com/1471-244X/13/163 Skapinakis et al. BMC Psychiatry 2013, 13:163
http://www.biomedcentral.com/1471-244X/13/163 44. Sareen J, Cox BJ, Clara I, Asmundson GJ: The relationship between anxiety
disorders and physical disorders in the U.S. National comorbidity survey. Depress Anxiety 2005, 21:193–202. 21. Reinert DF, Allen JP: The alcohol use disorders identification test: an
update of research findings. Alcohol Clin Exp Res 2007, 31:185–199. 22. The EuroQol Group: EuroQol–a new facility for the measurement of
health-related quality of life. Health Policy 1990, 16:199–208. 45. Stein MB, Cox BJ, Afifi TO, Belik SL, Sareen J: Does co-morbid depressive
illness magnify the impact of chronic physical illness? A population-
based perspective. Psychol Med 2006, 36:587–596. 23. Kontodimopoulos N, Pappa E, Niakas D, Yfantopoulos J, Dimitrakaki C,
Tountas Y: Validity of the EuroQoL (EQ-5D) instrument in a Greek general
population. Value Health 2008, 11:1162–1169. 46. Papathanassoglou ED, Karanikola MN, Kalafati M, Giannakopoulou M,
Lemonidou C, Albarran JW: Professional autonomy, collaboration with
physicians, and moral distress among European intensive care nurses. Am J Crit Care 2012. http://dx.doi.org/10.4037/ajcc2012205. 24. Stata Corporation: Stata Reference Manual, release 10.0. College Station,
Texas: Stata Press; 2007. 25. Helzer JE, Robins LN, McEvoy LT, Spitznagel EL, Stoltzman RK, Farmer A,
Brockington IF: A comparison of clinical and diagnostic interview
schedule diagnoses: physician reexamination of lay-interviewed cases in
the general population. Arch Gen Psychiatry 1985, 42:657–666. 47. Montgomery SM, Cook DG, Bartley MJ, Wadsworth ME: Unemployment
pre-dates symptoms of depression and anxiety resulting in medical
consultation in young men. doi:10.1186/1471-244X-13-163 doi:10.1186/1471-244X-13-163
Cite this article as: Skapinakis et al.: Prevalence and sociodemographic
associations of common mental disorders in a nationally representative
sample of the general population of Greece. BMC Psychiatry 2013 13:163. 32. Evans M, Kessler D, Lewis G, Peters TJ, Sharp D: Assessing mental health in
primary care research using standardized scales: can it be carried out
over the telephone? Psychol Med 2004, 34:157–162. 33. Pazarlis P, Mauri D, Cortinovis I, Katsigiannopoulos K, Alevizaki P, Koukourakis
G, Casazza G, Kamposioras K, Chatziioannou I, Milousis A, Papakonstantinou
A, Karathanasi I, Alexiou G, Proiskos A, Mitrogianni Z, Peponi C: Socio-
demographic status and alcohol drinking patterns among Greek healthy
adults. Cent Eur J Public Health 2006, 14:160–167. 34. Kokkevi A, Fotiou A, Richardson C: Drug use in the general population of
Greece over the last 20 years: results from nationwide household
surveys. Eur Addict Res 2007, 13:167–176. 35. Filippidis FT, Vardavas CI, Loukopoulou A, Behrakis P, Connolly GN, Tountas Y:
Prevalence and determinants of tobacco use among adults in Greece: 4
year trends. Eur J Public Health 2012. http://dx.doi.org/10.1093/eurpub/cks148. 36. Wittchen HU, Jacobi F: Size and burden of mental disorders in Europe - a
critical review and appraisal of 27 studies. Eur Neuropsychopharmacol
2005, 15:357–376. 37. Andrade L, Walters EE, Gentil V, Laurenti R: Prevalence of ICD-10 mental
disorders in a catchment area in the city of S o Paulo, Brazil. Soc
Psychiatry Psychiatr Epidemiol 2002, 37:316–325. 38. Jacobi F, Wittchen HU, Holting C, Hffler M, Pfister H, M ller N, Lieb R:
Prevalence, co-morbidity and correlates of mental disorders in the
general population: results from the German health interview and
examination survey (GHS). Psychol Med 2004, 34:597–611. Submit your next manuscript to BioMed Central
and take full advantage of:
• Convenient online submission
• Thorough peer review
• No space constraints or color figure charges
• Immediate publication on acceptance
• Inclusion in PubMed, CAS, Scopus and Google Scholar
• Research which is freely available for redistribution
Submit your manuscript at
www.biomedcentral.com/submit 39. Mavreas VG, Kontea M, Dikeos D, Neroutsos E, Sfikaki M, Aritzi S, Stefanis CN:
Results from the Athens centre. In Mental Ilness in General Health Care. Edited by Ustun and Sartorious. Chichester: John Wiley & Sons Ltd; 1995. Acknowledgements Int J Epidemiol 1999, 28:95–100. 48. Bambra C: Yesterday once more? Unemployment and health in the 21st
century. J Epidemiol Community Health 2010, 64:213–215. 26. Jordanova V, Wickramesinghe C, Gerada C, Prince M: Validation of two
survey diagnostic interviews among primary care attendees: a
comparison of CIS-R and CIDI with SCAN ICD-10 diagnostic categories. Psychol Med 2004, 34:1013–1024. 49. Bebbington PE, Brugha TS, Meltzer H, Jenkins R, Ceresa C, Farrell M, Lewis G:
Neurotic disorders and the receipt of psychiatric treatment. Psychol Med
2000, 30:1369–1376. 27. Brugha TS, Bebbington PE, Jenkins R: A difference that matters:
comparisons of structured and semi-structured diagnostic interviews of
adults in the general population. Psychol Med 1999, 29:1013–1020. 50. Lasalvia A, Zoppei S, Van Bortel T, Bonetto C, Cristofalo D, Wahlbeck K, Bacle
SV, Van Audenhove C, van Weeghel J, Reneses B, Germanavicius A,
Economou M, Lanfredi M, Ando S, Sartorius N, Lopez-Ibor JJ, Thornicroft G,
The ASPEN/INDIGO Study Group: Global pattern of experienced and
anticipated discrimination reported by people with major depressive
disorder: a cross-sectional survey. Lancet 2012. http://dx.doi.org/10.1016/
S0140-6736(12)61379-8. 28. Skapinakis P, Zissi A, Savidou M, Tseloni M, Chiou M: Prevalence and
sociodemographic associations of psychiatric morbidity in rural
communities of the North Aegean. Arch Hell Med 2007, 24(s1):30–36. 29. Tseloni A, Zissi A, Skapinakis P: Psychiatric morbidity and social capital in
rural communities of the Greek North Aegean islands. J Community
Psychol 2010, 38:1023–1041. 51. Prince M, Patel V, Saxena S, Maj M, Maselko J, Phillips MR, Rahman A: No
health without mental health. Lancet 2007, 370:859–877. 30. Stylianidis S, Skapinakis P, Pantelidou S, Chondros P, Avgoustaki A, Ziakoulis M:
Prevalence of common psychiatric disorders in an island region. Arch Hell Med 2010, 27:675–683. 52. Men T, Brennan P, Boffetta P, Zaridze D: Russian mortality trends for
1991–2001: analysis by cause and region. BMJ 2003, 327:964–970. 31. Mavreas VG, Beis A, Mouyias A, Rigoni F, Lyketsos GC: Prevalence of
psychiatric disorders in Athens. A community study. Soc Psychiatry 1986,
21:172–181. doi:10.1186/1471-244X-13-163 Submit your next manuscript to BioMed Central
and take full advantage of:
• Convenient online submission
• Thorough peer review
• No space constraints or color figure charges
• Immediate publication on acceptance
• Inclusion in PubMed, CAS, Scopus and Google Scholar
• Research which is freely available for redistribution
Submit your manuscript at
www.biomedcentral.com/submit Submit your next manuscript to BioMed Central
and take full advantage of:
• Convenient online submission
• Thorough peer review
• No space constraints or color figure charges
• Immediate publication on acceptance
• Inclusion in PubMed, CAS, Scopus and Google Scholar
• Research which is freely available for redistribution
Submit your manuscript at
www.biomedcentral.com/submit Submit your next manuscript to BioMed Central
and take full advantage of: Submit your next manuscript to BioMed Central
and take full advantage of: 40. Chishti P, Stone DH, Corcoran P, Williamson E, Petridou E, EUROSAVE
Working Group: Suicide mortality in the European Union. Eur J Public
Health 2003, 13:108–114. • Convenient online submission 41. Pappa E, Kontodimopoulos N, Niakas D: Validating and norming of the
Greek SF-36 health survey. Qual Life Res 2005, 14:1433–1438. 42. Bijl RV, Ravelli A, van Zessen G: Prevalence of psychiatric disorder in the
general population: results of The Netherlands mental health survey and
incidence study (NEMESIS). Soc Psychiatry Psychiatr Epidemiol 1998, 33:587–595. 43. Kessler RC, Berglund P, Demler O, Jin R, Merikangas KR, Walters EE: Lifetime
prevalence and age-of-onset distributions of DSM-IV disorders in the
national comorbidity survey replication. Arch Gen Psychiatry 2005,
62:593–602.
|
https://openalex.org/W2234942722
|
http://www.scielo.br/pdf/rbpi/v58n2/0034-7329-rbpi-58-02-00098.pdf
|
English
| null |
Public opinion and foreign policy in João Goulart's Brazil (1961-1964): Coherence between national and foreign policy perceptions?
|
Revista Brasileira de Política Internacional/Revista brasileira de política internacional
| 2,015
|
cc-by
| 9,603
|
* Universidade de São Paulo, Instituto de Relações Internacionais, São Paulo – SP, Brazil (felipeloureiro@usp.br).
** Universidade de São Paulo, Instituto de Relações Internacionais, São Paulo – SP, Brazil (felicianosa@usp.br).
*** Universidade de São Paulo, Instituto de Relações Internacionais, São Paulo – SP, Brazil (aschor@usp.br).
1 Studies on emergent democracies, such as Brazil, India, Mexico, and Turkey, include Faria (2008); Gonzalez
et al. (2013); Castillo et al. (2015); Kennedy and Dickenson (2012); and Kapur (2009). For a review of the
literature focused on the developed world, particularly the United States, see Burstein (2003) and Holsti (2004,
chs 1-3). For “emergent democracies”, see Culp and Plagemann (2013).
2 Historical accounts on the role played by public opinion in the making of US and UK foreign policies in the
twentieth century have contributed enormously for theory building in International Relations. See, for example,
Fenton (2012), Kusnitz (1984), Thompson (2010) and Wilkinson (2000). Public opinion and foreign policy in João Goulart’s
Brazil (1961-1964): Coherence between national and
foreign policy perceptions? Opinião pública e política externa no Brasil de João Goulart
(1961-1964): coerência entre percepções de políticas nacional
e internacional? http://dx.doi.org/10.1590/0034-7329201500206 FELIPE PEREIRA LOUREIRO*
FELICIANO DE SÁ GUIMARÃES**
ADRIANA SCHOR*** Rev. Bras. Polít. Int. 58 (2): 98-118 [2015] Artigo
Public opinion and foreign policy in João Goulart’s
Brazil (1961-1964): Coherence between national and
foreign policy perceptions?
Opinião pública e política externa no Brasil de João Goulart
(1961-1964): coerência entre percepções de políticas nacional
e internacional?
http://dx.doi.org/10.1590/0034-7329201500206
FELIPE PEREIRA LOUREIRO*
FELICIANO DE SÁ GUIMARÃES**
ADRIANA SCHOR***
Rev. Bras. Polít. Int. 58 (2): 98-118 [2015] Artigo 3 Report, ‘The Economic and Political Climate of Opinion in Latin America, and Attitudes Toward the Alliance
for Progress’, June 1963, Folder Latin America, General, 7/63-11/63, Box 216, National Security Files (hereafter
NSF), John F. Kennedy Library (hereafter JFKL). Introduction Scholars have been paying increasing attention to the relationship between
public opinion and foreign policy in emergent democracies, following a long
established academic tradition in the United States and Western Europe.1 Few
studies rely on a broader historical perspective though. This narrows down
significantly the set of comparisons that one can make, constraining the capacity
to better understand foreign policymaking in developing countries, and to advance
theories in a wider historical perspective.2 Recent studies on the relationship between public opinion and foreign policy
have shown that citizens tend to have coherent opinions across broad political
topics, such as redistributive, ideological, and foreign policy issues (Page and
Bouton 2006; Aldrich et al. 2006; Holsti 2004). In other words, a voter who 98 Public opinion and foreign policy in João Goulart’s Brazil (1961-1964) [...] favors liberal policies in the domestic arena will likely support similar foreign policy
actions. This is why scholars interested in understanding the role of public opinion
in foreign policy focus on how coherent is the perception between domestic and
foreign policy issues among citizens. This paper seeks to do the same taking João
Goulart’s Brazil (1961-1964) as case study. y
Widely acknowledged as one of the main emerging democracies of the
world, Brazil is an interesting case for an historical analysis (Whitehead 2002). In
particular, the Goulart administration is a crucial period of the short-lived Brazilian
postwar democracy (1946-1964). Scholars point out that the Goulart government
was marked by unprecedented social unrest and intense popular participation in the
public arena (Ferreira and Gomes 2014; Reis Filho 2014). The period was also a
critical moment for Brazil’s foreign policy. The Brazilian government moved away
from a closer relationship with the United States and advocated a more neutralist
role in the Cold War, particularly in the Western hemisphere (Storrs 1973). Revista Brasileira de Política Internacional Recent works on public opinion in Goulart’s Brazil have shown high public
support for socio-economic redistributive policies, such as agrarian and tax
reforms (Ferreira and Gomes 2014; Motta 2014). However, the opinion polls
used in these studies, conducted by the Brazilian Institute of Public Opinion and
Statistics (Instituto Brasileiro de Opinião Pública e Estatística, IBOPE), present
two limitations. First, they were restricted to few cities – normally São Paulo or
the state of Guanabara (currently the city of Rio de Janeiro), biasing the results
in favor of urban and more politicized voters. Second, they have no data on how
Brazilian citizens perceived the country’s foreign policy. In contrast, we employ a recently declassified national urban opinion poll
conducted in late 1962 by the Institute for Social and Economic Studies (Instituto
de Estudos Sociais e Econômicos, INESE) on behalf of the United States Information
Agency (USIA).3 This poll was conducted in all Brazilian cities with more than
10,000 habitants. It also contains some results for the same questions relative to
1961, which allows comparisons. Unfortunately, we were unable to find either
the original 1961 survey or the microdata for the 1962 opinion poll, limiting our
attempts to draw strong conclusions on the links between domestic and foreign
policy preferences. We try to overcome these limitations by cross referencing
the survey with other primary sources, and measuring the overlap coefficients
of answers across domestic and foreign policy issues. The available results brings
precious insight into Brazilian public opinion in the early 1960s. Moreover, this
is possibly the only source known up to this day that displays public opinion data
not only on socio-economic issues, but also on foreign policy initiatives. Two research questions inform this study. To what extent did the Brazilian
people support redistributive domestic policies, particularly agrarian and tax 99 Felipe Pereira Loureiro; Feliciano de Sá Guimarães; Adriana Schor reforms, as recent works on public opinion in Goulart’s Brazil have sustained? In case of a positive answer, was there any coherence between such support and
specific foreign policies strategies? In other words, to what extent did voters who
favored redistributive reforms also support, for example, a foreign policy recognized
as leftist, such as a nationalistic and anti-US foreign approaches? g
One expects to see, mutatis mutandis, coherency between domestic and
foreign policy perceptions. Revista Brasileira de Política Internacional Taking the context of the early 1960s Latin America
into consideration, it means “redistributive” supporters leaning towards Third
World or Communist countries in world affairs. Contrarily, “non-redistributive”
believers should support foreign policy preferences representing alignment with
the United States. Overall, we found that public opinion in Goulart’s Brazil was somewhat
coherent. The Brazilian society was, in its majority, not only supportive of
“redistributive” policies, but also in favor of an “autonomous” foreign policy
approach. However, the picture is not as clear as it might seem: the United States
enjoyed a good image in Brazil (while Communist countries presented a bad one),
and the number of Brazilians who supported a pro-US foreign policy was rapidly
increasing in the early 1960s. This suggests that the Brazilian public opinion could
be leaning towards incoherence on the eve of the 1964 military coup. This paper has four sections. Section one analyses the literature on public
opinion and foreign policy, and presents the methodology to measure coherence. Section two examines the historical context of the Goulart administration. Section
three analyses the data of the INESE-USIA opinion pool, comparing its results on
domestic policies to those related to Brazilian foreign policy and to the international
system. The last section presents the conclusions. Public opinion and foreign policy in João Goulart’s Brazil (1961-1964) [...] Public opinion and foreign policy in João Goulart’s Brazil (1961-1964) [...] Recently, however, researchers have shown that public opinion is relatively
stable and consistent in foreign policy perceptions across domestic and foreign
issues. The general idea is that voters are able to develop and hold fairly coherent
and structured views on foreign policy (Aldrich et al.1989; Aldrich et al. 2006). Page and Bouton (2006) argue that such change in the understanding is due
to a revolution in the cognitive sciences – that is, an overhaul on the scientific
understanding of how citizens receive, retain, and organize political information, as
well as on the degree of responsiveness to leaders preferences (Iyengar and Kinder
1987; Krosnick and Kinder 1990; Sniderman 1993). In this sense, although citizens are almost always in disadvantage compared to
political leaders and the media in terms of supply of information on international
politics, they tend to overcome this limitation by employing heuristics shortcuts
that allow them to make judgment about complex international issues (Baum
and Potter 2008). The literature shows two mechanisms through which citizens
are able to form their opinions: the first one is their traditional moral beliefs; the
second is their general positions or normative beliefs. Due to the lack of systematic
knowledge on issues of international relations, regular citizens use their more
accessible values and normative stances – usually related to domestic issues – as
source of information. They combine these two mechanisms with learning and
information from past experiences, daily life, media, and political campaign in order
to decide over international issues (Hurwitz and Peffley 1987; Popkin 1994). In
other words, by using shortcuts, even a low-informed citizen would behave very
similarly to a well-informed citizen (Castillo et al. 2015). Turning to Brazilian literature, the idea that citizens can be influenced by
foreign policy considerations when voting, or that they can structure broad political
views on foreign policy issues using domestic policy perceptions, is almost non-
existent. The general perspective is that Brazilian voters are either not interested
or not capable of organizing structured political views about foreign policy. The
reasons for such behavior range from lack of interest and low political participation
to isolated foreign policy formulation (Albuquerque et al. 1999; Manzur 1999;
Holzhacker 2006; Almeida et al. 2012; Bressian 2012; Souza 2002; Souza 2008). 4 The only study on the role of public opinion in Goulart’s Brazil foreign policy focuses more on newspapers
and the Legislative than public opinion per se. See Manzur (2009). Literature and methodology The dominant consensus on the relationship between foreign policy and
public opinion has evolved over time. The first wave of studies showed that public
opinion was not coherent on foreign policy issues. Studies have focused on the
stability or volatility of foreign policy perceptions, the level of knowledge about
foreign policy, and, more importantly, the coherence of belief structures between
domestic and foreign issues. In this regard, the idea was that voters did not have
a structured and stable framework of beliefs and interpretation on foreign policy
(Almond 1950; Verba et al. 1967). A second wave of studies partially changed these notions. Scholars argued
that collective behavior was to some extents coherent on foreign policy issues
(Holsti 2004; Delli-Carpini and Keeter 1996; Popkin 1994). Yet, it still prevailed
the perspective that voters knew almost nothing about international politics, or
that the general public was unable to analyze complex international issues (Zaller
1992; Jacobs and Shapiro 2000). 100 Public opinion and foreign policy in João Goulart’s Brazil (1961-1964) [...] Actually, Brazilian studies are more concerned about how organized groups and
the media influence foreign policy (supposedly monopolized by Itamaraty, Brazil’s
Foreign Ministry) than to properly access what is the electoral connection between
international politics and voters (Casarões 2012; Lessa and Gavião 2011; Faria
2008 and 2012). Furthermore, we have found no studies on how Brazilian citizens in the 1960s
perceived foreign policy issues, or to what extent their perceptions were coherent
to broader themes such as redistributive policies.4 Scholars who directly analyzed 101 Felipe Pereira Loureiro; Feliciano de Sá Guimarães; Adriana Schor Brazilian public opinion in the 1960s emphasized on domestic issues (land and
tax reforms, for example) rather than on foreign policy aspects, due to the kind
of questions posed by available surveys (Motta 2014). To answer our questions, we look at the 1962 Brazilian public opinion survey
through two axes related to domestic and foreign policy issues. First, we suggest a
redistributive/non-redistributive axis for domestic perceptions. Although scholars
generally analyze coherence taking partisanship, ideology and foreign policy
opinion as main variables, our survey does not show data on party or ideological
preferences, but only on redistributive issues.5 We understand redistributive policies
as those pursued through taxation, property expropriations, or social expenditure
and transfers. They are concerned with changing the existing allocation of resources
within the society, favoring the poorest (Lowi 1972). g
p
In this axis, we arranged public opinions perceptions on domestic reforms as
either “redistributive” or “non-redistributive”. “Redistributive” refers to those who
defend land reform and nationalization of foreign companies without proper or
full compensation to owners, as well as a progressive tax overhaul (wealthier paying
more taxes than the poor). Conversely, “non-redistributive” refers to those who
argue for no reforms at all (e.g., pro-status quo). Even though one could argue that
citizens who stood up for land reform and nationalization of foreign companies
with full compensation should have been interpreted as “non-redistributive”, the
redistributive impacts of such policies would have been significant, as it meant that
the State would compel owners to abdicate their properties on the society’s behalf. Second, we suggest an autonomist/alignment axis for foreign policy
perceptions. In general, the literature on how foreign policy opinion is
structured defines three types of voters’ attitudes towards foreign policy: militant
internationalist policies (unilateralism and use of force), cooperative internationalist
policies (active multilateralism), and isolationism (Hurwitz and Peffley 1987; Nau
and Baum 2012). 5 See Holsti (2004), Kull (2007), Nau and Baum (2012), and Page and Bouton (2006). Revista Brasileira de Política Internacional The early 1960s was a watershed in Brazil’s history. After years of intense
economic growth, the country was facing severe constraints, such as rising inflation
and deficits in the balance of payment (Abreu 2008). Implementing harsh
economic stabilization programs triggered social unrest mainly in urban areas,
where unionized labor initiated demonstrations and strikes. Workers struggled
to avoid having the burden on their shoulders, whether through wage constraints
or tax increases (Loureiro 2016). The level of social unrest in rural areas was also
unprecedented, particularly in the Northeast, Brazil’s most underdeveloped region. The fight for land and the extension of labor rights to the countryside were the main
grievances raised. Brazil was facing an intense redistributive struggle (Welch 1999). g
g
gg
Economic and social instability translated into the political arena. In August
1961, the recently inaugurated President Jânio Quadros resigned, opening a serious
political crisis. Vice-President João Goulart was impeded to step in, due to a faction
of the military, who regarded him with suspicion. Goulart was the leader of the
Brazilian Labor Party (Partido Trabalhista Brasileiro, PTB), and longtime protégé
of Getúlio Vargas, a former Brazilian president and dictator. The right-wing of the
military disregarded Goulart due to his close links with unionized labor, including
Communist-led unions. A legalist movement arose though, and it was prepared to
take up arms. To avoid a civil war, the Parliament decided to vote an amendment
to the 1946 Constitution, setting up a parliamentary regime. It was also agreed
that in mid-1965 the Brazilian people would decide whether the parliamentary
system should continue (Labaki 1986). However, Goulart managed to anticipate
this plebiscite to January 1963, when a large majority (82%) voted for restoring
his powers. In spite of this, socio-economic and political instability kept increasing,
leading to Goulart’s overthrow in March 1964 by a military coup (Skidmore 2007). Goulart’s rise to power in September 1961, even though constrained by a
parliamentary straitjacket, did not avert him to bring forth his political agenda. The president defended the so-called “basic reforms” (reformas de base), e.g.,
a series of proposals focused on creating a new development framework, and
meeting social demands for greater equality in property and income (Ferreira
2012). Among several issues the ones which stood out sere the proposals for
agrarian and tax reforms, aimed at addressing problems still acute in Brazil to this
day. Public opinion and foreign policy in João Goulart’s Brazil (1961-1964) [...] However, these types of attitudes are closely related to US foreign
policy and not to the context of developing countries. The Brazilian foreign policy
cannot be framed in these terms. The debate about Brazil’s foreign policy decisions gravitates around two
conceptual frameworks: alignment versus autonomy. Alignment represents strategic
proximity to the United States in regional and international fora. Broadly speaking,
it means a soft bandwagoning line through closer relations with Washington. On
the other hand, an autonomous approach means to distance Brazil from the Unites
States via soft balancing strategies, usually seeking closer links to either defiant
global powers or Third World coalitions (Lima 1994; Vigeviani and Cepaluni
2007; and Hirst 2013). As the Brazilian literature on public opinion coherence is still in its early
stages, the controversial and polarized political context of early 1960s’ Brazil is 102 Public opinion and foreign policy in João Goulart’s Brazil (1961-1964) [...] an interesting starting point to test the argument. However, coherence is context
dependent. The meaning of coherence in one scenario can be different from that
on another one. In this sense, to understand the terms of the INESE-USIA survey,
it is necessary to look at the debates on domestic and foreign policies taking place
in Brazil in the early 1960s. Next section addresses this issue. Revista Brasileira de Política Internacional Land distribution was very unequal, with few landowners holding the best
and largest rural properties, while the majority of Brazilian peasants were landless. 103 Felipe Pereira Loureiro; Feliciano de Sá Guimarães; Adriana Schor The Brazilian tax structure was also socially unfair, as taxation over consumption
instead of property and income was predominant. This type of tax system favored
wealthier citizens, stimulating income concentration (Loureiro 2012). Goulart and
some leftist groups defended an agrarian reform in which landowners would be
compensated according to the value employed for land taxation, and not to the
(higher) market value of the land. Also, landowners would receive compensations
not in money, but in long-term bonds. Of course this was strongly opposed by the
rural elites, who saw these proposals as outright expropriation. As to the tax reform,
Goulart planned to make a complete overhaul in the tax system, increasing direct
taxes (income and property taxes), and relaxing tolls on consumption (Bandeira
2010; Skidmore, 2007). A heated debate was also taking place as to how to proceed with foreign
companies which invested in public utilities. These firms were criticized due to the
low-quality services offered and lack of investments. The companies, on the other
hand, complained they could not invest properly because of official control on
tariffs. Expropriations pursued in the state of Rio Grande do Sul in 1959 and 1962
by Goulart’s brother-in-law, governor Leonel Brizola, a radical leftist nationalist,
triggered the debate. Brizola confiscated subsidiaries of two US companies in Rio
Grande do Sul (the American Foreign and Power, AMFORP, and the International
Telephone and Telegraph, ITT), respectively, paying token compensations in
return. The companies complained bitterly, and, as Brazil needed US foreign aid,
Goulart attempted to make a compromise, but pulled back (except for ITT) after
strong domestic criticisms. In the end, AMFORP’s expropriation would be settled
only by the military regime in 1964 (Saes and Loureiro 2014). y y
y
g
Goulart also maintained a fundamental shift in direction of Brazil’s foreign
policy, which had started in the short-lived Quadros administration. The so-called
“Independent Foreign Policy” of President Quadros (Política Externa Independente,
PEI) was based on the principle that the West-East divide could not be the main
guideline for Brazil’s international insertion. Instead, what should fulfill this role
was the increasing North-South socio-economic gap. Revista Brasileira de Política Internacional Even though Brazil did not
present itself as neutralist or non-aligned, the country did sustain the pursuit of
diplomatic relations pragmatically, focusing on how they could foster economic
development, and not grounded on ideological divisions. It meant, for instance,
that Brazil should be open to trade with and to receive investment from communist
countries, and not to align automatically with the United States. Besides, PEI
claimed the need to respect other nations’ right for self-determination. In practice,
Brazil would support the struggle for independence in Asia and Africa (a support
that was also made pragmatically, as this could bring new markets), and sustain
the principle of non-intervention in international relations (Storrs 1973; Vizentini
1995; Resende 2009). The Quadros and Goulart governments pursued concrete actions to
implement the Independent Foreign Policy. In May 1961, President Quadros 104 Public opinion and foreign policy in João Goulart’s Brazil (1961-1964) [...] sent a diplomatic mission to several Eastern European countries to settle trade
agreements, and ended up with accords that increased enormously the prospect
of trade with the Soviet bloc (Hershberg 2015). In November 1961, after thirteen
years without formal contacts, Brazil and the Soviet Union resumed diplomatic
relations. In January 1962, Foreign Minister Santiago Dantas led a group of Latin
American nations at the Punta del Este Conference, held in Uruguay, against
the US initiative to expel Cuba from the Organization of the American States
(OAS) (Weis 2001). And, finally, during the Cuban Missile Crisis in October
1962, despite the fact that Brazil supported the US quarantine against Cuba, the
Goulart government did not vote at the OAS meeting in favor of measures to allow
Washington to employ force if necessary against the island (Hershberg 2004). Scholars agree that the turn of Brazil’s foreign policy was important to deteriorate
US-Brazilian relations (Weis 2001; Rabe 2012; Loureiro 2014). Therefore, it was in this context of intense social struggles, support for basic
reforms, and ruptures in the country’s foreign policy that the INESE conducted
its survey in late 1962. At that time, Brazilians had just left the ballot box of the
October 1962 general election, choosing representatives for the National Congress,
and for half of the state governors and local legislators. In January 1963, Brazilians
would vote again – this time for the plebiscite to decide over the future of the
parliamentary regime (Skidmore 2007). 6 Report, ‘The Economic and Political Climate of Opinion’, Box 216, NSF, JFKL, p. iv. As an example of
another public opinion survey conducted by INESE on behalf of the USIA, see Report, ‘Brazil in Crisis Attitudes’,
Sept. 1961, Box 11, Country Project Files, compiled ca. 1953 – 1965 (hereafter CPF), Record Group (hereafter
RG) 306, National Archives and Records Administration, College Park (hereafter NARA). Revista Brasileira de Política Internacional Hence, there was hardly a better time to
assess the citizens’ opinion on domestic issues, as these were strongly explored in
public campaigns and in the media. Foreign policy issues were also on the spot
due to the Cuban Missile Crisis. In the lights of this favorable context, one needs
to ask: what were Brazilians’ views on domestic and foreign issues? Were they
coherent or incongruent? Next section addresses these points. Felipe Pereira Loureiro; Feliciano de Sá Guimarães; Adriana Schor Brazilians were interviewed on several economic and political issues, related
to domestic and international topics. All aggregate results were presented in
percentage points. Questions that referred to domestic issues were mainly focused
on mapping out the opinion regarding the country’s economic conditions and the
support for reforms proposed by the Goulart administration. The main character
of these reforms, as discussed above, was to redistribute wealth (through land
reform and nationalization of foreign capital) and income (through a tax reform). g
p
g
A first set of questions of the 1962 survey attempts to access public opinion
on what changes should take place in Brazil’s economy and society. There were
formulated as open-ended questions, which means interviewers had freedom
to elaborate multiple word answers, and not only to choose between “yes” or
“no” (closed-ended questions). Around 20% of those interviewed pointed out
that an agrarian reform was necessary. Other alternatives, ranging from changes
in education and housing policies to radical transformations in the economic
system, represented together less than 10% of the answers – with the exception
of “reducing the cost of living”, mentioned by 15%. It should be noted that 44%
of those interviewed did not have opinion, did not answer, or believed that no
reforms were needed, which is a very significant result. The great number of urban
Brazilians who did not know or did not have a position on central public issues
could probably be seen as a consequence of the country’s high illiteracy rates (37%
in 1965) – above that of Latin America’s average (32%) (Wattles 1966, 3). The second question reported asked what Brazilians thought about “a land
reform that would divide the large agricultural properties and distribute the
land among agricultural workers”. As Table 1 shows, 68% of those interviewed
in 1962 supported land reform (a 11.5% increase in relation to 1961). Among
these, roughly 42.6% argued that landowners should receive the full worth for
their properties as compensation, while almost the same proportion (41.2%) were
in favor of more leftist-oriented reforms: landowners were either to receive only
partially (33.8%) for their properties, or to get no compensation at all (7.4%). Broadly speaking, results for 1961 and 1962 present similar patterns, with just a
minor inversion: while in 1961 stronger leftist reformists slightly prevailed (44.3%),
in 1962 moderates were in a slender advantage. The 1962 INESE-USIA survey and data analysis The report of the 1961-1962 Brazilian opinion survey conducted on behalf
of the USIA was based on a sample of 1,593 (1961 survey) and 1,327 individuals
(1962 survey). Unfortunately, it does not discuss methodological procedures. The
report only points out that the results were “based primarily on area probability
samples of adults in urban areas” (cities of 10,000 inhabitants or over), and that
the questionnaire was produced by USIA’s Survey Research Division, having the
field work conducted by the Institute for Social and Economic Studies (INESE). The INESE had already made other surveys for the USIA. The institute was chosen
exactly because it constituted, according to the USIA, an “established, professional
research organization”.6 105 Felipe Pereira Loureiro; Feliciano de Sá Guimarães; Adriana Schor Felipe Pereira Loureiro; Feliciano de Sá Guimarães; Adriana Schor Taking into account those who opposed any type of land reform (11% of
the total sample in 1961, and 8% in 1962), one reaches the conclusion that the
majority of Brazilians had some respect for property rights (at least when it comes
to land tenure), either by means of full worth payment to landowners, or by simply
allowing owners to keep their lands. However, as the redistributive impacts of
even a moderate land reform would still be significant, we classify Brazilian public
opinion on land reform as “redistributive” – a trend strengthened in 1962. 106 Public opinion and foreign policy in João Goulart’s Brazil (1961-1964) [...] Table 1. Brazilian Public Opinion on Land Reform, Nationalization
of Foreign-owned Companies, and Tax Reform, % (1961-1962)
Land Reform
Nationalization of
Foreign-owned Companies
Tax
Reform
General position
1961
1962
1961
1962
1962
Favor
61,0
68,0
16,0
36,0
56,0
Oppose
11,0
8,0
57,0
31,0
16,0
Don't care
-
8,0
-
-
12,0
No opinion
28,0
16,0
27,0
33,0
16,0
Total
100,0
100,0
100,0
100,0
100,0
Type of Compensation*
1961
1962
1961
1962
1962
Full worth
42,6
42,6
31,3
44,4
-
Part worth
34,4
33,8
37,5
36,1
-
Nothing at all
9,8
7,4
18,8
8,3
-
No opinion
13,1
16,2
12,5
11,1
-
Total
100,0
100,0
100,0
100,0
-
Number of interviewers
1.593
1.327
1.593
1.327
1.327
Source: Report, ‘The Economic and Political Climate of Opinion’, Box 216, NSF, JFKL. Note: * Among those who favored land reform (61.0% and 68.0% of the total in 1961 and 1962, respectively)
or nationalization of foreign-owned companies (16.0% and 36.0% of the total in 1961 and 1962, respectively). Table 1. Brazilian Public Opinion on Land Reform, Nationalization
of Foreign-owned Companies, and Tax Reform, % (1961-1962) Source: Report, ‘The Economic and Political Climate of Opinion’, Box 216, NSF, JFKL. Note: * Among those who favored land reform (61.0% and 68.0% of the total in 1961 and 1962, respectively)
or nationalization of foreign-owned companies (16.0% and 36.0% of the total in 1961 and 1962, respectively). A similar picture emerges when one looks at opinions on tax reform (Table 1). The survey questioned whether Brazilians supported a change in legislation to
build a progressive tax structure – this is to say, to create a tax system “which
could make the better-off pay higher taxes”. Felipe Pereira Loureiro; Feliciano de Sá Guimarães; Adriana Schor The great majority (56%) pointed out
they favored a redistributive tax reform, while only 16% defended the status quo. The remaining 28% did not care, or did not have opinion about it, which is also a
high rate, even though significantly lower than the 44% one finds in open-ended
questions. Thus, with regards to tax reform, Brazilian public opinion may also be
ordered as “redistributive”. Turning to the nationalization of foreign-owned companies, another
much-debated issue in Goulart’s Brazil, one finds an extremely complex situation
(Table 1). There was a major turn of opinion during the first two years of the
Goulart administration. While in 1961 the absolute majority of Brazilians (51%)
opposed nationalization of foreign industries, in 1962 the picture evolved into a
stalemate: roughly one-third of those interviewed either favored, opposed or did
not have an opinion on the issue, with a slight advantage for those who argued 107 Felipe Pereira Loureiro; Feliciano de Sá Guimarães; Adriana Schor for nationalization (36%). Among these, there was a standoff between those who
wanted foreign companies to receive full-worth for their properties (44.4%), and
those who either advocated the payment of part-worth compensations (36.1%) or
no compensation at all (8.3%). Therefore, we classify the Brazilian public opinion
in this regard as slight “redistributive”. g
g
In sum, the data shows a fairly coherent picture of the Brazilian public
opinion during the Goulart administration, at least when looking at domestic
issues. If it is clear that there were many Brazilians who preferred moderate, and
not radical, socio-economic changes (land reform and nationalization of foreign
companies with full compensation to owners, for instance); on the other hand,
it is also evident that the majority wished for some type of redistributive reform. Adding moderate reformers to those who stood up for radical changes, one reaches
the conclusion that a “redistributive” position predominated in domestic topics,
and it was increasingly gaining ground. The second part of the report presents results on international issues. First,
Brazilians were asked about how they perceived Cold War superpowers (the United
States and the Soviet Union), as well as the leader of the 1959 Cuban Revolution,
Fidel Castro (Table 2). The data is clear in showing that Brazilians didn’t perceive
either Fidel or the Soviet Union in a positive light. Felipe Pereira Loureiro; Feliciano de Sá Guimarães; Adriana Schor Among those who did answer
the question (roughly 60% of the total), the absolute majority had either a “bad”
or a “very bad opinion” about Fidel Castro and the Soviet Union. Interestingly, the
perceptions on Fidel Castro deteriorated sharply between those two years. While
in 1961 half (50%) of the Brazilians presented a “bad” or a “very bad opinion” on
Castro, in 1962 this proportion jumped to almost 70%. Probably the fact Castro
announced the Communist character of the 1959 Revolution in late 1961, and
that Cuba and the USSR deepened their alliance in 1962, leading to the Cuban
Missile Crisis, might explain this deterioration. g
p
When it comes to the US image in Brazil, the situation is reversed. The large
majority of Brazilians held the United States in high regard, and such proportion
increased during the two first years of Goulart’s term. Among those who did have
a stand on the United States (roughly 70% of total respondents), 63.6% in 1961
and 69.4% in 1962 presented either a “good” or a “very good opinion” about
Washington. Similarly striking, the proportion of people who did not have an
opinion about the United States was not only lower than the Soviet Union, but
also decreased in comparison to 1961 (34% to 28%). Definitely, Brazilians had a
pro-US stance in the Cold War. According to our foreign policy axis, these rates
show an “alignment” outlook. 108 Public opinion and foreign policy in João Goulart’s Brazil (1961-1964) [...] Table 2. Brazilian Public Opinion on Fidel Castro, Soviet Union,
and United States, % (1961-1962)
General position
1961
1962
1962
1961
1962
1. Very good opinion
2,0
1,0
2,0
10,0
14,0
2. Good opinion
10,0
7,0
10,0
32,0
36,0
3. Neither good nor bad
16,0
12,0
16,0
20,0
18,0
4. Bad opinion
9,0
25,0
23,0
1,0
3,0
5. Very bad opinion
19,0
20,0
12,0
3,0
1,0
No opinion
25,0
18,0
37,0
34,0
28,0
Not asked*
19,0
17,0
–
–
–
Total
100,0
100,0
100,0
100,0
100,0
Those with opinion**
1961
1962
1962
1961
1962
Good (1-2)
21,4
12,3
19,0
63,6
69,4
Neither good nor bad (3)
28,6
18,5
25,4
30,3
25,0
Bad (4-5)
50,0
69,2
55,6
6,1
5,6
Number of interviewers
1.593
1.327
1.327
1.593
1.327
Source: Report, ‘The Economic and Political Climate of Opinion’, Box 216, NSF, JFKL. Felipe Pereira Loureiro; Feliciano de Sá Guimarães; Adriana Schor Notes: * Not asked because they have never heard of Castro; ** Of those who expressed opinion on Fidel Castro
(56% and 65% of the total in 1961 and 1962, respectively), the Soviet Union (63% of the total in 1962), and
the United States (66% and 72% of the total in 1961 and 1962, respectively). Table 2. Brazilian Public Opinion on Fidel Castro, Soviet Union,
and United States, % (1961-1962) Source: Report, ‘The Economic and Political Climate of Opinion’, Box 216, NSF, JFKL. Notes: * Not asked because they have never heard of Castro; ** Of those who expressed opinion on Fidel Castro
(56% and 65% of the total in 1961 and 1962, respectively), the Soviet Union (63% of the total in 1962), and
the United States (66% and 72% of the total in 1961 and 1962, respectively). This large favorable perception of the United States among Brazilians,
however, did not translate into a broad support of alignment with the West
(Table 3). Although 45% of the respondents in 1962 said they had very great
or considerable confidence “in the ability of the United States to provide wise
leadership for other countries in dealing with present world problems”, only 30%
in 1961, and 37% in 1962 favored an alliance with Washington in international
affairs. If we exclude those that did not have a position (roughly a quarter of total
respondents), the picture is not altered significantly. Indeed, a pro-US foreign
policy stance had a large support among the population with opinion, and this
support was increasing (from 39,5% to 48,1%). However, it still lagged a little
behind those that favored a neutralist approach in foreign affairs (56,6% in 1961,
and 50,6% in 1962). This suggests, albeit slightly, a more “autonomist” approach. 109 Felipe Pereira Loureiro; Feliciano de Sá Guimarães; Adriana Schor Table 3. Brazilian Public Opinion on the Orientation
of the Country’s Foreign Policy, % (1961-1962)
Brazil should side with
1961
1962
United States
30,0
37,0
Russia
3,0
1,0
Neither
43,0
39,0
No opinion
24,0
23,0
Total
100,0
100,0
Those with opinion*
1961
1962
United States
39,5
48,1
Russia
3,9
1,3
Neither
56,6
50,6
Total
100,0
100,0
Number of interviewers
1.593
1.327
Source: Report, ‘The Economic and Political Climate of Opinion’, Box 216, NSF, JFKL. Public opinion and foreign policy in João Goulart’s Brazil (1961-1964) [...] Public opinion and foreign policy in João Goulart’s Brazil (1961-1964) [...] A more difficult issue relates to the fact that the lack of the survey’s microdata
makes it harder to state whether a given Brazilian “reformist” was also supportive
of an “autonomous” foreign approach, and vice-versa. However, there are ways to
circumscribe this limitation (but not to overcome it). For each one of the selected
domestic variables, we calculated the overlap coefficient in relation to the group of
Brazilians who favored an “autonomous” foreign policy in 1961 and 1962.7 To find
out the intersection between a given pair of variables, we developed two distinct
scenarios: a worst-case scenario, in which we assumed that all of those with “no
opinion” never overlapped; and a best-case scenario, in which we supposed that
“no opinion” respondents overlapped perfectly (always taking the lower proportion
of “no opinion” respondents as guideline). In other words, our worst-case and
best-case scenarios worked, respectively, as low and high limits of the overlapping
area of a given pair of variables. As can be seen in Table 4, in the worst-case scenario, none of the respondents
who supported an autonomist foreign approach argued for a redistributive tax
reform or for the nationalization of foreign firms, and only 4% and 17.8% (in 1961
and 1962, respectively) defended an agrarian reform. On the other hand, in the
best-case scenario, only supporters of the nationalization of foreign firms did not
overlap with “autonomists” at all. In the remaining cases, the overlap coefficient
ranged from a low of 0.48 (tax reform) to a high of 0.76 (agrarian reform in 1962). The pair agrarian reform/autonomous foreign policy shows that between
17.8% and 76% of “autonomists” Brazilians in 1962 also supported agrarian
reform. Although one could argue that the actual overlapping might have been
something in between (47.5%), it is reasonable to expect that support was closer
to the best-case than to the worst-case scenario due to the existence of heuristic
shortcuts used by voters. That is, these shortcuts have probably reinforced
overlapping perceptions because an individual’s interpretation of policy issues
usually depends on preexisting belief systems and media cues – in this case the
domestic debate promoted by the government about redistributive reforms in
Brazil. 7 The overlap coefficient of a pair of variables (X,Y) is calculated by dividing the size of the intersection between
these variables by the size of the smaller variable. See the formula employed on Table 4. Felipe Pereira Loureiro; Feliciano de Sá Guimarães; Adriana Schor Notes: * Of those who expressed opinion on the orientation of Brazil’s foreign policy, e.g., pro-United States
(30% and 37% of the total in 1961 and 1962, respectively), pro-Soviet Union (3% and 1% of the total in 1961
and 1962, respectively), and neither countries (43% and 39% of the total in 1961 and 1962, respectively). Table 3. Brazilian Public Opinion on the Orientation
of the Country’s Foreign Policy, % (1961-1962) Source: Report, ‘The Economic and Political Climate of Opinion’, Box 216, NSF, JFKL. Notes: * Of those who expressed opinion on the orientation of Brazil’s foreign policy, e.g., pro-United States
(30% and 37% of the total in 1961 and 1962, respectively), pro-Soviet Union (3% and 1% of the total in 1961
and 1962, respectively), and neither countries (43% and 39% of the total in 1961 and 1962, respectively). Is it possible to argue that these numbers show an incoherent Brazilian
public opinion? We believe there are elements that suggest otherwise. Two
aspects in particular may explain why the United States enjoyed a good image
among Brazilians, even though the majority of the country’s citizens defended
redistributive domestic policies, and a slight majority argued for a neutralist foreign
policy approach. First, most citizens stood up for property rights, that is, many
advocated either the implementation of moderate structural reforms (reforms that
embodied a full-worth compensation to property owners), or the maintenance of
the status quo (no reforms at all). This is coherent with positions that expressed
admiration for the United States and disapproval of communist countries. Second,
in the early 1960s, Brazilian policymakers presented the country’s “Independent
Foreign Policy” (PEI) not as anti-US or anti-West. In fact, the PEI was defended
as a pragmatic – and, thus, supposedly non-ideological – response to Brazil’s dire
economic strains. Even though Brazil continued to uphold Western values and to
belong to the West, it needed to expand relations worldwide to foster economic
development (Storrs 1973; Vizentini 1995). In this sense, it is not incoherent that
most Brazilians had Washington in high esteem and, at the same time, a slight
majority supported a neutral (e.g., pragmatic) approach in foreign affairs. 110 Public opinion and foreign policy in João Goulart’s Brazil (1961-1964) [...] In other words, if a given Brazilian did not have opinion on redistributive
domestic issues, the probability that this same person would also have no opinion
on (more distant and complicated) international issues was high, thus making the
real overlap coefficient large. 111 Felipe Pereira Loureiro; Feliciano de Sá Guimarães; Adriana Schor Table 4. Brazil’s Public Opinion Overlap Coefficients for Redistributive
Domestic Policies and Autonomous Foreign Policy (1961-1962)†
Worst case*
Best case**
1961
1962
1961
1962
Land Reform and Autonomy
0,09
0,18
0,65
0,77
Tax Reform and Autonomy
–
0,00
–
0,46
Nationalization and Autonomy
0,00
0,00
0,00
0,00
Source: Report, ‘The Economic and Political Climate of Opinion’, Box 216, NSF, JFKL. Notes: † Overlap Coefficient: Overlap (X,Y) = |X ∩ Y| / min (|X|, |Y|); * No overlap between ‘no opinion’ or
‘don’t care’ respondents in the variables considered; ** Complete overlap between ‘no opinion’ or ‘don’t care’
respondents in the variables considered. Table 4. Brazil’s Public Opinion Overlap Coefficients for Redistributive
Domestic Policies and Autonomous Foreign Policy (1961-1962)† Table 4. Brazil’s Public Opinion Overlap Coefficients for Redistributive
Domestic Policies and Autonomous Foreign Policy (1961-1962)† Source: Report, The Economic and Political Climate of Opinion , Box 216, NSF, JFKL. Notes: † Overlap Coefficient: Overlap (X,Y) = |X ∩ Y| / min (|X|, |Y|); * No overlap between ‘no opinion’ or
‘don’t care’ respondents in the variables considered; ** Complete overlap between ‘no opinion’ or ‘don’t care’
respondents in the variables considered. Additional evidence also strengths this conclusion. One finds similar trends
looking at other public opinion surveys organized by the USIA in Brazil in the early
1960s. In January 1961, for example, a survey conducted in the country’s greatest
urban areas (the cities of Rio de Janeiro, São Paulo, Porto Alegre, Belo Horizonte,
and Recife) found out that 61% of all respondents supported agrarian reform,
while 39% thought Brazil should pursue a neutralist foreign approach (28% and
24% did not have opinion, respectively). In this case, the overlap coefficient ranges
from zero (worst-case) to 0.61 (best-case). Despite the more limited geographical
scope of this survey, it is highly significant the fact that, even before Goulart came
into power – bringing over proposals for agrarian reform, and intensifying the
country’s Independent Foreign Policy – there was already a significant support
for both policies in Brazil’s greatest urban areas.8 8 Report, ‘P/61 L, Study Opinion on Selected International and National Issues’, Jan. 1961, Box 11, CPF,
RG 306, NARA. Conclusions Brazil’s public opinion in the Goulart administration (1961-1964), at least
when it comes to urban centers and to the 1961-1962 period, presented some
degree of coherence regarding domestic and foreign issues. Overall, this matches the
recent conclusions brought out by the literature on the evolution of public opinion
in developed democracies, particularly in the United States. The 1962 survey poll
conducted by the INESE on behalf of the USIA suggests that Brazilians sponsored
a reformist position domestically and, to a lesser extent, argued for an autonomist
position in the international arena. Despite the lack of the survey’s microdata, we
argued that it is reasonable to expect that a large number of Brazilians reformist
also stood up for an autonomous foreign policy. This finding is supported by three
groups of evidences: first, the large overlap coefficients presented by the two major
proposals of redistributive reforms (particularly agrarian reform); second, the fact 112 Public opinion and foreign policy in João Goulart’s Brazil (1961-1964) [...] that is reasonable to assume that ‘no opinion’ respondents across domestic and
international variables tended to overlap; and, finally, the similar patterns found
in other (but less broad) surveys organized by the USIA in early 1960s Brazil. g
There are, however, two important constraints in this picture: first, if the
trend suggested in the 1961-1962 period is correct (e.g., greater support for
agrarian reform and lower support for neutralism), then Brazilians seemed to be
heading towards incoherence on the eve of the 1964 military coup. And, second,
even though many wanted redistributive reforms, a significant portion wished to
implement moderate changes, respecting property rights. This, along with the
pragmatic approach brought in by Goulart’s Independent Foreign Policy (PEI),
might explain why Brazilians who had a good opinion about the United States
could also have supported an autonomous foreign approach. pp
g
pp
The moderate perspective on domestic reforms and the country’s good
image of the United States tend to be neglected by the literature on public
opinion in Goulart’s Brazil. Scholars such as Rodrigo Motta (2014), who have
employed Brazil’s own survey pools (IBOPE), have argued that Brazilians wanted
redistributive reforms in the early 1960s (which is broadly correct), without posing
attention, however, to the fact that a significant portion of those who wished for
reforms wanted to implement them moderately. Bibliographic references Abreu, Marcelo de Paiva. The Brazilian Economy, 1930-1980. In The Cambridge History of
Latin America, by Leslie Bethell, 283-393, vol. 9. New York: Cambridge University Press, 2008. Albuquerque, et al. Os impactos da crise financeira sobre a opinião pública brasileira. Revista
Brasileira de Comércio Exterior ano 8, no. 58 (1999): 18-20. Aldrich, John et al. Foreign Policy and the Electoral Connection. Annual Review of Political
Science 9 (2006): 477-502. Aldrich, John et al. Foreign Affairs and Issue Voting: Do Presidential Candidates Waltz Before
a Blind Audience? American Political Science Review 83 (1989): 123-141. Almeida, Maria Hermínia et al. O Brasil, as Américas e o mundo: opinião pública e política externa
2010-2011. São Paulo: IRI-USP, 2012. Almond, Gabriel. The American People and Foreign Policy. New York: Harcourt Brace, 1950. Bandeira, Luiz Alberto Moniz. O Governo João Goulart: As Lutas Sociais no Brasil, 1961-1964. 8. São Paulo: UNESP, 2010. Baum, Matthew, Philip Potter. The Relationships Between Mass Media, Public Opinion,
and Foreign Policy: Toward a Theoretical Synthesis. Annual Review of Political Science
11 (2008): 39-65. Bressian, Regiane. A integração sul-americana e a superação da pobreza: uma abordagem pela
percepçao das elites. São Paulo: PhD Dissertation, University of São Paulo, 2012. Burstein, Paul. The Impact of Public Opinion on Public Policy: A Review and an Agenda. Political Research Quarterly 56, no. 1 (2003): 29-40. Casarões, Guilherme. A mídia e a política externa no Brasil de Lula. Austral: Revista Brasileira
de Estratégia e Relações Internacionais 1, no. 2 (2012): 211-236. Castillo, R. et al. To Know or not to Know? Realist and Liberal Theories on Foreign Affairs and
Public Opinion in Latin America. Latin America Policy, 6, no. 1 (2015), 2-18. Culp, Julian, Johannes Plagemann. Hooray for Global Justice? Emerging Democracies in a
Multipolar World. GIGA Working Papers, 242 (2013): 1-24. Delli-Carpini, Michael, Scott Keeter. What Americans Know about Politics and Why It Matters. New Haven, CT: Yale University Press, 1996. Faria, Carlos Aurélio Pimenta de. O Itamaraty e a política externa brasileira: do insulamento à
busca de coordenação dos atores governamentais e de cooperação com os agentes societários. Contexto Internacional 34, no. 1 (2012): 311-355. Faria, Carlos Aurélio Pimenta de. Opinião pública e política externa: insulamento, politização
e reforma na produção da política exterior do Brasil. Revista Brasileira de Política Internacional
51, no. 2 (2008): 80-97. Feldman, Stanley. Structure and Consistency in Public Opinion: the Role of Core Beliefs and
Values. Felipe Pereira Loureiro; Feliciano de Sá Guimarães; Adriana Schor Felipe Pereira Loureiro; Feliciano de Sá Guimarães; Adriana Schor Conclusions At the same time, authors who
have looked into Brazilian media and Federal Legislative, such as Mansur (2004),
have argued that a strong polarized debate was taking place between those that
she labels as “liberal-westerns” (liberais-ocidentalistas) and “national-revisionists”
(nacional-revisionistas). According to Mansur, while liberals were strongly anti-
communist, pro-market, and in favor of a strict alignment with Washington;
nationalists, on the other hand, favored strong redistributive reforms domestically
and an autonomous foreign approach in world affairs, pending, however, more
to the Soviet bloc than to the United States if it were necessary to choose sides. y
As it became clear, our findings show that the Brazilian public opinion
presented a very different picture of that portrayed by the country’s media and
political representatives. Public opinion on foreign policy issues in Goulart’s Brazil
could well be characterized as polarized, as Mansur suggests, but the relevant
polarization seems to have happened not between those that were pro-US and
pro-USSR, but those that were pro-Washington and pro-neutralism. Communist
countries and communist personalities, such as Fidel Castro, were not well regarded
by the Brazilian population in general. If this gulf between public opinion and
Brazilian political representatives in the early 1960s were in fact true, then major
interesting points are brought into the debate on the reasons behind the fall of
Brazil’s post-war democracy: why wasn’t the country’s political system able to
tune itself to domestic public opinion? Why have Brazilian politicians and media
been incapable of bringing this support for moderate reforms into the fore? Even
though these are points that surpass the boundaries of this paper, the answer does
not seem to rely on the incoherence of the Brazilian public opinion. 113 Bibliographic references American Journal of Political Science 32, no. 2 (1998): 416-440. Fenton, Laurence. Palmerston and The Times: Foreign Policy, the Press and Public Opinion in
Mid-Victorian Britain. London and New York: I.B. Tauris, 2012. 114 Public opinion and foreign policy in João Goulart’s Brazil (1961-1964) [...] Ferreira, Jorge. João Goulart. Uma Biografia. Rio de Janeiro: Civilização Brasileira, 2012. Ferreira, Jorge. João Goulart. Uma Biografia. Rio de Janeiro: Civilização Brasileira, 2012. Ferreira, Jorge, Ângela Maria Gomes. 1964 : O golpe que derrubou um presidente, pôs fim ao
regime democrático e instituiu a ditadura militar no Brasil. Rio de Janeiro: Record, 2014. Ferreira, Jorge, Ângela Maria Gomes. 1964 : O golpe que derrubou um presidente, pôs fim ao
regime democrático e instituiu a ditadura militar no Brasil. Rio de Janeiro: Record, 2014. Gonzalez, Guadalupe, et al. Mexico, the Americas, and the World, 2012-2013. Mexico, D.F.:
CIDE, 2013. Hershberg, James. Quadros, Foreign Aid (and Trade), and the Normalization of Brazilian-
Soviet-bloc Relations, 1961: The “Itinerant Ambassador” and the Explosion Over Germany. Paper Presented at the Annual Meeting of the Society for Historians of American Foreign Relations
(SHAFR), 2015. Hershberg, James. The United States, Brazil, and the Cuban Missile Crisis, 1962 (Part II). Journal of Cold War Studies 6, no. 3 (2004): 5-67. Hirst, Mônica. Understanding Brazil-United States Relations. Brasília: Fundação Alexandre de
Gusmão, 2013. Holsti, Ole. Public Opinion and American Foreign Policy. Ann Arbor: University of Michigan
Press, 2004. Holsti, Ole. Public Opinion and Foreign Policy: Challenges to the Almond-Lippmann
Consensus Mershon Series: Research Programs and Debates. International Studies Quarterly
36, no. 4 (1992): 439-466. Holzhacker, Denilze. Atitudes e Percepções das Elites e População sobre a Política Externa Brasileira
nos anos 90. São Paulo: PhD Dissertation, University of São Paulo, 2006. Hurwitz, Jon, Mark Peffley. How Are Foreign Policy Attitudes Structured? A Hierarchical
Model. The American Political Science Review 81, no. 4 (1987): 1099-1120. Hurwitz, Jon, Mark Peffley. Public Images of the Soviet Union: The Impact on Foreign Policy
Attitudes. The Journal of Politics 52, no 1 (1990): 3-28. Iyengar, Shanto, Donald Kinder. News That Matters. Chicago: University of Chicago Press, 1987. Jacobs, Lawrence, Robert Shapiro. Politicians Don’t Pander: Political Manipulation and the Loss
of Democratic Responsiveness. Chicago: University of Chicago Press, 2000. Kapur, Devesh. Public Opinion and Indian Foreign Policy. India Review 8, no. 3 (2009): 286-305. Kennedy, Ryan, Matt Dickenson. Turkish Foreign Policy and Public Opinion in the AKP Era. Bibliographic references Foreign Policy Analysis 9, no. 2 (2012): 171-188. Krosnick, Jon, Donald Kinder. Altering the Foundations of Support for the President Through
Priming. American Political Science Review 84, no. 2 (1990): 497-512. Kull, Steven. Opportunities for Bipartisan Consensus - What Both Republicans and Democrats
Want in US Foreign Policy. World Public Opinion Report, 2007: 1-31. Kusnitz, Leonard. Public Opinion and Foreign Policy: America’s China Policy, 1949-1979. Westport, CT: Greenwood Press, 1984. Labaki, Almir. A crise da renúncia e a solução parlamentarista. São Paulo: Brasiliense, 1986. Lessa, Mônica, Leandro Gavião. Política externa, mídia e propaganda nos governos Lula da
Silva (2003-2010). In A política externa brasileira na Era Lula: um balanço, by Adriano Freixo
et al. Rio de Janeiro: Apicuri, 2011. 115 Felipe Pereira Loureiro; Feliciano de Sá Guimarães; Adriana Schor Lima, Maria Regina. A Política Externa Brasileira e os Desafios Sul-Sul. Revista Brasileira de
Política Internacional 48, no. 1 (2005): 24-59. Lima, Maria Regina. Ejes analíticos y conflicto de paradigmas en la política exterior brasileña. América Latina Internacional 1 (1994): 27-46. Loureiro, Felipe. Empresários, Trabalhadores e Grupos de Interesse: a Política Econômica nos
Governos Jânio Quadros e João Goulart (1961-1964). São Paulo: PhD Dissertation, University
of São Paulo, 2012. Loureiro, Felipe. The Alliance For or Against Progress? US–Brazilian Financial Relations in the
Early 1960s. Journal of Latin American Studies 46, no. 2 (2014): 323-351. Loureiro, Felipe. Strikes in Brazil during the Government of João Goulart. Candian Journal of
Latin American and Caribbean Studies. Forthcoming (2016). Lowi, Theodore J. Four Systems of Policy, Politics, and Choice. Public Administration Review
32, no 4 (1972): 298-310. Maggiotto, Michael, e Eugene Wittkopf. American Public Attitudes Toward Foreign Policy. International Studies Quarterly 25, no. 4 (1981): 601-631. Manzur, Tânia. Opinião Pública e Política Exterior do Brasil (1961-1964). Curitiba: Juruá, 2009. Manzur, Tânia. Opinião pública e política externa do Brasil do Império a João Goulart: um
balanço historiográfico. Revista Brasileira de Política Internacional 42, no. 1 (1999): 30-61. Motta, Rodrigo Patto Sá. O golpe de 1964 e a ditadura nas pesquisas de opinião. Revista Tempo
20 (2014): 1-21. Nau, Henry, Matthew Baum. Foreign Policy Worldviews and US Standing in the World. APSA
2012 Annual Meeting Paper (2012). Page, Benjamin, Marshall Bouton. The Foreign Policy Disconnect: What Americans Want from
Our Leaders But Don’t Get. Chicago: University of Chicago Press, 2006. Page, Benjamin, Robert Shapiro. The Rational Public: Fifty Years of Trends in American Policy. Bibliographic references Chicago: University of Chicago Press, 1992. Popkin, Samuel. The Reasoning Voter. New York: University of Chicago Press, 1994. Rabe, Stephen G. The Killing Zone. The United States Wages War on Latin America. 1. New
York: Oxford University Press, 2012. Reis Filho, Daniel Aarão. Ditadura e democracia no Brasil: do golpe de 1964 à Constituição de
1988. Rio de Janeiro: Zahar, 2014. Resende, Carlos Augusto. Política Externa Independente: as relações com os Estados Unidos na
busca por autonomia. Brasília: PhD Dissertation, University of Brasília, 2009. Saes, Alexandre, Felipe Loureiro. What developing countries’ past energy policies can tell us
about energy issues today? Lessons from the expropriation of American Foreign and Power in
Brazil (1959–1965). Utilities Policy 29 (2014): 36-43. Skidmore, Thomas. Politics in Brazil, 1930-1964: An Experiment in Democracy. Oxford New
York: Oxford University Press, 2007. 116 Public opinion and foreign policy in João Goulart’s Brazil (1961-1964) [...] Sniderman, Paul. The New Look in Public Opinion Research. In Political Science: The State
of the Discipline, by A. W. Finifter, 219-46. Washington, D.C.: American Political Science
Association, 1993. Souza, Amaury. A Agenda Internacional do Brasil: um estudo sobre a comunidade brasileira de
política externa. Rio de Janeiro: Centro Brasileiro de Relações Internacionais, 2002. Souza, Amaury. O Brasil na Região e no Mundo: Percepções da Comunidade Brasileira de Política
Externa. Rio de Janeiro: Centro Brasileiro de Relações Internacionais, 2008. Storrs, Keith. Brazil’s Independent Foreign Policy, 1961-1964: Background, Tenets, Linkage to
Domestic Politics, and Aftermath. Ithaca, NY: PhD Dissertation, Cornell University, 1973. Thompson, J. M. The Impact of Public Opinion on Theodore Roosevelt’s Foreign Policy. Cambridge,
UK: PhD Dissertation, University of Cambridge, 2010. Verba, Sidney et al. Public Opinion and the War in Vietnam. American Political Science Review
61 (1967): 317-33. Vigeviani, Tullo, Gabriel Cepaluni. A política externa de Lula da Silva: a estratégia da autonomi
pela diversificação. Contexto Internacional 29, no. 2 (2007): 273-335. Vizentini, Paulo Fagundes. Relações Internacionais e Desenvolvimento: o Nacionalismo e a Política
Externa Independente, 1951-1964. Petrópolis: Vozes, 1995. Wattles, Marshall. Literacy and Development in Latin America. Paris, UNESCO/EAO,
(1966): 1-16. Weis, Michael. The Twilight of Pan-Americanism: The Alliance for Progress, Neo-Colonialism,
and Non-Alignment in Brazil, 1961-1964. The International History Review 23, no. 2 (2001):
322-344. Welch, Cliff. The Seed Was Planted. The São Paulo Roots of Brazil’s Rural Labor Movement,
1924–1964. University Park: The Pennsylvania State University Press, 1999. Whitehead, Laurence. Emerging Market Democracies: East Asia and Latin America. Bibliographic references Baltimore
and London: The John Hopkins University Press, 2002. Wilkinson, S. Perceptions of Public Opinion. British Foreign Policy Decisions About Nazi Germany,
1933-1938. Oxford: PhD Dissertation, University of Oxford, 2000. Zaller, John. The Nature and Origins of Mass Opinion. Cambridge, UK: Cambridge University
Press, 1992. Abstract This paper analyses public opinion during the João Goulart government in Brazil (1961-1964),
focusing on public perceptions on domestic and foreign policies. We employ a recently declassified
public opinion survey conducted on behalf of United States Information Agency (USIA) in
urban areas. We found that the Brazilian public opinion was somewhat coherent, supporting
redistributive reforms domestically and a neutralist approach in foreign affairs. Keywords: Brazilian Foreign Policy; João Goulart’s administration; Public Opinion; Domestic
issues in Brazil. Acknowledgements This research has been funded by Fundação de Amparo à Pesquisa do Estado de São Paulo
(FAPESP). We thank Denise Holzhacker, Fernando Mourón, two anonymous reviewers of the
RBPI, and seminar participants at the Institute of International Relations of the University
of São Paulo for their valuable comments and suggestions. Faults and limitations are our
responsibility. 117 Felipe Pereira Loureiro; Feliciano de Sá Guimarães; Adriana Schor Resumo O artigo analisa a opinião pública durante o governo de João Goulart no Brasil (1961-1964),
focando nas percepções sobre políticas nacional e internacional. Emprega-se uma pesquisa
de opinião pública recentemente liberada feita a pedido da United States Information Agency
(USIA) em áreas urbanas. Conclui-se que a opinião pública era razoavelmente coerente, apoiando
reformas domésticas distributivas e uma abordagem neutralista em questões internacionais. Palavras-chave: Política Externa Brasileira; Governo João Goulart; Opinião Pública; Assuntos
domésticos no Brasil; Received: September 29, 2015
Accepted: November 30, 2015 Received: September 29, 2015
Accepted: November 30, 2015 118
|
https://openalex.org/W2582331105
|
https://digital.csic.es/bitstream/10261/148084/1/Aparicio_et_al_2017.pdf
|
English
| null |
Horizontal and Vertical Distributions of Transparent Exopolymer Particles (TEP) in the NW Mediterranean Sea Are Linked to Chlorophyll a and O2 Variability
|
Frontiers in microbiology
| 2,017
|
cc-by
| 10,064
|
Horizontal and Vertical Distributions
of Transparent Exopolymer Particles
(TEP) in the NW Mediterranean Sea
Are Linked to Chlorophyll a and O2
Variability Eva Ortega-Retuerta*, Maria M. Sala, Encarna Borrull, Mireia Mestre, Fran L. Aparicio,
Rachele Gallisai, Carolina Antequera, Cèlia Marrasé, Francesc Peters, Rafel Simó and
Josep M. Gasol Eva Ortega-Retuerta*, Maria M. Sala, Encarna Borrull, Mireia Mestre, Fran L. Aparicio,
Rachele Gallisai, Carolina Antequera, Cèlia Marrasé, Francesc Peters, Rafel Simó and
Josep M. Gasol Biologia Marina i Oceanografia, Consejo Superior de Investigaciones Científicas, Institut de Ciències del Mar, Barcelona,
Spain ORIGINAL RESEARCH
published: 31 January 2017
doi: 10.3389/fmicb.2016.02159 Keywords: transparent exopolymer particles, chlorophyll a, bacteria, carbon, Mediterranean Sea Edited by:
Kai Ziervogel,
University of New Hampshire, USA Reviewed by:
Lars-Eric Heimbürger,
Mediterranean Institute of
Oceanography (CNRS), France
Daniel Conrad Ogilvie Thornton,
Texas A&M University, USA
Astrid Anne-Marie Gärdes,
Leibniz Center for Tropical Marine
Ecology, Germany *Correspondence:
Eva Ortega-Retuerta
ortegaretuerta@icm.csic.es Specialty section:
This article was submitted to
Aquatic Microbiology,
a section of the journal
Frontiers in Microbiology Specialty section:
This article was submitted to
Aquatic Microbiology,
a section of the journal
Frontiers in Microbiology Specialty section:
This article was submitted to
Aquatic Microbiology,
a section of the journal
Frontiers in Microbiology
Received: 12 September 2016
Accepted: 22 December 2016
Published: 31 January 2017 Received: 12 September 2016
Accepted: 22 December 2016
Published: 31 January 2017 Transparent Exopolymer Particles (TEP) are relevant in particle and carbon fluxes in the
ocean, and have economic impact in the desalination industry affecting reverse osmosis
membrane fouling. However, general models of their occurrence and dynamics are not
yet possible because of the poorly known co-variations with other physical and biological
variables. Here, we describe TEP distributions in the NW Mediterranean Sea during
late spring 2012, along perpendicular and parallel transects to the Catalan coast. The
stations in the parallel transect were sampled at the surface, while the stations in the
perpendicular transect were sampled from the surface to the bathypelagic, including the
bottom nepheloid layers. We also followed the short-term TEP dynamics along a 2-day
cycle in offshore waters. TEP concentrations in the area ranged from 4.9 to 122.8 and
averaged 31.4 ± 12.0 µg XG eq L−1. The distribution of TEP measured in transects
parallel to the Catalan Coast correlated those of chlorophyll a (Chla) in May but not
in June, when higher TEP-values with respect to Chla were observed. TEP horizontal
variability in epipelagic waters from the coast to the open sea also correlated to that of
Chla, O2 (that we interpret as a proxy of primary production) and bacterial production
(BP). In contrast, the TEP vertical distributions in epipelagic waters were uncoupled
from those of Chla, as TEP maxima were located above the deep chlorophyll maxima. The vertical distribution of TEP in the epipelagic zone was correlated with O2 and BP,
suggesting combined phytoplankton (through primary production) and bacterial (through
carbon reprocessing) TEP sources. However, no clear temporal patterns arose during
the 2-day cycle. In meso- and bathypelagic waters, where phytoplanktonic sources are
minor, TEP concentrations (10.1 ± 4.3 µg XG eq l−1) were half those in the epipelagic, but
we observed relative TEP increments coinciding with the presence of nepheloid layers. These TEP increases were not paralleled by increases in particulate organic carbon,
indicating that TEP are likely to act as aggregating agents of the mostly inorganic particles
present in these bottom nepheloid layers. Study Site and Sampling y
p
g
Samples were taken during the cruises NEMO1, NEMO2, and
SUMMER2 in Mediterranean waters between the Catalan Coast
and north of Majorca Island on board the Spanish RV “García del
Cid” (Figure 1). Transects parallel to the Catalan coast (following
the bathymetry line at 40 m bottom depth) between Barcelona
and Blanes were conducted in May 10th (transect 1) and June
11th (transect 2, Figure 1). During these transects, surface (2 m)
samples were taken every hour from the underway continuous
flow with the ship moving at ∼7 knots, so that each sample
was taken approximately at every 12 km. A coast-to-offshore
transect was performed during NEMO1, from May 11th to 20th,
including stations located in the shelf (Stations 1 and 2), slope
(stations 3, 4, and 5) and basin (stations 6, 8, and 9). Station 7
was sampled during NEMO2, 1 month later (June 12th). Water
samples in these transects were collected using a rosette (12
Niskin bottles with external spring, 12 L each) coupled to a Sea-
Bird Conductivity-Temperature-Depth profiler, a WET Labs C-
Star transmissometer and a SeaPoint optical backscatter sensor. Up to six depths were sampled from each station, from surface
to bottom (down to 2300 m) waters including the surface, the
O2 maximum, the deep chlorophyll maximum (DCM) when
present, mesopelagic waters, and bottom nepheloid layers. y
TEP were first observed, and have mostly been described,
associated with phytoplankton blooms, in the field (Alldredge
et al., 1993; Van Oostende et al., 2012) or in mesocosms and
controlled chambers (Engel et al., 2015). From these studies
we know that phytoplankton are a major source of TEP and
TEP precursors in the sea. However, the relationship between
phytoplankton and TEP varies depending on phytoplankton
composition and physiology (Passow, 2002; Klein et al.,
2011), and environmental variables such as nutrient availability
(Mari et al., 2001), turbulence (Pedrotti et al., 2010), or UV
irradiation (Ortega-Retuerta et al., 2009a). Therefore, even when
phytoplankton are the likely main source for TEP, this is not
necessarily translated into predictable relationships in the field
between TEP and chlorophyll a (Chla), the most used proxy for
phytoplankton biomass or production. Elucidating the sources of
variability in the TEP-Chla relationships would help predicting
the occurrence and dynamics of TEP in the ocean. Citation: Ortega-Retuerta E, Sala MM,
Borrull E, Mestre M, Aparicio FL,
Gallisai R, Antequera C, Marrasé C,
Peters F, Simó R and Gasol JM (2017)
Horizontal and Vertical Distributions of
Transparent Exopolymer Particles
(TEP) in the NW Mediterranean Sea
Are Linked to Chlorophyll a and O2
Variability. Front. Microbiol. 7:2159. doi: 10.3389/fmicb.2016.02159 January 2017 | Volume 7 | Article 2159 Frontiers in Microbiology | www.frontiersin.org 1 TEP in the Mediterranean Sea Ortega-Retuerta et al. INTRODUCTION TEP concentrations (up to 11,000 µg Xeq. L−1 in surface
waters) were observed in Adriatic Sea samples (Passow, 2002). Exceptionally high in the Mediterranean Sea are the relative
TEP concentrations with respect to Chla concentrations; higher
TEP/Chla ratios than in other ocean basins have been taken to
suggest that TEP are an important fraction of the particulate
organic matter pool, and likely important drivers of carbon and
particle fluxes in this oligotrophic sea. Transparent Exopolymer Particles (TEP) are defined as a
subclass of gel-like organic particles, mainly composed by acidic
polysaccharides, that are stainable with Alcian Blue (Alldredge
et al., 1993). These particles are widespread in aquatic ecosystems,
and their study in the ocean has biogeochemical and applied
interests. Due to their high stickiness, TEP act as gluing agents
for other particles to form larger aggregates susceptible to sink
in the water column, hence stimulating the biological carbon
pump (Passow et al., 2001; Burd and Jackson, 2009). However,
TEP themselves have low density, and when unballasted, they
can ascend through the water column (Azetsu-Scott and Passow,
2004) and accumulate in the sea surface microlayer (Wurl et al.,
2011) where they can constitute a major source of primary
aerosols (Orellana et al., 2011). The study of TEP has also gained
interest in the water desalination industry since they are major
agents of reverse osmosis membrane fouling (Berman, 2013). Given the ecological and economic relevance of TEP, there is a
need to improve the knowledge about how these substances are
distributed in the field and what factors affect their dynamics. Here, we report for the first time TEP distributions in the
Catalan Sea (NW Mediterranean). Our specific goals were: (1)
to determine the potential drivers of TEP from a wide range of
physicochemical and biological variables and (2) to examine the
variability in the TEP-Chla relationship across multiple spatial
and temporal scales. Frontiers in Microbiology | www.frontiersin.org Study Site and Sampling In addition to the role of phytoplankton, there are other
sources of TEP in the sea, such as macroalgae (Thornton,
2004) or zooplankton (Prieto et al., 2001). Also bacteria are
known to modify TEP distributions in the sea in various ways:
They colonize and degrade TEP that are released by other
organisms (Bar-Zeev et al., 2011; Taylor et al., 2014), thus acting
as TEP sinks. Bacteria can also directly release TEP (Ortega-
Retuerta et al., 2010) so they constitute TEP sources themselves. Finally, bacterial interactions with phytoplankton mediate TEP
release (Van Oostende et al., 2013) and induce changes on their
formation rates and properties such as their stickiness (Rochelle-
Newall et al., 2010). The relative importance of these mechanisms
governing the TEP dynamics in aquatic habitats, specifically in
the Mediterranean Sea, remains unexplored. The
2-day
lagrangian
study
(SUMMER2
cruise)
was
conducted aboard R/V “García del Cid” from 22nd to 24th
May at ca. 45 nautical miles from the coast, within the core of
a cyclonic eddy over a water-column depth of ca. 2000 m. A
Lagrangian drifter was deployed to track the movement of the
upper 15-m water layer. Each drifter consisted of a spherical
floatable enclosure that contained a GPS and an emitter, from
which 10 m cylindrical drogues hanged 5 m below the sphere. The drifters sent their position every 30 min, and all ship
operations were conducted next to them. Samples were taken
with the rosette every 4 h at six depths from surface to 200 m. Chemical and Biological Analyses Chemical and Biological Analyses
TEP were analyzed following the colorimetric method proposed
by Passow and Alldredge (1995). Samples (250–500 mL) were
filtered through 25 mm diameter 0.4 µm pore size Polycarbonate
filters (DHI) at low pressure (100 mm Hg). The filters were The published information on TEP distributions in the
Mediterranean Sea is particularly scarce (Prieto et al., 2006;
Ortega-Retuerta et al., 2010; Bar-Zeev et al., 2011). The few
published studies, however, concur in that TEP stocks are
high when compared to other oceans. For instance, maximum January 2017 | Volume 7 | Article 2159 2 Ortega-Retuerta et al. TEP in the Mediterranean Sea FIGURE 1 | Study area. Blue symbols, Coastal transects. symbols,
Coast-to-open sea transect. Green symbol, 2-day cycle. Analyses of dissolved inorganic nutrient concentrations
[nitrate (NO3), nitrite (NO2), phosphate (PO4), and silicate
(SiO2)], were done by standard segmented flow analyses with
colorimetric detection (Hansen and Grasshoff, 1983) using an
Seal Analytical AA3 High Resolution AutoAnalyzer. Particulate organic carbon (POC) was measured by filtering
1000 mL of seawater on pre-combusted GF/F glass fiber filters
(4 h, 450◦C). The filters were frozen in liquid nitrogen and kept
at −80◦C until analysis. Prior to analysis, the filters were dried at
60◦C for 24 h. Then the filters were dried again and analyzed with
a C:H:N autoanalyser (Perkin-Elmer 240). For bacterial abundance samples, 1.8 ml were preserved with
1% paraformaldehyde + 0.05% glutaraldehyde (final conc.) and
frozen in liquid nitrogen until processed in the lab. Bacterial
abundance (BA) was analyzed by flow cytometry (FACSCalibur
cytometer, Becton and Dickinson) after staining with SYBRGreen
I (Molecular probes). Bacteria were detected by their signature in
a plot of side scatter vs. FL1 (green fluorescence) as explained in
Gasol and del Giorgio (2000). FIGURE 1 | Study area. Blue symbols, Coastal transects. symbols,
Coast-to-open sea transect. Green symbol, 2-day cycle. Bacterial Production (BP) was estimated using the 3H-leucine
incorporation method described by Kirchman et al. (1985). Three 1.2-mL aliquots and two trichloroacetic acid (TCA)-
killed controls (5% final concentration) of each sample were
incubated with 40 nmol L−1,3H-leucine (epipelagic samples) or
80 nmol L−1,3H-leucine (meso- and bathy-pelagic samples). The
incubations were carried out in a water bath at in situ temperature
in the dark. The incorporation was stopped by adding cold
TCA (5% final concentration) to the vials, and samples were
kept at −20◦C until processing as described by Smith and
Azam (1992). Chemical and Biological Analyses Radioactivity was then counted on a Beckman
scintillation counter. Leucine incorporation rates were converted
into carbon production using the conversion factor of 1.55 kg C
produced per mole of leucine incorporated and considering no
isotope dilution (Simon and Azam, 1989). Given that TEP are frequently enriched in fucose (Zhou et al.,
1998), we determined fucosidase activity using a fluorogenic
substrate, as in Sala et al. (2016). Each sample (350 µl) was
pipetted in quadriplicates into 96 black well-plates, with 50
µl of the substrate 4-methylumbelliferyl β-D-fucoside (Sigma-
Aldrich) at a final concentration of 125 µM. Fluorescence was
measured immediately after addition of the substrate and after
incubations, at in situ temperature and in the dark, for 15,
30 min, 1, 3, and 5 h. The measurements were done with a
Modulus Microplate (DISMED, Turner BioSystems) at 450 nm
excitation and 365 nm emission wavelengths. The increase of
fluorescence units during the period of incubation was converted
into enzymatic activity with a standard curve prepared with
4-methylumbelliferone (MUF, Sigma-Aldrich). FIGURE 1 | Study area. Blue symbols, Coastal transects. symbols,
Coast-to-open sea transect. Green symbol, 2-day cycle. stained with 500 µL of Alcian Blue (0.02%, pH 2.5) for 5 s and
rinsed with MilliQ water. The filters were soaked in 80% sulfuric
acid for 3 h and the absorbance of the extract was determined
at 787 nm in a Varian Cary spectrophotometer. Duplicates were
taken for each sample. Previous analyses have shown a CV of 13%
between TEP replicated measurements with this method (details
not shown). We have calculated an average range of ±30.9%
between duplicates in our dataset. Duplicate blanks (empty filters
stained with alcian blue) were also taken at every station. The
Alcian Blue dye solution was calibrated just before the cruise
using a standard solution of xanthan gum processed with a tissue
grinder and subsequently filtered through two sets of filters (five
points in triplicate). Frontiers in Microbiology | www.frontiersin.org Horizontal TEP Distribution along the
Catalan Coast FIGURE 2 | Variations of TEP (blue symbols) and chlorophyll a (green
symbols) concentration in the coastal transects performed in May
(A) and June (B). In the coastal transects, TEP concentrations ranged from 27.4 to
122.8 µg XG eq L−1, and were overall higher in June (average 83.7
± 23.9 µg XG eq L−1) than in May (average 56.1 ± 25.8 µg XG
eq L−1; Table 1). In contrast, Chla concentrations were overall
higher in May (0.65 ± 0.55 µg L−1) than in June (0.24 ± 0.14 µg
L−1) and BP rates were similar in the two transects (0.18 ± 0.16
µg C L−1 h−1, ranging from 0.91 to 10.97, in May, 0.14 ± 0.21
µg C L−1 h−1, ranging from 0.50 to 14.79 in June). Dissolved
inorganic nitrogen (DIN, nitrate+nitrite+ammonia) averaged
0.65 µM in May, ranging from 0.19 to 1.28 µM, and averaged 0.41
µM in June, ranging from 0.21 to 1.02 µM. Dissolved phosphate
concentrations averaged 0.07 µM in May, ranging from 0.05
to 0.10 µM, and averaged 0.06 µM in June, ranging from 0.05
to 0.10 µM (Supplementary Figure 2). In May, TEP showed
maxima in waters near Barcelona and north of the outflow of the
Tordera River (Figure 2A). In these locations DIN concentration
was 1.2 µM and phosphate concentration was 0.093 µM, two
to eight-fold higher than in the rest of the stations. TEP-values
were significantly correlated to Chla concentration (r = 0.93, p
= 0.0003, n = 7) and marginally correlated to BP (r = 0.72,
p = 0.06, n = 7, Figure 3). In June, TEP distributions showed
maxima south of the outflow of the Tordera River, and were
uncorrelated to Chla (Figure 2B) nor to BP. The TEP/Chla ratios
were markedly higher in June (434.0 ± 197.5) than in May (136.7
± 91.0, Table 1). FIGURE 2 | Variations of TEP (blue symbols) and chlorophyll a (green
symbols) concentration in the coastal transects performed in May
(A) and June (B). FIGURE 2 | Variations of TEP (blue symbols) and chlorophyll a (green
symbols) concentration in the coastal transects performed in May
(A) and June (B). Statistical Analyses We used the Statistica 7.0 software package to test the
potential drivers of TEP distributions across the different spatial
and timescales. We performed pairwise Pearson correlations
between TEP concentrations and the following physico-chemical
and biological variables: Temperature, salinity, turbidity, O2,
nutrients (NO3, PO4, SiO4), particulate organic carbon (POC)
and nitrogen (PON), chlorophyll a (Chla), bacterial abundance Chla concentration was determined by filtering 150 mL of
seawater on GF/F filters (Whatman), extracting the pigment in
acetone (90% v:v) in the dark at 4◦C for 24 h, and measuring
fluorescence with a Turner Designs fluorometer. January 2017 | Volume 7 | Article 2159 Frontiers in Microbiology | www.frontiersin.org 3 TEP in the Mediterranean Sea Ortega-Retuerta et al. (BA), bacterial production (BP), and extracellular fucosidase
activity. Data were log10-transformed and Bonferroni-corrected
when needed. to basin transition (station 6, Figure 4A). The same horizontal
patterns were observed for Chla, O2, BP (Figure 4B), rendering
significant correlations with TEP (r = 0.7, p < 0.05, n = 9). TEP were likewise related to POC (r = 0.9, p < 0.01, n = 9)
and to the ratio between BP and O2(r = 0.7, p < 0.05, n = 9), TEP were likewise related to POC (r = 0.9, p < 0.01, n = 9)
and to the ratio between BP and O2(r = 0.7, p < 0.05, n = 9),
FIGURE 2 | Variations of TEP (blue symbols) and chlorophyll a (green
symbols) concentration in the coastal transects performed in May
(A) and June (B). FIGURE 3 | Scatterplot between Chla and TEP concentrations in the
coastal transects performed in May (filled symbols) and June (open
symbols). TEP were related to Chla in May (r = 0.93, p < 0.001, n = 7). Frontiers in Microbiology | www.frontiersin.org Drivers of TEP Vertical Distribution We assessed which were the environmental drivers of TEP
vertical distributions in epipelagic waters performing Pearson
correlation tests between TEP concentrations and a number of
physical (temperature, salinity, turbulence), chemical (nutrients,
O2), and biological (Chla, BP, bacterial abundance, bacterial
fucosidase activity) variables (Table 2). Since vertical distribution
patterns of TEP and biological variables such as Chla and BP
differed among shelf, slope and basin stations, we separated
shelf stations from the others for the analysis. TEP was
never significantly correlated to Chla. In shelf waters, TEP
was not related to any other physicochemical or biological
parameter. In slope and basin waters, TEP was significantly
and positively correlated to O2, BP and the ratio BP/O2, and
significantly negatively correlated to nutrients and N/P ratios
(Table 2). Horizontal TEP Distribution from Coastal
to Open Sea Waters FIGURE 3 | Scatterplot between Chla and TEP concentrations in the
coastal transects performed in May (filled symbols) and June (open
symbols). TEP were related to Chla in May (r = 0.93, p < 0.001, n = 7). The concentration of TEP in the studied coast-to-open sea
transect ranged from 4.9 to 54.2 µg XG eq L−1 with a mean
concentration of 18.7 ± 11.4 µg XG eq L−1. We
calculated
depth-averaged
TEP
concentrations
in
epipelagic waters (0–200 m) in an attempt to look at horizontal
distribution patterns. Depth-averaged epipelagic TEP ranged
from 9.9 to 24.9 µg XG eq L−1 (Figure 4A). TEP concentrations
were highest near the coast and 60 km offshore, at the slope TABLE 1 | Ranges of TEP concentration and TEP/Chla ratios in the
different transects, depth profiles and diel cycles presented here. TEP (µg XG eq L−1)
TEP/Chla
n
Coastal transect May
27.4–92.1
57.7–283.4
7
Coastal transect June
49.8–122.8
98.4–706.7
7
Epipelagic
4.9–54.2
18.1–316.8
36
Meso- and bathy-pelagic
5.2–19.0
–
23
2-day cycle
5.7–55.9
15.3–1217.6
78
n, number of samples. Frontiers in Microbiology | www.frontiersin.org
4 TABLE 1 | Ranges of TEP concentration and TEP/Chla ratios in the
different transects, depth profiles and diel cycles presented here. FIGURE 3 | Scatterplot between Chla and TEP concentrations in the
coastal transects performed in May (filled symbols) and June (open
symbols). TEP were related to Chla in May (r = 0.93, p < 0.001, n = 7). January 2017 | Volume 7 | Article 2159 4 TEP in the Mediterranean Sea Ortega-Retuerta et al. which can be considered a proxy of bacterial reprocessing of
photosynthetically fixed carbon. generally peaked at the surface (slope stations 3 and 5) or at the
subsurface (slope station 4 and basin stations). These subsurface
maxima were located between 25 and 55 m and showed values
from 33.0 to 54.2 µg XG eq L−1 (mean concentration 39.6
± 10.3 µg XG eq L−1). TEP maxima were always located
shallower than the DCM and coincided with O2 maxima, and
nutrient minima (Figure 5). At the slope, TEP maxima were also
coincident with BP and POC maxima, while the Chla maximum
was always deeper. By contrast, in the basin, BP and POC
coincided with Chla while TEP and O2 peaked at shallower
depths. TEP Vertical Distribution in Mesopelagic
and Bathypelagic Waters FIGURE 4 | Depth-averaged TEP concentration (µg XG eq L−1) and
Chla (µg L−1) (A) and BP (µg C L−1 h−1) and O2 (mg L−1) (B) in epipelagic
waters of the coast to open sea transect. FIGURE 4 | Depth-averaged TEP concentration (µg XG eq L−1) and
Chla (µg L−1) (A) and BP (µg C L−1 h−1) and O2 (mg L−1) (B) in epipelagic
waters of the coast to open sea transect. yp
g
TEP concentrations were half lower in meso- and bathypelagic
waters with respect to the epipelagic, showing minima between
TABLE 2 | Results of Pearson correlations between TEP concentration
and different physical, chemical, and biological variables measured in
epipelagic waters during the NEMO cruises at the slope and basin
stations. Dependent variable
Independent variable
r
p
SLOPE AND BASIN STATIONS (n = 25)
TEP
Temperature
0.55
0.004
Turbidity
0.68
0.000
O2
0.83
0.000
NO3
−0.82
0.000
PO4
−0.58
0.005
N/P
−0.56
0.008
SiO4
−0.73
0.003
POC
0.79
0.000
Chla
ns
Bact. Ab. ns
Bact. Prod. 0.70
0.000
Bact. Fuc. ns
BP/O2
0.68
0.000
r, correlation coefficient; p, level of significance. TEP concentrations were half lower in meso- and bathypelagic
waters with respect to the epipelagic, showing minima between TABLE 2 | Results of Pearson correlations between TEP concentration
and different physical, chemical, and biological variables measured in
epipelagic waters during the NEMO cruises at the slope and basin
stations. Dependent variable
Independent variable
r
p
SLOPE AND BASIN STATIONS (n = 25)
TEP
Temperature
0.55
0.004
Turbidity
0.68
0.000
O2
0.83
0.000
NO3
−0.82
0.000
PO4
−0.58
0.005
N/P
−0.56
0.008
SiO4
−0.73
0.003
POC
0.79
0.000
Chla
ns
Bact. Ab. ns
Bact. Prod. 0.70
0.000
Bact. Fuc. ns
BP/O2
0.68
0.000
r, correlation coefficient; p, level of significance. TABLE 2 | Results of Pearson correlations between TEP concentration
and different physical, chemical, and biological variables measured in
epipelagic waters during the NEMO cruises at the slope and basin
stations. TABLE 2 | Results of Pearson correlations between TEP concentration
and different physical, chemical, and biological variables measured in
epipelagic waters during the NEMO cruises at the slope and basin
stations. FIGURE 4 | Depth-averaged TEP concentration (µg XG eq L−1) and
Chla (µg L−1) (A) and BP (µg C L−1 h−1) and O2 (mg L−1) (B) in epipelagic
waters of the coast to open sea transect. TEP Vertical Distribution in Epipelagic
Waters TEP vertical distribution patterns were variable among stations. Shelf waters exhibited mixed temperature and salinity profiles
(Figure 5) and other chemical and biological variables were also
quite uniform in the vertical profile. In these stations, Chla
ranged from 0.10 to 0.38 µg L−1, BP ranged from 0.012 to
0.078 µg C L−1 h−1, and POC ranged from 3.7 to 7.2 µM. TEP
vertical distributions were also quite homogenous from surface
to the bottom, averaging 24.9 ± 3.0 and 19.7 ± 2.5 µg XG
eq L−1 in stations 1 and 2, respectively (Figure 5). Conversely,
well-developed DCM were detected in slope and basin waters
between 50 and 60 m, with Chla concentrations ranging from
0.41 (station 5, slope) to 1.73 (station 6, basin) µg L−1. The
TEP/Chla ratios ranged from 16.8 (station 6, DCM) to 316.0
(station 3, surface). They were generally higher at the surface
and lower at the DCM. Bacterial production ranged from 0.005
to 0.065 µg C L−1 h−1 and the vertical distribution varied
between stations: at the slope stations, BP was highest at the
surface and subsurface, while in basin stations BP showed a
bimodal profile, with peaks at the surface and at the DCM. POC concentrations, that ranged from 2.9 to 11.5 µM, showed
similar distributions than TEP in the slope stations (with surface
or subsurface peaks) but covaried with Chla, with maxima
at the DCM, in the basin stations. TEP also showed marked
vertical changes in slope and basin waters (Figure 5): they DISCUSSION TEP dynamics can affect particle aggregation rates because of
stickiness, and particle sinking rates due to their low density. Thus, determining and predicting TEP dynamics is crucial if
we want to accurately estimate carbon and particle fluxes. The TEP dynamics can affect particle aggregation rates because of
stickiness, and particle sinking rates due to their low density. Thus, determining and predicting TEP dynamics is crucial if
we want to accurately estimate carbon and particle fluxes. The FIGURE 6 | Diel variations of depth-averaged TEP (blue circles),
Chlorophyll a (green triangles), O2 (purple triangles) concentration. TEP Vertical Distribution in Mesopelagic
and Bathypelagic Waters January 2017 | Volume 7 | Article 2159 Frontiers in Microbiology | www.frontiersin.org 5 Ortega-Retuerta et al. TEP in the Mediterranean Sea GURE 5 | Vertical profiles of Temperature (◦C), salinity (practical salinity units), turbidity (FTU) and NO3 (µM) (left panels) and O2 (mg L−1), TEP
oncentration (µg XG eq L−1), Chla (µg L−1) and BP (µg C L−1 h−1) (right panels) in epipelagic waters of station 2 (shelf, upper panels), station 4
lope, middle panels), and station 8 (basin, bottom panels). FIGURE 5 | Vertical profiles of Temperature (◦C), salinity (practical salinity units), turbidity (FTU) and NO3 (µM) (left panels) and O2 (mg L−1), TEP
concentration (µg XG eq L−1), Chla (µg L−1) and BP (µg C L−1 h−1) (right panels) in epipelagic waters of station 2 (shelf, upper panels), station 4
(slope, middle panels), and station 8 (basin, bottom panels). TEP in meso- and bathypelagic waters were only significantly
related to turbidity (r = 0.54, p < 0.01 n = 23), indicative of
the relevance of bottom nepheloid layers (BNL). Remarkably,
TEP were uncorrelated with POC in these layers. No significant 4.9 and 11.2 µg XG eq L−1. By contrast, relative TEP increases
were observed in waters near the bottom in all slope and basin
stations, parallel to increases in turbidity (proxy of total particle
concentration). January 2017 | Volume 7 | Article 2159 Frontiers in Microbiology | www.frontiersin.org 6 TEP in the Mediterranean Sea Ortega-Retuerta et al. TEP-values observed in this study are within the range of data
published for Mediterranean Sea waters (Prieto et al., 2006;
Ortega-Retuerta et al., 2010; Bar-Zeev et al., 2011) and at the
lower range of other coastal ocean areas (Klein et al., 2011; Van
Oostende et al., 2012). By contrast, the TEP/Chla ratios were in
the upper range of those previously published for other ocean
basins (Prieto et al., 2006), except for those measured in the
Mediterranean Sea (Ortega-Retuerta et al., 2010; Bar-Zeev et al.,
2011). correlation was found between TEP and BP, but a correlation with
bacterial fucosidase activity (r = 0.45, p < 0.05 n = 23) could be
observed. Diel TEP Variations The same surface water mass was sampled over time during the
2-day lagrangian study in the SUMMER2 cruise, as confirmed by
plotting depth-averaged temperature and salinity values of each
CTD cast (Supplementary Figure 1), except for the last two casts
were a warmer water mass was likely sampled. We did not detect
a recurrent diel pattern of TEP or of any other biological variables
such as Chla or BP (Figure 6), even though these variables varied
highly during the cycle. Chla showed subsurface maxima of
0.75–1.33 µg L−1 between 40 and 54 m. TEP concentrations
ranged ten-fold, from 5.7 to 55.9 µg XG eq L−1 (average 34.1
± 5.7 µg XG eq L−1). TEP maxima were situated between 25
and 47 m with highest values during the night (10 p.m. and
2 a.m.). These maxima were always shallower than the DCM,
and were again vertically coincident with O2 and BP maxima. TEP concentrations were not significantly correlated to Chla,
but a significant correlation was observed with O2 (r = 0.70,
p = 0.000, n = 78), BP (r = 0.48 p = 0.000, n = 78), and
the ratio BP/O2 (r = 0.44, p = 0.0001, n = 78). We looked
at temporal variations of TEP, Chla, O2, and BP using depth-
averaged values for all the epipelagic depths. Diel dynamics of
TEP, Chla, O2, and BP were not coupled, but we detected Chla
and O2 increases that were followed by a TEP increase after
4–12 h (Figure 6). However, lagged correlations between these
variables, using either maxima or integrated values, were not
significant. TEP Variation in the Coastal Transects TEP maxima were observed near the city of Barcelona and near
the outflow of the Tordera River, areas with higher nutrient
concentrations. Coincidentally, there are two desalination plants
located at the mouth of the Llobregat (next to Barcelona) and
Tordera Rivers, whose feedwater intake pipes are located between
800 and 2200 m from the coast and at 30 m depth, although
they are not always operative. Given that TEP are directly linked
to membrane fouling, it is important to be able to predict
TEP occurrence in these locations. In May, TEP variations in
these coastal waters could be explained by variations in Chla,
suggesting a direct linkage between phytoplankton and TEP at
this geographical scale. However, this relationship was absent in
June (Figure 2). The higher TEP/Chla ratios in June were due
to both higher TEP concentrations than in May and lower Chla
concentrations than in May. Indeed, primary productivity (PP)
is on average two-fold higher in May than in June in coastal
NW Mediterranean Sea waters (Gasol et al., 2016). Bacteria did
not seem to be a significant TEP source in June as no significant
correlations between TEP and BP or TEP and the BP/Chla ratios
were observed. Hence, TEP may accumulate in the sea surface
at the beginning of the stratification period, similar to other
phytoplankton-derived organic matter (Avril, 2002; Vila-Reixach
et al., 2012; Romera-Castillo et al., 2013). These substances may
be not taken up immediately by bacterioplankton due to nutrient
defficiency, which is common in summer (Sala et al., 2002;
Pinhassi et al., 2006). What we could resolve in this analysis was
that the distribution of TEP in coastal waters could be predicted
from Chla, a variable that is frequently monitored, but only at
certain periods of the year. TEP Vertical Distribution in Epipelagic
Waters Waters
In contrast to horizontal TEP distributions, vertical TEP
distributions could not be predicted from Chla at the vertical
scale. In this case, TEP maxima were always located shallower
than the DCM. TEP maxima above the DCM and near the
surface have been reported in previous studies in the Alboran
Sea (García et al., 2002), the Eastern Mediterranean Sea (Bar-
Zeev et al., 2011), and along a West to East transect in the
Mediterranean Sea (Ortega-Retuerta et al., 2010). Furthermore,
in a recent study (Kodama et al., 2014), TEP maxima were
associated to layers of maximum O2 and nutrient minima, similar
to what we observed in our study. Correlation analyses confirmed
the Chla-TEP decoupling. From the whole set of variables, those
that explained TEP distribution in both slope and basin waters
were O2 (positively related to TEP) and nitrate (negatively related
to TEP, Table 2). This suggests that TEP are a direct product
of PP, which is frequently decoupled from Chla concentration
and whose highest rates are located at depths shallower than
the DCM (Estrada et al., 1993). This suggests that Chla is not
the best proxy of phytoplankton biomass or PP at the vertical
scale. Indeed, increases in Chla concentration at the DCM
reflect photoacclimation to low light levels through increases
in the Chla/carbon ratio, and do not necessarily match highest
phytoplankton biomass (Delgado et al., 1992; Gernez et al., 2011). Given that we lacked direct PP estimates, in this study we
consider O2 as a proxy of PP in this area. We are aware,
however, that O2 concentration in the ocean is a result of
biological processes as well as physical processes, such as
ventilation in the upper mixed layer. But a non-significant
correlation between O2 and temperature indicates that O2
distribution in the area majorly reflects biological processes. Furthermore, in surface waters of our study, TEP and O2
were not significantly correlated, which supports our view that
the O2-TEP correlation is through PP with O2 being a proxy
of the latter. Similarly, the absence of negative correlations
between O2 and bacterial abundance or production suggests
that respiration by heterotrophs is not the main driver of O2
distributions. In contrast to horizontal TEP distributions, vertical TEP
distributions could not be predicted from Chla at the vertical
scale. In this case, TEP maxima were always located shallower
than the DCM. TEP Horizontal Distribution from Coastal
to Offshore Waters Our dataset also allowed concluding that Chla is a good predictor
of TEP concentrations at the horizontal scale from the coast
to the open sea, since significant positive relationships were
observed between these two variables. This information is
important because it would allow estimating TEP concentrations,
for instance, using remote sensing Chla values. It is worthy to
mention that, if we restrict the analysis to surface values only
instead of depth-averaged values, TEP and Chla were also related
(r = 0.63 p = 0.06, n = 9), while not remarkable covariations
were found between TEP and other biological variables. This
reinforces the possibility of using remote sensing Chla data to
estimate the geographical distributions of TEP in the area. FIGURE 6 | Diel variations of depth-averaged TEP (blue circles),
Chlorophyll a (green triangles), O2 (purple triangles) concentration. FIGURE 6 | Diel variations of depth-averaged TEP (blue circles),
Chlorophyll a (green triangles), O2 (purple triangles) concentration. FIGURE 6 | Diel variations of depth-averaged TEP (blue circles),
Chlorophyll a (green triangles), O2 (purple triangles) concentration. At the horizontal scale, TEP was also positively related to
O2, POC, and BP. The covariance between all these variables January 2017 | Volume 7 | Article 2159 Frontiers in Microbiology | www.frontiersin.org 7 TEP in the Mediterranean Sea Ortega-Retuerta et al. and utilization of TEP, or (3) dependence of both TEP and
bacteria on other factors, namely organic compounds released
by phytoplankton during primary production. We performed
multiple regression analyses with TEP as the dependent variable
and O2 and BP as independent variables. Both variables
significantly explained TEP variability (r2 = 0.71, p = 0.000,
n = 25), with partial coefficients of 0.72 for O2 and 0.46 for
BP, respectively. Additionally, we correlated the raw residuals
of the O2-TEP regression against BP, resulting significant and
positive (r = 0.38, p < 0.05, n = 25). Both analyses lead us
to conclude that both PP (with O2 as surrogate) and bacterial
activity (with BP as surrogate) has a significant influence on
vertical TEP distributions. Finally, the correlation between TEP
and the BP/O2 ratio was significant and positive. The positive
correlation indicates that the higher the bacterial processing of
organic matter originated in primary production, the higher TEP
concentrations we observed, suggesting then a synergy between
carbon fixation and bacterial reutilization of this fixed C to
generate TEP. TEP Vertical Distribution in Epipelagic
Waters TEP maxima above the DCM and near the
surface have been reported in previous studies in the Alboran
Sea (García et al., 2002), the Eastern Mediterranean Sea (Bar-
Zeev et al., 2011), and along a West to East transect in the
Mediterranean Sea (Ortega-Retuerta et al., 2010). Furthermore,
in a recent study (Kodama et al., 2014), TEP maxima were
associated to layers of maximum O2 and nutrient minima, similar
to what we observed in our study. Correlation analyses confirmed
the Chla-TEP decoupling. From the whole set of variables, those
that explained TEP distribution in both slope and basin waters
were O2 (positively related to TEP) and nitrate (negatively related
to TEP, Table 2). This suggests that TEP are a direct product
of PP, which is frequently decoupled from Chla concentration
and whose highest rates are located at depths shallower than
the DCM (Estrada et al., 1993). This suggests that Chla is not
the best proxy of phytoplankton biomass or PP at the vertical
scale. Indeed, increases in Chla concentration at the DCM
reflect photoacclimation to low light levels through increases
in the Chla/carbon ratio, and do not necessarily match highest
phytoplankton biomass (Delgado et al., 1992; Gernez et al., 2011). Nutrients may also have an impact on TEP distributions as
suggested by the negative correlations with NO3, PO4, and SiO4. TEP are enriched in carbon respect to N and P (Mari et al.,
2001), and different experimental studies have demonstrated
a higher TEP release rate when nutrients are limiting (Mari
et al., 2005; Pedrotti et al., 2010). Also, the N/P ratios are likely
important in determining TEP production and degradation, as
suggested by a significant correlation between the N/P ratio
and TEP concentration in epipelagic waters. Although how
nutrient stoichiometry affects TEP dynamics is unclear and
probably dependent on the composition of the initial microbial
community (Gärdes et al., 2012); our negative correlation
suggests that the lower the relative proportion of P, the higher
the TEP; which contrasts with previous experimental results
(Engel et al., 2015). However, the likely limitation by phosphorus
in our system (e.g., Sala et al., 2002; Pinhassi et al., 2006) in
contrast to the Atlantic area studied by Engel et al. (2015) could
differently affect TEP production. Indeed, it has been proposed
that extracellular organic carbon production is highest under
P limitation (Mauriac et al., 2011). TEP Horizontal Distribution from Coastal
to Offshore Waters at the horizontal scale suggests physical forces that drive them
all. We observed maxima of TEP in epipelagic waters near
the coast and at the interface between slope and basin waters. In this area, which is located next to the Catalan front, a
salinity doming is frequently observed (Font et al., 1988; Estrada
and Salat, 1989), and relatively high Chla and PP values are
found due to increased deep ocean nutrient availability at
shallower depths (Estrada, 1996; Pedrós-Alió et al., 1999). This
is, however, not evident in our dataset since some basin stations
(i.e., stations 8 and 9) were sampled during stormy conditions
that rendered more mixed temperature and salinity profiles,
while the rest were sampled in sunny days and well-stratified
profiles. Frontiers in Microbiology | www.frontiersin.org TEP Vertical Distribution in Epipelagic
Waters TEP properties also vary
depending on whether they are released during active growth or
during bloom senescence (Mari et al., 2001), with implications
for the fate of these particles in seawater (degradation vs. export). In the NW Mediterranean, PP usually peaks at the end of
winter and spring (Gasol et al., 2016), so we sampled during
the transition of spring to summer, which coincides with the
beginning of nutrient depletion and associated decreases in PP. Thus, we expected TEP to accumulate and be prone to further
export. Given that we lacked direct PP estimates, in this study we
consider O2 as a proxy of PP in this area. We are aware,
however, that O2 concentration in the ocean is a result of
biological processes as well as physical processes, such as
ventilation in the upper mixed layer. But a non-significant
correlation between O2 and temperature indicates that O2
distribution in the area majorly reflects biological processes. Furthermore, in surface waters of our study, TEP and O2
were not significantly correlated, which supports our view that
the O2-TEP correlation is through PP with O2 being a proxy
of the latter. Similarly, the absence of negative correlations
between O2 and bacterial abundance or production suggests
that respiration by heterotrophs is not the main driver of O2
distributions. TEP Vertical Distribution in Meso- and
Bathy-Pelagic Waters g
In meso- and bathypelagic waters TEP distributions were only
explained by turbidity changes. Specifically, we could detect
TEP increases associated with near-bottom particle-rich layers
(BNL). Surprisingly, these TEP increases were not paralleled by
increases in POC. This suggests that these BNL were composed The observed significant positive correlation with bacterial
production would suggest one of the following mechanisms:
(1) bacteria act as a source of TEP, (2) bacterial colonization January 2017 | Volume 7 | Article 2159 8 TEP in the Mediterranean Sea Ortega-Retuerta et al. TABLE 3 | Review of TEP-Chla relationships available in the literature. January 2017 | Volume 7 | Article 2159 Frontiers in Microbiology | www.frontiersin.org TEP Vertical Distribution in Meso- and
Bathy-Pelagic Waters Scale
Year
Season
Area
Depth
TEP range
Chl range
Correlation
r
References
Horizontal
1997
Spring
Gulf of Cadiz
10 m
+
0.60
Prieto et al., 2006
Spring
25 m
+
0.33
Spring
Euphotic
63.8–202.4
+
0.29
Spring
Aphotic
Spring
NW Alboran Sea
+
0.92
2006
Spring
Bay of Biscay
0–160 m
1.41–31.7*1
0.8–2.1
+
0.61
Harlay et al., 2010
2006–2008
Spring
Celtic Sea
0–80 m
22–1101
0.39-1.44
+
0.42
Van Oostende et al., 2012
2012
Spring
NW Mediterranean Sea
0–200 m
9.9–24.9
+
0.75
This study
1999–2002
All year
Adriatic Sea
Surface
4–14000
1–14
+
Radic et al., 2005
Vertical
2008–2009
Spring–summer
Eastern mediterranean Sea
51–290
0.02–0.99
–
−0.69
Bar-Zeev et al., 2011
2009
Spring
North Pacific
Mixed layer
5.1–65.4
0.30–1.72
ns
Wurl et al., 2011
2009
Summer
Hawaii
Mixed layer
<2.5–60.2
0.38–0.63
ns
2009
Fall
Arctic
Mixed layer
7.9–117.9
0.57–7.80
ns
2009-2010
Summer–winter
Pearl River Estuary
88.7–1586
na
ns
Sun et al., 2012
2012
Spring
NW mediterranean Sea
0-2300
4.9–54.9
0–1.73
ns
This study
2013
Spring
NW Pacific
Epipelagic
35.3–47.2
0.05–0.26
ns
Kodama et al., 2014
Hor-vert
2007
Spring
Mediterranean Sea
0–200 m
4.5–94.3
0–1.78
ns
Ortega-Retuerta et al., 2010
2014
Spring
Mediterranean Sea
Surface-deep
ns
Mazuecos, 2015, thesis
2005
Summer
Southern ocean
0-200 m
0–48.9
0.01–5.36
+
0.52
Ortega-Retuerta et al., 2009b
2010–2011
All
Tropical and subtropical World oceans
Surface-deep
ns
Mazuecos, 2015, thesis
Summer
Gibraltar Strait
0–75 m
169.3
+
0.73
Prieto et al., 2006
Temporal
1997–1998
All
Tokio Bay
0–10 m
14–1774
<5–30
+
0.65
Ramaiah and Furuya, 2002
1998–2000
All
Dona Paula Bay
Surface
1.3–149.1
1.2–12.3
ns
Bhaskar and Bhosle, 2006
1999–2000
All
NW Mediterranean Sea
Surface-DCM
Nd*2
0–2.8
+
0.71
Beauvais et al., 2003
2012
Spring-summer
English channel
Surface and microlayer
254–1301
0.5–5.5
ns
Taylor et al., 2014
Spring-summer
English channel
Surface
36.9–1735.1
9
+*3
0.87
Klein et al., 2011
2002–2004
Summer–winter
Aegean Sea
Surface to 4 m (bottom)
101–259
0.1–7
+
0.19
Scoullos et al., 2006
Horizontal-vertical scales (hor-vert in the Table) are considered when analyses are made using data that covered wide geographical areas but also include deep vertical profiles. ns, not significant. na, not available. *1Relative units (samples not calibrated). *2TEP
l
d b
i
i
ti
rontiers in Microbiology | www.frontiersin.org
9
Janua January 2017 | Volume 7 | Article 2159 9 TEP in the Mediterranean Sea Ortega-Retuerta et al. CONCLUSIONS We showed that the TEP-Chla relationship in the ocean is
variable and mainly depends on the time and spatial scale
studied. TEP can be predicted from Chla distributions at the
horizontal scale, which opens the possibility to estimate surface
TEP using remote sensing Chla; but this relationship is not
evident at the vertical scale, nor at a short timescale, and
also likely varies seasonally. Since our dataset is limited, in an
attempt to compare our results to other areas, we compiled Supplementary Figure 1 | Bivariate plot between depth-averaged (upper
mixed layer) temperature (abscises) and salinity (ordinates) in the 13
sampling stations along the diel cycle. Supplementary Figure 2 | Variations of dissolved inorganic nitrogen (DIN,
orange circles), dissolved inorganic phosphorus (DIP, green-yellow
triangles), and estimated bacterial production with factor 1.55 kgC mol
leucine (BP, purple triangles) in the coastal transects in May (A) and
June (B). SUPPLEMENTARY MATERIAL The Supplementary Material for this article can be found
online
at:
http://journal.frontiersin.org/article/10.3389/fmicb. 2016.02159/full#supplementary-material AUTHOR CONTRIBUTIONS EO, MS, FP, RS, CM, and JG designed the work. EO, CM, FA, CA,
RG, EB, and MM sampled and performed laboratory analyses and
processed the data. EO wrote the manuscript with the help and
inputs of all co-authors. TEP Diel Variations To our knowledge this is the first time that high frequency
(every 4 h) and short-term (i.e., 2-day) TEP changes have been
monitored in the field during a lagrangian study. However, we
did not find a recurrent pattern of any of the variables measured. In our case, this lack of diel patterns of microbial biomass/activity
likely explained the absence of recurrent TEP diurnal or
nocturnal maxima. Additionally, although we confirmed in
this study the vertical decoupling between Chla and TEP and
the better coincidence with O2 concentrations, the results of
short-term variations of these variables were less conclusive. We got some hints about short-term temporal decoupling,
where TEP peaks followed Chla and O2 peaks, but further
work, with longer diel sampling, is needed to explore this
issue. This work was funded by projects funded by the Spanish Ministry
of
Science
STORM
(CTM2009-09352/MAR),
SUMMER
(CTM2008-03309/MAR), DOREMI (CTM2012-34294), REMEI
(CTM2015-70340-R), ANIMA (CTM2015-65720-R), PEGASO
(CTM2012-37615), and Grup consolidat de Recerca de la
Generalitat de Catalunya (2014SGR/1179). We thank the captain
and crew of the R/V “García del Cid” and the Marine Technology
Unit for their assistance in the field. We also thank the reviewers
for helping improve the manuscript. EO was supported by
Beatriu de Pinós (Generalitat de Catalunya) and Juan de la
Cierva (Spanish Ministry of Economy) postdoctoral fellowships. TEP Vertical Distribution in Meso- and
Bathy-Pelagic Waters mostly of mineral particles that could be coated or aggregated by
TEP. A previous study in the NW Mediterranean (Puig et al.,
2013) observed the presence of fine particles in the BNL. Via
microscope visualizations, they showed that organic matter in
the BNL was mainly composed by “amorphous aggregates,” and
suggested that these aggregates had lower sinking rates than
phytoplankton cells or other solid organic particles. This is in
line with our findings, where TEP, which are low density particles,
may have a longer residence time in the BNL than the rest of POC
compounds. In bathypelagic waters, TEP cannot have a direct
phytoplanktonic source as light is absent. However, a bacterial
source was not evident either since no significant correlations
were observed between TEP and BA or BP. Interestingly, TEP
concentrations in that layer were positively correlated to bacterial
fucosidase activity. Since TEP are enriched in fucose (Zhou
et al., 1998), this may reflect bacterial degradation of TEP in
deep waters. Therefore, bacteria would act as a sink instead of a
source for TEP, and a probable non-local TEP source must exist,
material either sunk from epipelagic waters, resuspended from
the sediment, or advected offthe shelf. information on the various TEP-Chla relationships published in
the literature (Table 3). The previous results mainly concur with
our observations: TEP patterns mimic Chla patterns horizontally,
but they are vertically decoupled, and TEP concentration maxima
are frequently found at depths shallower than the DCM. Hence,
we propose O2 concentrations and bacterial production as
predictive variables for vertical TEP distributions in the ocean. Further, we suggest looking at the spatial and temporal variations
of TEP together with primary productivity measurements and
orienting further work to elucidate what is the specific role of
bacterioplankton at explaining geographical and vertical TEP
distributions in the ocean. Azetsu-Scott, K., and Passow, U. (2004). Ascending marine particles: significance
of transparent exopolymer particles (TEP) in the upper ocean. Limnol.
Oceanogr. 49, 741–748. doi: 10.4319/lo.2004.49.3.0741
Bar-Zeev, E., Berman, T., Rahav, E., Dishon, G., Herut, B., Kress, N., et al. (2011).
Transparent exopolymer particle (TEP) dynamics in the eastern Mediterranean
Sea. Mar. Ecol. Progr. Ser. 431, 107–118. doi: 10.3354/meps09110
Beauvais, S., Pedrotti, M. L., Villa, E., and Lemée, R. (2003). Transparent
exopolymer particle (TEP) dynamics in relation to trophic and hydrological REFERENCES Using flow cytometry for counting natural
planktonic bacteria and understanding the structure of planktonic bacterial
communities. Sci. Mar. 64, 197–224. doi: 10.3989/scimar.2000.64n2197 Gasol, J. M., Cardelús, C., Morán, X. A. G., Balagué, V., Massana, R., Pedrós-Alió,
C., et al. (2016). Seasonal patterns in phytoplankton photosynthetic parameters
and primary production in a coastal NW Mediterranean site. Sci. Mar. 80,
63–77. doi: 10.3989/scimar.04480.06E Pedrotti, M. L., Peters, F., Beauvais, S., Vidal, M., Egge, J., Jacobsen, A., et al. (2010). Effects of nutrients and turbulence on the production of transparent
exopolymer particles: a mesocosm study. Mar. Ecol. Progr. Ser. 419, 57–69. doi: 10.3354/meps08840 Pinhassi, J., Gómez-Consarnau, L., Alonso-Sáez, L., Sala, M., Vidal, M., Pedrós-
Alió, C., et al. (2006). Seasonal changes in bacterioplankton nutrient
limitation and their effects on bacterial community composition in the NW
Mediterranean Sea. Aquat. Microb. Ecol. 44, 241–252. doi: 10.3354/ame044241 Gernez, P., Antoine, D., and Huot, Y. (2011). Diel cycles of the particulate beam
attenuation coefficient under varying trophic conditions in the northwestern
Mediterranean Sea: observations and modeling. Limnol. Oceanogr. 56, 17–36. doi: 10.4319/lo.2011.56.1.0017 Hansen, H. P., and Grasshoff, K. (1983). “Automatic chemical analysis,” in Methods
of Seawater Analysis, 2nd Edn., eds K. Grassohoff, M. Ehrhardt, and K. Kremling (Weinheim: Verlag Chemie), 368–376. Prieto,
L.,
Navarro,
G.,
Cózar,
A.,
Echevarría,
F.,
and
García,
C. M. (2006). Distribution of TEP in the euphotic and upper mesopelagic
zones of the southern Iberian coasts. Deep Sea Res. II 53, 1314–1328. doi: 10.1016/j.dsr2.2006.03.009 Harlay, J., Borges, A. V., Van Der Zee, C., Delille, B., Godoi, R. H. M., Schiettecatte,
L. S., et al. (2010). Biogeochemical study of a coccolithophore bloom in the
northern Bay of Biscay (NE Atlantic Ocean) in June 2004. Progr. Oceanogr. 86,
317–336. doi: 10.1016/j.pocean.2010.04.029 Prieto, L., Sommer, F., Stibor, H. N., and Koeve, W. (2001). Effects of planktonic
copepods on transparent exopolymeric particles (TEP) abundance and size
spectra. J. Plankton Res. 23, 515–525. doi: 10.1093/plankt/23.5.515 Puig, P., Madron, X. D. D., Salat, J., Schroeder, K., Martín, J., Karageorgis, A. P., et al. (2013). Thick bottom nepheloid layers in the western Mediterranean
generated by deep dense shelf water cascading. Progr. Oceanogr. 111, 1–23. doi: 10.1016/j.pocean.2012.10.003 Kirchman, D., Knees, E., and Hodson, R. (1985). Leucine incorporation and
its potential as a measure of protein-synthesis by bacteria in natural aquatic
systems. Appl. Environ. Microbiol. 49, 599–607. Klein, C., Claquin, P., Pannard, A., Napoléon, C., Le Roy, B., and Véron, B. (2011). REFERENCES Alldredge, A. L., Passow, U., and Logan, B. E. (1993). The abundance and
significance of a class of large, transparent organic particles in the ocean. Deep
Sea Res. I 40, 1131–1140. doi: 10.1016/0967-0637(93)90129-Q Avril,
B. (2002). DOC
dynamics
in
the
northwestern
Mediterranean
Sea
(DYFAMED
site). Deep
Sea
Res. II
49,
2163–2182. doi: 10.1016/S0967-0645(02)00033-4 Frontiers in Microbiology | www.frontiersin.org January 2017 | Volume 7 | Article 2159 10 Ortega-Retuerta et al. TEP in the Mediterranean Sea by
Phaeocystis
globosa
under
N-
or
P-limitation:
a
controlling
factor
of
the
retention/export
balance. Harmful
Algae
4,
895–914. doi: 10.1016/j.hal.2004.12.014 conditions in the NW Mediterranean Sea. Mar. Ecol. Progr. Ser. 262, 97–109. doi: 10.3354/meps262097 conditions in the NW Mediterranean Sea. Mar. Ecol. Progr. Ser. 262, 97–109. doi: 10.3354/meps262097 Berman, T. (2013). Transparent exopolymer particles as critical agents in aquatic
biofilm formation: implications for desalination and water treatment. Desal. Water Treat. 51, 1014–1020. doi: 10.1080/19443994.2012.713585 Mauriac, R., Moutin, T., and Baklouti, M. (2011). Accumulation of DOC in Low
Phosphate Low Chlorophyll (LPLC) area: is it related to higher production
under high N:P ratio? Biogeosciences 8, 933–950. doi: 10.5194/bg-8-933-2011 Bhaskar, P. V., and Bhosle, N. B. (2006). Dynamics of transparent exopolymeric
particles (TEP) and particle-associated carbohydrates in the Dona Paula bay,
west coast of India. J. Earth Sys. Sci. 115, 403–413. doi: 10.1007/BF02702869 Mazuecos, I. P. (2015). Exopolymer Particles in the Ocean: Production by
Microorganisms, Carbon Export and Mesopelagic Respiration. PhD, University
of Granada, 281. Burd, A. B., and Jackson, G. A. (2009). Particle aggregation. Ann. Rev. Mar. Sci. 1,
65–90. doi: 10.1146/annurev.marine.010908.163904 Orellana, M. V., Matrai, P. A., Leck, C., Rauschenberg, C. D., Lee, A. M., and Coz, Orellana, M. V., Matrai, P. A., Leck, C., Rauschenberg, C. D., Lee, A. M., and Coz,
E. (2011). Marine microgels as a source of cloud condensation nuclei in the high
Arctic. Proc. Natl. Acad. Sci. 108, 13612–13617. doi: 10.1073/pnas.1102457108 E. (2011). Marine microgels as a source of cloud condensation nuclei in the high
Arctic. Proc. Natl. Acad. Sci. 108, 13612–13617. doi: 10.1073/pnas.1102457108 Delgado, M., Latasa, M., and Estrada, M. (1992). Variability in the size-fractionated
distribution of the phytoplankton across the Catalan front of the north-west
Mediterranean. J. Plankton Res. 14, 753–771. doi: 10.1093/plankt/14.5.753 Ortega-Retuerta, E., Duarte, C. M., and Reche, I. (2010). Significance of
bacterial
activity
for
the
distribution
and
dynamics
of
transparent
exopolymer particles in the Mediterranean Sea. Microb. Ecol. 59, 808–818. REFERENCES doi: 10.1007/s00248-010-9640-7 Engel, A., Borchard, C., Loginova, A., Meyer, J., Hauss, H., and Kiko, R. (2015). Effects of varied nitrate and phosphate supply on polysaccharidic and
proteinaceous gel particle production during tropical phytoplankton bloom
experiments. Biogeosciences 12, 5647–5665. doi: 10.5194/bg-12-5647-2015 Ortega-Retuerta, E., Passow, U., Duarte, C. M., and Reche, I. (2009a). Effects
of ultraviolet B radiation on (not so) transparent exopolymer particles. Biogeosciences 6, 3071–3080. doi: 10.5194/bg-6-3071-2009 Estrada, M. (1996). Primary production in the northwestern Mediterranean. Sci. Mar. 60, 55–64. Ortega-Retuerta, E., Reche, I., Pulido-Villena, E., Agustí, S., and Duarte, C. M. (2009b). Uncoupled distributions of transparent exopolymer particles (TEP)
and dissolved carbohydrates in the Southern Ocean. Mar. Chem. 115, 59–65. doi: 10.1016/j.marchem.2009.06.004 Estrada, M., Marrasé, C., Latasa, M., Berdalet, E., Delgado, M., and Riera, T. (1993). Variability of deep chlorophyll maximum characteristics in the Northwestern
Mediterranean. Mar. Ecol. Prog Ser. 92, 289–300. doi: 10.3354/meps092289 Estrada, M., and Salat, J. (1989). Phytoplankton assemblages of deep and surface
water layers in a Mediterranean frontal zone. Sci. Mar. 53, 203–214. Passow,
U. (2002). Transparent
exopolymer
particles
(TEP)
in
aquatic
environments. Progr. Oceanogr. 55,
287–333. doi: 10.1016/S0079-6611(02)00138-6 Font, J., Salat, J., and Tintore, J. (1988). Permanent features of the circulation in the
Catalan Sea. Oceanol. Acta. 51–57. Special issue. Passow, U., and Alldredge, A. L. (1995). A dye-binding assay for the
spectrophotometric measurement of transparent exopolymer particles (TEP). Limnol. Oceanogr. 40, 1326–1335. doi: 10.4319/lo.1995.40.7.1326 García, C. M., Prieto, L., Vargas, M., Echevarría, F., García-Lafuente, J., Ruiz, J.,
et al. (2002). Hydrodynamics and the spatial distribution of plankton and TEP
in the Gulf of Cádiz (SW Iberian Peninsula). J. Plankton Res. 24, 817–833. doi: 10.1093/plankt/24.8.817 Passow, U., Shipe, R. F., Murray, A., Pak, D. K., Brzezinski, M. A., and Alldredge,
A. L. (2001). The origin of transparent exopolymer particles (TEP) and their
role in the sedimentation of particulate matter. Cont. Shelf Res. 21, 327–346. doi: 10.1016/S0278-4343(00)00101-1 Gärdes, A., Ramaye, Y., Grossart, H. P., Passow, U., and Ullrich, M. S. (2012). Effects of Marinobacter adhaerens HP15 on polymer exudation by
Thalassiosira weissflogii at different N:P ratios. Mar. Ecol. Prog. Ser. 461, 1–14. doi: 10.3354/meps09894 Pedrós-Alió, C., Calderon-Paz, J. I., Guixa-Boixereu, N., Estrada, M., and Gasol,
J. M. (1999). Bacterioplankton and phytoplankton biomass and production
during summer stratification in the northwestern Mediterranean Sea. Deep Sea
Res. I 46, 985–1019. doi: 10.1016/S0967-0637(98)00106-X Gasol, J., and del Giorgio, P. (2000). REFERENCES Dynamics of soluble extracellular polymeric substances and transparent
exopolymer particle pools in coastal ecosystems. Mar. Ecol. Progr. Ser. 427,
13–27. doi: 10.3354/meps09049 Radic, T., Kraus, R., Fuks, D., Radic, J., and Pecar, O. (2005). Transparent
exopolymeric particles’ distribution in the northern Adriatic and their relation
to microphytoplankton biomass and composition. Sci. Total Environ. 353,
151–161. doi: 10.1016/j.scitotenv.2005.09.013 Kodama, T., Kurogi, H., Okazaki, M., Jinbo, T., Chow, S., Tomoda, T., et al. (2014). Vertical distribution of transparent exopolymer particle (TEP) concentration
in the oligotrophic western tropical North Pacific. Mar. Ecol. Progr. Ser. 513,
29–37. doi: 10.3354/meps10954 Ramaiah, N., and Furuya, K. (2002). Seasonal variations in phytoplankton
composition and transparent exopolymer particles in a eutrophicated coastal
environment. Aquat. Microb. Ecol. 30, 69–82. doi: 10.3354/ame030069 Rochelle-Newall, E. J., Mari, X., and Pringault, O. (2010). Sticking properties of
transparent exopolymeric particles (TEP) during aging and biodegradation. J. Plankton Res. 32, 1433–1442. doi: 10.1093/plankt/fbq060 Mari, X., Beauvais, S., Lemée, R., and Pedrotti, M. L. (2001). Non-Redfield C: N
ratio of transparent exopolymeric particles in the northwestern Mediterranean
Sea. Limnol. Oceanogr. 46, 1831–1836. doi: 10.4319/lo.2001.46.7.1831 Romera-Castillo, C., Álvarez-Salgado, X. A., Galí, M., Gasol, J. M., and Marrasé,
C. (2013). Combined effect of light exposure and microbial activity on distinct Mari, X., Rassoulzadegan, F., Brussaard, C. P. D., and Wassmann, P. (2005). Dynamics
of
transparent
exopolymeric
particles
(TEP)
production January 2017 | Volume 7 | Article 2159 Frontiers in Microbiology | www.frontiersin.org 11 TEP in the Mediterranean Sea Ortega-Retuerta et al. dissolved organic matter pools. A seasonal field study in an oligotrophic
coastal system (Blanes Bay, NW Mediterranean). Mar. Chem. 148, 44–51. doi: 10.1016/j.marchem.2012.10.004 Van Oostende, N., Harlay, J., Vanelslander, B., Chou, L., Vyverman, W.,
and
Sabbe,
K. (2012). Phytoplankton
community
dynamics
during
late spring coccolithophore blooms at the continental margin of the
Celtic Sea (North East Atlantic, 2006–2008). Progr. Oceanogr. 104, 1–16. doi: 10.1016/j.pocean.2012.04.016 Sala, M. M., Aparicio, F., Balagué, V., Boras, J. A., Borrull, E., Cardelús, C.,
et al. (2016). Contrasting effects of ocean acidification on the microbial
food web under different trophic conditions. ICES J. Mar. Sci. 73, 670–679. doi: 10.1093/icesjms/fsv130 Van Oostende, N., Moerdijk-Poortvliet, T. C., Boschker, H. T., Vyverman,
W., and Sabbe, K. (2013). Release of dissolved carbohydrates by Emiliania
huxleyi and formation of transparent exopolymer particles depend on
algal life cycle and bacterial activity. Environ. Microbiol. 15, 1514–1531. doi: 10.1111/j.1462-2920.2012.02873.x Sala, M. M., Peters, F., Gasol, J. REFERENCES M., Pedrós-Alió C., Marrasé C., and Vaqué
D. (2002). Seasonal and spatial variations in the nutrient limitation of
bacterioplankton growth in the Northwestern Mediterranean. Aquat. Microb. Ecol. 27, 47–56. doi: 10.3354/ame027047 Vila-Reixach, G., Gasol, J. M., Cardelús, C., and Vidal, M. (2012). Seasonal
dynamics and net production of dissolved organic carbon in an oligotrophic
coastal environment. Mar. Ecol. Progr. Ser. 456, 7–19. doi: 10.3354/meps09677 Scoullos, M., Plavši´c, M., Karavoltsos, S., and Sakellari, A. (2006). Partitioning
and distribution of dissolved copper, cadmium and organic matter in
Mediterranean marine coastal areas: the case of a mucilage event. Estuarine
Coast. Shelf Sci. 67, 484–490. doi: 10.1016/j.ecss.2005.12.007 Wurl, O., Miller, L., and Vagle, S. (2011). Production and fate of transparent
exopolymer
particles
in
the
ocean. J. Geophys. Res. Oceans
116. doi: 10.1029/2011JC007342 Simon, M., and Azam, F. (1989). Protein content and protein synthesis
rates of planktonic marine bacteria. Mar. Ecol. Prog. Ser. 51, 201–213. doi: 10.3354/meps051201 Zhou, J., Mopper, K., and Passow, U. (1998). The role of surface-active
carbohydrates
in
the
formation
of
transparent
exopolymer
particles
by bubble adsorption of seawater. Limnol. Oceanogr. 43, 1860–1871. doi: 10.4319/lo.1998.43.8.1860 p
Smith, D. C., and Azam, F. (1992). A simple, economical method for measuring
bacterial protein synthesis rates in seawater using 3H-leucine. Mar. Microb. Food Webs 6, 107–114. Conflict of Interest Statement: The authors declare that the research was
conducted in the absence of any commercial or financial relationships that could
be construed as a potential conflict of interest. Sun, C. C., Wang, Y. S., Li, Q. P., Yue, W. Z., Wang, Y. T., Sun, F. L., et al. (2012). Distribution characteristics of transparent exopolymer
particles in the Pearl River estuary, China. J. Geophys. Res. Biogeosci. 117. doi: 10.1029/2012JG001951 Taylor, J. D., Cottingham, S. D., Billinge, J., and Cunliffe, M. (2014). Seasonal microbial community dynamics correlate with phytoplankton-
derived polysaccharides in surface coastal waters. ISME J. 8, 245–248. doi: 10.1038/ismej.2013.178 Copyright © 2017 Ortega-Retuerta, Sala, Borrull, Mestre, Aparicio, Gallisai,
Antequera, Marrasé, Peters, Simó and Gasol. This is an open-access article
distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the
original author(s) or licensor are credited and that the original publication in this
journal is cited, in accordance with accepted academic practice. No use, distribution
or reproduction is permitted which does not comply with these terms. Thornton, D. C. O. (2004). Frontiers in Microbiology | www.frontiersin.org January 2017 | Volume 7 | Article 2159 REFERENCES Formation of transparent exopolymeric particles
(TEP)
from
macroalgal
detritus. Mar. Ecol. Prog. Ser. 282,
1–12. doi: 10.3354/meps282001 January 2017 | Volume 7 | Article 2159 Frontiers in Microbiology | www.frontiersin.org 12
|
https://openalex.org/W4306167011
|
https://www.isprs-ann-photogramm-remote-sens-spatial-inf-sci.net/X-4-W3-2022/65/2022/isprs-annals-X-4-W3-2022-65-2022.pdf
|
English
| null |
OPTIMIZING URBAN MONITORING BETWEEN STATIONARY, OPPORTUNISTIC VEHICULAR, AND HYBRID SENSING
|
ISPRS annals of the photogrammetry, remote sensing and spatial information sciences
| 2,022
|
cc-by
| 8,052
|
1.
INTRODUCTION backs in their spatial and temporal coverage. Several studies re-
vealed already the benefits of OVS in the sensing coverage over
SS, but only at a given spatio-temporal granularity depending
on phenomena being monitored in those studies (Chen et al.,
2017, Gao et al., 2016, O’Keeffe et al., 2019). However, the
required spatio-temporal granularity varies with the properties
of phenomena being monitored and the applications in mind. For instance, a high spatial density is needed for monitoring
noise, whereas temperature can be captured with lower spatial
granularity (Anjomshoaa et al., 2021). In the temporal domain,
monitoring street surface quality is less sensitive to temporal
granularity compared to monitoring traffic flow. Besides, task
initiators allocate budgets to the sensing, reflecting the value of
the generated information. These budgets determine how many
sensors can be deployed for sensing tasks, creating an optim-
ization problem for the coverage (Asprone et al., 2021): Given
the requirements of a certain task, in terms of spatial granular-
ity, temporal granularity, and budget limitations, the task is to
determine which sensing method is superior to the other and
should be chosen by an agency. Over the past decades, with the advance in the technology
of low-cost sensors and wireless communication, wireless
sensor networks (WSNs) have evolved to a new data collec-
tion paradigm for urban monitoring at a large scale (Zhao and
Guibas, 2004, Duckham, 2013). The new paradigm plays a
significant role in smart city domain by empowering city com-
puting for decision makers (Anjomshoaa et al., 2018, Lee et
al., 2020). Compared with traditional stationary environmental
monitoring stations, the low-cost sensors can be deployed in
larger numbers at finer spatial granularity (Mao et al., 2012). WSNs have been applied for example in weather monitoring in
the wild (Barrenetxea et al., 2008), and air pollution monitoring
in urban environments (Boubrima, 2019, Mao et al., 2012). Low-cost sensors can be deployed in stationary sensing (SS)
to collect data at selected locations over long periods of time,
thus, with static and limited spatial coverage (O’Keeffe et al.,
2019, Anjomshoaa et al., 2021). Alternatively, those sensors
can also be mounted on mobile devices, e.g., smartphones (Ji
et al., 2016), drones (Yanmaz et al., 2018), and vehicles (Apte
et al., 2017, Lee and Gerla, 2010), or be directly carried by
humans (Ma et al., 2020). ∗Corresponding author Commission IV, WG IV/9 KEY WORDS: Opportunistic sensing, Stationary sensor networks, Vehicular sensor networks, Coverage, Urban sensing ABSTRACT: Urban monitoring based on wireless sensor networks is a recent paradigm that exploits a large number of low-cost sensors deployed
in certain places or/and on mobile devices to collect data ubiquitously at a large scale. In this study, we explore and compare
the coverage of stationary and opportunistic vehicular sensing methods with respect to the requirements of a task at hand. We
distinguish spatial granularity, temporal granularity, and budget constraints. First we compare the spatio-temporal coverage of
stationary sensing and opportunistic vehicular sensing for various tasks, which demonstrates that these two sensing methods are
suitable for different tasks. Then we propose a hybrid sensing deployment framework integrating a genetic algorithm to achieve
the maximum spatio-temporal coverage for specific tasks. Experiments with a real-world vehicle trajectory dataset demonstrate
that the proposed hybrid sensing framework achieves the maximum spatio-temporal coverage in various tasks. Our results provide
fundamental guidelines on network planning for urban monitoring applications. ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume X-4/W3-2022
7th International Conference on Smart Data and Smart Cities (SDSC), 19–21 October 2022, Sydney, Australia ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume X-4/W3-2022
7th International Conference on Smart Data and Smart Cities (SDSC), 19–21 October 2022, Sydney, Australia This contribution has been peer-reviewed. The double-blind peer-review was conducted on the basis of the full paper.
https://doi.org/10.5194/isprs-annals-X-4-W3-2022-65-2022 | © Author(s) 2022. CC BY 4.0 License. OPTIMIZING URBAN MONITORING BETWEEN STATIONARY, OPPORTUNISTIC
VEHICULAR, AND HYBRID SENSING W. Hu 1,∗, S. Winter 1, K. Khoshelham 1 1 Department of Infrastructure Engineering, The University of Melbourne, Parkville, VIC 3010, Australia
(wenyan1@student., winter@, k.khoselham@)unimelb.edu.au 1 Department of Infrastructure Engineering, The University of Melbourne, Parkville, VIC 3010, Australia
(wenyan1@student., winter@, k.khoselham@)unimelb.edu.au Commission IV, WG IV/9 1.
INTRODUCTION 65 ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume X-4/W3-2022
7th International Conference on Smart Data and Smart Cities (SDSC), 19–21 October 2022, Sydney, Australia ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume X-4/W3-2022
7th International Conference on Smart Data and Smart Cities (SDSC), 19–21 October 2022, Sydney, Australia ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume X-4/W3-2022
7th International Conference on Smart Data and Smart Cities (SDSC), 19–21 October 2022, Sydney, Australia ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume X-4/W3-2022
7th International Conference on Smart Data and Smart Cities (SDSC), 19–21 October 2022, Sydney, Australia on the known number and placements of stationary sensors to
obtain a coverage (Zhang and Fok, 2017, Zygowski and Jaekel,
2018). To the best of our knowledge, no study explores how
to co-deploy stationary sensors in HS. The challenge in co-
deployment is to achieve the optimal spatio-temporal cover-
age for a limited number of sensors with an appropriate split
between stationary and mobile sensors. The research questions
that need to be addressed include a) can HS achieve higher
spatio-temporal coverage than SS and OVS for a given task? b) what is the optimal configuration of SS and OVS in hybrid
sensing? To answer these questions, we propose a HS frame-
work that optimizes the deployment of HS to achieve an ex-
pected optimal spatio-temporal coverage according to historical
trajectories of candidate vehicles and task requirements. into unit cells, where the required spatial granularity (Gs) de-
termines the cell base, and the cell height is determined by the
required temporal granularity (Gt). The task requires the cov-
erage of each unit cell by at least one sensor. For instance, in
Figure 1(a) the target area is spatially partitioned into four cells,
denoted as S = {A, B, C, D}), and the time axis is divided into
eight unit intervals, denoted as T = {t1, ...t8}. We say, Gs is
2 × 2 (Ji et al., 2016, Wu et al., 2019), Gt is T/8, and the total
number of cells is 32. The main contributions of this study can be summarized as fol-
lows: • We demonstrate that SS can achieve a higher spatio-
temporal coverage over OVS for some tasks, and vice
versa. 2.
METHODOLOGY In this section, first, we present the formal definition of terms
regarding the sensing task and spatio-temporal coverage (Sec-
tion 2.1). Then, a framework for HS deployment is developed
to obtain the maximum spatio-temporal coverage (Section 2.2). Ii,j =
(
1
if si is covered by at least one sensor during tj
0
otherwise
(1) (1) (
(1) 1.
INTRODUCTION • We propose a HS framework integrating an objective func-
tion, a genetic algorithm, and a stationary sensor deploy-
ment mechanism to optimize HS for urban monitoring
with the maximum spatio-temporal coverage. The frame-
work solves an NP-hard problem heuristically, resulting in
the solution with a high sensing coverage. Figure 1. Spatial and spatiotemporal coverage ratios of
stationary, opportunistic vehicular and hybrid sensing, with two
sensors, Gs = 2 × 2 ({A, B, C, D}), Gt = T /8, and candidate
vehicles V = {v1, v2, v3}. Definition 2 : The number of sensors N . In a sensing task,
only a limited number of sensors can be used to complete the
task due to budget limitations. For instance, in Figure 1(b)
shows the base cells that each vehicle vi visits at each time tj . In
this case, there are three candidate vehicles (V = {v1, v2, v3}),
but only two vehicles might get recruited due to budget limita-
tions, such that the number of sensors is N = 2 in this monitor-
ing task. Figure 1(c) shows all three possible combinations of
choosing two vehicles out of the three. Similarly, Figure 1(d)
shows results for two stationary sensors deployed alternatively
in A, B, C or D. Figure 1(e) shows results for one stationary
sensor and one vehicular sensor. • We evaluate SS, OVS, and HS on a substantial trajectory
dataset from a global city, and use the case study also
to verify the reliability and validity of our proposed HS
framework. Accordingly, this paper is organized as follows. In Section 2,
the definition of spatio-temporal coverage and the proposed hy-
brid sensing method are introduced. In Section 3, we evalu-
ate the different sensing methods and present the computational
results obtained by applying the methods in real scenarios. The
major conclusions and research limitations are presented in
Section 4. Definition 3 : Unit cell coverage Ii,j . I i,j represents whether
the unit cell si is covered in time period tj or not. The equation
is defined as: This contribution has been peer-reviewed. The double-blind peer-review was conducted on the basis of the full paper.
https://doi.org/10.5194/isprs-annals-X-4-W3-2022-65-2022 | © Author(s) 2022. CC BY 4.0 License. 1.
INTRODUCTION Mobile sensors are able to monitor
certain phenomena in places traversed by their hosts, i.e., at
varying locations. In opportunistic vehicular sensing (OVS), a
common kind of mobile sensing, mobile sensors are deployed
on a set of existing vehicles without any influence on the routes
of these vehicles (Anjomshoaa et al., 2021). Thus, OVS im-
proves the spatial coverage compared to SS, but at the cost of
temporal coverage (Boubrima, 2019). To the best of our knowledge, this issue has not been addressed
in the literature. To fill the gap, this paper addresses a num-
ber of research questions. The initial one is: For which tasks
does the spatio-temporal coverage of one sensing method ex-
ceed the other? To answer the question, we estimate and com-
pare the spatio-temporal coverage of SS and OVS for tasks with
various requirements. The next research questions address the
optimal deployment of a limited number of sensors to achieve
a high spatio-temporal coverage, which has been a key prob-
lem in urban monitoring. Existing studies mainly focus either
on SS (Mao et al., 2012, Chakrabarty et al., 2002) or vehicu-
lar sensing (He et al., 2015, Asprone et al., 2021, Zhao et al.,
2015, Cortes et al., 2004) to achieve maximum coverage. Al-
though several studies have discussed hybrid sensing (HS), con-
sisting of stationary and mobile sensors, these studies aim at
optimizing vehicle selection or moving route planning based The concept of coverage is a fundamental metric to evaluate
the sensing quality (Chen et al., 2017, Ghosh and Das, 2008,
Zhao et al., 2015), and it includes both the spatial and tem-
poral domains. SS and OVS each have their merits and draw- his contribution has been peer-reviewed. The double-blind peer-review was conducted on the basis of the full paper. https://doi.org/10.5194/isprs-annals-X-4-W3-2022-65-2022 | © Author(s) 2022. CC BY 4.0 License. The only constraint is the number of sensors N: The only constraint is the number of sensors N: The only constraint is the number of sensors N: In OVS, to calculate SC, vehicle trajectories are analysed to
obtain the vehicles trajectory matrix Mv. For example, in Fig-
ure 1(b), vehicle v1 drove through base cells B and D in time
period t5, i.e., SC5 = 2, RSC 5 = 0.5. Similarly, SC and
RSC can be computed for multiple sensors. For example, Fig-
ure 1(c) shows the spatial coverage ratio of two vehicles at each
time period. In SS, RSC is constant during the sensing time T
and only depends on the number of sensors and their (fix) dis-
tribution over all base cells. For instance, if sensors are placed
in base cells A and B, the RSC is always 0.5 (Figure 1(d)). In
the following we assume an optimal distribution, i.e., no more
than one sensor per base cell. In a similar way, the RSC of HS
formed by one stationary sensor and one vehicle is shown in
Figure 1(e). Subject To: |V ′| + |S′| = N. Note that the optimal sensing deployment does not always re-
quire a mix of stationary and vehicular sensors. Single-mode
sensing deployment might also satisfy the objective. 2.2.2
Genetic algorithm design
In this study, we mainly
focus on how many stationary sensors and/or vehicular sensors
should be deployed, and where the stationary sensors should
be placed. Vehicle recruitment optimization has been discussed
in previous works (He et al., 2015, Zhao et al., 2015). In this
study, we only focus on the deployment of stationary sensors
ignoring vehicle selection. Definition 5: Spatio-temporal coverage ratio RST C. RST C
is defined as the average coverage ratio of base cells during time
period T (T = {t1, t2, ...tn}). Formally, it is defined as: Each base cell has two states, with or without stationary sensors. As the time complexity of working out the optimal solution is
O(2|S|), the problem above is an NP-hard problem. Thus, to
solve the problem, it is necessary to trade off between com-
plexity and optimality. The genetic algorithm (GA), a classic
evolutionary algorithm for optimization, is inspired by the pro-
cess of natural selection of the fittest (Goldberg and Holland,
1988, Holland et al., 1992). On one hand, the algorithm can
find various alternative solutions in the solution population. Given: • A total of m base cells S = {s1, s2, ..., sm}, depending on
Gs; • A total of m base cells S = {s1, s2, ..., sm}, depending on
Gs; • q candidate vehicles in the pool from which |V ′| vehicles
can be recruited for the sensing task, V = {v1, v2, ... vq}
and |V ′| ≤q. • N, the number of sensors (stationary or/and vehicular
sensors) that can be afforded for a given budget; Figure 3 shows the process of GA. First, a certain number (pop-
ulation size) of potential solutions, each representing a place-
ment scheme for stationary sensors, are randomly generated as
the initial population (i.e., the first generation). Then in sub-
sequent iterations, all the individuals in the current generation
crossover in pairs to generate new individuals, simulating the
genetic recombination process. For instance, as shown in Fig-
ure 2, the genes in s4 to s6 are crossed for these two solutions,
generating two new individuals. Note that crossover may gen-
erate invalid solutions if the sensor number constraint is viol-
ated. To keep the number of sensors fixed in each solution,
any increase or decrease in the number of stationary sensors
in the crossover gene segment is compensated for by remov-
ing or adding a stationary sensor from the other gene segments. • the historical vehicle trajectory matrix Mv according to q,
Gs and Gt; The only constraint is the number of sensors N: On
the other hand, the combination of directional search and ran-
dom search in the genetic algorithm provides a good trade-off
between finding the optimal solution and limiting the search
space. Owing to these characteristics, the genetic algorithm
has been extensively applied to objective optimization prob-
lems (Kim et al., 2008, Tian et al., 2016). It has also been
applied to optimize urban monitoring for selecting vehicles in
OVS (He et al., 2015). In the genetic algorithm, three key issues
need to be addressed: genetic representation, population initial-
ization, and genetic operators (including crossover, mutation,
and selection). RST C =
P|T |
j=1 RSC j
|T|
(4) (4) As shown in Figure 1(c) and (d), RST C of v2 and v3 is 0.666 in
OVS, while RST C of two stationary sensors in any two of the
four base cells is always 0.5. This contribution has been peer-reviewed. The double-blind peer-review was conducted on the basis of the full paper.
https://doi.org/10.5194/isprs-annals-X-4-W3-2022-65-2022 | © Author(s) 2022. CC BY 4.0 License. 2.2
Hybrid sensing framework For a given number of sensors, the best spatio-temporal cover-
age may not be obtained by SS or OVS alone but with a com-
bination of the two. To achieve a maximum spatio-temporal
coverage with a limited number of sensors, an optimized HS
deployment framework is proposed. We formulate the HS de-
ployment problem as an optimization problem (Section 2.2.1). Then, a genetic algorithm is proposed to solve the optimization
problem (Section 2.2.2). Genetic representation encodes the candidate solutions of an
optimization problem into variable arrays called gene strings. For the optimization problem of HS deployment, each potential
solutions is represented as a gene string whose length equals to
the number of base cells. For each character in the string, we
use 1 to indicate a stationary sensor is placed in the base cell
and 0 otherwise. For instance, in Figure 2, Gs is 3 × 3 (i.e., the
number of base cells is 9). The string (0,1,1,0,1,0,0,1,1) stands
for stationary sensors being placed in base cells s2, s3, s5, s8
and s9. The fitness (f) of each individual gene string is repres-
ented by RST C of the corresponding combination of stationary
and vehicular sensors. The larger the fitness f, the better the
individual. 2.2.1
Problem formulation
For the problem formulation,
the input includes Gs, Gt, N, and the historical trajectories of
candidate vehicles. So 2.1
Definitions Definition 4: Spatial coverage ratio RSC. Ideally, each unit
cell is expected to have at least one sensor (or one vehicle
passing) to monitor the requested phenomenon. RSC is defined
as the proportion of spatial coverage at every time period. The
number of grid cells covered at time period tj is denoted as
SCj, and defined as follows: In this section, we give definitions regarding the sensing task
and spatio-temporal coverage of sensing. A sensing task usually has its certain task requirements and
budget limitations. The task – i.e., the nature of the phe-
nomenon being monitored and the application in mind – de-
termines the required spatio-temporal granularity. The budget
limitations restrict the number of sensors – stationary sensors as
well as vehicular sensors. Therefore, a task is defined as mon-
itoring a phenomenon in a target area S over a certain sensing
cycle T by either SS, or OVS, or both. SCj =
|S|
X
i=1
Ii,j
(2) (2) Based on Equation 2, RSC at a time period tj can be represen-
ted as follows: Definition 1: Spatio-temporal granularity. According to the
requirements of the task at hand, the target area can be divided RSC j = SCj
|S|
(3) (3) This contribution has been peer-reviewed. The double-blind peer-review was conducted on the basis of the full paper. https://doi.org/10.5194/isprs-annals-X-4-W3-2022-65-2022 | © Author(s) 2022. CC BY 4.0 License. 6 66 ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume X-4/W3-2022
7th International Conference on Smart Data and Smart Cities (SDSC), 19–21 October 2022, Sydney, Australia then the objective of the problem is to achieve maximum spatio-
temporal coverage RST C: Objective: Select a subset of base cells S′ ⊂S to place sta-
tionary sensors, and a set of candidate vehicles V ′ ⊂V such
that: S′ = argmax
S′⊂S
RST C(S). S′ = argmax
S′⊂S
RST C(S). This contribution has been peer-reviewed. The double-blind peer-review was conducted on the basis of the full paper. https://doi.org/10.5194/isprs-annals-X-4-W3-2022-65-2022 | © Author(s) 2022. CC BY 4.0 License. 67 ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume X-4/W3-2022
7th International Conference on Smart Data and Smart Cities (SDSC), 19–21 October 2022, Sydney, Australia Figure 2. An example of representation, crossover and mutation in the genetic algorithm. Figure 2. An example of representation, crossover and mutation in the genetic algorithm. For instance, in Figure 2, after crossover, the first new indi-
vidual will have an additional stationary sensor, so a randomly
selected gene character encoded as 1 (here s9) is changed to 0. Similarly, the second new individual will have one less sensor,
so the gene in s9 changes to 1 from 0. This step guarantees
the number of stationary sensors remain unchanged for each
new individual, thereby avoiding invalid solutions. After cros-
sover, each new individual has a certain probability of mutation
depending on mutation rate, i.e., the content of a solution is
changed randomly, to simulate the genetic drift in nature. For
instance, in Figure 2, the gene characters from s3 to s6 are re-
versed after mutation. At the end of the iteration, we employ
the tournament selection method (Miller et al., 1995) to keep
only a proportion of better individuals as the new generation
for the next iteration and discard the rest by ranking individuals
in order of their fitness. To be specific, in the tournament, sev-
eral individuals in the generation are randomly selected to form
a group. After a certain number of groups are formed, solutions
with the highest fitness (RST C) in each group remain to gen-
erate the next generation population and others are discarded
(see Section 3.2.2 for detailed parameter settings). Finally, in
the last generation (final iteration), the best individual with the
highest fitness is chosen as the best solution representing the
optimized HS deployment. changing phenomena in this city is important for policymakers
and city managers to make decisions, for scientists to explore
urban changes, and for industries to develop novel services. 1 The minimum bounding box of the target area is divided into 10 rows
and 10 columns resulting in 100 cells 3. CASE STUDY AND RESULTS Table 1. Data and experimental setup. This section presents the results of the proposed methods ap-
plied to a case study of Shanghai, China. Firstly, the study area
and data are introduced in Section 3.1, and then the results of
the case study are shown in Section 3.2. 3.2
Results Results are presented in four parts. First, the RST C of SS and
OVS are calculated and compared (Section 3.2.1). Second, the
performance of the proposed HS framework is shown by com-
paring its solution with SS and OVS in terms of spatio-temporal
coverage (Section 3.2.2). Then, the reliability and validity of
HS proposed by our framework is evaluated (Section 3.2.3). This contribution has been peer-reviewed. The double-blind peer-review was conducted on the basis of the full paper.
https://doi.org/10.5194/isprs-annals-X-4-W3-2022-65-2022 | © Author(s) 2022. CC BY 4.0 License. then the objective of the problem is to achieve maximum spatio-
temporal coverage RST C: To explore the spatio-temporal coverage of OVS, we use a
large-scale dataset of taxi GNSS trajectories in Shanghai, which
are provided by Qiangsheng, a large and city-owned taxi com-
pany based in Shanghai. The dataset contains the GNSS tra-
jectories of more than 12,000 taxis during the period of April 1
to April 7 (Sunday to Saturday) (Table 1). The GNSS sampling
frequency is about 10 s. Each GNSS recording is denoted by a
tuple (taxi ID, timestamp, longitude and latitude). To facilitate the analysis, we conduct data preprocessing ac-
cording to the following five steps: • Step 1: Removing all trajectory points that are located out-
side of the target area. • Step 2: Converting coordinates to the World Geodetic
System 1984 as a common spatial reference system. • Step 3: Assuming a sensing frequency of 60 s, i.e., res-
ampling each trajectory at 60 s intervals. • Step 4: To explore the RST C of SS and OVS in differ-
ent sensing tasks, a number of sensing tasks are generated. We generate 7000 different tasks with Gs ranging from
10 × 101 to 160 × 160, Gt from 1 h to 12 h and N ranging
from 1 to 1000. Table 1 summarises the requirements of
these tasks. Figure 3. The flowchart of a genetic algorithm. • Step 5: Combining historical vehicle trajectories with Gs
and Gt, vehicle trajectory matrix is generated for each
task. • Step 5: Combining historical vehicle trajectories with Gs
and Gt, vehicle trajectory matrix is generated for each
task. Table 1. Data and experimental setup. Dataset
Taxi trajectories
Period
1-7 April 2018 (Sunday to Saturday)
Time interval after data preprocessing
1 min
# taxis in the dataset
12025
Gs (base cells)
10×10, 20×20, 40×40, 80×80
and 160 × 160
Gt (h)
1, 2, 3, 4, 6, 8 and 12
N
1 to 1,000 Table 1. Data and experimental setup. Dataset
Taxi trajectories
Period
1-7 April 2018 (Sunday to Saturday)
Time interval after data preprocessing
1 min
# taxis in the dataset
12025
Gs (base cells)
10×10, 20×20, 40×40, 80×80
and 160 × 160
Gt (h)
1, 2, 3, 4, 6, 8 and 12
N
1 to 1,000 Figure 3. The flowchart of a genetic algorithm. 3.1 Study area and data However, when the number of sensors
reaches the number of base cells in the target area, RST C of SS
reaches the maximum (i.e., RST C = 1) and does not increase
with the increase of sensors any further (Figure 4 (a) to (c)). creasingly advantageous. For instance, when Gs is 80 × 80 or
160×160, due to the limited number of sensors, OVS is advant-
ageous, obtaining a higher spatio-temporal coverage (Figure 4
(e)). 3.2.2
Performance of the hybrid sensing framework
To
demonstrate the performance of the HS framework, we assume
historical trajectories of candidate vehicles from the past four
days (1-4 April 2018) are known. We compare the RST C of
the designed HS with that of SS and OVS for various tasks. Six tasks with different Gs and N are selected from the pool
of tasks generated in the previous experiment (Section 3.2.1) as
shown in Table 2, consisting of three tasks (Task 1 to 3) where
SS achieved a higher spatio-temporal coverage and three tasks
(Task 4 to 6) where OVS achieved a higher spatio-temporal cov-
erage. For SS, RST C depends only on Gs and N, thus it is the
same in each iteration. For OVS, since any candidate vehicle
could be selected for the task, we randomly select vehicles, and
calculate RST C in each iteration. We then take the maximum
RST C over all iterations. For the hybrid approach, the largest
RST C in each generation is the result of the iteration. After
many tests, the following parameter settings were empirically
found appropriate: the number of iterations is 200; the muta-
tion rate is 0.25; the immigration rate is 0.2; the population size
is 120; the number of groups and the number of members in
each group is 20 and 10 for tournament selection, respectively. By comparing the RST C of SS and OVS, it is revealed that OVS
is not always better than SS (Figure 4). When the spatial gran-
ularity of a task is coarse, a small number of stationary sensors
can cover the entire target area, while OVS cannot ensure that
vehicles pass each base cell at each time interval due to the ar-
bitrary nature of the vehicle trajectories. 3.1 Study area and data In the case study, the study area (i.e., the target area of the sens-
ing task) is the Shanghai land administrative region (within lat-
itude 30◦42′ to 31◦52′ N, and longitude 120◦52′ to 121◦58′ E),
the most populous city in China. It is an international center of
economics, trade, science and technology, and has 16 admin-
istrative districts with a total area of 6340.5 km2. Monitoring 1 The minimum bounding box of the target area is divided into 10 rows
and 10 columns resulting in 100 cells his contribution has been peer-reviewed. The double-blind peer-review was conducted on the basis of the full paper. https://doi.org/10.5194/isprs-annals-X-4-W3-2022-65-2022 | © Author(s) 2022. CC BY 4.0 License. 68 ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume X-4/W3-2022
7th International Conference on Smart Data and Smart Cities (SDSC), 19–21 October 2022, Sydney, Australia (a) 10 × 10
(b) 20 × 20
(c) 40 × 40
(d) 80 × 80
(e) 160 × 160
Figure 4. The ratio of spatio-temporal coverage in stationary and opportunistic vehicular sensing for various tasks. (c) 40 × 40 (b) 20 × 20 (a) 10 × 10 (a) 10 × 10 (b) 20 × 20 (c) 40 × 40 (e) 160 × 160 (d) 80 × 80 (d) 80 × 80 (e) 160 × 160 he ratio of spatio-temporal coverage in stationary and opportunistic vehicular sensing for various tasks. 3.2.1
Analysis of RST C in SS and OVS
The RST C of SS
and OVS for various tasks is calculated using seven days vehicle
trajectories (i.e., the sensing cycle for all tasks is seven days)
(Table 1) based on Equations 1 to 4. Figure 4 shows the RST C
of 7000 tasks with various spatial, temporal, and budget task
requirements. The x-axis, y-axis and z-axis indicate N, Gt and
RST C, respectively. The green and blue scatters correspond
to SS and OVS, respectively. Meanwhile, the circles represent
tasks where the RST C of OVS is superior to that of SS, while
the triangles represent tasks where SS can achieve a higher
spatio-temporal coverage. Intuitively, for a fixed number of
sensors, RST C is higher when the spatio-temporal granularity is
coarser. In terms of the same spatio-temporal granularity, when
the number of sensors is increased, the RST C of both SS and
OVS becomes larger. This contribution has been peer-reviewed. The double-blind peer-review was conducted on the basis of the full paper.
https://doi.org/10.5194/isprs-annals-X-4-W3-2022-65-2022 | © Author(s) 2022. CC BY 4.0 License. 3.1 Study area and data defined according to the HS deployment, whereas vehicles are
randomly selected. To evaluate the generalization of the frame-
work for tasks with multiple sensing cycles, RST C of the HS
is calculated in three sensing cycles. For HS-II where vehicles
are randomly selected, we conducted 200 tests to calculate the
mean value of RST C to reduce bias. In terms of OVS, we select
the same vehicles for the selected tasks, and calculate RST C of
OVS in three sensing cycles. For SS, coverage is independent
of vehicle trajectories, so it does not change. Figure 6(a)-(c) show RST C of SS, OVS and HS during the de-
ployment (based on vehicle trajectories from 1-4 April 2018)
and sensing task execution (based on vehicle trajectories from
5-7 April 2018). The green bars correspond to SS; the blue bars
with and without grey lines correspond to OVS at deployment
and at sensing task execution, respectively; the pure red bars
correspond to HS at deployment, the red bars with grey lines
and grey grids indicate HS-I and HS-II during task execution,
respectively. Error bars on red bars with grey grid represent the
standard deviation of RST C of HS-II. Overall, HS can obtain
higher spatio-temporal coverage than OVS and SS, regardless
of employing HS-I or HS-II, except for Task 5. More import-
antly, when the vehicle selection is based on historical traject-
ories (during deployment), the predicted spatio-temporal cover-
age is very close to the actual coverage obtained from current
trajectories (during task execution), demonstrating the valid-
ity of the predictions obtained from historical trajectories. In
addition, comparing RST C values for different sensing cycles
reveals that RST C does not change significantly for different
sensing cycles. This demonstrates that the proposed framework
is not sensitive to the sensing cycle, so it can be applied to tasks
with various sensing cycles. For Task 5, RST C of HS is very
similar to that of OVS during the optimal deployment, so spatio-
temporal coverage obtained by HS-I is still similar to OVS in
task execution. However, RST C of HS-II is slightly lower than
that of OVS. This is because the selection of vehicles and the
selection of stationary sensor locations is a coupled process in
optimal deployment. The placement of stationary sensors is op-
timized based on vehicle selection, which also shows that HS-I
is consistently superior to HS-II. 3.1 Study area and data For instance, for a spa-
tial granularity of 10 × 10 base cells (which is equivalent here
to a total of 75 base cells since 25 base cells are located on wa-
ter or outside the target area), 75 stationary sensors are enough
to monitor the whole area, while the RST C of OVS with 1000
vehicles can only reach close to 0.9 (Figure 4 (a)). On one hand,
this is because vehicles can only travel along roads, but in rural
areas where roads are sparse, there may be fewer roads or even
no roads in some of the base cells. On the other hand, there
are fewer vehicles in the suburbs than in the city centre. Thus,
few vehicles or no vehicle pass by these base cells. However,
for tasks with the denser spatial granularity, OVS becomes in- Table 2 shows RST C of six tasks. In Task 1, the result of HS
is consistent with SS, i.e., the maximum spatio-temporal cov-
erage is achieved by deploying only stationary sensors and no
vehicular sensors are needed. Compared to SS, RST C of HS
improves by 20.31 % for Task 3, and 17.33 % for Task 2. For
tasks where OVS can obtain a higher spatio-temporal coverage
than SS, HS does not always outperform OVS. For instance,
OVS can obtain a higher spatio-temporal coverage without sta-
tionary sensors for Task 6, so no stationary sensor is deployed his contribution has been peer-reviewed. The double-blind peer-review was conducted on the basis of the full paper. https://doi.org/10.5194/isprs-annals-X-4-W3-2022-65-2022 | © Author(s) 2022. CC BY 4.0 License. 69 ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume X-4/W3-2022
7th International Conference on Smart Data and Smart Cities (SDSC), 19–21 October 2022, Sydney, Australia Table 2. Six tasks and the ratio of spatio-temporal coverage in stationary, opportunistic vehicular and hybrid sensing. 3.1 Study area and data To sum up, the proposed op-
timization framework for HS deployment produces valid estim-
ates of spatio-temporal coverage and can be generalized to tasks
with various sensing cycles. 3.2.3
Validation
When we apply the proposed HS frame-
work to optimize the sensor deployment, we predict the spatio-
temporal coverage based on historical vehicle trajectories. But
due to the arbitrary nature of vehicle trajectories, the predicted
spatio-temporal coverage may not reflect the actual sensing
coverage during the task execution. Therefore, we need a val-
idation of the assumption that vehicles selected based on their
historic trajectories would achieve a sensing coverage close to
predictions. To validate this assumption, we compare the pre-
dicted coverage from historical trajectories with the actual cov-
erage obtained from current trajectories. In Section 3.2.2, the deployment of HS was made using tra-
jectories over four days (1-4 April 2018). We will now valid-
ate the coverage estimates using trajectories over the next three
days (5-7 April 2018), supposing that the sensing task is sched-
uled for any of these days. We will focus on the reliability
and validity of these estimates, assuming that sensing cycles
of these tasks are one day (5 April 2018), two days (5-6 April
2018) and three days (5-7 April 2018). The actual RST C in the
sensing cycle is compared with the estimated RST C from the
HS deployment framework for Tasks 2 to 5 where mixed-mode
sensing achieves higher RST C. 3.1 Study area and data Task 1
Task 2
Task 3
Task 4
Task 5
Task 6
Gs
10 × 10
20 × 20
40 × 40
40 × 40
80 × 80
160 × 160
Gt
6 h
6 h
6 h
6 h
6 h
6 h
N
75
200
600
300
300
300
RST C of SS
1
0.75
0.64
0.32
0.09
0.02
RST C of OVS
0.64
0.53
0.56
0.47
0.37
0.26
RST C of HS
1
0.88
0.77
0.55
0.37
0.25
# vehicles in HS
0
23
92
155
293
300
# stationary sensors in HS
75
177
508
145
7
0 Six tasks and the ratio of spatio-temporal coverage in stationary, opportunistic vehicular and hybrid sensing. by the framework proposed. Meanwhile, for Task 5, only a few
stationary sensors are placed, and RST C of optimized HS is al-
most the same as RST C of OVS. However, this does not mean
that OVS is superior to HS for tasks where OVS outperforms
SS. For Task 4, HS improves RST C by 17.02 % compared to
OVS. To sum up, when Gs is coarse, with an adequate number
of sensors SS can achieve higher spatio-temporal coverage than
OVS and the optimized HS allocates only stationary sensors. However, when the provided sensors cannot cover all base cells,
the optimized HS can achieve a higher spatio-temporal cover-
age than single-mode sensing, i.e., SS or OVS. When Gs is fine,
it is not always feasible to provide stationary sensors covering
all base cells due to the limited budget. Hence, the advantages
of SS diminish, ane more vehicular sensors will be needed. For
some tasks with a very fine Gs requirement, OVS is slightly
better than HS. The framework we proposed can determine not
only how many stationary and vehicular sensors are needed in
HS (Table 2), but also which base cells stationary sensor should
be placed in. Figure 5 shows the optimal placement of the sta-
tionary sensors for the six tasks determined by the genetic al-
gorithm. The green color means a stationary sensor needs to be
placed in the base cell, while the red color means the base cell
does not need a stationary sensor. It can be seen that station-
ary sensors are mainly deployed close to the boundaries of the
target area – areas not well served by vehicles. This contribution has been peer-reviewed. The double-blind peer-review was conducted on the basis of the full paper.
https://doi.org/10.5194/isprs-annals-X-4-W3-2022-65-2022 | © Author(s) 2022. CC BY 4.0 License. 4.
CONCLUSIONS (a) 1-day sensing cycle
(b) 2-day sensing cycle
(c) 3-day sensing cycle
Figure 6. Comparison of the ratio of spatio-temporal coverage between deployment and task execution for different
sensor deployment in different tasks. Figure 6. Comparison of the ratio of spatio-temporal coverage between deployment and task execution for differen
sensor deployment in different tasks. tunistic vehicular sensors, thereby reaching maximum spatio-
temporal coverage. Evaluations show that the spatio-temporal
coverage of optimized HS designed by our framework outper-
forms SS and OVS by up to 20.31 %. In tasks where SS or
OVS can achieve a higher spatio-temporal coverage, the pro-
posed HS framework also finds the single-mode sensing as the
optimal solution. The validity and generalization ability for
various sensing cycles are also evaluated. The results show
that the performance of the proposed HS framework is super-
ior or similar to the single-mode sensing. Our study provides
fundamental guidelines on sensor network planning for urban
monitoring applications. tunistic vehicular sensors, thereby reaching maximum spatio-
temporal coverage. Evaluations show that the spatio-temporal
coverage of optimized HS designed by our framework outper-
forms SS and OVS by up to 20.31 %. In tasks where SS or
OVS can achieve a higher spatio-temporal coverage, the pro-
posed HS framework also finds the single-mode sensing as the
optimal solution. The validity and generalization ability for
various sensing cycles are also evaluated. The results show
that the performance of the proposed HS framework is super-
ior or similar to the single-mode sensing. Our study provides
fundamental guidelines on sensor network planning for urban
monitoring applications. but if there are other vehicle trajectories, we can still use the ap-
proach to deploy stationary sensors, and then randomly select
vehicles carrying sensors (i.e., HS-II in Section 3.2.3). Second,
we ignored some technical limitations of sensors during the pro-
cess of urban monitoring, e.g., signal instability and inaccuracy
caused by sensing devices and sensing conditions. Third, we
do not take into account the difference in price between placing
stationary sensors and installing vehicular sensors, and the later
maintenance. In the future, we will investigate on the optimal
solution for combination of vehicles and stationary sensors,
and incentive reward mechanism to achieve maximum spatio-
temporal coverage in terms of OVS. The results of our study need to be interpreted in consideration
of several limitations. First, we mainly focus on co-deploying
stationary sensors in the HS framework. 4.
CONCLUSIONS Although it is shown
that the results are still better than SS and OVS, the optimal
solution for combination of vehicles and stationary sensors can-
not be produced. However, our approach also has its advantage. In real-world, we may not have candidate vehicle trajectories, 4.
CONCLUSIONS This study explores the spatio-temporal coverage of SS and
OVS for various tasks, and proposes a HS framework to im-
prove the sensing coverage. Firstly, we calculated and com-
pared the spatio-temporal coverage of SS and OVS for various
tasks with various spatial, temporal and budget requirements. The results provided a new insight showing that the spatio-
temporal coverage of SS may exceed OVS for certain tasks. This study also developed a HS deployment framework integ-
rating a genetic algorithm to co-deploy stationary and oppor- Two deployment schemes according to the deployment de-
signed by the proposed framework are evaluated. HS-I is the
scheme where the selection of vehicles executing the sensing
task and the placement of stationary sensors are defined accord-
ing to the HS deployment; HS-II is the scheme where only the
placement of stationary sensors and the number of vehicles are 70 ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume X-4/W3-2022
7th International Conference on Smart Data and Smart Cities (SDSC), 19–21 October 2022, Sydney, Australia (a) Task 1
(b) Task 2
(c) Task 3
(d) Task 4
(e) Task 5
(f) Task 6
Figure 5. The places of stationary sensors in hybrid sensing for six tasks. (a) 1-day sensing cycle
(b) 2-day sensing cycle
(c) 3-day sensing cycle
Figure 6. Comparison of the ratio of spatio-temporal coverage between deployment and task execution for different
sensor deployment in different tasks. (a) Task 1
(b) Task 2
(c) Task 3
(d) Task 4
(e) Task 5
(f) Task 6
Figure 5. The places of stationary sensors in hybrid sensing for six tasks. (c) Task 3 (b) Task 2 (a) Task 1
(d) Task 4 (a) Task 1
(d) Task 4 (c) Task 3 (b) Task 2 (e) Task 5 (f) Task 6 (e) Task 5 (d) Task 4 Figure 5. The places of stationary sensors in hybrid sensing for six tasks. Figure 5. The places of stationary sensors in hybrid sensing for six tasks. Figure 5. The places of stationary sensors in hybrid sensing for six tasks. gu e 5. e p aces o stat o a y se so s
yb d se s g o s
tas s. (a) 1-day sensing cycle
(b) 2-day sensing cycle
(c) 3-day sensing cycle
Figure 6. Comparison of the ratio of spatio-temporal coverage between deployment and task execution for different
sensor deployment in different tasks. This contribution has been peer-reviewed. The double-blind peer-review was conducted on the basis of the full paper.
https://doi.org/10.5194/isprs-annals-X-4-W3-2022-65-2022 | © Author(s) 2022. CC BY 4.0 License. REFERENCES Anjomshoaa, A., Duarte, F., Rennings, D., Matarazzo, T. J., de-
Souza, P., Ratti, C., 2018. City scanner: Building and schedul-
ing a mobile sensing platform for smart city services. IEEE In-
ternet of Things Journal, 5(6), 4567–4579. This contribution has been peer-reviewed. The double-blind peer-review was conducted on the basis of the full paper. https://doi.org/10.5194/isprs-annals-X-4-W3-2022-65-2022 | © Author(s) 2022. CC BY 4.0 License. 71 ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume X-4/W3-2022
7th International Conference on Smart Data and Smart Cities (SDSC), 19–21 October 2022, Sydney, Australia ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume X-4/W3-2022
7th International Conference on Smart Data and Smart Cities (SDSC), 19–21 October 2022, Sydney, Australia Kim, K., Murray, A. T., Xiao, N., 2008. A multiobjective evol-
utionary algorithm for surveillance sensor placement. Environ-
ment and Planning B: Planning and Design, 35(5), 935–948. Anjomshoaa, A., Santi, P., Duarte, F., Ratti, C., 2021. Quantify-
ing the Spatio-Temporal Potential of Drive-by Sensing in Smart
Cities. Journal of Urban Technology, 28(1-2), 199-216. Lee, R., Jang, R.-y., Park, M., Jeon, G.-y., Kim, J.-k., Lee, S.-h.,
2020. Making IoT data ready for smart city applications. 2020
IEEE International Conference on Big Data and Smart Com-
puting (BigComp), 605–608. Apte, J. S., Messier, K. P., Gani, S., Brauer, M., Kirchstetter,
T. W., Lunden, M. M., Marshall, J. D., Portier, C. J., Vermeu-
len, R. C., Hamburg, S. P., 2017. High-resolution air pollution
mapping with Google street view cars: exploiting big data. En-
vironmental Science & Technology, 51(12), 6999–7008. Lee, U., Gerla, M., 2010. A survey of urban vehicular sensing
platforms. Computer Networks, 54(4), 527-544. Asprone, D., Di Martino, S., Festa, P., Starace, L. L. L., 2021. Vehicular crowd-sensing: a parametric routing algorithm to
increase spatio-temporal road network coverage. International
Journal of Geographical Information Science, 1–29. Ma, J., Tao, Y., Kwan, M.-P., Chai, Y., 2020. Assessing
mobility-based real-time air pollution exposure in space and
time using smart sensors and GPS trajectories in Beijing. An-
nals of the American Association of Geographers, 110(2), 434–
448. Barrenetxea, G., Ingelrest, F., Schaefer, G., Vetterli, M., Cou-
ach, O., Parlange, M., 2008. Sensorscope: Out-of-the-box en-
vironmental monitoring. 2008 International Conference on In-
formation Processing in Sensor Networks (ipsn 2008), IEEE,
332–343. Mao, X., Miao, X., He, Y., Li, X.-Y., Liu, Y., 2012. Citysee:
Urban CO2 monitoring with sensors. 2012 Proceedings IEEE
INFOCOM, IEEE, 1611–1619. Boubrima, A., 2019. REFERENCES Deployment and scheduling of wireless
sensor networks for air pollution monitoring. PhD thesis, Uni-
versit´e de Lyon. Miller, B. L., Goldberg, D. E. et al., 1995. Genetic al-
gorithms, tournament selection, and the effects of noise. Com-
plex Aystems, 9(3), 193–212. Chakrabarty, K., Iyengar, S. S., Qi, H., Cho, E., 2002. Grid cov-
erage for surveillance and target location in distributed sensor
networks. IEEE Transactions on Computers, 51(12), 1448–
1453. O’Keeffe, K. P., Anjomshoaa, A., Strogatz, S. H., Santi, P.,
Ratti, C., 2019. Quantifying the sensing power of vehicle fleets. Proceedings of the National Academy of Sciences, 116(26),
12752–12757. Chen, Y., Yonezawa, T., Nakazawa, J., Tokuda, H., 2017. Eval-
uating the spatio-temporal coverage of automotive sensing for
smart cities. 2017 Tenth International Conference on Mobile
Computing and Ubiquitous Network (ICMU), 1–5. Tian, J., Gao, M., Ge, G., 2016. Wireless sensor network node
optimal coverage based on improved genetic algorithm and bin-
ary ant colony algorithm. EURASIP Journal on Wireless Com-
munications and Networking, 2016(1), 1–11. Cortes, J., Martinez, S., Karatas, T., Bullo, F., 2004. Coverage
control for mobile sensing networks. IEEE Transactions on Ro-
botics and Automation, 20(2), 243–255. Wu, L., Xiong, Y., Wu, M., He, Y., She, J., 2019. A task assign-
ment method for sweep coverage optimization based on crowd-
sensing. IEEE Internet of Things Journal, 6(6), 10686–10699. Duckham,
M.,
2013. Decentralized
Spatial
Computing. Springer, Heidelberg. Yanmaz, E., Yahyanejad, S., Rinner, B., Hellwagner, H., Bett-
stetter, C., 2018. Drone networks: Communications, coordina-
tion, and sensing. Ad Hoc Networks, 68, 1–15. Gao, Y., Dong, W., Guo, K., Liu, X., Chen, Y., Liu, X., Bu, J.,
Chen, C., 2016. Mosaic: A low-cost mobile sensing system for
urban air quality monitoring. IEEE INFOCOM 2016 - The 35th
Annual IEEE International Conference on Computer Commu-
nications, 1–9. Zhang, Q., Fok, M. P., 2017. A two-phase coverage-enhancing
algorithm for hybrid wireless sensor networks. Sensors, 17(1),
117. Zhao, D., Ma, H., Liu, L., Li, X.-Y., 2015. Opportunistic cov-
erage for urban vehicular sensing. Computer Communications,
60, 71–85. Ghosh, A., Das, S. K., 2008. Coverage and connectivity issues
in wireless sensor networks: A survey. Pervasive and Mobile
Computing, 4(3), 303–334. Zhao, F., Guibas, L. J., 2004. Wireless Sensor Networks. El-
sevier, Amsterdam. Goldberg, D. E., Holland, J. H., 1988. Genetic algorithms and
machine learning. Zygowski, C., Jaekel, A., 2018. Path planning for maximiz-
ing area coverage of mobile nodes in wireless sensor networks. This contribution has been peer-reviewed. The double-blind peer-review was conducted on the basis of the full paper.
https://doi.org/10.5194/isprs-annals-X-4-W3-2022-65-2022 | © Author(s) 2022. CC BY 4.0 License. REFERENCES 2018 IEEE 9th Annual Information Technology, Electronics and
Mobile Communication Conference (IEMCON), IEEE, 12–17. He, Z., Cao, J., Liu, X., 2015. High quality participant recruit-
ment in vehicle-based crowdsourcing using predictable mobil-
ity. 2015 IEEE Conference on Computer Communications (IN-
FOCOM), 2542–2550. Holland, J. H. et al., 1992. Adaptation in natural and artificial
systems: an introductory analysis with applications to biology,
control, and artificial intelligence. MIT Press. Ji, S., Zheng, Y., Li, T., 2016. Urban sensing based on hu-
man mobility. Proceedings of the 2016 ACM International Joint
Conference on Pervasive and Ubiquitous Computing, 1040–
1051. 72
|
https://openalex.org/W4362481699
|
https://figshare.com/articles/journal_contribution/Supplementary_Figure_3_from_A_Locked_Dimeric_CXCL12_Variant_Effectively_Inhibits_Pulmonary_Metastasis_of_CXCR4-Expressing_Melanoma_Cells_Due_to_Enhanced_Serum_Stability/22496931/1/files/39954948.pdf
|
Latin
| null |
Supplementary Figure 2 from A Locked, Dimeric CXCL12 Variant Effectively Inhibits Pulmonary Metastasis of CXCR4-Expressing Melanoma Cells Due to Enhanced Serum Stability
| null | 2,023
|
cc-by
| 150
|
wtCXCL12
CXCL122
250
100
[kDa]
10
50
25
20
15
time(h)
12
0
3
6
0
0
0
12
0
3
6
0
0
0
Serum
concentration(μM)
time(h)
Supplementary Figure 3
10
1
10
10
10
5
10
12
0
3
6
0
0
0
-
+
+
+
+
-
-
10
1
10
10
10
5
10
12
0
3
6
0
0
0
-
+
+
+
+
-
- wtCXCL12
CXCL122
250
100
[kDa]
10
50
25
20
15
time(h)
12
0
3
6
0
0
0
12
0
3
6
0
0
0
Serum
concentration(μM)
time(h)
Supplementary Figure 3
10
1
10
10
10
5
10
12
0
3
6
0
0
0
-
+
+
+
+
-
-
10
1
10
10
10
5
10
12
0
3
6
0
0
0
-
+
+
+
+
-
- CXCL122 Supplementary Figure 3
|
https://openalex.org/W4307869861
|
https://www.frontiersin.org/articles/10.3389/fimmu.2022.956982/pdf
|
English
| null |
The efficacy and safety analysis of first-line immune checkpoint inhibitors in pulmonary sarcomatoid carcinoma
|
Frontiers in immunology
| 2,022
|
cc-by
| 5,311
|
The efficacy and safety analysis
of first-line immune checkpoint
inhibitors in pulmonary
sarcomatoid carcinoma OPEN ACCESS
EDITED BY
Alessandro Russo,
A.O. Papardo, Italy
REVIEWED BY
Chengzhi Zhou,
Clinical Management Department of
National Respiratory Medical Center,
China
Qiang Guo,
Hubei University of Medicine, China
*CORRESPONDENCE
Anwen Liu
awliu666@163.com
Yong Li
liyongcsco@email.ncu.edu.cn
†These authors have contributed
equally to this work
SPECIALTY SECTION
This article was submitted to
Cancer Immunity and Immunotherapy,
a section of the journal
Frontiers in Immunology
RECEIVED 30 May 2022
ACCEPTED 04 October 2022
PUBLISHED 31 October 2022 Zhimin Zeng 1,2,3†, Xiaoying Qian 4,5†, Fanrong Liu 6†,
Yong Wang 4,5, Yong Yuan 7, Chen Fang 4,5, Xinwei Zhang 4,5,
Shangkun Yuan 4,5, Renfang Chen 4,5, Biao Yu 4,5, Tong Wang 4,5,
Yan Yin 8, Yong Li 4,5* and Anwen Liu 1,2,3* Zhimin Zeng 1,2,3†, Xiaoying Qian 4,5†, Fanrong Liu 6†,
Yong Wang 4,5, Yong Yuan 7, Chen Fang 4,5, Xinwei Zhang 4,5,
Shangkun Yuan 4,5, Renfang Chen 4,5, Biao Yu 4,5, Tong Wang 4,5,
Yan Yin 8, Yong Li 4,5* and Anwen Liu 1,2,3* 1Department of Oncology, The Second Affiliated Hospital of Nanchang University, Nanchang, Jiangxi, China,
2Jiangxi Key Laboratory of Clinical Translational Cancer Research, Nanchang, Jiangxi, China, 3Radiation
Induced Heart Damage Institute of Nanchang University, Nanchang, Jiangxi, China, 4Department of Medical
Oncology, The First Affiliated Hospital of Nanchang University, Nanchang, China, 5Medical Innovation
Center, The First Affiliated Hospital of Nanchang University, Nanchang, China, 6Department of Pathology,
The Second Affiliated Hospital of Nanchang University, Nanchang, China, 7Department of Thoracic Surgery,
The Second Affiliated Hospital of Nanchang University, Nanchang, China, 8Department of Pathology, The
First Affiliated Hospital of Nanchang University, Nanchang, China Background: Pulmonary sarcomatoid carcinoma (PSC) is a rare and aggressive
disease without standardized treatment strategies. The efficacy of second-line
or beyond immune checkpoint inhibitors (ICIs) has been proven in recent
studies, whereas the evidence for first-line immunotherapy for PSC is still
limited to case reports and remains poorly understood. Zeng Z, Qian X, Liu F, Wang Y, Yuan Y,
Fang C, Zhang X, Yuan S, Chen R,
Yu B, Wang T, Yin Y, Li Y and Liu A
(2022) The efficacy and safety analysis
of first-line immune checkpoint
inhibitors in pulmonary
sarcomatoid carcinoma. Front. Immunol. 13:956982. doi: 10.3389/fimmu.2022.956982 Materials and methods: This was a multicenter, retrospective analysis of 21
patients with a histological diagnosis of PSC who received ICI as first-line
therapy from January 2019 to March 2022. The expression of PD-L1 was
evaluated by immunohistochemistry (IHC) using the monoclonal antibody
22C3. TYPE Original Research
PUBLISHED 31 October 2022
DOI 10.3389/fimmu.2022.956982 TYPE Original Research
PUBLISHED 31 October 2022
DOI 10.3389/fimmu.2022.956982 TYPE Original Research
PUBLISHED 31 October 2022
DOI 10.3389/fimmu.2022.956982 COPYRIGHT
© 2022 Zeng, Qian, Liu, Wang, Yuan,
Fang, Zhang, Yuan, Chen, Yu, Wang, Yin,
Li and Liu. This is an open-access article
distributed under the terms of the
Creative Commons Attribution License
(CC BY). The use, distribution or
reproduction in other forums is
permitted, provided the original
author(s) and the copyright owner(s)
are credited and that the original
publication in this journal is cited, in
accordance with accepted academic
practice. No use, distribution or
reproduction is permitted which does
not comply with these terms. KEYWORDS pulmonary sarcomatoid carcinoma, immune checkpoint inhibitors, immunotherapy,
anlotinib, camrelizumab, tislelizumab Introduction University in China. The patients diagnosed with advanced PSC
(III/IV) who received first-line immunotherapy from January
2019 to March 2022 were enrolled. Patients with pulmonary
interstitial disease, systemic immunosuppression, autoimmune
disease, or second primary malignancy were excluded from this
study. Written informed consent was not required, as this was a
retrospective review study. The last follow-up time was 30 April
2022. Each patient’s relevant clinical data were collected from
hospital electronic medical records, including sex, age, smoking
status, Eastern Cooperative Oncology Group (ECOG) score,
histology, clinical stage, and distant metastasis. The best tumor
response was evaluated according to RECIST version 1.1. Progression-free survival (PFS) was the time from the date of
immunotherapy to the date of disease progression or death. Overall survival (OS) was the time from the date of
immunotherapy to death from any cause or the last follow-up. Pulmonary sarcomatoid carcinoma (PSC), a rare and
aggressive disease, accounts for less than 1% of non-small cell
lung cancer (NSCLC) (1). It often is at advanced stages during
diagnosis and is highly heterogeneous (2). No standardized
treatment strategies exist for PSC, and conventional
chemotherapy also has limited efficacy (3). In addition to
potential targeted molecular therapy from genetic sequencing,
immune checkpoint inhibitors (ICIs) using programmed cell
death 1 (PD-1)/PD-L1 antibodies are considered to be one of the
most promising immunotherapy strategies (4–7). Studies have shown higher frequencies of genetic mutations
and PD-L1–positive expression in PSC than in conventional
NSCLC, and PD-L1 positivity might lead to survival benefits
from immunotherapy or even a favorable response in those
harboring actionable mutations (4, 8, 9). Immunotherapy by
immune checkpoint blockade is emerging, and the efficacy of
second-line or beyond immunotherapy for PSC has been proven
in recent studies (10, 11). Furthermore, several studies suggested
the remarkable response of PSC to first-line immunotherapy,
while the reports are limited to the case (12, 13). The efficacy and safety analysis
of first-line immune checkpoint
inhibitors in pulmonary
sarcomatoid carcinoma Low and high PD-L1 expressions were defined using the tumor
proportion score (TPS), with cutoffs of 1 and 50%, respectively. COPYRIGHT
© 2022 Zeng, Qian, Liu, Wang, Yuan,
Fang, Zhang, Yuan, Chen, Yu, Wang, Yin,
Li and Liu. This is an open-access article
distributed under the terms of the
Creative Commons Attribution License
(CC BY). The use, distribution or
reproduction in other forums is
permitted, provided the original
author(s) and the copyright owner(s)
are credited and that the original
publication in this journal is cited, in
accordance with accepted academic
practice. No use, distribution or
reproduction is permitted which does
not comply with these terms. Results: All eight patients had PD-L1 positivity who underwent PD-L1 expression
assessment, and six patients (6/8, 75.0%) had high PD-L1 expression. Among the
21 PSC patients, seven received tislelizumab, six received camrelizumab, four
received sintilimab, three received pembrolizumab, and one received
durvalumab. Among them, 18 PSCs received combination therapy, whereas
another three PSCs received immunotherapy alone. Out of the 21 PSC
patients, 12 (57.1%) achieved a partial response (PR), and five patients had
stable disease (SD) as the best response, whereas four PSCs experienced
dramatic progressive disease (PD). The median progression-free survival (PFS)
was 9.2 (95% CI [4.3, 14.1]) months, and the median OS was 22.8 (95% CI [4.0,
41.5]) months. Among the three treatment groups (immunotherapy alone,
immunotherapy combined with anlotinib, and chemoimmunotherapy), the
median PFS was 8.0, 9.4, and 9.6 months, and the median OS was 19.0, 22.8, frontiersin.org Frontiers in Immunology 01 Zeng et al. 10.3389/fimmu.2022.956982 and 30.6 months, respectively. There was no difference in PFS and OS between
the three treatment regimen groups (P = 0.86 and P = 0.34, respectively) and
different immunotherapies (P = 0.10 and P = 0.23, respectively). No serious
adverse events (grade ≥3) were noted. Conclusion: First-line immunotherapy has promising therapeutic potential in
the treatment of PSC. More studies are warranted to confirm these findings. PD-L1 expression The expression of PD-L1 protein was evaluated by
immunohistochemistry (IHC) performed on 4-µm formalin-
fixed paraffin-embedded (FFPE) tissue sections using a Dako
PD-L1 22C3 pharmDx kit (Dako, Carpinteria, CA). PD-L1
protein expression was determined by using the tumor
proportion score (TPS), and the cutoffs for low and high
expressions were 1 and 50%, respectively (14). The low number of patients who have undergone
immunotherapy as a first-line treatment strategy makes it
challenging to evaluate the specific safety and efficacy of first-line
immunotherapy for PSC. The purpose of this paper is to report the
largest study offirst-line immunotherapy for PSC treatment to date. Clinical characteristics In the end, 21 PSC patients who were treated with first-line
immunotherapy were enrolled. The clinical characteristics are
summarized in Table 1. The median age was 65 (range: 39–95)
years, and all the men had a history of smoking (15/21, 71.4%). In five patients (23.8%), the tumors were located in the left lung,
whereas most of the tumors (76.2%) were in the right lung. All
patients with PSC were diagnosed at advanced stages (III/IV). The majority of these were already in the IVB stage (61.9%) and
had distant metastasis when they were diagnosed. Bone was the
most common site of metastases (33.3%). Eleven out of 21
patients (52.4%) with good physical status had an ECOG
performance status score of 0. Not surprisingly, pleomorphic
carcinoma (PLC) was the most common pathological type of
PSC at 80.9%, three cases were spindle cell carcinoma (SCC),
and another case was carcinosarcoma (CS). Treatment and outcome of
immunotherapy Immunotherapy was given as a first-line treatment in all PSC
patients, and the related data are presented in Table 2. The
majority (18/21, 85.7%) were given combination therapies. The
other three received camrelizumab (2/12, 16.7%) or sintilimab
(1/12, 8.3%) monotherapy. Among those with combination treatment, 14 were given
anlotinib, a multitargeted tyrosine kinase inhibitor, combined
with immunotherapy, including tislelizumab (7/21, 33.3%),
camrelizumab (4/21, 19.0%), sintilimab (2/21, 9.5%), and
pembrolizumab (1/21, 4.8%). Of the seven patients who
received tislelizumab combined with anlotinib treatment, two
achieved partial response (PR) as the best response, one of them
with tumor shrinkage of approximately 72%, and four achieved
stable disease (SD), whereas rapid progression occurred in
another patient for PFS of 1.7 months and OS of 5.4 months
(Figure 1B). All patients who received camrelizumab
combination treatment achieved PR as their best response and
a survival time of more than 14 months (Figure 1B). It is
noteworthy that one female PSC patient harboring the BRAF
V600E mutation at the IVB stage refused dabrafenib plus
trametinib treatment because it was expensive, received
camrelizumab combination treatment, had a PFS of 22.4
months, and is still alive at the time of this last follow-up. In
another female case with TP53 mutations, the lesions suggested
a PR that endured for more than 14.8 months without
progression. The PFS was 8 months in the patient who
accepted camrelizumab combination treatment, and the
continuation of the original treatment led to sustained stability
for more than 21.2 months. The last PSC patient with TP53
mutation achieved PR and had a PFS and OS of 9.4 and 22.8
months, respectively. Unfortunately, two patients with stage IVB
had relatively poor physical health (ECOG ≥1), and despite
treatment with sintilimab plus anlotinib, their lesion rapidly
progressed and they died within 3 months (Figure 1B). One of
them had SCC with high PD-L1 expression and KRAS mutation. A man with a smoking history at the IVB stage harboring ATM,
CREBBP, KRAS, and TP53 mutation and who received
pembrolizumab in combination with anlotinib achieved PR;
the PFS was 10.5 months and OS was 14.2 months. TABLE 1 Clinical characteristics of PSC patients. TABLE 1 Clinical characteristics of PSC patients. Characteristics
No. Study design and patients Kaplan–Meier curves were drawn to analyze the survival of
PSC patients. The corresponding 95% confidence interval (95%
CI) was calculated. Statistical tests were performed in IBM SPSS
version 25 (IBM Corp, Armonk, NY). p < 0.05 was considered
statistically significant. A multicenter, retrospective study was conducted at two
tertiary medical institutions in the Nanchang region of China,
namely The First and Second Affiliated Hospital of Nanchang Frontiers in Immunology frontiersin.org 02 Zeng et al. 10.3389/fimmu.2022.956982 10.3389/fimmu.2022.956982 *Including duodenum and pancreas.
ECOG, Eastern Cooperative Oncology Group; PLC, pleomorphic carcinoma; SCC,
spindle cell carcinoma; CS, carcinosarcoma; RLN, retroperitoneal lymph node. Results PD-L1 expression was assessed by IHC in the eight samples
available. Six patients (6/8, 75.0%) had high expression (TPS >
49%) and the others had low expression (TPS 1–49%)
(Figure 1A). Genetic testing was performed on 16 patients. Only one patient had an actionable mutation, for BRAF
V600E. No EGFR, ALK, or MET mutations were found. Treatment and outcome of
immunotherapy of patients (%)
Median age (range)
65 (39–95)
Gender
Men
15 (71.4)
Women
6 (28.6)
Smoking status
Never
6 (28.6)
Former/current
15 (71.4)
Primary location
Left
5 (23.8)
Right
16 (76.2)
ECOG at diagnosis
0
11 (52.4)
≥1
10 (47.6)
Histology
PLC
17 (80.9)
SCC
3 (14.3)
CS
1 (4.8)
Clinical stage
IIIA–IVA
8 (38.1)
IVB
13 (61.9)
Distant metastasis
Bone
7 (33.3)
Adrenal
5 (23.8)
Liver
5 (23.8)
RLN
5 (23.8)
Other*
3 (14.3) Another four patients took the combination of platinum-
based chemotherapy with pembrolizumab (2/21, 9.5%),
sintilimab (1/21, 4.8%), or durvalumab (1/21, 4.8%). Both frontiersin.org Frontiers in Immunology 03 Zeng et al. 10.3389/fimmu.2022.956982 TABLE 2
Characteristics of first-line immunotherapy. Characteristics
No. of patients (%)
PD-L1 expression
<1%/undetected
13 (61.9)
1–49%
2 (9.5)
≥50%
6 (28.6)
Immunotherapy strategies
Monotherapy
3 (14.3)
Immune+ anlotinib
14 (66.7)
Chemoimmunotherapy
4 (19.0)
Type of immunotherapy
Tislelizumab
7 (33.3)
Camrelizumab
6 (28.6)
Sintilimab
4 (19.0)
Pembrolizumab
3 (14.3)
Durvalumab
1 (4.8)
Best response
PR
12 (57.2)
SD
5 (23.8)
PD
4 (19.0)
Adverse events
No
16 (76.2)
Yes
5 (23.8)
ECOG, Eastern Cooperative Oncology Group; PR, partial response; SD, stable disease;
PD, progressive disease. TABLE 2
Characteristics of first-line immunotherapy. patients treated with the pembrolizumab combination achieved
PR (Figure 1B). One patient’s lesion continued remission for
11.7 months, while the PFS was not achieved. More surprisingly,
the other patient had a PFS of 10 months and an OS of up to 30.6
months. A male PSC patient at the IVA stage received sintilimab
combined with chemotherapy and achieved PR with tumor
shrinkage of approximately 67%; the PFS was 9.6 months and
the OS was not reached. On the other hand, in one patient
treated with durvalumab, rapid progression occurred (only
1.8 months). Notably, three patients were reluctant to receive
chemotherapy in favor of monotherapy, two were PD-L1
positive (one with high PD-L1 expression of 95% and the
other with 20%), and another did not have PD-L1 testing
since there was no gene mutation. Two were treated with
camrelizumab monotherapy. The OS was 12.6 months in one
patient (PD-L1 95%) and more than 25.4 months in the other. Sintilimab monotherapy was adopted as a first-line treatment for
one CS patient with PD-L1 expression of 20%, who had a PFS of
more than 5 months and whose OS was not reached. Survival analysis The median follow-up time for this cohort was 8.5 (range:
0.3–30.6) months, and only seven of the 21 PSC patients (33.3%)
were deceased at the last follow-up time. In the whole sample,
the median PFS was 9.2 (95% CI [4.3, 14.1]) months (Figure 2A),
and the median OS was 22.8 (95% CI [4.0, 41.5]) months
(Figure 2B). Twelve out of 21 (57.1%) PSC patients achieved a
PR and five patients had SD as the best response, whereas four
PSCs experienced dramatic progress with first-line
immunotherapy, although one of them was PD-L1 TPS > 50%
(Figure 1A). Regrettably, we did not do further molecular testing
or autopsies due to the family members’ refusal. ECOG, Eastern Cooperative Oncology Group; PR, partial response; SD, stable disease;
PD, progressive disease. ECOG, Eastern Cooperative Oncology Group; PR, partial response; SD, stable disease;
PD, progressive disease. We further analyzed the differences between different
treatment regimens and immunotherapies on PFS and OS in
PSC patients. Among the three treatment groups
(immunotherapy alone, immunotherapy combined with
anlotinib, and chemoimmunotherapy), the median PFS was
8.0, 9.4, and 9.6 months, and the median OS was 19.0, 22.8,
and 30.6 months, respectively. Nevertheless, there was no
difference in PFS and OS between the three treatment A
B
FIGURE 1
Waterfall plot of the best response rate according to PD-L1 expression in tumor cells (A) and different immunotherapy strategies (B) among PSC
patients who received first-line immunotherapy treatment. PD-L1, programmed cell death ligand-1; PSC, pulmonary sarcomatoid carcinoma. A FIGURE 1
Waterfall plot of the best response rate according to PD-L1 expression in tumor cells (A) and different immunotherapy strategies (B) among PSC
patients who received first-line immunotherapy treatment. PD-L1, programmed cell death ligand-1; PSC, pulmonary sarcomatoid carcinoma. 04 frontiersin.org Frontiers in Immunology Zeng et al. 10.3389/fimmu.2022.956982 A
B
FIGURE 2
Survival analysis of first-line immunotherapy treatment in PSC. Kaplan–Meier analysis of PFS (A) and OS (B) in 21 PSC patients treated with first-
line immunotherapy. PSC, pulmonary sarcomatoid carcinoma; PFS, progression-free survival; OS, overall survival. B FIGURE 2
Survival analysis of first-line immunotherapy treatment in PSC. Kaplan–Meier analysis of PFS (A) and OS (B) in 21 PSC patients treated with first-
line immunotherapy. PSC, pulmonary sarcomatoid carcinoma; PFS, progression-free survival; OS, overall survival. regimens groups (P = 0.86 and 0.34, respectively; Figure 3). Survival analysis Similarly, results have shown no differences in PFS and OS
between different immunotherapies (P = 0.10 and 0.23,
respectively; Figure 3). strategies for PSC. Evidence indicates that 80.0% of PLCs with
high PD-L1 expression, which is correlated with longer PFS and
OS than low/negative/unknown PD-L1 expression, suggests that
high PD-L1 expression may enable more benefit from
immunotherapy (20). Another study showed that PD-L1
positivity in the tumor led to a better ORR of 58.8% in PSCs
treated with second-line or beyond immunotherapy than PD-L1
negativity, and the median OS was 12.7 months in the whole
sample. The authors thought that it was necessary to identify the
efficacy of this promising immunotherapy as a first-line
treatment for PSC (10). The overall objective response rate (ORR) was 57.2%, and a
disease control rate (DCR) of 81.0% was achieved (Table 2). Remarkably, one patient who received first-line immunotherapy
achieved tumor shrinkage of approximately 55%, even with the
BRAF V600E mutation; the PFS was 22.4 months and the OS
was not attained. We speculate that first-line immunotherapy
was effective in achieving continuous remission. Along with accumulating research on immune therapies in
PSC, the efficacy of first-line treatment has been gradually
unveiled. Pembrolizumab monotherapy treatment in three
PSC patients showed a response, with a PFS of more than 11.0
months (21). Similarly, first-line camrelizumab combined with
platinum-based chemotherapy, although followed by
camrelizumab monotherapy for serious adverse events,
achieved partial remission for more than 20.0 months (12). Interestingly, obvious tumor shrinkage was found in advanced-
stage patients treated with chemoimmunotherapy, whereas PD-
L1 expression was found in 1% (13). Consistent with the findings
described above, our study makes significant advancements in
first-line immunotherapy, including monotherapy and
combination treatment. Adverse events Five patients (23.8%) experienced adverse events (Table 2),
including myocarditis, myelosuppression, nausea, vomiting, and
diarrhea, whereas no serious adverse events (grade ≥3) were
noted. None of the patients discontinued immunotherapy due to
adverse effects. Discussion Continuing efforts to find novel therapeutic strategies for
PSC are imperative. Although the efficacy of immunotherapy
after disease progression has been demonstrated (10), first-line
immunotherapy in PSCs remains poorly understood. To the best
of our knowledge, the present study is the largest to report the
efficacy of immunotherapy for PSC as first-line immune therapy. Studies have shown that harboring the BRAF V600E
mutation was associated with greater clinical benefit from ICIs
and significantly prolonged PFS in NSCLC (22). Similarly, in our
study, one PLC patient with the BRAF V600E mutation taking
camrelizumab plus anlotinib as first-line treatment without
targeted therapies had a PFS of 22.4 months. The results
indicated that first-line immunotherapy is a potential choice to
induce durable clinical benefits for PSC patients with the BRAF Immunotherapy has substantially contributed to the
treatment of conventional NSCLC and has improved patient
outcomes (15–19), providing promising new treatment frontiersin.org Frontiers in Immunology 05 Zeng et al. 10.3389/fimmu.2022.956982 A
B
C
D
FIGURE 3
PFS (A, C) and OS (B, D) curves of PSC patients with different treatment regimens and immunotherapy. PFS, progression-free survival; OS,
overall survival; PSC, pulmonary sarcomatoid carcinoma. A
B C
D D C D C D FIGURE 3
PFS (A, C) and OS (B, D) curves of PSC patients with different treatment regimens and immunotherapy. PFS, progression-free survival; OS,
overall survival; PSC, pulmonary sarcomatoid carcinoma. mutation experienced limited treatment efficacy, with rapid
progression, in contrast with that in one PLC patient with
multigene mutations, including KRAS and TP53 mutations,
who obtained a PR of more than 7.2 months. Two more PLC
patients with TP53 single mutations also achieved PR for more
than 8.0 months. The reasons for this are not well known;
limited by a small sample size, KRAS mutation was associated
with poor efficacy from immunotherapy for PSC, which is
unlikely to hold true. However, this result is controversial, as
several studies have suggested that those harboring KRAS
mutations had a better response to immunotherapy (5, 28). V600E mutation. Intriguingly, camrelizumab combined with
anlotinib as first-line treatment demonstrated excellent effects
in four PLC patients. This promising treatment strategy has been
proven by a previous study showing that anti-PD-1 treatment,
including camrelizumab with anlotinib, has favorable antitumor
activity even in previously treated advanced NSCLC (16, 23). Discussion Disease progression is inevitable, and little is known about
what contributes to dramatic progress without any response to
immunotherapy; however, we cannot ignore the four patients
who showed progress in our study with advanced IVB stages at
immunotherapy initiation. A phase Ib trial (NCT03628521) regarding combined
sintilimab and anlotinib as first-line therapy in patients with
advanced NSCLC saw decent success, with a median PFS of 15.0
months (29). Another retrospective study revealed that
sintilimab plus anlotinib in NSCLC patients with previous
systemic treatment failure yielded a favorable response (23). However, this situation did not occur in our study: Two PSC KRAS and TP53 mutations have been demonstrated to be
associated with high PD-L1 expression in NSCLC (24–26). Furthermore, Lococo et al. indicated that PSC harboring
KRAS mutations portended a dismal prognosis (27). The
results of our study are consistent with those; patients
harboring KRAS or TP53 mutations indeed had high PD-L1
expression, whereas one SCC patient with a single KRAS Frontiers in Immunology frontiersin.org 06 Zeng et al. 10.3389/fimmu.2022.956982 University and the Medical Research Ethics Committee of the
First Affiliated Hospital of Nanchang University. patients who took sintilimab plus anlotinib achieved rapid
progression as their best response and died within 3.0 months. Similarly, the combined use of durvalumab and chemotherapy
in a male PLC patient at stage IVB exhibited unsatisfactory
results, although this combination of treatment strategies
showed a promising effect in a previous investigation (30). Conclusions Our study sheds light on the promising therapeutic potential
offirst-line immunotherapy in the treatment of PSC. Given these
findings, a prospective study is warranted to explore the efficacy
of immunotherapy with or without chemotherapy, and it is
necessary to determine the reason for the dramatic progression
at initial immunotherapy treatment. Author contributions The limitations of this study are its small sample size, and we
did not find an exact predictive factor or mechanism of dramatic
progression. Additionally, this study had different treatment
regimens and several immunotherapies with limited numbers
of people on each treatment, and although the results showed no
differences between groups, a larger sample of studies is needed
to support this. Although remarkable outcomes have been
obtained, they may suffer from selection bias (a low rate of
incidence). Several clinical trials including first-line toripalimab
combined therapy (NCT04725448) in the treatment of patients
with advanced PSC are in progress. We hope that those studies
can reveal more detail about the value of immunotherapy and
help guide treatment decisions. One challenge is to be able to
predict which patients are most likely to derive benefit from
immune therapies and the mechanism of progression at initial
treatment. Considering the rarity and complexity of PSC, clinical
studies involving multiple centers, even globally, are needed to
improve prognosis. AL and YL co-designed the research. ZZ, XQ, and FL made a
significant contribution to the data integration and analysis. ZZ
provided patients’ information and data analysis. XQ drew up
the manuscript. FL and YW processed the figures and tables. YYu, CF, XZ, and SY collected the outcome. RC, BY, TWand
YYi followed up on the case. All authors contributed to the
article and approved the submitted version. Publisher’s note The original contributions presented in the study are
included in the article/supplementary material. Further
inquiries can be directed to the corresponding author. All claims expressed in this article are solely those of the
authors and do not necessarily represent those of their affiliated
organizations, or those of the publisher, the editors and the
reviewers. Any product that may be evaluated in this article, or
claim that may be made by its manufacturer, is not guaranteed
or endorsed by the publisher. Conflict of interest The authors declare that the research was conducted in the
absence of any commercial or financial relationships that could
be construed as a potential conflict of interest. Funding This study was supported by grants from the National
Natural Science Foundation of China (No.81560379,
81460292, 81660315), the Surface project of the Natural
Science Foundation of Jiangxi Province (No.20181BAB205046,
No.20202BAB216031), Technology Supporting Program of
Jiangxi Province (No.2015BBG70236), The Key Project of
Education Department of Jiangxi Province (No.GJJ170012),
and The Graduate Student Innovation Special Fund Project of
Jiangxi Province (No. YC2022-s207). References Immune
checkpoint blockade is associated with durable responses in pulmonary
sarcomatoid carcinoma. Clin Lung Cancer (2019) 20(3):e242–e6. doi: 10.1016/
j.cllc.2018.12.013 6. Nakagomi T, Goto T, Hirotsu Y, Shikata D, Yokoyama Y, Higuchi R, et al. New therapeutic targets for pulmonary sarcomatoid carcinomas based on their
genomic and phylogenetic profiles. Oncotarget (2018) 9(12):10635–49. doi: 10.18632/oncotarget.24365 22. Negrao MV, Skoulidis F, Montesion M, Schulze K, Bara I, Shen V, et al. Oncogene-specific differences in tumor mutational burden, pd-L1 expression, and
outcomes from immunotherapy in non-small cell lung cancer. J Immunother
Cancer (2021) 9(8):e002891. doi: 10.1136/jitc-2021-002891 7. Qin J, Chen B, Li C, Yan J, Lu H. Genetic heterogeneity and predictive
biomarker for pulmonary sarcomatoid carcinomas. Cancer Genet (2020) 250-
251:12–9. doi: 10.1016/j.cancergen.2020.11.004 23. Wang P, Fang X, Yin T, Tian H, Yu J, Teng F. Efficacy and safety of anti-Pd-
1 plus anlotinib in patients with advanced non-Small-Cell lung cancer after
previous systemic treatment failure-a retrospective study. Front Oncol (2021)
11:628124. doi: 10.3389/fonc.2021.628124 8. Mayenga M, Assie JB, Monnet I, Massiani MA, Tabeze L, Friard S, et al. Durable responses to immunotherapy of non-small cell lung cancers harboring met
exon-14-Skipping mutation: A series of 6 cases. Lung Cancer (2020) 150:21–5. doi: 10.1016/j.lungcan.2020.09.008 24. Schoenfeld AJ, Rizvi H, Bandlamudi C, Sauter JL, Travis WD, Rekhtman N,
et al. Clinical and molecular correlates of pd-L1 expression in patients with lung
adenocarcinomas. Ann Oncol (2020) 31(5):599–608. doi: 10.1016/j.annonc.2020. 01.065 9. Liu X, Wang F, Xu C, Chen X, Hou X, Li Q, et al. Genomic origin and
intratumor heterogeneity revealed by sequencing on carcinomatous and
sarcomatous components of pulmonary sarcomatoid carcinoma. Oncogene
(2021) 40(4):821–32. doi: 10.1038/s41388-020-01573-9 10. Domblides C, Leroy K, Monnet I, Mazières J, Barlesi F, Gounant V, et al. Efficacy of immune checkpoint inhibitors in lung sarcomatoid carcinoma. J Thorac
Oncol (2020) 15(5):860–6. doi: 10.1016/j.jtho.2020.01.014 25. Hu C, Zhao L, Liu W, Fan S, Liu J, Liu Y, et al. Genomic profiles and their
associations with tmb, pd-L1 expression, and immune cell infiltration landscapes in
synchronous multiple primary lung cancers. J Immunother Cancer (2021) 9(12):
e003773. doi: 10.1136/jitc-2021-003773 11. Babacan NA, Pina IB, Signorelli D, Prelaj A, Garassino MC, Tanvetyanon T. Relationship between programmed death receptor-ligand 1 expression and
response to checkpoint inhibitor immunotherapy in pulmonary sarcomatoid
carcinoma: A pooled analysis. Clin Lung Cancer (2020) 21(5):e456–e63. doi: 10.1016/j.cllc.2020.02.022 26. Shirasawa M, Yoshida T, Shimoda Y, Takayanagi D, Shiraishi K, Kubo T,
et al. Ethics statement This study was reviewed and approved by the institutional
ethics committee of the Second Affiliated Hospital of Nanchang frontiersin.org Frontiers in Immunology 07 Zeng et al. 10.3389/fimmu.2022.956982 References inhibitors for treatment of non-small cell lung cancer. JAMA Oncol (2020) 6
(12):1952–6. doi: 10.1001/jamaoncol.2020.5012 1. Yendamuri S, Caty L, Pine M, Adem S, Bogner P, Miller A, et al. Outcomes of
sarcomatoid carcinoma of the lung: A surveillance, epidemiology, and end results
database analysis. Surgery (2012) 152(3):397–402. doi: 10.1016/j.surg.2012.05.007 1. Yendamuri S, Caty L, Pine M, Adem S, Bogner P, Miller A, et al. Outcomes of
sarcomatoid carcinoma of the lung: A surveillance, epidemiology, and end results
database analysis. Surgery (2012) 152(3):397–402. doi: 10.1016/j.surg.2012.05.007 17. Luo M, Wang F, Zhang H, To KKW, Wu S, Chen Z, et al. Mitomycin c
enhanced the efficacy of pd-L1 blockade in non-small cell lung cancer. Signal
Transduct Target Ther (2020) 5(1):141. doi: 10.1038/s41392-020-0200-4 2. Rahouma M, Kamel M, Narula N, Nasar A, Harrison S, Lee B, et al. Pulmonary sarcomatoid carcinoma: An analysis of a rare cancer from the
surveillance, epidemiology, and end results database. Eur J Cardiothorac Surg
(2018) 53(4):828–34. doi: 10.1093/ejcts/ezx417 18. Osmani L, Askin F, Gabrielson E, Li QK. Current who guidelines and the
critical role of immunohistochemical markers in the subclassification of non-small
cell lung carcinoma (Nsclc): Moving from targeted therapy to immunotherapy. Semin Cancer Biol (2018) 52(Pt 1):103–9. doi: 10.1016/j.semcancer.2017.11.019 3. Bae H, Min H, Lee S, Kim D, Chung D, Lee J, et al. Palliative chemotherapy
for pulmonary pleomorphic carcinoma. Lung Cancer (Amsterdam Netherlands)
(2007) 58(1):112–5. doi: 10.1016/j.lungcan.2007.05.006 19. Boeri M, Milione M, Proto C, Signorelli D, Lo Russo G, Galeone C, et al. Circulating mirnas and pd-L1 tumor expression are associated with survival in
advanced nsclc patients treated with immunotherapy: A prospective study. Clin
Cancer Res (2019) 25(7):2166–73. doi: 10.1158/1078-0432.Ccr-18-1981 4. Zhou F, Huang Y, Cai W, Li J, Su C, Ren S, et al. The genomic and
immunologic profiles of pure pulmonary sarcomatoid carcinoma in Chinese
patients. Lung Cancer (2021) 153:66–72. doi: 10.1016/j.lungcan.2021.01.006 20. Lee J, Choi Y, Jung HA, Lee SH, Ahn JS, Ahn MJ, et al. Outstanding clinical
efficacy of pd-1/Pd-L1 inhibitors for pulmonary pleomorphic carcinoma. Eur J
Cancer (2020) 132:150–8. doi: 10.1016/j.ejca.2020.03.029 5. Dong ZY, Zhong WZ, Zhang XC, Su J, Xie Z, Liu SY, et al. Potential
predictive value of Tp53 and kras mutation status for response to pd-1 blockade
immunotherapy in lung adenocarcinoma. Clin Cancer Res (2017) 23(12):3012–24. doi: 10.1158/1078-0432.Ccr-16-2554 21. Sukrithan V, Sandler J, Gucalp R, Gralla R, Halmos B. References Differential immune-related microenvironment determines programmed cell
death protein-1/Programmed death-ligand 1 blockade efficacy in patients with
advanced nsclc. J Thorac Oncol (2021) 16(12):2078–90. doi: 10.1016/j.jtho.2021. 07.027 12. Kong F, Wang W, Gong L, Wu W, Liu Y. Anti-Pd-1 antibody camrelizumab
plus doxorubicin showed durable response in pulmonary sarcomatoid carcinoma:
Case report and literature review. J Clin Pharm Ther (2020) 45(6):1489–96. doi: 10.1111/jcpt.13234 27. Lococo F, GandolfiG, Rossi G, Pinto C, Rapicetta C, Cavazza A, et al. Deep
sequencing analysis reveals that kras mutation is a marker of poor prognosis in
patients with pulmonary sarcomatoid carcinoma. J Thorac Oncol (2016) 11
(8):1282–92. doi: 10.1016/j.jtho.2016.04.020 13. Taniguchi H, Takemoto S, Ozasa M, Honda N, Suyama T, Umeyama Y,
et al. Remarkable response to pembrolizumab with platinum-doublet in pd-L1-
Low pulmonary sarcomatoid carcinoma: A case report. Thorac Cancer (2021) 12
(7):1126–30. doi: 10.1111/1759-7714.13890 28. Cinausero M, Laprovitera N, De Maglio G, Gerratana L, Riefolo M,
Macerelli M, et al. Kras and erbb-family genetic alterations affect response to pd-
1 inhibitors in metastatic nonsquamous nsclc. Ther Adv Med Oncol (2019)
11:1758835919885540. doi: 10.1177/1758835919885540 14. Adam J, Le Stang N, Rouquette I, Cazes A, Badoual C, Pinot-Roussel H, et al. Multicenter harmonization study for pd-L1 ihc testing in non-Small-Cell lung
cancer. Ann Oncol (2018) 29(4):953–8. doi: 10.1093/annonc/mdy014 29. Chu T, Zhong R, Zhong H, Zhang B, Zhang W, Shi C, et al. Phase 1b
study of sintilimab plus anlotinib as first-line therapy in patients with
advanced nsclc. J Thorac Oncol (2021) 16(4):643–52. doi: 10.1016/j.jtho.2020. 11.026 15. Aguilar EJ, Ricciuti B, Gainor JF, Kehl KL, Kravets S, Dahlberg S, et al. Outcomes to first-line pembrolizumab in patients with non-Small-Cell lung cancer
and very high pd-L1 expression. Ann Oncol (2019) 30(10):1653–9. doi: 10.1093/
annonc/mdz288 30. Leighl NB, Laurie SA, Goss GD, Hughes BGM, Stockler M, Tsao MS, et al. Cctg Br34: A randomized phase ii trial of durvalumab and tremelimumab +/-
platinum-based chemotherapy in patients with metastatic non-small cell lung
cancer. J Thorac Oncol (2021) 17(4):434–45. doi: 10.1016/j.jtho.2021.10.023 16. Shankar B, Zhang J, Naqash AR, Forde PM, Feliciano JL, Marrone KA, et al. Multisystem immune-related adverse events associated with immune checkpoint Frontiers in Immunology frontiersin.org 08
|
https://openalex.org/W2169889046
|
https://research.rug.nl/files/2520949/2011IntJPeptAbts.pdf
|
English
| null |
Easy and Rapid Purification of Highly Active Nisin
|
International Journal of Peptides
| 2,011
|
cc-by
| 8,067
|
University of Groningen University of Groningen Easy and Rapid Purification of Highly Active Nisin
Abts, André; Mavaro, Antonino; Stindt, Jan; Bakkes, Patrick J.; Metzger, Sabine; Driessen,
Arnold J.M.; Smits, Sander H.J.; Schmitt, Lutz Easy and Rapid Purification of Highly Active Nisin Published in:
International Journal of Peptides DOI:
10.1155/2011/175145 IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from
it. Please check the document version below. Document Version
Publisher's PDF, also known as Version of record IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from
it. Please check the document version below. Document Version
Publisher's PDF, also known as Version of record Document Version
Publisher's PDF, also known as Version of record Document Version
Publisher's PDF, also known as Version of record Publication date:
2011 Link to publication in University of Groningen/UMCG research database Citation for published version (APA):
Abts, A., Mavaro, A., Stindt, J., Bakkes, P. J., Metzger, S., Driessen, A. J. M., Smits, S. H. J., & Schmitt, L. (2011). Easy and Rapid Purification of Highly Active Nisin. International Journal of Peptides, 2011(1), 1-9. Article 175145. https://doi.org/10.1155/2011/175145 Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the
author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons). Citation for published version (APA):
Abts, A., Mavaro, A., Stindt, J., Bakkes, P. J., Metzger, S., Driessen, A. J. M., Smits, S. H. J., & Schmitt, L.
(2011). Easy and Rapid Purification of Highly Active Nisin. International Journal of Peptides, 2011(1), 1-9.
Article 175145. https://doi.org/10.1155/2011/175145 Andr´e Abts,1 Antonino Mavaro,1 Jan Stindt,1 Patrick J. Bakkes,1 Sabine Metzger,2
Arnold J. M. Driessen,3 Sander H. J. Smits,1 and Lutz Schmitt1 1Institute of Biochemistry, Heinrich Heine University D¨usseldorf, Universit¨atsstrabe 1, 40225 D¨usseldorf, Germany
2Biological and Medical Research Center, Heinrich Heine University D¨usseldorf, Universit¨atsstrabe 1, 40225 D¨usseldorf, Germany
3Department of Molecular Microbiology, Groningen Biomolecular Sciences and Biotechnology Institute,
Zernike Institute for Advanced Materials and the Kluyver Centre for the Genomics of Industrial Microorganisms,
University of Groningen, Nijenborgh 7, 9747 AG Groningen, The Netherlands Correspondence should be addressed to Lutz Schmitt, lutz.schmitt@hhu.de Received 8 February 2011; Accepted 20 June 2011 Received 8 February 2011; Accepted 20 June 2011 Received 8 February 2011; Accepted 20 June 2011 Academic Editor: Piero Andrea Temussi Copyright © 2011 Andr´e Abts et al. This is an open access article distributed under the Creative Commons Attribution License,
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Copyright © 2011 Andr´e Abts et al. This is an open access article distributed under the Creative Commons A
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is Nisin is an antimicrobial peptide produced and secreted by several L. lactis strains and is specifically active against Gram-positive
bacteria. In previous studies, nisin was purified via cation exchange chromatography at low pH employing a single-step elution
using 1 M NaCl. Here, we describe an optimized purification protocol using a five-step NaCl elution to remove contaminants. The obtained nisin is devoid of impurities and shows high bactericidal activity against the nisin-sensitive L. lactis strain NZ9000. Purified nisin exhibits an IC50 of ∼3 nM, which is a tenfold improvement as compared to nisin obtained via the one-step elution
procedure. Copyright strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of th
or copyright holder(s), unless the work is under an open content license (like Creative Commons). The publication may also be distributed here under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license. More information can be found on the University of Groningen website: https://www.rug.nl/library/open-access/self-archiving-pure/taverne-
amendment. Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately
and investigate your claim. Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately
and investigate your claim. Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to
and investigate your claim. Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the
number of authors shown on this cover page is limited to 10 maximum. Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the
number of authors shown on this cover page is limited to 10 maximum. Download date: 24-10-2024 Hindawi Publishing Corporation
International Journal of Peptides
Volume 2011, Article ID 175145, 9 pages
doi:10.1155/2011/175145 Research Article
Easy and Rapid Purification of Highly Active Nisin Andr´e Abts,1 Antonino Mavaro,1 Jan Stindt,1 Patrick J. Bakkes,1 Sabine Metzger,2
Arnold J. M. Driessen,3 Sander H. J. Smits,1 and Lutz Schmitt1 1. Introduction On agar plates, zones of growth inhibition of the indicator
organism can be easily visualized, and these so-called halo
assays allow determination of the minimum inhibitory
concentration (MIC) of the tested AMP [8]. Alternatively,
bacterial growth can be monitored in liquid culture in
dependence of the AMP concentration, by measuring the
optical density. This method allows the simple determination
of both the MIC and the IC50, that is, the concentration of
AMP that inhibits cell growth by 50% [9]. g
y
Nisin from L. lactis, like almost all naturally produced
AMPs, can be purified directly from the culture medium
[10, 11]. It is a cationic peptide, and therefore, commonly
purified using cation exchange chromatography (cIEX) at
acidic pH, using high salt concentration for elution, typically
a single-step elution with 1 M NaCl [12–16]. By using nisin
purified via such a method, the IC50 and MIC values were
determined for a variety of bacteria, such as various L. lactis strains, Enterococcus faecium, Bacillus anthracis, Bacillus
subtilis, Staphylococcus aureus, and Streptococcus mutans [9,
17]. Whereas E. faecium and S. mutans were equally sensitive
to nisin exhibiting an IC50 of 5 μM and MIC of 12.5 μM, L. lactis HP was much more sensitive to nisin, with an IC50
of 14 nM and MIC of 32 nM [9]. Nisin shows promising
activity towards clinical isolates of the Methicillin-resistant
Staphylococcus aureus (MRSA) bacterium, Streptococcus pyo-
genenes, and several of the most severe human pathogens,
including the multiresistant Streptococcus pneumoniae and
vancomycin-resistant E. faecium or E. faecalis, against which
new effective antibiotics are most urgently needed [18–
20]. In all these, studies purified nisin was used and the
bactericidal activity of nisin was measured by determining
the MIC or IC50. 2.1.2. Purchased Nisin. Commercial nisin (Sigma) is avail-
able as a lyophilized powder containing ∼2.5% (w/w) nisin. 1.3 g of nisin powder (corresponding to 32.5 mg nisin) was
dissolved in 100 mL 50 mM lactic acid pH 3 and filtered
through a 0.45 μm membrane filter (Pall Corporation). The nisin solution was applied to a 5 mL HiTrap SP HP
cation exchange column (GE Healthcare) at a flow rate
of 2 mL/min, whereas elution was performed at a flow
rate of 1 mL/min. Nisin purification, precipitation, and
concentration determination were carried out as described
above. Here, we describe a rapid and easy nisin purification
protocol, optimized to yield active, monomeric nisin. 1. Introduction Nisin, encoded by the nisA gene, is 2 International Journal of Peptides 2 of the overnight culture was used to inoculate 2 l of
GM17, and cells were grown at 30◦C. At an OD600 of
0.8, the culture was supplemented with nisin (Sigma) to a
final concentration of 1 ng/ml cell culture to ensure nisin
production and growth was continued over night. Next,
cells were removed by centrifugation (30 min., 6000xg at
4◦C). The supernatant was used and diluted with 1 l of
50 mM lactic acid pH 3 resulting in a 3 l solution with
a pH of 5. This solution was loaded on a 5 mL HiTrap
SP HP cation exchange (cIEX) column (GE Healthcare)
using a flow rate of 4 mL/min. Due to the large volume,
this step is optimally performed overnight. Protein elution
was monitored by measuring the absorbance at 215 nm. Unfortunately, it is not possible to detect nisin at 280 nm,
because it does not contain any aromatic amino acids;
therefore, the 215 nm was chosen as wavelength. The column
was subsequently washed with 50 mM lactic acid pH 3
until a stable baseline was reached to remove nonspecifically
bound material. Peptides were eluted by increasing the
NaCl concentration stepwise using a flow rate of 1 mL/min,
resulting in elution fractions with 200 mM (Step I), 400 mM
(Step II), 600 mM (Step III), 800 mM (Step IV) and 1 M
(Step V) NaCl. To remove NaCl, protein in the elution
fractions was precipitated with 20% (v/v) trichloroacetic acid
(TCA) overnight at 4◦C. Precipitated protein was washed two
times with ice-cold acetone to remove residual TCA. Finally,
the protein pellet was suspended in 50 mM lactic acid pH
3. Nisin concentrations were determined by a colorimetric
assay (Pierce BCA Protein Assay Kit, Thermo scientific)
by measuring the absorbance at 584 nm according to the
protocol of the manufacturer. molecules [6]. This highly specific interaction with lipid II is
reflected by the low nano- to micromolar concentrations of
nisin, sufficient to permeabilize the membrane of the target
cells [7]. The bactericidal activity of AMPs is generally measured
by quantifying growth inhibition of an AMP-sensitive target
organism, grown either on agar plates or in liquid culture. 1. Introduction a (methyl-)lanthionine ring containing peptide that is ribo-
somally synthesized as a prepeptide consisting of 57 amino
acids. The NisA prepeptide is modified posttranslationally
by the dehydratase NisB, which selectively dehydrates Serine
and Threonine residues, and by the cyclase NisC, which
catalyzes lanthionine ring formation. NisT finally transports
the modified prenisin across the cell membrane, and mature,
biologically active nisin is produced upon cleavage of the
leader peptide by the extracellular, membrane-anchored
protease NisP. Mature nisin harbors three dehydrated amino
acids (one dehydrobutyrine and two dehydroalanines), one
lanthionine ring, and four methyl-lanthionine rings. In
particular, these intramolecular rings are important for the
biological activity of nisin [5]. It is worth mentioning that
nisin induces its own synthesis via interaction with the two-
component regulatory system NisRK. The capacity to produce antimicrobial peptides (AMPs) is
widespread among Gram-positive bacteria. These AMPs are
directed against competitive microorganisms in order to
generate a selective advantage for the producer organism [1]. AMPs can be divided in three major classes. Class I consists
of the so-called lantibiotics, which are posttranslationally
modified peptides containing (methyl-)lanthionines, and
have a typical size of <5 kDa. Class II comprises heat-stable,
nonmodified peptides of 37–58 amino acids (typical size of
<10 kDa) with the leader peptide, as for Class I AMPs, being
removed during maturation. Class III contains AMPs that
are heat labile and that are generally larger in size (roughly
30 kDa). Especially heat-stable peptides secreted by lactic
acid bacteria have been studied extensively because of their
potential use as natural preservatives in the food industry. Nisin is active against Gram-positive bacteria and exerts
two killing mechanisms. Firstly, nisin inhibits cell-wall syn-
thesis by binding to lipid II, an essential membrane-anchored
cell-wall precursor, and secondly, nisin permeabilizes the
target membrane. The binding of nisin to lipid II induces
membrane integration of nisin resulting in the formation
of a pore, likely composed of eight nisin and four lipid II One of the best characterized AMPs is nisin, which is
secreted by Lactococcus lactis and is a member of the Class
I AMPs (for comprehensive reviews, see [2, 3]). Although
nisin has been used as a food preservative for more than
fifty years, no significant bacterial resistance against nisin
has been reported [4]. 1. Introduction For
this purpose, nisin was either produced from L. lactis strain
NZ9700 or purchased in the form of a lyophilized powder
that contains 2.5% (w/w) nisin. Nisin from both sources was
purified and tested for bactericidal activity using the nisin-
sensitive L. lactis NZ9000 strain. During the purification
low molecular weight contaminants are removed, which
results in purified nisin with high specific activity. This
allows a more accurate and reproducible determination of
the biological activity of nisin. 2.2. Tricine-SDS-PAGE. Tricine-SDS-PAGE was essentially
carried out as described in [21]. For analysis, 16 μL sample
was supplemented with 4 μL 5x SDS sample buffer (0.2 M
Tris-HCl, pH 6.8, 10% (w/v) SDS, 40% (v/v) glycerol, 0.02%
(w/v) bromophenol blue, and 10 mM DTT) and loaded
on a tricine gel consisting of a stacking gel containing 5%
acrylamide and a separation gel containing 16% acrylamide. The gel was run at 100 V for 2 hours, and proteins were
detected via silver staining. For all purification fractions,
3.2 μg of total protein was analyzed. 2.2. Tricine-SDS-PAGE. Tricine-SDS-PAGE was essentially
carried out as described in [21]. For analysis, 16 μL sample
was supplemented with 4 μL 5x SDS sample buffer (0.2 M
Tris-HCl, pH 6.8, 10% (w/v) SDS, 40% (v/v) glycerol, 0.02%
(w/v) bromophenol blue, and 10 mM DTT) and loaded
on a tricine gel consisting of a stacking gel containing 5%
acrylamide and a separation gel containing 16% acrylamide. The gel was run at 100 V for 2 hours, and proteins were
detected via silver staining. For all purification fractions,
3.2 μg of total protein was analyzed. 3.1. Purification of Nisin 3.1.1. Lyophilized Commercial Nisin. To determine the
antimicrobial activity of commercially available nisin, we
used a lyophilized powder, which contains ∼2.5% (w/w)
nisin. To further purify nisin, we initially performed SP
Sepharose cation exchange chromatography using 1 M NaCl
to elute the bound nisin. Subsequent SDS-PAGE analysis
of the eluate revealed a major protein band corresponding
to a peptide with a molecular mass of about 3.5 kDa (data
not shown) in line with the calculated mass of 3354 Da
for mature nisin. However, several higher molecular weight
components were also present. It is of note that these con-
taminants are not readily visualized by Coomassie Brilliant
Blue staining, whereas these impurities are clearly detected
by silver staining. Nisin purified via this method showed
antimicrobial activity against L. lactis NZ9000 exhibiting an
IC50 of 30 ± 12 nM. 2.4. IC50 Determination of Nisin. To investigate the antimi-
crobial activity of nisin, the nisin sensitive L. lactis strain
NZ9000 [22] was grown in a 96-well plate in GM17 medium. The total volume in each well was 200 μL, consisting of 50 μL
sample and 150 μL GM17 containing L. lactis NZ9000 cells
(starting OD600 = 0.1). Samples were prepared by diluting
nisin obtained from various cIEX elution fractions in 50 mM
lactic acid pH 3 to yield final protein concentrations ranging
from 0.15 nM to 300 nM in the wells. 50 mM lactic acid pH
3 without protein served as positive growth control. Cells
were grown at 30◦C, and the optical density was monitored at
620 nm every 20 min for a period of 8 hours (96 plate reader
BMG). To determine the IC50 values, the optical density
was normalized and plotted against the log of the nisin
concentration. Data were evaluated according to 50
The presence of contaminants prompted us to optimize
the purification of nisin. We first tested elution with a
linear gradient (50 times the column volume) ranging from
0–1 M NaCl. This approach, however, resulted in a broad
peak eluting throughout the NaCl gradient and further
analysis revealed no improvement when compared to the
single-step 1 M NaCl elution (data not shown). In contrast,
a substantial improvement was achieved when a five-step
NaCl step gradient was used to elute nisin from the cIEX
column (Figure 1(a)). Bound protein eluted at every step
as evidenced by the elution profile and subsequent tricine-
SDS-PAGE analysis (Figure 2(a)). 3.1. Purification of Nisin The Step II elution fraction
contained the bulk of nisin as evident by the major protein
band with a corresponding molecular mass of ∼3.5 kDa,
whereas substantially lower amounts of nisin were detected
in elution fractions I, III, and IV. In the Step V fraction
no nisin was visible. The Step I and II elution fractions
contained exclusively nisin, while elution fractions III–V
contained predominantly higher molecular weight com-
pounds (ranging from 6 kDa–70 kDa). In the latter fractions
compounds with molecular weights of ∼8 kDa and ∼10 kDa
were most prominent. The total protein concentration of the
elution fractions was determined to be: 2.1 mg/mL for Step
I (200 mM NaCl), 7.5 mg/mL for Step II (400 mM NaCl),
1.9 mg/mL for Step III (600 mM NaCl), 0.5 mg/mL for Step
IV (800 mM NaCl) and 0.4 mg/mL for Step V (1 M NaCl)
(Table 1). In Step II, ∼60% of the total purified nisin eluted. Thus, it appears that 400 mM NaCl is sufficient to elute the
vast majority of the nisin molecules. More importantly, the
nisin eluting under these conditions is essentially devoid of
contaminants. Y = OD min +(OD max −OD min)
1 + 10((log(IC50)−X)∗slope)
. (1) (1) The ODmax value describes the normalized OD600 value in
the starting plateau. The ODmin value corresponds to the
normalized OD600 of the end plateau value. Y stands for
the normalized optical density value, and X represents the
logarithmic concentration of the peptide. The IC50 value
is calculated as the value of the peptide concentration
used where the growth inhibition (or OD600) is 50%. This
corresponds to the inflection point of the resulting curves. The ODmax value describes the normalized OD600 value in
the starting plateau. The ODmin value corresponds to the
normalized OD600 of the end plateau value. Y stands for
the normalized optical density value, and X represents the
logarithmic concentration of the peptide. The IC50 value
is calculated as the value of the peptide concentration
used where the growth inhibition (or OD600) is 50%. This
corresponds to the inflection point of the resulting curves. 2.5. Growth Inhibition Visualized by a Halo Assay. The
antimicrobial activity of the different nisin preparations
was assessed by means of a halo assay. 2.1. Production and Purification of Nisin 2.1. Production and Purification of Nisin 2.1.1. Nisin Produced by L. Lactis. A culture of the L. lactis
strain NZ9700 was grown overnight in M17 medium con-
taining 0.5% (w/v) glucose (GM17) at 30◦C. Next, 100 mL 2.3. MALDI-TOF Mass Spectrometry. Samples obtained
from cIEX chromatography were precipitated with TCA and
washed with acetone as described above. The protein pellets 3 International Journal of Peptides International Journal of Peptides by the presence of clear zones (halos) resulting from growth
inhibition. were then dried for 15 min. at 30◦C in a vacuum concentrator
(Eppendorf concentrator plus). Dried pellets were stored
at −20◦C until analysis. For mass spectrometric analysis,
the samples were dissolved in water with 1% (v/v) formic
acid. Measurements were performed using a MALDI-TOF
instrument (Voyager-DE STR, Applied Biosystems) with a
nitrogen laser (λ = 337 nm) operating in reflector mode with
25 kV acceleration voltage. The samples were prepared by
the standard dried-droplet procedure, by applying 0.5 μL of
2,5-dihydroxybenzoic acid (DHB) matrix solution (10 mg in
1 mL water) plus 0.5 μL of sample solution. The droplet was
dried by a gentle flow of air. An external calibration with the
monomer ion of des-Pro-Bradykinin, Sub P, Bombesin, and
Melittin was used. The spectrum was obtained by averaging
200 laser shots. were then dried for 15 min. at 30◦C in a vacuum concentrator
(Eppendorf concentrator plus). Dried pellets were stored
at −20◦C until analysis. For mass spectrometric analysis,
the samples were dissolved in water with 1% (v/v) formic
acid. Measurements were performed using a MALDI-TOF
instrument (Voyager-DE STR, Applied Biosystems) with a
nitrogen laser (λ = 337 nm) operating in reflector mode with
25 kV acceleration voltage. The samples were prepared by
the standard dried-droplet procedure, by applying 0.5 μL of
2,5-dihydroxybenzoic acid (DHB) matrix solution (10 mg in
1 mL water) plus 0.5 μL of sample solution. The droplet was
dried by a gentle flow of air. An external calibration with the
monomer ion of des-Pro-Bradykinin, Sub P, Bombesin, and
Melittin was used. The spectrum was obtained by averaging
200 laser shots. 3.1. Purification of Nisin Purified nisin
obtained from the different cIEX elution fractions were
supplemented (16 μL) with 4 μL of 5x SDS sample buffer
and tricine-SDS-PAGE was carried out as described above. After electrophoresis the gel was incubated for 30 min in
an aqueous solution containing 20% (v/v) isopropanol and
10% (v/v) acetic acid. Subsequently, the gel was washed two
times for 30 min in ddH2O. At this stage, the gel was kept
in ddH2O at 8◦C until usage. Finally, the gel was overlaid
with GM17-agar (0.5% w/v agar) containing L. lactis NZ9000
cells at an OD600 of 0.1. After solidification, the overlaid
gel was incubated overnight at 30◦C to allow for bacterial
growth. The bactericidal activity of nisin is readily visualized 3.1.2. Nisin Produced by L. Lactis. An alternative to pur-
chasing nisin is to produce it in the laboratory, since L. lactis strains that secrete nisin in large amounts are readily International Journal of Peptides 4 4 2000
1500
1000
1000
500
0
0
100
200
300
800
600
400
200
0
Volume (mL)
Concentration (NaCl) (mM)
I
II
III
IV
V
Abs 215 (mAu)
(a)
0
2500
50
100
150
200
250
Volume (mL)
2000
1500
1000
500
0
1000
800
600
400
200
0
Concentration (NaCl) (mM)
I
II
III
IV
V
Abs 215 (mAu)
(b)
Figure 1: Purification of nisin via cation exchange chromatography. The elution profiles of the purification of commercial nisin (a) and
nisin produced by L. lactis NZ9700 (b) are shown. In both cases, nisin is eluted from the column using a five-step gradient with 200 mM
(Step I), 400 mM (Step II), 600 mM (Step III), 800 mM (Step IV) and 1 M NaCl (Step V). The different elution steps and corresponding
NaCl concentrations are indicated by the dashed line and the right y-axis, respectively. Protein was detected by measuring the absorbance at
215 nm. 2000
1500
1000
1000
500
0
0
100
200
300
800
600
400
200
0
Volume (mL)
Concentration (NaCl) (mM)
I
II
III
IV
V
Abs 215 (mAu)
(a) 0
2500
50
100
150
200
250
Volume (mL)
2000
1500
1000
500
0
1000
800
600
400
200
0
Concentration (NaCl) (mM)
I
II
III
IV
V
Abs 215 (mAu)
(b) 0
Concentration (NaCl) (mM) Concentration (NaCl) (mM) (b) (a) Figure 1: Purification of nisin via cation exchange chromatography. 3.1. Purification of Nisin The elution profiles of the purification of commercial nisin (a) and
nisin produced by L. lactis NZ9700 (b) are shown. In both cases, nisin is eluted from the column using a five-step gradient with 200 mM
(Step I), 400 mM (Step II), 600 mM (Step III), 800 mM (Step IV) and 1 M NaCl (Step V). The different elution steps and corresponding
NaCl concentrations are indicated by the dashed line and the right y-axis, respectively. Protein was detected by measuring the absorbance at
215 nm 10
15
25
M
I
II
III
IV
V
M
I
II
III
IV
V
(a)
(b)
Figure 2: Tricine-SDS-PAGE analysis of the cIEX purification of nisin. Purification of commercial nisin (a) and nisin secreted by the L. lactis NZ9700 strain (b). M, marker proteins; I, elution with 200 mM NaCl; II, elution with 400 mM NaCl; III, elution with 600 mM NaCl;
IV, elution with 800 mM NaCl; V, elution with 1 M NaCl. Protein was visualized by silver staining. The three lowest marker proteins are
indicated with molecular weights (kDa). M
I
II
III
IV
V
(a) 10
15
25
M
I
II
III
IV
V
(b) (b) (a) Figure 2: Tricine-SDS-PAGE analysis of the cIEX purification of nisin. Purification of commercial nisin (a) and nisin secreted by the L. lactis NZ9700 strain (b). M, marker proteins; I, elution with 200 mM NaCl; II, elution with 400 mM NaCl; III, elution with 600 mM NaCl;
IV, elution with 800 mM NaCl; V, elution with 1 M NaCl. Protein was visualized by silver staining. The three lowest marker proteins are
indicated with molecular weights (kDa). Table 1: IC50 values of each nisin containing fraction eluted from the cIEX column. Values are combined data from at least three independent
nisin purifications and subsequent inhibition experiments. Table 1: IC50 values of each nisin containing fraction eluted from the cIEX column. Values are combined data from at least three independent
nisin purifications and subsequent inhibition experiments. Purchased nisin
NZ9700 secreted nisin
Elution step
IC50 (nM)
Yield (mg)
IC50 (nM)
Yield (mg)
Step I (200 mM)
35.1 ± 0.1
4.2
n.i. 0.06
Step II (400 mM)
2.6 ± 0.1
15.1
11.2 ± 0.3
5.90
Step III (600 mM)
6.9 ± 0.2
3.8
n.i. 3.29
Step IV (800 mM)
27.0 ± 0.2
1.1
n.i. 0.14
Step V (1 M)
n.i. 0.8
n.i. 3.1. Purification of Nisin This difference in purity may
relate to differences in the loaded material. The lyophilized
nisin powder (also containing denatured milk solids) was
dissolved in 50 mM lactic acid, whereas nisin produced by L. lactis NZ9700 was applied to the cationic exchange column
as a 1 : 2 mixture of GM17 culture medium and 50 mM lactic
acid, respectively. 3.3. Mass Spectrometry. To assess and confirm the presence
of nisin in the individual cIEX elution fractions (Step II–
V) of the lyophilized nisin purification, we applied MALDI-
TOF mass spectrometry. The 400 mM NaCl elution fraction
contained only one peptide with a molecular mass of
3355.09 Da (Figure 4), which is in agreement with the
calculated mass of 3354.07 Da for nisin. Peak integration of
the total mass spectrum revealed that the 400 mM elution
fraction contains >98% of nisin, indicating that this fraction
is essentially devoid of contaminants. Nisin was also found
in the 600 mM and 800 mM elution fractions, whereas nisin
was not detected in the 1 M NaCl fraction. These results are
in agreement with tricine-SDS-PAGE analysis (Figure 2(a)). g
y
g
However, the 0.6–1 M NaCl elution fraction contained
several peptides with higher molecular masses. Subsequent
tandem MS analysis of these peptides yielded in-sequence
tags of eight amino acids and six amino acids, which unfor-
tunately could not be assigned to specific proteins. This was
due to the fact that the obtained sequence tags were too short
and when blasted gave multiple different protein hits (data
not shown). Nevertheless, the obtained sequence tags did not
match with the primary sequence of (pre)nisin. Therefore,
it can be excluded that the peptide contaminants with
molecular masses of ∼6–8 kDa are derived from prenisin. We considered the possibility that the compounds with
molecular weights of 6–8 kDa may represent the unprocessed
form of nisin. Immature nisin, that is, nisin still containing
the leader sequence, has a molecular weight of 5.9 kDa
and may arise if inefficient leader cleavage by the protease
NisP occurs. We therefore performed Western-blot analysis
using a polyclonal antibody raised against the nisin leader
sequence. Purified prenisin secreted by a L. lactis strain
lacking NisP was used as a positive control for Western-
blot analysis [23]. For all elution fractions, no signals were
observed suggesting that the observed compounds are not
derived from prenisin (data not shown). 3.4. Antimicrobial Activity of Purified Nisin. 3.1. Purification of Nisin 0.45
n i = no inhibition observed under the experimental setup International Journal of Peptides 5 gel was overlaid with GM17-agar containing nisin sensitive
bacteria (see Section 2). The biological activity of nisin is
visualized by the growth inhibition zones (halos) at the
position where nisin is present. The results for the lyophilized
and the laboratory produced nisin are shown in Figures
3(a) and 3(b), respectively. After overnight incubation at
30◦C halos were observed for elution Step I–IV for the
lyophilized nisin, while for the produced nisin, a halo was
only observed for elution fraction II. For both purifications,
the highest level of growth inhibition was observed for
fraction II. Importantly, the zones of inhibition are located
only at the position of the 3.5 kDa nisin peptide (Figure 3). Thus, purified nisin was biologically active and no growth
inhibitory activity is associated with the higher molecular
weight compounds. available. We used L. lactis NZ9700 grown in GM17 medium
to produce nisin (see Section 2). Culture supernatant con-
taining nisin was subjected to cIEX chromatography using
the same five-step NaCl elution gradient as described above. Here, a high absorbance at 215 nm occurred at the Step I
elution (Figure 1(b)), but this did not correspond to nisin
or other proteins as evidenced by silver-staining following
SDS-PAGE analysis (Figure 2(b), lane 2). Likely, this high
absorbance is due to ingredients from the growth medium,
which contains large amounts of peptone, tryptone, and
yeast extract. This was confirmed by a run with only GM17
media (data not shown). The remainder of the elution
profile is similar to that of the lyophilized nisin purification,
with two major absorbance peaks observed for the Step
II and III fractions (Figure 1(b)). Analysis of the protein
content of the different elution fractions revealed that nisin
is exclusively found in the 400 mM NaCl elution fraction
(Figure 2(b)). The total protein concentration of the 400 mM
elution fraction was 2.9 mg/mL. However, whereas nisin is
most prominent in this fraction, components with a MW ∼
6 kDa, ∼10 kDa, and ∼12 kDa are also present. The 0.6–1 M
elution fractions on the other hand contained compounds
with molecular weights ranging from 8 kDa–70 kDa, similar
to those observed for the purification of lyophilized nisin
(Figures 2(a) and 2(b)). When compared to the purification
of lyophilized nisin, nisin purified from GM17 medium
still contained contaminants. 3.1. Purification of Nisin Notably, the growth inhibition zone is only visible at a position of ∼3.5 kDa. 15
10
I
II
III
IV
V
kDa
(
)
(a) I
II
III
IV
V
(b) (a) (b) Figure 3: Bactericidal activity of the various nisin purification fractions. Equal amounts of protein of the different elution fractions (Step
I–V) from the purification of commercial nisin (a) and from nisin secreted by the L. lactis NZ9700 strain (b) were run on a tricine-SDS-PA
gel and overlaid with nisin-sensitive L. lactis NZ9000 cells (see Section 2). The position of marker proteins with known molecular weight
(kDa) are indicated on the left. The growth inhibition zones are visible as dark areas. Lanes I–V represent the five different elution fractions
of the cation exchange chromatography. For both purifications, maximum growth inhibition is observed for the Step II elution fraction
(400 mM NaCl). Notably, the growth inhibition zone is only visible at a position of ∼3.5 kDa. 100
90
80
70
60
50
40
30
20
10
0
3355.09
999
3799.4
6599.8
9400.2
12200.6
15001
646
Intensity (%)
Mass (m/z)
Figure 4: MALDI-TOF mass spectrometry analysis of purified
nisin. Mass spectrum of the Step II elution fraction (400 mM NaCl)
from the lyophilized nisin purification (for corresponding tricine-
SDS-PAGE analysis, see Figure 2(a), lane II). 100
90
80
70
60
50
40
30
20
10
0
3355.09
999
3799.4
6599.8
9400.2
12200.6
15001
646
Intensity (%)
Mass (m/z) the corresponding fraction obtained from the purification
of lyophilized nisin. We attribute this difference to the
contaminants that are still present (Figure 2(b), lane II). Nevertheless, for both purifications the highest bactericidal
activity is associated with the fraction that contains the
highest amount of nisin (Figures 2-3 and 5). Taken together
the data indicate that nisin obtained from the Step II elution
fraction has the highest specific activity. g
p
y
To determine whether NaCl used for elution has an effect
on nisin activity, we repeated the experiment and adjusted
the concentration of NaCl after elution in every fraction
to 500 mM either by dilution with buffer without salt or
by adding buffer and salt. Precipitated protein from these
fractions was subsequently used for growth experiments
as described above. 3.1. Purification of Nisin To quantitatively
assess the growth inhibitory activity of nisin obtained from
the different purification fractions, a liquid culture assay was
performed using L. lactis NZ9000 as reporter organism. The
optical density of the L. lactis NZ9000 cultures after 5 hours
of growth was plotted against the total protein concentration
of the nisin purification fractions. Results are shown for
both, the lyophilized (Figure 5(a)) and the produced nisin
(Figure 5(b)), respectively. The resulting growth curves and
the calculated IC50 values are shown in Figure 5 and Table 1,
respectively. 3.2. Biological Activity of Nisin. To test the biological activity
of the purified nisin, we used L. lactis NZ9000 as indicator
organism. L. lactis NZ9000 is a derivative of the plasmid-
cured L. lactis MG1363 and contains the nisRK genes inserted
in the chromosomal pepN locus [24]. This strain is com-
monly used as the host for nisin-induced expression system
(NICE) purposes [25]. However, since this strain lacks the
nisin immunity genes nisIFEG, it is sensitive to nisin [22]. The antimicrobial activities associated with the different
elution fractions obtained as described above were tested in a
so-called halo assay. For this, the various fractions containing
nisin were analyzed by tricine-SDS-PAGE and the tricine For lyophilized nisin, all cIEX elution fractions exhibited
growth inhibitory activity, however, with very distinct IC50
values (Figure 5(a) and Table 1). The Step II elution
fraction displayed the highest bactericidal activity with an
IC50 of 2.6 ± 0.1 nM. The other elution fractions showed
substantially higher IC50 values. Whereas the Step V elution 6 International Journal of Peptides 6 15
10
I
II
III
IV
V
kDa
(
)
(a)
kDa
15
10
I
II
III
IV
V
(
)
(b)
Figure 3: Bactericidal activity of the various nisin purification fractions. Equal amounts of protein of the different elution fractions (Step
I–V) from the purification of commercial nisin (a) and from nisin secreted by the L. lactis NZ9700 strain (b) were run on a tricine-SDS-PA
gel and overlaid with nisin-sensitive L. lactis NZ9000 cells (see Section 2). The position of marker proteins with known molecular weight
(kDa) are indicated on the left. The growth inhibition zones are visible as dark areas. Lanes I–V represent the five different elution fractions
of the cation exchange chromatography. For both purifications, maximum growth inhibition is observed for the Step II elution fraction
(400 mM NaCl). 3.1. Purification of Nisin In all cases, the IC50 values of the
“salt experiment” were slightly higher than when measured
directly after elution, indicating that more nisin is needed to
inhibit cell growth by 50% (data not shown). This indicates
that residual salt does not have a major influence on the
activity of nisin. It can, therefore, be excluded that the
differences in IC50 values of the various nisin containing
elution fractions are induced by the amounts of NaCl used
to elute nisin from the column. Figure 4: MALDI-TOF mass spectrometry analysis of purified
nisin. Mass spectrum of the Step II elution fraction (400 mM NaCl)
from the lyophilized nisin purification (for corresponding tricine-
SDS-PAGE analysis, see Figure 2(a), lane II). fraction had only an inhibitory effect at the highest tested
concentrations, the Step I, III, and IV elution fractions
exhibited an IC50 value of 35.1 ± 0.1 nM, 6.9 ± 0.2 nM, and
27.0 ± 0.2 nM, respectively. Thus, the 400 mM NaCl elution
fraction contains not only the bulk of nisin, it also contains
nisin that displayed the highest specific activity. 4. Discussion Data was fitted and evaluated according to (1). Commonly, nisin and other cationic AMPs are purified using
a single 1 M NaCl elution step, from a cIEX column at an
acidic pH [12–14, 16, 26]. production and secretion. Recently, in nisin-non-producing
L. lactis, nisin resistance was shown to be conferred by a
specific nisin resistance gene (nsr), which encodes a 35 kDa
nisin resistance protein (NSR). NSR proteolytically removes
the last six amino acids of nisin, thereby reducing its
bactericidal activity by a factor of 100 [27]. The activity of AMPs is strictly dependent on the target
organism. For example, the IC50 value of nisin ranges from
14 nM for L. lactis HP to 5 μM for Vancomycin-resistant E. faecium, with respective MICs of 32 nM and 12.5 μM [9]. In all these studies, it is noticeable that there is a large
variation in the sensitivity to nisin between isogenic strains of
Gram-positive bacteria, whereas some bacteria are inherently
resistant to nisin. There are several mechanisms by which
bacteria can become resistant to an antibiotic. The most
prominent example is the enzymatic destruction or mod-
ification of the antibiotic, thereby rendering it ineffective. β-Lactamases, for example, degrade the β-lactam ring of
penicillins. A second important mechanism of resistance is
shielding of the target such that the antibiotic cannot get
access to it—for example, by cell-surface alterations (capsules
S-layers) or by active extrusion by efflux pumps. Moreover,
the resistance of the AMP producer organism towards its
secreted AMP (autoimmunity) is typically based on ATP-
binding cassette (ABC) transporters which expel the AMP
from the membrane. The level of intrinsic resistance and the employed
mechanisms of antibiotic resistance may differ greatly among
microorganisms. It is, therefore, difficult, if not impossible,
to directly compare IC50 values for a given AMP when
comparing strains. Moreover, the purity of the AMP prepa-
ration is of great importance to determine accurately the
bactericidal activity of the AMP. y
In this study, we optimized the purification of the AMP
nisin and determined its IC50 values against the nisin-
sensitive L. lactis strain NZ9000. Nisin typically purified
via a 1 M NaCl one-step elution yields high levels of active
nisin (see above and [28]) but contains a substantial amount
of contaminants. We show that these contaminants, which
are mainly proteinaceous in nature, are largely removed by
using a five-step NaCl elution. 4. Discussion AMPs produced by Gram-positive bacteria form a unique
group secreted peptides [1]. Their uniqueness of especially
the lantibiotic group of AMPs, resides in the posttransla-
tional modifications, such as dehydration of amino acids
and intramolecular thioether bridges. One of the best-
characterized AMP is nisin, a compound used for more
than 40 years in up to 80 countries as an effective agent to
combat food-borne pathogens. Nisin has been purified and
its antimicrobial activity verified in numerous of studies [9]. A similar observation can be made for nisin purified
from the medium (Figure 5(b)). However, here only the
400 mM elution fraction shows bactericidal activity, which is
in agreement with tricine-SDS-PAGE analysis (Figure 2(b))
and the halo assay (Figure 3(b)). The calculated IC50 of
11.2 ± 0.3 nM is, however, ∼4-fold higher than that of International Journal of Peptides 7 1
0.8
0.6
0.4
0.2
−1
0
1
2
Log (nisin (nM))
OD 600
(a)
1
0.8
0.6
0.4
0.2
OD 600
−1
0
1
2
Log (nisin (nM))
(b)
Figure 5: IC50 determination of the nisin purification fractions. Growth inhibition experiments were performed with nisin obtained from
the different elution fractions of the purifications of commercial nisin (a) and of nisin secreted by the L. lactis NZ9700 strain (b). The log
of the used nisin concentration of each elution fraction is plotted against the normalized optical density of L. lactis NZ9000 after five hours
of growth. Shown are the inhibition curves for the NaCl elution fractions of 200 mM (o), 400 mM (△), 600 mM (∇), 800 mM (♦), and 1 M
(∗). Data was fitted and evaluated according to (1). 1
0.8
0.6
0.4
0.2
−1
0
1
2
Log (nisin (nM))
OD 600
(a) 1
0.8
0.6
0.4
0.2
OD 600
−1
0
1
2
Log (nisin (nM))
(b) (a) (b) Figure 5: IC50 determination of the nisin purification fractions. Growth inhibition experiments were performed with nisin obtained from
the different elution fractions of the purifications of commercial nisin (a) and of nisin secreted by the L. lactis NZ9700 strain (b). The log
of the used nisin concentration of each elution fraction is plotted against the normalized optical density of L. lactis NZ9000 after five hours
of growth. Shown are the inhibition curves for the NaCl elution fractions of 200 mM (o), 400 mM (△), 600 mM (∇), 800 mM (♦), and 1 M
(∗). References [17] W. M. de Vos, J. W. Mulders, R. J. Siezen, J. Hugenholtz, and
O. P. Kuipers, “Properties of nisin Z and distribution of its
gene, nisZ, in Lactococcus lactis,” Applied and Environmental
Microbiology, vol. 59, no. 1, pp. 213–218, 1993. [1] T. R. Klaenhammer, “Genetics of bacteriocins produced by
lactic acid bacteria,” FEMS Microbiology Reviews, vol. 12, no. 1-3, pp. 39–85, 1993. [18] E. M. Bavin, A. S. Beach, R. Falconer, and R. Friedmann,
“Nisin in experimental tuberculosis,” The Lancet, vol. 259, no. 6699, pp. 127–129, 1952. [2] C. Chatterjee, M. Paul, L. Xie, and W. A. van der Donk,
“Biosynthesis and mode of action of lantibiotics,” Chemical
Reviews, vol. 105, no. 2, pp. 633–683, 2005. [19] W. Brumfitt, M. R. J. Salton, and J. M. T. Hamilton-Miller,
“Nisin, alone and combined with peptidoglycan-modulating
antibiotics: activity against methicillin-resistant Staphylococ-
cus aureus and vancomycin-resistant enterococci,” Journal of
Antimicrobial Chemotherapy, vol. 50, no. 5, pp. 731–734, 2002. [3] J. Lubelski, R. Rink, R. Khusainov, G. N. Moll, and O. P. Kuipers, “Biosynthesis, immunity, regulation, mode of action
and engineering of the model lantibiotic nisin,” Cellular and
Molecular Life Sciences, vol. 65, no. 3, pp. 455–476, 2008. [4] M. Enserink, “Promising antibiotic candidate identified,”
Science, vol. 286, no. 5448, pp. 2245–2247, 1999. py
[20] E. Severina, A. Severin, and A. Tomasz, “Antibacterial
efficacy of nisin against multidrug-resistant Gram-positive
pathogens,” Journal of Antimicrobial Chemotherapy, vol. 41,
pp. 341–347, 1998. [5] O. P. Kuipers, M. M. Beerthuyzen, R. J. Siezen, and W. M. De
Vos, “Characterization of the nisin gene cluster nisABTCIPR
of Lactococcus lactis. Requirement of expression of the nisA
and nisI genes for the development of immunity,” European
Journal of Biochemistry, vol. 216, no. 1, pp. 281–291, 1993. [21] H. Schagger, “Tricine-SDS-PAGE,” Nature Protocols, vol. 1, pp. 16–22, 2006. [22] O. P. Kuipers, P. G. de Ruyter, M. Kleerebezem, and W. M. De Vos, “Quorum sensing-controlled gene expression in lactic
acid bacteria,” Journal of Biotechnology, vol. 64, no. 1, pp. 15–
21, 1998. [6] H. E. Hasper, B. De Kruijff, and E. Breukink, “Assembly and
stability of nisin-Lipid II pores,” Biochemistry, vol. 43, no. 36,
pp. 11567–11575, 2004. [7] H. E. van Heusden, B. De Kruijff, and E. Breukink, “Lipid
II induces a transmembrane orientation of the pore-forming
peptide lantibiotic nisin,” Biochemistry, vol. 41, no. 40, pp. 12171–12178, 2002. [23] H. B. van den Berg van Saparoea, P. J. Bakkes, G. N. Abbreviations IC50: 50% Inhibitory concentration
MIC: Minimal inhibitory concentration
MW: Molecular weight
AMP: Antimicrobial peptide. IC50: 50% Inhibitory concentration IC50: 50% Inhibitory concentration
MIC: Minimal inhibitory concentration MIC: Minimal inhibitory concentration [13] H. J. Choi, C. I. Cheigh, S. B. Kim, and Y. R. Pyun, “Production
of a nisin-like bacteriocin by Lactococcus lactis subsp. lactis
A164 isolated from Kimchi,” Journal of Applied Microbiology,
vol. 88, no. 4, pp. 563–571, 2000. MW: Molecular weight AMP: Antimicrobial peptide. A. Abts and A. Mavaro, contributed equally to this work A. Abts and A. Mavaro, contributed equally to this work [15] R. Rink, J. Wierenga, A. Kuipers et al., “Production of dehy-
droamino acid-containing peptides by Lactococcus lactis,”
Applied and Environmental Microbiology, vol. 73, no. 6, pp. 1792–1796, 2007. International Journal of Peptides 8 8 autoimmunity proteins are of great importance for L. lactis
cells that produce nisin (e.g., L. lactis NZ9700). [8] I. Wiedemann, E. Breukink, C. Van Kraaij et al., “Specific
binding of nisin to the peptidoglycan precursor lipid II com-
bines pore formation and inhibition of cell wall biosynthesis
for potent antibiotic activity,” Journal of Biological Chemistry,
vol. 276, no. 3, pp. 1772–1779, 2001. p
g
AMPs get more and more into the focus of biochemical,
biophysical, and medical studies due to their antimicrobial
activity against a wide variety of bacteria. Here, we report
an easy and rapid protocol for the purification of highly
active nisin, purified either directly from the culture medium
or from a commercially available lyophilized powder. Our
studies demonstrate the importance of obtaining AMP
preparations with high specific activity. A pure, homoge-
nous, and biologically active preparation will ensure reliable
determination of the efficacy of AMPs towards their micro-
bial target(s). Due to the similar chemical and biophysical
properties of lantibiotics, our manner of purification may
also apply to AMPs other than nisin. [9] T. J. Oman and W. A. Van Der Donk, “Insights into the mode
of action of the two-peptide lantibiotic haloduracin,” ACS
Chemical Biology, vol. 4, no. 10, pp. 865–874, 2009. [10] L. D. Kluskens, A. Kuipers, R. Rink et al., “Post-translational
modification of therapeutic peptides by NisB, the dehydratase
of the lantibiotic nisin,” Biochemistry, vol. 44, no. 38, pp. 12827–12834, 2005. [11] R. Rink, J. Wierenga, A. Kuipers et al., “Dissection and mod-
ulation of the four distinct activities of nisin by mutagenesis
of rings A and B and by C-terminal truncation,” Applied and
Environmental Microbiology, vol. 73, no. 18, pp. 5809–5816,
2007. [12] C. I. Cheigh, M. C. Kook, S. B. Kim, Y. H. Hong, and
Y. R. Pyun, “Simple one-step purification of nisin Z from
unclarified culture broth of Lactococcus lactis subsp. lactis
A164 using expanded bed ion exchange chromatography,”
Biotechnology Letters, vol. 26, no. 17, pp. 1341–1345, 2004. Acknowledgments The authors thank members of the group for valuable
discussions. This work was supported by DFG grant no. Schm1279/10-1 to L. Schmitt and grants from the Heinrich
Heine University [16] M. Uteng, H. H. Hauge, I. Brondz, J. Nissen-Meyer, and G. Fimland, “Rapid two-step procedure for large-scale purifica-
tion of pediocin-like bacteriocins and other cationic antimi-
crobial peptides from complex culture medium,” Applied and
Environmental Microbiology, vol. 68, no. 2, pp. 952–956, 2002. Heine University. 4. Discussion Notably, 400 mM NaCl (Step
II) was sufficient to elute the bulk of the nisin molecules,
while the majority of contaminants remain bound to the
column. In this manner, a nisin preparation was obtained
that exhibited a high specific activity. When tested against
the nisin-sensitive L. lactis NZ9000, this highly active nisin
exhibited an IC50 of 2.6 ± 0.1 nM, which is a 10-fold
improvement as compared to the nisin obtained via the one-
step elution. The potent bactericidal activity of nisin against
L. lactis NZ9000 lacking NisI and NisFEG suggests that these In L. lactis NZ9700, cells autoimmunity is mediated by
the scavenger protein NisI and the ABC transporter NisFEG
[5]. L. lactis cells lacking the NisIFEG defense system (e.g.,
L. lactis NZ9000) are sensitive to nisin and can be used as
indicator organism to measure the biological activity of nisin. The nisI and nisFEG genes are part of the nis operon and
are expressed in concert with the genes involved in nisin Author Contributions [14] J. Lubelski, R. Khusainov, and O. P. Kuipers, “Directionality
and coordination of dehydration and ring formation during
biosynthesis of the lantibiotic nisin,” Journal of Biological
Chemistry, vol. 284, no. 38, pp. 25962–25972, 2009. References Moll,
and A. J. M. Driessen, “Distinct contributions of the nisin
biosynthesis enzymes NisB and NisC and transporter NisT 9 International Journal of Peptides 9 to prenisin production by Lactococcus lactis,” Applied and
Environmental Microbiology, vol. 74, no. 17, pp. 5541–5548,
2008. [24] P. G. de Ruyter, O. P. Kuipers, and W. M. de Vos, “Controlled
gene expression systems for Lactococcus lactis with the food-
grade inducer nisin,” Applied and Environmental Microbiology,
vol. 62, no. 10, pp. 3662–3667, 1996. [25] I. Mierau and M. Kleerebezem, “10 Years of the nisin-
controlled gene expression system (NICE) in Lactococcus
lactis,” Applied Microbiology and Biotechnology, vol. 68, no. 6,
pp. 705–717, 2005. [26] L. D. Kluskens, S. A. Nelemans, R. Rink et al., “Angiotensin-(1-
7) with thioether bridge: an angiotensin- converting enzyme-
resistant, potent angiotensin-(1-7) analog,” Journal of Phar-
macology and Experimental Therapeutics, vol. 328, no. 3, pp. 849–855, 2009. [27] Z. Sun, J. Zhong, X. Liang, J. Liu, X. Chen, and L. Huan, “Novel mechanism for nisin resistance via proteolytic
degradation of nisin by the nisin resistance protein NSR,”
Antimicrobial Agents and Chemotherapy, vol. 53, no. 5, pp. 1964–1973, 2009. [28] N. S. Choi, S. Y. Jeong, H. J. Yang et al., “Activity assay for nisin-
like acidic bacteriocins using an optimal pH-conditioned gel
matrix,” Analytical Biochemistry, vol. 397, no. 2, pp. 259–261,
2010.
|
https://openalex.org/W4220713989
|
https://www.repository.cam.ac.uk/bitstream/1810/335775/2/fmars-09-830927.pdf
|
English
| null |
Anthropogenic Mixing in Seasonally Stratified Shelf Seas by Offshore Wind Farm Infrastructure
|
Frontiers in marine science
| 2,022
|
cc-by
| 21,520
|
Anthropogenic Mixing in Seasonally
Stratified Shelf Seas by Offshore
Wind Farm Infrastructure
Robert M. Dorrell 1*, Charlie J. Lloyd 1, Ben J. Lincoln 2, Tom P. Rippeth 2, John R. Taylor 3,
Colm-cille P. Caulfield 3,4, Jonathan Sharples 5, Jeff A. Polton 6, Brian D. Scannell 2,
Deborah M. Greaves 7, Rob A. Hall 8 and John H. Simpson 2 Robert M. Dorrell 1*, Charlie J. Lloyd 1, Ben J. Lincoln 2, Tom P. Rippeth 2, John R. Taylor 3,
Colm-cille P. Caulfield 3,4, Jonathan Sharples 5, Jeff A. Polton 6, Brian D. Scannell 2,
Deborah M. Greaves 7, Rob A. Hall 8 and John H. Simpson 2 1 Energy and Environment Institute, University of Hull, Hull, United Kingdom, 2 School of Ocean Sciences, Bangor University,
Bangor, United Kingdom, 3 Department of Applied Mathematics and Theoretical Physics, University of Cambridge,
Cambridge, United Kingdom, 4 BP Institute, University of Cambridge, Cambridge, United Kingdom, 5 School of Environmental
Sciences, University of Liverpool, Liverpool, United Kingdom, 6 National Oceanography Center, Joseph Proudman Building,
Liverpool, United Kingdom, 7 School of Engineering, Computing and Mathematics, University of Plymouth, Plymouth,
United Kingdom, 8 Centre for Ocean and Atmospheric Sciences, School of Environmental Sciences, University of East Anglia,
Norwich, United Kingdom The offshore wind energy sector has rapidly expanded over the past two decades,
providing a renewable energy solution for coastal nations. Sector development has
been led in Europe, but is growing globally. Most developments to date have been in
well-mixed, i.e., unstratified, shallow-waters near to shore. Sector growth is, for the first
time, pushing developments to deep water, into a brand new environment: seasonally
stratified shelf seas. Seasonally stratified shelf seas, where water density varies with
depth, have a disproportionately key role in primary production, marine ecosystem and
biogeochemical cycling. Infrastructure will directly mix stratified shelf seas. The magnitude
of this mixing, additional to natural background processes, has yet to be fully quantified. If large enough it may erode shelf sea stratification. Therefore, offshore wind growth may
destabilize and fundamentally change shelf sea systems. However, enhanced mixing may
also positively impact some marine ecosystems. This paper sets the scene for sector
development into this new environment, reviews the potential physical and environmental
benefits and impacts of large scale industrialization of seasonally stratified shelf seas
and identifies areas where research is required to best utilize, manage, and mitigate
environmental change. Edited by:
Frédéric Cyr,
Fisheries and Oceans Canada,
Canada Reviewed by:
Andrew M. Fischer,
University of Tasmania, Australia
Peter Hamlington,
University of Colorado Boulder,
United States *Correspondence:
Robert M. Dorrell
r.dorrell@hull.ac.uk Specialty section:
This article was submitted to
Physical Oceanography,
a section of the journal
Frontiers in Marine Science Received: 07 December 2021
Accepted: 24 January 2022
Published: 22 March 2022 REVIEW REVIEW
published: 22 March 2022
doi: 10.3389/fmars.2022.830927 published: 22 March 2022
doi: 10.3389/fmars.2022.830927 Keywords: offshore wind energy, shelf seas, marine biogeochemistry, stratification, turbulent mixing Citation: Renewable energy solutions, including offshore wind, are prerequisite for clean growth and thus
the reduction of greenhouse gas emissions needed to mitigate against climate change. Offshore
wind energy in shelf seas has seen a rapid increase over the past decade (Díaz and Soares, 2020; Xu
et al., 2020), motivated by: high-quality and reliable energy (wind) resources (Esteban et al., 2011);
space availability and site accessibility for installation of large, efficient, turbine systems (Sun et al.,
2012); rapidly maturing, reliable and energy-efficient technologies (Jansen et al., 2020); and reduced
visual impact on populated areas (Wen et al., 2018). Government programmes have helped drive
development of renewable offshore wind energy from offshore wind farm arrays, of tens increasing Dorrell RM, Lloyd CJ, Lincoln BJ,
Rippeth TP, Taylor JR, Caulfield CCP,
Sharples J, Polton JA, Scannell BD,
Greaves DM, Hall RA and Simpson JH
(2022) Anthropogenic Mixing in
Seasonally Stratified Shelf Seas by
Offshore Wind Farm Infrastructure. Front. Mar. Sci. 9:830927. doi: 10.3389/fmars.2022.830927 March 2022 | Volume 9 | Article 830927 Frontiers in Marine Science | www.frontiersin.org Infrastructure Mixing in Shelf Seas Dorrell et al. to hundreds, of offshore wind turbines (OWT) supported by
various fixed foundation designs with new floating foundations
being designed to access deeper water sites. Northwest Europe
has led sector development, with the UK leading in gigawatt
(GW) operational capacity to date (Global Wind Energy Council,
2020). The sector has grown rapidly, with technological advances
reducing the Levelised Cost of Electricity to a point where price is
competitive with alternative energy solutions (Shen et al., 2020). Thus, to meet demand, global development of offshore wind
energy in shelf seas is predicted to grow from 35 GW operational
in 2020 to 243 GW operational by 2030 (Figure 1A). biogeochemical cycling (Simpson and Sharples, 2012). Expansion
into this new environment means that offshore wind farms
will increasingly come into conflict with its environmental
functioning, controlled by natural mixing of water column
stratification (Figure 1C). Addressing engineering challenges, both fixed and floating
foundations are being developed to enable expansion into deeper
waters. Fixed foundations, which span the entire water depth,
include monopiles, gravity bases and jacket constructions (see,
e.g., Figure 2 and
Esteban et al., 2019; Díaz and Soares,
2020; Jiang, 2021b). However, floating foundations are crucial
to deep water, > 50 meters, deployment. Citation: This is critical to enable balance
of key global societal goals, i.e., to ensure access to affordable,
reliable, sustainable and modern energy and to conserve and
sustainably use the oceans, seas and marine resources (United
Nations, 2015). (
g
)
For the first time, large scale industrialization of seasonally
stratified marine environments is planned. Over two decades
of research has already focused on the direct impacts of
offshore wind farm development on well-mixed shallow water
marine ecosystems, from: benthic habitats (Dannheim et al.,
2020), fisheries (Gray et al., 2005) to seabirds (Exo et al.,
2003). Whilst this research is translatable with sector growth,
the seasonally stratified regime offers a fundamentally new
challenge: the introduction of infrastructure will lead to
enhanced “anthropogenic” mixing of stratified waters. Enhanced
mixing may lead to profound impacts on shelf sea dynamics
and thus marine ecosystem functioning. The aim of this
paper is to investigate the scope of these potential impacts. Section 2 reviews the ecosystem and physical functioning
of stratified shelf seas, highlighting the interface of physical
and biogeochemical processes where offshore wind farm scale
and submesoscales coincide. Section 3 then describes our
current understanding of the impact of offshore infrastructure
on unstratified and stratified waters. Section 4 discusses
current research challenges, the potential impact of offshore
wind on stratified shelf seas and the sector requirements
needed to ensure acceleration of renewable energy and its
sustainable development. It is concluded that offshore wind farm
infrastructure may have significant, and long lasting, effects on To date, most offshore wind farms have been installed in
the near-shore shallow water regions, up to 50 m depth, of
shelf-seas (Figure 1B). Near-shore shallow-water installations
have been preferred due to the cost reduction from ease
of access for installation, grid connection and operation and
maintenance (Jacobsen et al., 2019). With sector plans for an
additional 208 GW of operational capacity in the next decade,
and targets of 1.4 TW total by 2050 (Offshore Renewable
Energy Action Coalition, 2020), near-shore and shallow-water
sites are rapidly becoming limited. The scale of expansion
of offshore wind energy means the sector is now expanding
into deeper water sites further from shore (Soares-Ramos
et al., 2020). The transition from near-shore and shallow-
water environments to deeper water further from shore marks
a fundamental change in the marine environment. Citation: Learning from the
petroleum industry (Schneider and Senders, 2010), designs
include tension-leg platforms (Uzunoglu and Soares, 2020),
spar-pendulum (Cottura et al., 2021) and spar-buoy platforms
(Jacobsen and Godvik, 2021), and semi-submersible platforms
(Castro-Santos et al., 2020). Using the submerged structural
buoyancy and mooring forces to balance atmospheric thrust
and wave loads, floating foundations typically have large draft,
e.g., spar platforms, or large cross sectional area, e.g., semi-
submersible platforms (Butterfield et al., 2007). Thus, with
sector development requiring larger turbines that need bigger
rotors, which are subject to greater atmospheric loads, the
draft and diameter of fixed and floating foundations will need
to increase. The dynamics of atmospheric wakes from OWT
are already of key interest, given their control on available
wind power from turbine to array scale (Howland et al.,
2019). However, the dynamics of sub sea surface wakes from
foundations in well-mixed, and in particular, stratified waters is
poorly understood. Despite this, the > 20 m minimum draft of
current floating foundations is already large enough to penetrate
the thermocline and directly mix seasonally stratified shelf seas
(Figure 2). With over 80% of the global population living within 100 km
of the ocean, shelf seas have significant economic and social value,
including fishing, shipping, carbon storage (“blue” carbon) and
recreation. Despite comprising 8% of the total area of the global
ocean (Figure 1A), shelf seas support 15–30% of global ocean
biological production (Wollast, 1998). This biological production
ultimately supports > 90% of the world’s fish landings (Pauly
et al., 2002) and plays a disproportionately important role in
the absorption of CO2 from the atmosphere (Roobaert et al.,
2019). Thus, high biological productivity means shelf seas are
key components of global biogeochemical cycles, supporting
societally important bioresources and also the biological uptake
and storage of carbon in the marine environment. However,
interplay of social and economic drivers already places significant
stress on shelf seas (Kröger et al., 2018). Further industrialization
of shelf seas will enhance these stresses, with the potential for
significant long term environmental impact. Shelf sea dynamics
directly control primary production: the growth of microscopic
marine plankton. However, from OWT scale to coastal scale,
the impact of offshore wind development on shelf seas has yet
to be fully considered. Therefore, future offshore wind energy
development must be grounded in advanced understanding of
impact on shelf sea dynamics. Frontiers in Marine Science | www.frontiersin.org Citation: Shallow
waters are typically well-mixed; however deeper waters may
be subject to seasonal stratification, where density varies
vertically with depth (Figures 1A,B). Stratified waters are a
vital part of shelf seas, controlling primary production and March 2022 | Volume 9 | Article 830927 Frontiers in Marine Science | www.frontiersin.org 2 Infrastructure Mixing in Shelf Seas Dorrell et al. FIGURE 1 | Offshore wind and seasonally stratified shelf seas. (A) Shelf sea development; the limited extent of well-mixed waters, defined using the “h/u3
M2” criterion,
is highlighted against projected global offshore wind energy growth (Global Wind Energy Council, 2020). (B) NEMO model of the Northwest Europe summer potential
energy anomaly, φ, a measure of the amount of stratification (Guihou et al., 2018). (C) Fishing hotspots in the Celtic sea caused by topographically-enhanced mixing
of stratified waters (Sharples et al., 2013). In (A,B) seas are partitioned into regions prone to seasonal stratification and those remaining well mixed, based on: (A)
h/u3
M2 = 220 s3 m−2 (Simpson and Sharples, 2012); and (B) φ = 20 J m−3 (Gowen et al., 1995). In (B,C) offshore wind farms are separated into active sites, sites still
in construction or development and identified zones for future development (4C Offshore, 2021). FIGURE 1 | Offshore wind and seasonally stratified shelf seas. (A) Shelf sea development; the limited extent of well-mixed waters, defined using the “h/u3
M2” criterion,
is highlighted against projected global offshore wind energy growth (Global Wind Energy Council, 2020). (B) NEMO model of the Northwest Europe summer potential
energy anomaly, φ, a measure of the amount of stratification (Guihou et al., 2018). (C) Fishing hotspots in the Celtic sea caused by topographically-enhanced mixing
of stratified waters (Sharples et al., 2013). In (A,B) seas are partitioned into regions prone to seasonal stratification and those remaining well mixed, based on: (A)
h/u3
M2 = 220 s3 m−2 (Simpson and Sharples, 2012); and (B) φ = 20 J m−3 (Gowen et al., 1995). In (B,C) offshore wind farms are separated into active sites, sites still
in construction or development and identified zones for future development (4C Offshore, 2021). March 2022 | Volume 9 | Article 830927 Frontiers in Marine Science | www.frontiersin.org Infrastructure Mixing in Shelf Seas Dorrell et al. Citation: FIGURE 2 | Existing and emerging offshore wind engineering solutions (Díaz and Soares, 2020; Jiang, 2021b), including fixed monopile and jacket foundations and
floating semi-submersible and spar-buoy foundations, in comparison to shelf sea regimes from coastline to open ocean (Simpson and Sharples, 2012). FIGURE 2 | Existing and emerging offshore wind engineering solutions (Díaz and Soares, 2020; Jiang, 2021b), including fixed monopile and jacket foundations and
floating semi-submersible and spar-buoy foundations, in comparison to shelf sea regimes from coastline to open ocean (Simpson and Sharples, 2012). 2.1. Distribution and Seasonal Cycle of
Stratification fragile shelf sea ecosystems. Criteria for Environmental Impact
Assessments, must therefore be revised and updated to enable the
sustainable growth, and acceleration, of renewable offshore wind
energy development. The focus of offshore wind development is now shifting to the
central regions of temperate shelf seas; away from the generally
tidally energetic coast, and regions of freshwater influence
(Figure 1B). In these central shelf-sea regions the water column
structure of temperate shelf seas undergoes a seasonal cycle in
response to changes in heat exchange at the surface. In spring
and summer, some areas of the temperate shelf seas become
thermally stratified whilst neighboring areas remain well mixed. Here it has been established that the first order control on the
water column structure is the balance between the stratifying
influence of surface heating and turbulent mixing due to the tides
(Simpson and Hunter, 1974; Holt and Proctor, 2008). Within the
regions of seasonal stratification the energy sources and pathways
to mid-water column mixing remains an area of active research
(Lincoln et al., 2016; Inall et al., 2021). In regions of shallow
water and strong tidal currents (of order meters per second), the
rate of buoyancy input due to surface heating is insufficient for
the establishment of persistent stratification, and in consequence
the water column remains homogeneous. However, in regions of
deeper water and/or weaker tidal currents (and associated lower
levels of turbulence), surface heating dominates and seasonal
stratification develops. Away from coastal regions, shelf seas
generally exhibit tidal currents of order of tens of cm per second,
meaning that stratification are typically found when water depths 2. OCEANOGRAPHY OF STRATIFICATION
IN SHELF SEAS Shelf seas lie on the continental shelf between the coast and
the continental slope, where at 200 m water depth the sea
floor slopes down to the deep ocean. Despite only accounting
for 0.5% of ocean volume shelf seas play a key role in the
Earth system, dissipating > 70% of the tidal energy (Egbert and
Ray, 2000) and are disproportionately important in supporting
ocean biological production (Wollast, 1998), fish landings (Pauly
et al., 2002) and the absorption of CO2 from the atmosphere
(Roobaert et al., 2019). Biological production is underpinned
by the growth of microscopic marine phytoplankton, which
is tightly controlled by the timing and strength of seasonal
stratification. The seasonally stratified zones in shelf seas act as
an important net sink of carbon (Thomas et al., 2004). This
makes the physical and biogeochemical processes described here
a key dynamic component of the global carbon cycle (Bauer
et al., 2013) linking the atmospheric, terrestrial, and oceanic
carbon pools. March 2022 | Volume 9 | Article 830927 Frontiers in Marine Science | www.frontiersin.org 4 Infrastructure Mixing in Shelf Seas Dorrell et al. FIGURE 3 | Seasonal time-series of mooring observations from 2014 in the Celtic Sea (49◦24 N 8◦36 W), collected as part of the UK Natural Environment Research
Council (NERC) funded CaNDyFloSS project (NE/K00168X/1) (Scannell, 2020). (A) Measured thermocline current speed over 20–50 m mean depth; average speed
(blue line) and range (shaded region) over 2 tidal periods. (B) Surface heat exchange. (C) Vertical temperature structure. (D) Surface and bottom layer temperatures. FIGURE 3 | Seasonal time-series of mooring observations from 2014 in the Celtic Sea (49◦24 N 8◦36 W), collected as part of the UK Natural Environment Research
Council (NERC) funded CaNDyFloSS project (NE/K00168X/1) (Scannell, 2020). (A) Measured thermocline current speed over 20–50 m mean depth; average speed
(blue line) and range (shaded region) over 2 tidal periods. (B) Surface heat exchange. (C) Vertical temperature structure. (D) Surface and bottom layer temperatures. are greater than 80 m. In stratified regions, a warmer surface
water layer 5–40 m thick overlies a deeper cooler water layer. The two layers are separated by a region of strong vertical
temperature gradient, the thermocline, which forms a barrier to
vertical exchange of heat, salt, nutrients and momentum. Simpson and Bowers, 1981), consistent values have continued to
be estimated for a range of shelf seas globally. 2. OCEANOGRAPHY OF STRATIFICATION
IN SHELF SEAS Examples include
the Gulf of Maine and Bay of Fundy, Garrett et al. (1978) and
Loder and Greenberg (1986); the South China (Tong et al., 2010)
and Yellow Sea, Lie (1989) and Ren et al. (2014); the Patagonian
Shelf, Glorioso and Flather (1995); the northwest European Shelf
Seas, Pingree and Griffiths (1978), Sheehan et al. (2018) and
the Bering Sea, Schumacher et al. (1979). The robustness of the
critical value highlights the key role of the tides in determining
the position of shelf sea fronts and provided the first quantitative
link between the dissipation of tidal energy and ocean mixing. Tidal mixing fronts separate regions of seasonal stratification
from well mixed regions. Simpson and Hunter (1974) use an
energetics argument to derive a single parameter to predict the
positions of these fronts. By considering only vertical exchange
processes and assuming the surface input of heat was the only
stratifying influence, and that tidal currents are the only source
of energy driving mixing, they showed that the first order
determinate for the position of shelf sea fronts is given by the ratio The strength of stratification may be quantified in terms of the
potential energy anomaly, φ (J m−3), which describes the energy
required to fully mix a stratified water column (Simpson and
Bowers, 1981), h
u3
M2
,
(1) (1) φ = g
h
Z 0
−h
(ρ −ρ)zdz,
(2) (2) where u3
M2 is the principle lunar M2 tidal current amplitude,
and h is water depth (Figure 1A). In Figure 1A h/u3
M2 is
calculated from bathymetry and M2 tide data taken from TPX09
global tidal atlas (Egbert and Erofeeva, 2002), which applies a
generalized inverse method, assimilating satellite altimeter data,
into a global barotropic tidal model; here model resolution
limits precise location of stratified fronts from global data. In
terms of area, regions of seasonal stratification dominate the
continental shelf seas (Figure 1A). Whilst the critical value for
the ratio characterizing the position of tidal mixing fronts was
initially estimated for the Irish Sea (Simpson and Hunter, 1974; where h is the water depth and ρ the water density, ρ denotes the
density calculated using the mean water temperature and salinity
(Holt and Proctor, 2008). Geographical variation of φ across the
NW European Shelf Seas is plotted on the map in Figure 1B. Frontiers in Marine Science | www.frontiersin.org 2.2. Ecosystem Response to Stratification The organic products of the spring and summer primary
production
sink
into
the
deeper
waters,
where
bacteria
remineralise the organic material (nutrients and carbon)
back to the inorganic components. Remineralisation removes
oxygen from the deep water. In addition, the barrier role
of the thermocline limits the replenishment of that oxygen
from the atmosphere (Mahaffey et al., 2020). Together both
processes
determine
the
dissolved
oxygen
concentrations
available to benthic and pelagic organisms. Thus, high shelf sea
biological productivity means shelf seas are key components of
global biogeochemical cycles, supporting societally important
bioresources, and also the biological uptake and storage of
carbon in the marine environment. In turbulent unstratified regions, primary production occurs
mainly during summer months when sunlight is strong. However, plankton are continuously mixed from sea surface to
bed by turbulence, and spend much of their time below a depth
where light intensity is sufficient for growth. In contrast, stratified
waters provide ideal conditions for phytoplankton growth in
spring. As stratification forms phytoplankton become trapped
in the well lit surface layer, with the thermocline acting as a
barrier to mixing. Phytoplankton retained in the surface layer
enjoy the abundance of light, and exhibit rapid growth forming
the annual “spring bloom,” a biological abundance visible from
space that forms the year’s first supply of significant new organic
fuel. As the phytoplankton grow, they fix inorganic carbon in the
surface water into organic carbon, which causes the sea surface to
replenish its dissolved carbon concentration by absorbing CO2
from the atmosphere. The timing of the spring bloom is so
significant that zooplankton and fish larvae have evolved to use
it as a food source (Platt et al., 2003), with further implications
higher up the food-web, e.g., for shrimp survival (Ouellet et al.,
2011) and seabird breeding success (Frederiksen et al., 2006). 2.3. Shelf Sea Mixing Processes
Currents in shelf seas provide energy for stirring the water
column and are generally dominated by tidal motions, with
episodic contributions by the wind in the upper water column. An example of the current variability from the Celtic Sea, a typical
shelf sea location, is presented in Figure 3A, where velocities vary
from 0.1 to 0.7 m s−1. For this location, the semi-diurnal tide
lunar M2 produces two high and low tides a day with 4 peaks
in current speed. The interaction with the principle solar tidal
component, S2, produces the 14 day spring-neap cycle. 2. OCEANOGRAPHY OF STRATIFICATION
IN SHELF SEAS A typical seasonal stratification cycle shows a time-series
of warming and cooling, varying with water depth (Figure 3). Surface mixed layer temperatures warm from April into the
summer in response to a net positive buoyancy input due to
surface heating. The observed surface temperatures of 15−20 ◦C
are typical for temperate shelf seas in midsummer, while the deep March 2022 | Volume 9 | Article 830927 Frontiers in Marine Science | www.frontiersin.org 5 Infrastructure Mixing in Shelf Seas Dorrell et al. the depth of which varies from 10 to 40 m across the hundreds
of kms that it extends over the shelf (Figure 4). The SCM
plays a vital role in supporting the pelagic food web during
summer. Estimates based on observations of primary production
rates within the SCM suggest that subsurface carbon fixation
accounts for up to 50% of annual primary production in the
seasonally stratified North Sea (Richardson et al., 2000; Weston
et al., 2005). An extrapolation using microstructure-based nitrate
flux estimates also gives the same approximate figure (Rippeth
et al., 2009). The persistence of production in the SCM is
dependent on a vertical flux of nutrients from the deep nutrient-
rich water below (Sharples and Tett, 1994). In consequence,
the processes responsible for mixing across the thermocline,
discussed in Section 2.5, are key to delivering the limiting
nutrients to the euphotic zone and sustaining the SCM (Sharples
et al., 2001a,b, 2007; Williams et al., 2013a). Episodic mixing
associated with storm initiated inertial oscillations (Burchard
and Rippeth, 2009; Lincoln et al., 2016) have been shown to
drive significantly enhanced nutrient fluxes (Williams et al.,
2013b). Local enhancement of primary productivity is evident in
regions of steep topography, where tidally induced internal waves
elevate mixing and nutrient fluxes. Chlorophyll concentration in
the SCM is greatly elevated at the shelf break (Sharples et al.,
2007), and mid-shelf sand banks (Sharples et al., 2013). The
dependence of the marine food web and fisheries on upward
flux of nutrients is evident in the distribution of seasonal fishing
hot-spots (Figure 1C). water remains close to its winter temperature of 10 ◦C. Over
this period the strength of stratification grows with a surface to
bed temperature difference exceeding 10 ◦C by August, which
gives rise to a density difference of 0.3 kg m−3. Frontiers in Marine Science | www.frontiersin.org 2. OCEANOGRAPHY OF STRATIFICATION
IN SHELF SEAS The stratification
weakens into autumn as surface cooling leads to convection
and storms drive turbulent mixing, such that the water column
becomes well mixed during the winter. During the stratified period the deeper water is isolated from
the surface layer by a thermocline. A slow warming of the deep-
water results from mixing down of heat from the sea surface. The
rate of warming, set by thermocline mixing, varies geographically
and is important as it determines the timing of the autumnal
breakdown of stratification (Rippeth, 2005), and transport of
nutrient rich deep water up to the surface layer. Regional variations on controls of the distribution, and
seasonal cycles of stratification, do exist. For example, wind
driven upwelling on narrow shelves, subject to weak tides, can
dictates patterns of stratification (see, e.g., Austin and Barth,
2002). However, offshore wind sector development has not yet
extended to such regions. 2.2. Ecosystem Response to Stratification
Primary production of organic matter by phytoplankton forms,
directly or indirectly, the primary food source for almost all
marine organisms. Phytoplankton growth requires CO2, sunlight
and nutrients, the availability of which are determined by water
column structure, with profound implications for the biological
functioning of the shelf seas. 2.2. Ecosystem Response to Stratification High chlorophyll
concentration in the SCM (subsurface chlorophyll maximum) indicate phytoplankton production extending hundreds of kms across the shelf. Enhancements in
concentration over rough topography, such as Jones Bank, are a result of elevated turbulent mixing, driving nutrient fluxes which correspond directly to hotspots of
marine biodiversity and thus fisheries (Figure 1C). ever smaller scales until their energy is dissipated either to heat,
or to potential energy via mixing. range from ε = 10−7 −10−5 W m−3 (Rippeth, 2005), 2–3 orders
of magnitude smaller than rates commonly found in shallow well
mixed waters (Simpson et al., 1996). Empirical estimates of the
bulk mixing efficiency of the barotropic tide in stratified waters,
are very low, RfBT ≈0.0037 (Simpson et al., 1978), as most
turbulence is produced in the well mixed bottom layer, so that
no mixing is possible. In addition, strong density gradients in the
thermocline inhibit vertical mixing, and as such rates of vertical
mixing observed in shelf seas are comparable to using a hand
mixer in a swimming pool. In the absence of convection, a three-way local balance is
assumed (assuming both at least quasi-stationarity and that
transport terms can be ignored): P = B + ε,
(3) (3) where P is the (total) production of turbulent kinetic energy
(TKE), B is the buoyancy production (mixing), and ε is the TKE
dissipation rate (heat). The production of turbulence in stratified waters is inhibited
by buoyancy forces arising from vertical density gradients, which
are quantified by the Brunt-Väisälä, or buoyancy, frequency,
N, where The efficiency of mixing by turbulence can be quantified
by the flux Richardson number Rf
=
B/P and is widely
assumed to have a value Rf ≪1 in a stratified fluid. Since ε
is a commonly measured turbulence metric, it is often used to
infer the rate of mixing using the closely related flux dissipation
coefficient, defined in terms of the buoyancy production as
Ŵ = B/ε. A value of Ŵ ≈0.2 (i.e., Rf ≈1/6) is routinely
applied, and has been verified for the shelf sea thermocline by
a number of different observational approaches (Inall et al.,
2000; Oakey and Greenan, 2004; Palmer et al., 2008; Bluteau
et al., 2013), though it has been found to vary in other regimes
(Monismith et al., 2018). 2.2. Ecosystem Response to Stratification Wind
driven currents are also observed in the top 50 m, and take the
form of inertial oscillations, which have a latitude dependent
period, which is 14.9 h at the mooring location. During the spring bloom, the availability of nutrients in
the surface layer becomes exhausted, and further production is
limited by nutrient supply. Despite this limitation on plankton
growth, a persistent and significant level of primary production
is sustained at depth, throughout the period of seasonal
stratification. This sub-surface phytoplankton layer located in the
stratified thermocline water is a ubiquitous feature and is known
as the “subsurface chlorophyll maximum” (SCM) (Pingree et al.,
1982). In shelf seas, the SCM occupies a 10–30 m thick layer, Friction at the seabed and the sea surface generate vertical
current shear in the flow, and turbulent eddies which cascade to March 2022 | Volume 9 | Article 830927 6 Infrastructure Mixing in Shelf Seas Dorrell et al. FIGURE 4 | Observational data from the Celtic Sea collected in summer 2008, supported by NERC’s Oceans2025 Programme (Sharples et al., 2013). Section of
temperature (line contours) and chlorophyll concentration (colors) measured using a Scanfish CTD towed along the path show on the map. High chlorophyll
concentration in the SCM (subsurface chlorophyll maximum) indicate phytoplankton production extending hundreds of kms across the shelf. Enhancements in
concentration over rough topography, such as Jones Bank, are a result of elevated turbulent mixing, driving nutrient fluxes which correspond directly to hotspots of
marine biodiversity and thus fisheries (Figure 1C). FIGURE 4 | Observational data from the Celtic Sea collected in summer 2008, supported by NERC’s Oceans2025 Programme (Sharples et al., 2013). Section of
temperature (line contours) and chlorophyll concentration (colors) measured using a Scanfish CTD towed along the path show on the map. High chlorophyll
concentration in the SCM (subsurface chlorophyll maximum) indicate phytoplankton production extending hundreds of kms across the shelf. Enhancements in
concentration over rough topography, such as Jones Bank, are a result of elevated turbulent mixing, driving nutrient fluxes which correspond directly to hotspots of
marine biodiversity and thus fisheries (Figure 1C). FIGURE 4 | Observational data from the Celtic Sea collected in summer 2008, supported by NERC’s Oceans2025 Programme (Sharples et al., 2013). Section of
temperature (line contours) and chlorophyll concentration (colors) measured using a Scanfish CTD towed along the path show on the map. Frontiers in Marine Science | www.frontiersin.org 2.2. Ecosystem Response to Stratification N =
s
−g
ρ
∂ρ
∂z ,
(4) (4) describes the frequency at which a displaced parcel of fluid will
oscillate in a stratified system and is thus a measure of the
stability of stratified waters. Conversely, the vertical current shear,
S = ∂u/∂z, is a measure of the extraction of energy from the
mean flow, and therefore power available to overcome buoyancy
forces and generate turbulence. The generation of instabilities in
stratified water is quantified using measurements of the buoyancy A consequence of shear production at the seabed by
barotropic tidal currents is that measured rates of turbulence are
extremely low in the seasonal thermocline, orders of magnitude
lower than at the boundaries (Figure 5B). Mean dissipation rates March 2022 | Volume 9 | Article 830927 Frontiers in Marine Science | www.frontiersin.org 7 Infrastructure Mixing in Shelf Seas Dorrell et al. FIGURE 5 | Marginal stability and energy dissipation in seasonally stratified shelf seas. (A) The equivalence between buoyancy frequency N2 (blue) and vertical current
shear S2 (red), temporally averaged (over two tidal cycles). (B) The coincident temporally averaged profile of TKE dissipation rate ε. Measurements from direct
observations in the Western Irish Sea in June 2002 (Rippeth, 2005). FIGURE 5 | Marginal stability and energy dissipation in seasonally stratified shelf seas. (A) The equivalence between buoyancy frequency N2 (blue) and vertical current
shear S2 (red), temporally averaged (over two tidal cycles). (B) The coincident temporally averaged profile of TKE dissipation rate ε. Measurements from direct
observations in the Western Irish Sea in June 2002 (Rippeth, 2005). frequency, N, and vertical current shear, S, to calculate the
gradient Richardson Number winner, with all schemes under representing the thermocline
properties and suggested that physical processes are still missing. Candidate mechanisms to account for the deficit in mid water
mixing include internal waves generated by stratified flow over
topography and wind generated inertial currents. Rig = −g
ρ
∂ρ
∂z
∂u
∂z
2 ≡N2
S2 . (5) (5) Internal tides propagate at the thermocline in response to
tidal currents flowing over steep topography (Rippeth, 2005; Inall
et al., 2021). These waves generate strongly sheared currents
about the thermocline which can lead to the development of
shear instability, draining energy to turbulence which then
supports mixing in the quiescent mid-water region. 2.2. Ecosystem Response to Stratification Although the
energy in the internal tide is much less than the barotropic tide,
an empirical estimate of bulk mixing RfIT ≈0.056 (Stigebrandt
and Aure, 1989) shows it to be highly efficient compared with
the barotropic tide RfBT ≈0.0037 (Simpson et al., 1978), on
account of the turbulence coinciding with the stratification. A consequence of the differing bulk mixing efficiencies is
that whilst the barotropic tide dominates the budget for the
dissipation of barotropic tidal energy in the seasonally stratified
shelf seas to the west of the UK, the shelf break generated
internal tide is estimated to dominate diapcynal mixing (Rippeth
et al., 2005). With advances in computational performance and
associated resolution increase, simulation of internal tides are
now routine for European Shelf simulations (Guihou et al., 2018)
implemented operationally (Graham et al., 2018; Tonani et al.,
2019). Though, as hydrostatic simulations, the non-hydrostatic
mixing processes must still be parameterised. Both theory (Howard, 1961; Miles, 1961) and observations
(Silvester et al., 2014; Lincoln et al., 2016) show that shear
instability and internal wave breaking occurs when Rig ≲0.25. Average of measurements, over two tidal cycles, show N2 and
S2 are approximately equal implying a (close to) marginally
stable water column where Rig ≈1 (Figure 5A). Therefore, any
additional sources of shear can generate instabilities and mixing. This marginal stability has been widely observed across shelf seas
(van Haren et al., 1999; Rippeth, 2005; Palmer et al., 2008) but it
does not universally explain the observations (Palmer et al., 2013)
suggesting additional sources of shear may, at times, control
system dynamics. One dimensional vertical exchange models using turbulence
closure schemes (also commonly referred to as RANS models)
fail to represent the measured rates of mixing through tidal
and wind driven boundary processes (Simpson et al., 1996). The deficit in the predicted mid-water ε points to either an
incorrect parameterisation of the small scale physics away from
the boundaries or to the absence of key physical processes in
the model. A recent study (Luneva et al., 2019) evaluated a
range of alternative one dimensional turbulence closure schemes
for the Northwest European shelf seas, packaged within the
Generic Length Scale two-equation formulation (Umlauf and
Burchard, 2005). Evaluating the schemes against profile data
(28,000 profile in total) confirmed that there was no outright Frontiers in Marine Science | www.frontiersin.org 2.4. Submesoscale Mixing The size of many existing and proposed offshore wind farms
coincides with the ocean submesoscale (∼1 −20 km). The March 2022 | Volume 9 | Article 830927 8 Infrastructure Mixing in Shelf Seas Dorrell et al. FIGURE 6 | Near-surface chlorophyll concentration (mg m−3) from 20 April, 2020, inferred from the NASA MODIS Aqua satellite. (A) Chlorophyll concentration for
Northwestern European waters. (B) Detail of the north North sea highlighting submesoscale instabilities at, and below, the scale of offshore wind farm development. The image was made from data downloaded from http://oceancolor.gfsc.nasa.gov and 4C Offshore (2021). FIGURE 6 | Near-surface chlorophyll concentration (mg m−3) from 20 April, 2020, inferred from the NASA MODIS Aqua satellite. (A) Chlorophyll concentration for
Northwestern European waters. (B) Detail of the north North sea highlighting submesoscale instabilities at, and below, the scale of offshore wind farm development. The image was made from data downloaded from http://oceancolor.gfsc.nasa.gov and 4C Offshore (2021). submesoscales increase the water column stratification and tend
to reduce the depth of the surface mixed layer. Submesoscales
can also be generated as currents move around topographic
features. For example, submesoscale eddies can be generated by
flow past islands (Marmorino et al., 2018), along continental
slopes (Gula et al., 2015; Molemaker et al., 2015), and over
seamounts (Srinivasan et al., 2019). It is possible that the
enhanced drag experienced by the flow through a submesoscale
offshore wind farm could similarly generate submesoscale eddies. The enhanced mixing and drag within wind farms could also
influence submesoscale instabilities and eddies, although these
hypotheses remain untested. imprint of submesoscale eddies can be seen in satellite images of
chlorophyll concentration in the North Sea as seen in Figure 6
(note that the future development zones shown for reference will
likely consist of multiple smaller windfarms). The submesoscale
range is characterized by Rossby numbers, Ro ≡u/(fL) ∼1,
where u and L are characteristic horizontal velocity and length
scales associated with submesoscales and f is the Coriolis
parameter. This makes submesocales dynamically distinct from
mesoscales (∼100 km) where Ro ≪1. While the Earth’s rotation
is important on submesoscales, it does not constrain the flow
as strongly as it does for larger scales, leading to a unique set
of physical processes. Submesoscale currents are generally more
energetic in regions with large horizontal density contrasts which
includes coastal waters. Submesoscales are important partly because they interact
with vertical mixing processes. Frontiers in Marine Science | www.frontiersin.org 2.5. Mixing Across Density Interfaces As noted, stratified shelf seas sometimes exhibit a two-layer
density structure with relatively homogeneous mixed layers at the
top and bottom separated by a stratified thermocline (Figure 3). There are two paradigms for mixing, at the small scales of
density interfaces; ‘scouring’ and ‘overturning’ (see Caulfield,
2021 and references therein). The mixing across the density
interfaces and the ultimate fate of the density interface depends
strongly on which regime the turbulence is in. The mixing
regimes are controlled by the relative strength of stratification
and turbulence. This can be quantified by the kinetic energy
associated with three-dimensional turbulent eddies and the
potential energy associated with the density interface. When
the stratification at the density interface is sufficiently strong,
turbulent eddies do not have kinetic energy to overturn the
density interface. Instead, if there is a source of turbulence in
the surrounding mixed layers, turbulent eddies will ‘scour’ the
density interface, pulling characteristic wisps of fluid into the
mixed layers (Figure 7B). On the other hand, when the kinetic
energy associated with the turbulent eddies is large enough,
turbulence is able to ‘overturn’ the density interface (Figure 7A). The scouring regime requires a source of turbulence external
to the density interface. In natural stratified shelf seas, turbulence
generated by bottom friction and surface forcing (e.g., wind,
waves, and/or convection) generate turbulence below and above
mid-water density interfaces. Turbulence associated with flow
past OWTs could also play this role. Indeed, as discussed in
Section 3.2, horizontal shear can be very effective at generating
and maintaining layers in stratified flows. On the other hand,
energetic turbulence driven by the flow past OWTs could be
strong enough to overturn and mix natural density interfaces. Thus, it is not clear whether OWTs would generate turbulence
in the overturning or scouring regime. The distinction will likely
depend on the strength of stratification, the speed of flow through
offshore wind farms, and the geometry of wind turbines. As will
be expanded on in the following section more research is needed
to understand mixing generated by offshore wind infrastructure
and their impact from the scale of OWTs to the shelf sea scale. The fate of the density interface is intimately tied to the
mixing regime; scouring tends to sharpen density interfaces,
while overturning tends to mix the interface into a more diffuse
state. 2.5. Mixing Across Density Interfaces The mixing regime can be quantified by considering the
budget for the buoyancy frequency in sorted density coordinates
(Taylor and Zhou, 2017; Zhou et al., 2017): ∂N2
∗
∂t
= ∂2κe
∂z2∗
N2
∗
| {z }
A
+ 2∂κe
∂z∗
∂N2
∗
∂z∗
|
{z
}
B
+ κe
∂2N2
∗
∂z2∗
| {z }
C
. (6) 2.4. Submesoscale Mixing Submesoscales are characterized
by timescales that range from hours to ∼1 day. As a
result, the tendency for submesoscales to increase the vertical
stratification of the water column is fast enough to compete
with vertical mixing driven by winds and tides. At the same
time, submesoscales can induce strong vertical circulations and
locally enhance the exchange between the surface mixed layer
and thermocline (Mahadevan and Tandon, 2006). Gula et al. (2015) provide an in-depth review of the connections between
submesoscales and ocean mixing. It is not clear if, or how, offshore wind farms might
interact
with
submesoscale
currents. However,
since
the spatial extent of wind farms typically lie within the
submesoscale
range,
array
scale
flow
patterns
generated
by enhanced drag or mixing within offshore wind farms
could be subject to the physical processes that are active
on the submesoscale. Below we very briefly consider some
of these processes. See Thomas et al. (2008), McWilliams
(2016), and Gula et al. (2022) for in-depth reviews of
submesoscale dynamics. Submesoscales can have a strong impact on biogeochemistry
(Lévy et al., 2012; Mahadevan, 2016). For example, shoaling
of the surface mixed layer depth and suppression of turbulent
mixing induced by submesoscales can trigger phytoplankton
blooms in otherwise light-limited conditions (Taylor and Ferrari,
2011; Mahadevan et al., 2012; Taylor, 2016). In nutrient-limited
conditions, submesoscales can upwell nutrient-rich waters to the
euphotic zone, enhancing primary production (Mahadevan and
Archer, 2000; Lévy et al., 2001). Finally, downwelling circulation
and the suppression of turbulent mixing can enhance the export Submesoscales
are
typically
generated
by
dynamical
instabilities or flow/topography interactions. The instability
mechanisms
include
ageostrophic
versions
of
instabilities
that also exist at larger scales (e.g., baroclinic instability)
and instabilities that are unique to the submesoscale (e.g.,
inertial/centrifugal and symmetric instability). Although the
details and energy pathways differ, the net effect of submesoscale
instabilities is to convert potential energy associated with
horizontal density gradients into kinetic energy. In the process, March 2022 | Volume 9 | Article 830927 9 Infrastructure Mixing in Shelf Seas Dorrell et al. where κ is the molecular diffusivity and ⟨·⟩z∗denotes a spatial
average for a fixed z∗. Note that this formulation assumes
that density is controlled by one variable (temperature or
salinity). The key advantage of using sorted height coordinates
is that κe is strictly positive. 3. FLOW STRUCTURE INTERACTIONS (6) The dynamics of flows past a range of structures has seen
significant research over the last 50 years, both due to immediate
real-world applications and the increase in computational
resources and experimental measurement fidelity (Williamson,
1996). Here work studying the dynamics of unstratified and
stratified flow past infrastructure, relevant to the offshore wind
sector, are integrated and reviewed. Here z∗is the height of a vertically sorted isopycnal, adiabatically
sorted to monotonically decrease with depth; N2
∗is the buoyancy
frequency, Equation (4), in sorted height coordinates; and ρ is the
potential density, where ρ0 is a reference density. The ‘effective’
diffusivity is (Nakamura, 1996; Winters and D’Asaro, 1996) Here z∗is the height of a vertically sorted isopycnal, adiabatically
sorted to monotonically decrease with depth; N2
∗is the buoyancy
frequency, Equation (4), in sorted height coordinates; and ρ is the
potential density, where ρ0 is a reference density. The ‘effective’
diffusivity is (Nakamura, 1996; Winters and D’Asaro, 1996) 2.4. Submesoscale Mixing A purely laminar flow has κe =
κ, and turbulence will lead to κe
>
κ by distorting the
isopycnals, increasing the density surface and thereby increasing
the density gradients. FIGURE 7 | Cartoon illustrating overturning (A) and scouring (B) regimes at a
density interface (Caulfield, 2021). The relative sizes of the terms on the right hand side of
Equation (6) can be used to diagnose the mixing regimes. Terms
B and C in Equation (6) represent translation and diffusion
of the sorted density profile, respectively, and since κe > 0
this diffusion acts to spread out density interfaces. Term A in
Equation (6) can be positive or negative, and this term dictates
whether turbulence at a density interface is in the scouring
or overturning regimes. For example, consider a flow in the
overturning regime where shear-driven turbulence is generated
at a density interface (Figure 7A). If the flow above and below the
interface is relatively quiescent, κe could exhibit a local maximum
at the density interface with ∂2κe/∂z2
∗< 0. In this case term A
will act to reduce the stratification at the interface. On the other
hand, in the scouring regime (Figure 7B) strong stratification will
suppress mixing at the interface where κe will be relatively small. In this case strong mixing on either side of the density interface
can result in a flow with ∂2κe/∂z2
∗> 0, in which case term A will
act to increase N2
∗and sharpen the interface. FIGURE 7 | Cartoon illustrating overturning (A) and scouring (B) regimes at a
density interface (Caulfield, 2021). of particulate organic matter from the surface mixed layer
(Omand et al., 2015; Taylor et al., 2020). Frontiers in Marine Science | www.frontiersin.org 3.1. Unstratified Flow κe = κ
∂z∗
∂ρ
2
|∇ρ|2
z∗,
(7) (7) Flows past cylindrical structures are well studied, owing to their
geometrical simplicity and vast engineering importance; such Frontiers in Marine Science | www.frontiersin.org March 2022 | Volume 9 | Article 830927 10 Infrastructure Mixing in Shelf Seas Dorrell et al. FIGURE 8 | Typical flow past a cylinder. (A) Shows the finite-depth 3D flow past a cylinder over an erodible bed, depicting the shedding Karman vortices (KV). Note,
although not discussed herein, the run-up/depression at the water surface upstream/downstream of the cylinder and formation of horseshoe vortices cause scour at
the bed, which may destabilize structures (Matutano et al., 2013). (B) Shows the three shear layers comprising the cylinder flow: BL denotes the boundary layer and
SL denotes the shear layer. (C) Shows the typical Red −St relationship for flow past a cylinder. Shaded regions represent the 2D laminar vortex shedding regime
(49 ≲Red ≲192) and the trans-critical flow regime (2 × 105 ≲Red ≲6 × 106). Upper curve represents flow for a smooth cylinder, lower curve for a rough cylinder. Curves are based upon the experiments reviewed by Lienhard et al. (1966). The red line indicates the expected minimum Reynolds number of an OWT. FIGURE 8 | Typical flow past a cylinder. (A) Shows the finite-depth 3D flow past a cylinder over an erodible bed, depicting the shedding Karman vortices (KV). Note,
although not discussed herein, the run-up/depression at the water surface upstream/downstream of the cylinder and formation of horseshoe vortices cause scour at
the bed, which may destabilize structures (Matutano et al., 2013). (B) Shows the three shear layers comprising the cylinder flow: BL denotes the boundary layer and
SL denotes the shear layer. (C) Shows the typical Red −St relationship for flow past a cylinder. Shaded regions represent the 2D laminar vortex shedding regime
(49 ≲Red ≲192) and the trans-critical flow regime (2 × 105 ≲Red ≲6 × 106). Upper curve represents flow for a smooth cylinder, lower curve for a rough cylinder. Curves are based upon the experiments reviewed by Lienhard et al. (1966). The red line indicates the expected minimum Reynolds number of an OWT. The typical Red of shelf sea currents past offshore wind
foundations (Figure 2) is at least Red ≳105, estimated assuming
a small 5 m diameter monopile with a minimum tidal velocity
of 0.02 m s−1 (Vindenes et al., 2018). 3.1. Unstratified Flow This minimum Reynolds
number is sub-critical (see Figure 8), indicating that shear layer
instabilities may be present, manifesting as Kelvin-Helmholtz
(KH) type instabilities. Three dimensionality is a dominant
feature of cylinder wakes at these high Reynolds numbers. Critical transition occurs between 2 × 105 < Red < 5 × 105,
and is associated with boundary layer transition to turbulence
which causes the separation point to occur further downstream
on the cylinder surface. For a smooth cylinder, asymmetric
separation-reattachment of the boundary layer either side of the
cylinder causes a sudden increase in vortex shedding frequency
(Figure 8). Super-critical flow is associated with Red > 5 × 105,
where symmetric separation bubbles and turbulent boundary
layers are present on both sides of the cylinder. Roughness
effects are felt primarily in the critical transition regime; surface
roughness causes earlier transition to turbulence and bypasses the
asymmetric regime (Figure 8B). studies are directly analogous to flow past OWT foundations,
such as monopiles. Unsteady vortices shed by cylinders can lead
to vibration, acoustic noise, resonance, and ultimately structural
failure. Shed vortices form coherent wakes that are spatially vast,
and can be detected several hundred diameters downstream of
their source, depending on the background flow conditions (e.g.,
turbulence properties). Cylinder wakes fundamentally alter flow
conditions as a source of anthropogenic turbulence (mixing),
particularly coherent along the cylinder axis. Dynamics are
complex due to the interaction of at least three shear layers; the
boundary layer, shear layer, and wake (Figure 8). In unstratified
waters, flow structure is dependent on the cylinder roughness,
end conditions, freestream conditions, and the ratio of inertial
to viscous forces, the cylinder Reynolds number Red, Red = u∞d
ν
,
(8) (8) where u∞denotes the freestream velocity, d the cylinder
diameter and ν the fluid viscosity. The dependence of the flow
structure on Red arises due to transitions in the different shear
layers. The most well known instability arising from the flow
past a cylinder is the Karman vortex (KV) (Williamson, 1996),
which develops for Red ≳49 (below which the flow is steady and
laminar). The Karman instability is associated with alternating
2D vortices shed from either side of the cylinder, aligned with
the cylinder axis, and is a consistent feature of even high
Reynolds number flows. The unsteady KV is characterized by the
dimensionless frequency, the Strouhal number St = fKVd/u∞,
where fKV is the frequency of vortex shedding. Frontiers in Marine Science | www.frontiersin.org 3.1. Unstratified Flow St varies with Red
depending on shear layer transitions, as shown in Figure 8C. The
KV is a dominant feature of flow past a cylinder (Figure 9A). y
g
With distance x downstream, the far-wake x/d
≳
50
is particularly sensitive to freestream conditions. The spectral
energy
of
the
KV
(energy
associated
with
fKV)
decays
downstream of the cylinder and the wake width grows like √x
(under low levels of freestream turbulence) and is approximately
self-similar (Ghosal and Rogers, 1997). A secondary vortex
street emerges at a lower frequency than the KV, arising due
to the merging of vortex pairs, or via hydrodynamic instability
of the mean flow, depending on the flow Reynolds number
(Jiang, 2021a). Coherent structures in the far-wake can also arise
due to non-linear interactions between freestream structures March 2022 | Volume 9 | Article 830927 Frontiers in Marine Science | www.frontiersin.org 11 Infrastructure Mixing in Shelf Seas Dorrell et al. FIGURE 9 | Sediment plumes generated by cylindrical structures. (A) Wake and turbid plume showing the coherent Karman street from the lee of a metocean mast,
image courtesy D. van der Zande. (B) Satellite imaging of turbid plumes show they extend over multiple kilometers, individual monopile diameters are d = 4.3 m,
image courtesy R. Forster (Forster, 2018). With average monopile spacing of 1 km, interactions of the turbid plumes are clearly visible. FIGURE 9 | Sediment plumes generated by cylindrical structures. (A) Wake and turbid plume showing the coherent Karman street from the lee of a metocean mast,
image courtesy D. van der Zande. (B) Satellite imaging of turbid plumes show they extend over multiple kilometers, individual monopile diameters are d = 4.3 m,
image courtesy R. Forster (Forster, 2018). With average monopile spacing of 1 km, interactions of the turbid plumes are clearly visible. and the KV. Cimbala and Krein (1990) and Williamson and
Prasad (1993) found that the interaction between the KV and
freestream waves could lead to resonant peaks in spectral
energy associated to their non-linear interaction, indicative
of hydrodynamic instability. These peaks can be detected far
downstream of the cylinder, x/d
>
300, indicating that
non-linear dynamics have a large effect on the wake even
far downstream. It is therefore unsurprising to observe non-
linear interaction between wake effects from multiple monopiles
in offshore wind farms (Figure 9). 3.1. Unstratified Flow Under high freestream
turbulence the wake spreads more rapidly, depending on the
background turbulent intensity and the turbulent integral length
scale. Eames et al. (2011) demonstrated that the growth rate of the
wake increases from ∼√x to ∼x when the wake deficit velocity
is approximately equal to the background turbulent intensity,
and the integral length scale of the turbulence is comparable
to the cylinder diameter. When subject to high turbulence a
cylinder wake will dissipate downstream more rapidly and diffuse
into the background turbulence. In contrast, when background
turbulence levels are low the turbulence generated in the wake of
a cylinder can persist hundreds of diameters downstream. farms and the deep water future development sites. It is to
be expected that in deep water sites, with lower background
turbulence, that wakes may be even larger. Frontiers in Marine Science | www.frontiersin.org 3.2. Stratified Flow (2014) March 2022 | Volume 9 | Article 830927 12 Infrastructure Mixing in Shelf Seas Dorrell et al. FIGURE 10 | Layering in a linearly stratified fluid emerging from a turbulent cloud, with the resulting density profile across the intrusions (A), from Maffioli et al. (2014). A Dye visualization of the zig-zag instability is presented in (B), from the experiments of Billant and Chomaz (2000a). FIGURE 10 | Layering in a linearly stratified fluid emerging from a turbulent cloud, with the resulting density profile across the intrusions (A), from Maffioli et al. (2014). A Dye visualization of the zig-zag instability is presented in (B), from the experiments of Billant and Chomaz (2000a). intrusions developed from local regions of mixed fluid, which
grew until their length scales were approximately balanced by
l ∼˜u/N, where ˜u represents a characteristic velocity scale of the
turbulence. The buoyancy frequency may also be used to define
a Froude number, the ratio of inertial to gravitational forcing,
where L is some length scale over which flow velocity and N
are averaged, that such flows are prone to a class of ‘zig-zag’ instabilities. These
instabilities imprint the ˜u/N vertical scale on the flow, and can be
connected directly to the inherently nonlinear layered structures
that develop at finite amplitude (Lucas et al., 2017). Furthermore,
several studies (Deloncle et al., 2008; Waite and Smolarkiewicz,
2008; Augier et al., 2015) have demonstrated that the breakdown
of these vertical vortices in a stratified fluid introduce a new,
inherently stratified route to turbulence in a stratified fluid, and
hence substantially enhanced mixing. Fr =
˜u
NL. (9) (9) Whilst work on layering has been restricted to small scale
experiments and comparatively low Reynolds numbers (at least
by oceanographic standards), it has provided oceanographers
with vital information on the fine-scale density structure of
the ocean. However, it is unclear how turbulence generated at
high Reynolds number, by flows past offshore wind foundations,
will interact with the essentially two-layer density profile of
seasonally stratified shelf seas. In addition to the high Reynolds
number, offshore wind foundations are a similar width to
the typical thermocline thickness. 3.2. Stratified Flow Relatively few studies have investigated stratified flow interaction
with vertically oriented cylinders, analogous to proposed
offshore wind foundation deployment in seasonally-stratified
shelf seas. This may be in part due to similarity between
unstratified and stratified flow past vertical cylinders at very
low Reynolds, Red
≲45 (Meunier, 2012), where wakes are
inherently two-dimensional. However, differences arise when
three-dimensionality is present in the wake, which is certainly
characteristic of the high Reynolds number flows associated
with offshore wind farm infrastructure. Three-dimensionality
in the wake of vertical cylinders is important as it can lead
to fundamental reorganization of stratified fluids. Layering can
emerge where such a flow can lead to multiple intermittent
regions of fairly constant density neighboring thin interfaces with
steep density gradients (Bosco and Meunier, 2014). This layering
behavior may arise from any process that produces spatially
periodic mixing in the vertical direction (Thorpe, 2016). In current shallow-water offshore wind farms, where levels of
turbulence are high, wakes have been observed at least 1 km in
length (e.g., Figure 9). The formation of turbid plumes, where
suspended sediment is trapped and transported in KVs and
secondary vortex streets, is correlated with high levels of TKE
(Grashorn and Stanev, 2016). As the wakes pass through the wind
farm plumes are observed to spread and interact (Figure 9B). Wake spreading indicates that the effects of monopile wakes are
not limited to a short narrow region downstream. Although KV
are expected to decay for x/d ≈60 (Jiang and Cheng, 2019),
observed plume interaction at large lengths, x/d > 150, suggests
wake-wake interactions are important. However, conditions vary
considerably between the current shallow coastal offshore wind Cylinder wakes are prone to layering due to their increase in
both horizontal shear, normal to the direction of stratification,
and organized vertical vorticity, parallel to the direction of
stratification, through the formation of KV in particular. The
susceptibility of a stratified flow, perturbed by moving cylinders,
to layering is dependent on the mechanisms that generate layers,
which vary considerably depending on the type of forcing (e.g.,
oscillatory vs. continuous stirring, Thorpe, 2016). Experiments
have demonstrated that layers can develop from the ‘cloud’ of
stratified turbulence that results from dragging vertical cylinders
through a stratified flow, equivalent to flow through a wind
farm (Figure 10). In the experiments of Maffioli et al. Frontiers in Marine Science | www.frontiersin.org 3.2. Stratified Flow A good example case study of the impact of finite
depth obstacles on stratified flow, is flow past a horizontal
cylinder. At high Froude numbers dynamics are similar to
unstratified flow. But as the Froude number decreases several new
regimes occur (Boyer et al., 1989). Under stratification internal
waves can be generated not only by the structure but also the KV. This in turn can increase drag, and thus the amount of mixing,
by up to 100% (Arntsen, 1996). The drag coefficient of spheres
in stratified flows also varies by up to 100%, when the vortex
shedding frequency tends to the buoyancy frequency (Cocetta
et al., 2021). A variety of fixed and floating foundations may be
susceptible to such additional drag, where vortices shed are not
necessarily aligned with the direction of stratification (e.g., jackets
and semi-submersible foundations, Figure 2). natural topological effects. In addition to reduced stratification,
Floeter et al. (2017) also measured local upwelling at the edges
of the wind farms, similar to those observed near islands in
stratified waters (e.g., Simpson et al., 1982). Shallow water model
studies of the effects of offshore wind farms on local oceanic
circulation patterns have shown arrays of infrastructure induce
strong horizontal shear in wind stress leading to local regions
of upwelling and downwelling (Broström, 2008; Paskyabi and
Fer, 2012), consistent with the observations of Floeter et al. (2017). (
)
Field measurements were also taken by Schultze et al. (2020a)
who measured the stratified wake from an offshore monopile
at the leading edge, with respect to local flow, of the DanTysk
wind farm (Figure 12). The monopile wake spread to a width
10 times the (6 m) diameter of the monopile and had reduced
the potential energy anomaly by up to 65 % at a downstream
distance of x/d ≈20 (Figure 12). The full distance required
to return to pre-monopile conditions was not captured by the
survey, even after approximately 300 m, a distance over 50 times
the 6 m monopile diameter. The survey clearly demonstrates
that turbulence generated by monopiles reduces stratification. Complementing the survey, Schultze et al. (2020a) used Large
Eddy Simulations (LES) to model flow past a single monopile
which was simulated under different levels of background
stratification. TKE dissipation rate was found to be up to
two orders of magnitude larger in the thermocline than when
monopiles were not present. 3.2. Stratified Flow This contrasts with previous
work where turbulence was either generated by structures with
length scales two orders of magnitude smaller than the density
gradient length scale (Maffioli et al., 2014), or (as in the ‘zig-
zag’ instability studies mentioned above) the unstable horizontal
shear/vertical vorticity is embedded initially in a linearly stratified
fluid with close to constant buoyancy frequency. Where KV
shed from offshore wind foundations are of the same order
diameter as the density gradient length scale, flow mixing and
density profile reorganization may be fundamentally different,
and this generic stratified flow geometry with a range of key
characteristic length scales is very poorly understood, even in
highly idealized circumstances. Thus, in the experiments of Maffioli et al. (2014), packets of
turbulent fluid grew until they collapsed under gravity (Fr ∼1)
and spread outwards as pancakes, triggering horizontally
propagating internal waves. The outward spread of intrusions
generates a layered density profile, where well-mixed intrusions
are neighboured by thin regions of strong density gradient. It
is important to note here that the cylinders are not critical to
the reorganization of stratification; they only act as the source of
turbulence, via horizontal shear, in the stratified flow. Addition
of purely horizontal shear to simulations of stratified shear
flow can lead to coherent vertical vortices (Basak and Sarkar,
2006), analogous to the KV and KH instabilities of a cylinder
wake. The coherent vertical vortices exhibit pairing, tearing,
and amalgamation, resulting in a complex braided vorticity
structure that ultimately leads to vertical variability. Layers
can subsequently develop as intrusions, coupled with internal
waves (Basak and Sarkar, 2006), once again with a characteristic
vertical scale of the order of ˜u/N, when these two quantities are
estimated appropriately. Indeed, provided the flow Reynolds number is sufficiently
large, there is accumulating evidence, that horizontal shear (and
hence vertical vorticity) in vertically stratified fluids inevitably
forms layers on this scale. Dating back to the first theoretical
analysis of Billant and Chomaz (2000b), there is clear evidence Expansion of the offshore wind sector to deeper waters is
predicated on development of floating foundations, which are
finite in depth and anchored to the seabed. Finite depth structures March 2022 | Volume 9 | Article 830927 13 Infrastructure Mixing in Shelf Seas Dorrell et al. induce complex dynamics within stratified fluids, especially if
structures intersect sharp density gradients, i.e., the thermocline
(Figure 2). 3.2. Stratified Flow High TKE dissipation rate persisted
far downstream of the cylinder and, at x/d > 40, was still
an order of magnitude larger than without monopiles at the
thermocline depth. Further, the TKE dissipation rate was greater
than that generated by the bottom boundary layer. Through considering reflectional symmetry, flow past floating
foundations may be analogous to oceanic and atmospheric flow
past sea-mounts and hills. Stratified flow past such obstacles
displaces fluid and leads to both vertically and horizontally
propagating internal waves, as well as a significant downstream
wake (Figure 11). If an obstacle is wide compared to its height, most fluid
impinging on the obstacle passes over it, otherwise flow can
pass around it. For low Froude numbers, where L is scaled
with obstacle depth, flow blockage by internal waves creates
additional drag, which is a very effective momentum sink on
the impinging flow. Flow blockage can increase the drag force
by 1–2 orders of magnitude compared to unstratified flow over
the same obstacle (Smith, 1978; Castro et al., 1990; Cummins
et al., 1994). In all cases understanding drag increase is critical to
evaluate both the hydrodynamic loads placed on foundations and
anchors, and the mixing in stratified shelf seas, associated with
the flow past offshore wind foundations. Such wave generation
and propagation also suggests the possibility that the effect of
flow structures can be both local and non-local, as the emitted
waves may transport momentum flux significant distances until
they ‘break’. During a period of stronger stratification, Schultze et al. (2020a) conducted a second survey at the opposite end of the
DanTysk wind farm. The second survey was less conclusive than
their previous measurements, finding that no clear signal from
the wake could be separated from background variability. This
may be because stratification was strong enough to suppress the
growth and interactions of the wakes, although (Schultze et al.,
2020a) only sampled the wake at distances greater than 200 m (or
33.3 diameters) from the monopile. It is therefore unclear if this
is a result consistent with the earlier survey in Figure 12. Frontiers in Marine Science | www.frontiersin.org 3.3. Mixing of Stratified Shelf Seas by
Offshore Wind Foundations Internal waves can be observed in the lee of the structure propagating
vertically and horizontally, clear in both (A) the horizontal buoyancy gradient bx = −∂ρ/∂x and (B) vertical buoyancy gradient bz = −∂ρ/∂z. The wake is also
particularly apparent in the bz field, demonstrating the existence of significant mixing. N is the background buoyancy frequency and R denotes the height of the hill. FIGURE 11 | Stratified flow over a hill, taken from the experiments of Dalziel et al. (2011). Internal waves can be observed in the lee of the structure propagating
vertically and horizontally, clear in both (A) the horizontal buoyancy gradient bx = −∂ρ/∂x and (B) vertical buoyancy gradient bz = −∂ρ/∂z. The wake is also
particularly apparent in the bz field, demonstrating the existence of significant mixing. N is the background buoyancy frequency and R denotes the height of the hill. Here the potential production of TKE PD per unit volume,
generated by the addition of a 10 m diameter monopile,
is estimated using the natural conditions, averaged over the
thermocline thickness, at the CaNDYFLOSS experiment site
(Figure 3). Power generated by flow past such a potential
foundation is strongly dependent on the flow velocity, with low
frequency oscillations associated with neap-spring tide cycles and
higher frequencies associated to daily tidal cycles (Figure 13). The potential power added to the thermocline, PD, varies
between approximately 2 × 10−6 W m−3 to 4 × 10−4 W m−3
with an average value of 5.9 × 10−5 W m−3. Over the same
period, at the same site, the natural background dissipation
rate (εB) at the thermocline is 3.5 × 10−5 W m−3 (Scannell
et al., 2021). In this approximate calculation, the additional
(average) turbulence production is therefore 69% greater
than background dissipation without monopiles present: PD
≈1.69 εB. background shear. Crucially cD is also dependent on the Froude
number, especially for highly stratified flows (see Section 3.2). At present there is no parametrisation of the Froude number
dependence of the drag coefficient for vertical infrastructure in
stratified flows. We therefore assume cD = 1 as per the “high
drag” case of Carpenter et al. (2016), although this estimate does
not account for the potential effects of the Froude number. A wind farm comprises many turbines, potentially supported
by a range of foundations. 3.3. Mixing of Stratified Shelf Seas by
Offshore Wind Foundations Leading order arguments for the mixing induced by offshore
wind foundations have been explored by Rennau et al. (2012) and
Carpenter et al. (2016), here they are reviewed. Models start by
assuming that the turbulence produced by foundations is equal
to the power lost to drag. Over an arbitrary vertical layer of fluid,
L, the power lost to drag is given by To date there has been only two limited studies observing
offshore wind foundation induced mixing of stratified waters
(Floeter et al., 2017; Schultze et al., 2020a). These studies
have been restricted to developments in aperiodically stratified
regions of freshwater influence, in shallow water (depths of
approximately 40 and 24 m, respectively). Neither studies have
investigated the potential impacts on seasonally, or permanently,
stratified shelf seas. PD = 1
2ρcDdL⟨|u|3⟩L,
(10) (10) Floeter et al. (2017) performed surveys on two wind farms in
the German Bight, North Sea. Water property transects through
the wind farm revealed a consistent weakening of stratification
near the centre. Effects extended into the surrounding area by
approximately half the diameter of an ambient tidal excursion. However, it was unclear how much of this was due to
‘infrastructure’ turbulence from turbine foundations rather than where cD denotes a drag coefficient and ⟨|u|3⟩L the velocity
magnitude cubed, averaged over the layer L. Whilst, Rennau et al. (2012) and Carpenter et al. (2016) assumed that cD is a constant
in reality it is variable and dependent on the Reynolds number,
surface roughness and structure geometry. The drag coefficient
is also likely a function of depth and time and will vary with March 2022 | Volume 9 | Article 830927 14 Infrastructure Mixing in Shelf Seas Dorrell et al. FIGURE 11 | Stratified flow over a hill, taken from the experiments of Dalziel et al. (2011). Internal waves can be observed in the lee of the structure propagating
vertically and horizontally, clear in both (A) the horizontal buoyancy gradient bx = −∂ρ/∂x and (B) vertical buoyancy gradient bz = −∂ρ/∂z. The wake is also
particularly apparent in the bz field, demonstrating the existence of significant mixing. N is the background buoyancy frequency and R denotes the height of the hill. FIGURE 11 | Stratified flow over a hill, taken from the experiments of Dalziel et al. (2011). 3.3. Mixing of Stratified Shelf Seas by
Offshore Wind Foundations The approximate flow direction
is indicated by the blue arrow, and neighboring monopiles are marked by black circles. The wake is clearly visible by the blue regions, and approximately enclosed by
the black lines. (B) Shows the spatial development of the temperature field in the flow before and the wake after the monopile. Evolution of the temperature field is
obtained by averaging data in the regions enclosed by the black lines in (A). Data are from Schultze et al. (2020b). buoyancy field may be expressed, from Equation (3), by on the dissipation rate. Secondly, setting Ŵ ≃0.2 may well
be an under-estimate of the vigorous overturning mixing likely
to be triggered in the wake of such infrastructure (Caulfield,
2021). Nevertheless, these simple conservative estimates show
that foundations produce turbulence at levels that will clearly
affect the leading order balance of TKE transport, even when
normalized by the total area between offshore wind foundations. Thus, offshore wind has the potential to impact directly the
stability of seasonally-stratified shelf seas (Figure 5). P = PD + PB = (1 + Ŵ)εD + (1 + Ŵ)εB,
(12) (12) where subscript D represents contributions from structures
(drag) and subscript B represents all other (background)
contributions to TKE production (P) and dissipation rate (ε). For simplicity, the turbulent flux coefficient can be assumed
equal for each component of B and ε such that Ŵ = B/ε =
BB/εB = BD/εD. In deriving (12) it is also assumed that OWT
infrastructure contributes linearly to P, and is simply related to
the structure-induced TKE dissipation rate by PD = (1+Ŵ)εD. In
the present example we have shown that PD ≈1.69εB, such that The
dissipation
rate
estimate
indicates
that
OWT
infrastructure is likely to affect the leading order balance of
TKE transport in vicinity of the windfarm, but the implications
of this on regional-scale fluid dynamics is more challenging to
evaluate. To address this, Carpenter et al. (2016) derived similar
models for TKE production as Equation (11), and constructed
arguments based on estimates of the different timescales in the
flow. In particular, timescales of OWT-induced mixing, and
timescales of flow parcels convecting through the windfarm. By comparing different estimates of these timescales Carpenter
et al. (2016) theorised the extent that the water column would
mix as a parcel passed through the array. 3.3. Mixing of Stratified Shelf Seas by
Offshore Wind Foundations Turbines are separated by a distance of
approximately 10D (Howland et al., 2019), where D is the rotor
diameter. The average production of TKE per unit volume PD
over a layer of thickness L is therefore: PD = 1
2
ρ0cDd⟨|u|3⟩L
100D2
. (11) (11) Here it is assumed that the power lost to drag at a turbine
foundation acts to increase TKE production over the full area
occupied by a foundation. In reality this production will be
localized to the comparatively narrow wake, and will vary by
several orders of magnitude in space. Under several very strong assumptions, the relation between
TKE produced by offshore wind foundations and how it is
balanced by viscous dissipation and work performed on the March 2022 | Volume 9 | Article 830927 Frontiers in Marine Science | www.frontiersin.org 15 Infrastructure Mixing in Shelf Seas Dorrell et al. FIGURE 12 | Observations of the turbulent stratified wake downstream of a monopile, derived from the survey conducted by Schultze et al. (2020a). The monopile of
diameter 6 m, is located at 55.16◦N 7.07◦E on the leading edge, relative to local flow, of the DanTysk wind farm. The wind farm is marked by the diamond in the inset
figure of (A). (A) Depicts the potential energy anomaly, φ around the monopile. φ is normalized by its background reference value, φref. The approximate flow direction
is indicated by the blue arrow, and neighboring monopiles are marked by black circles. The wake is clearly visible by the blue regions, and approximately enclosed by
the black lines. (B) Shows the spatial development of the temperature field in the flow before and the wake after the monopile. Evolution of the temperature field is
obtained by averaging data in the regions enclosed by the black lines in (A). Data are from Schultze et al. (2020b). FIGURE 12 | Observations of the turbulent stratified wake downstream of a monopile, derived from the survey conducted by Schultze et al. (2020a). The monopile of
diameter 6 m, is located at 55.16◦N 7.07◦E on the leading edge, relative to local flow, of the DanTysk wind farm. The wind farm is marked by the diamond in the inset
figure of (A). (A) Depicts the potential energy anomaly, φ around the monopile. φ is normalized by its background reference value, φref. Frontiers in Marine Science | www.frontiersin.org 3.3. Mixing of Stratified Shelf Seas by
Offshore Wind Foundations It was concluded that
small developments in shallow waters (Bard 1 and Global Tech 1,
German Bight) were unlikely to affect stratification, where TKE
production from the bed and free surface dominate. However,
despite the simplification of the models and assumptions
made, large scale developments in deep water were recognized
to have potential for significant impact (Carpenter et al.,
2016). (13) PD = (1 + Ŵ)εD ≈1.69εB. (13) Under the simplest conventional assumption that Ŵ
= 0.2,
as typical at the pycnocline in seasonally stratified waters (see
Section 2), we obtain εD
≈1.4εB, such that the total TKE
dissipation rate, ε
=
εB + εD, is at least 140 % higher at
the thermocline when OWT infrastructure is present. As noted,
several sweeping assumptions have been made to arrive at
this estimate, which inevitably has a large amount of implicit
uncertainty. It is reasonable to suppose that this estimate
is likely to be a conservative lower bound on the effect of
OWT infrastructure on turbulent dissipation and mixing at the
thermocline for at least two reasons. The first is that there is clear
potential for a strongly nonlinear effect of such infrastructure Under the simplest conventional assumption that Ŵ
= 0.2,
as typical at the pycnocline in seasonally stratified waters (see
Section 2), we obtain εD
≈1.4εB, such that the total TKE
dissipation rate, ε
=
εB + εD, is at least 140 % higher at
the thermocline when OWT infrastructure is present. As noted,
several sweeping assumptions have been made to arrive at
this estimate, which inevitably has a large amount of implicit
uncertainty. It is reasonable to suppose that this estimate
is likely to be a conservative lower bound on the effect of
OWT infrastructure on turbulent dissipation and mixing at the
thermocline for at least two reasons. The first is that there is clear
potential for a strongly nonlinear effect of such infrastructure March 2022 | Volume 9 | Article 830927 16 Infrastructure Mixing in Shelf Seas Dorrell et al. FIGURE 13 | Mean TKE production by a d = 10 m offshore monopile. TKE production is based on direct velocity measurements averaged over the thermocline,
20–50 m in the Celtic Sea between 22/06/2014 and 20/08/2014, Figure 3 (Scannell et al., 2021). Turbulence production (Equation 11) is closed by assuming cD = 1,
ρ0 = 1.000 kgm−3 and D = 200 m. 3.3. Mixing of Stratified Shelf Seas by
Offshore Wind Foundations The power averaged over the full time period is marked in red. FIGURE 13 | Mean TKE production by a d = 10 m offshore monopile. TKE production is based on direct velocity measurements averaged over the thermocline,
20–50 m in the Celtic Sea between 22/06/2014 and 20/08/2014, Figure 3 (Scannell et al., 2021). Turbulence production (Equation 11) is closed by assuming cD = 1,
ρ0 = 1.000 kgm−3 and D = 200 m. The power averaged over the full time period is marked in red. Rennau et al. (2012) adopted regional-scale numerical models
to assess the effects of offshore wind development in the Baltic
Sea, where tidal currents are minimal and stratification are
driven by dense saline currents beneath fresher water. They
modified a two-equation (RANS) turbulence closure scheme to
capture the enhanced mixing arising from OWT installations
and found that current installations did not have a significant
effect on regional stratification (though cautioned that future
development, covering more of the Western Baltic Sea, could lead
to significant impacts, such as reduced bottom salinity). However,
these closure schemes are known to be overly dissipative
in stratified flows (Hewitt et al., 2005), and demonstrated
to generate overly diffuse pycnocline structures in realistic
simulations (Luneva et al., 2019). These schemes are therefore
improperly conditioned, at present, to address the impact of
OWT infrastructure in stratified seas. vessels have comparatively small drafts, ∼10 m (Golbraikh and
Beegle-Krause, 2020; Nylund et al., 2020). Further, whilst offshore
oil and gas platforms are similar to offshore wind infrastructure,
spanning the thermocline, only ∼6, 500 platforms have been
installed, globally, over the last 75 years (Schneider and Senders,
2010). Therefore, planned offshore wind developments will
add new and large scale infrastructure sources of turbulent
mixing in seasonally stratified seas. This discussion reviews
our current knowledge gaps, frames potential impacts on
shelf sea dynamics and thus marine ecosystem functioning
and highlights routes for sustainable growth of the offshore
wind sector. 4.1. Infrastructure Mixing g
The mixing of stratified waters by offshore wind infrastructure is
poorly understood. As evidenced in Section 3.3 there is a dearth
of research on high Reynolds number stratified flow past vertical
structures, which is vital for understanding and parameterizing
‘infrastructure’ mixing processes in natural environments. The
problem is particularly complex due to its scale; laboratory
experiments and fully resolved numerical simulations are limited
to relatively low Reynolds number flows, certainly by comparison
to the real oceanographic flows. In all scenarios of fixed or
floating OWTs, fine scale vertical density structure and enhanced
mixing is anticipated due to horizontal shear generated by the
flow past the infrastructure. The O
10−2 −100
m length scales
associated with the resulting turbulence will be unresolved by
low resolution numerical simulations. Wakes from individual
structures may persist for O
102 −103
m downstream. From
individual wind turbines to a single offshore wind farm,
infrastructure adds a wide range of length scales of O
101 −104
m. Further, multiple wind farms are distributed at shelf-wide
scales. The vast range of length scales present in flow past offshore
wind farms necessitates a variety of modeling techniques. Cazenave et al. (2016) adopted a similar oceanographic model
to Rennau et al. (2012) and attempted to explicitly resolve the
structure-induced wakes shed from wind turbines using local
grid refinement. They modeled the effects of windfarms in the
essentially well mixed (unstratified) Irish Sea, where dissipation
rates in the water column are already high (Simpson et al., 1996). However, they did not modify the closure scheme to account
for wake generated turbulence. Whilst visualizations of turbine
wakes by Cazenave et al. (2016) are in qualitative agreement with
sediment plumes observations (Figure 9), accurate prediction
of wake associated mixing in stratified waters still require
advances in fundamental physical process understanding to
enable development of appropriate turbulence closure schemes
(Rennau et al., 2012). Frontiers in Marine Science | www.frontiersin.org 4. DISCUSSION The scale of planned offshore wind energy industry is much
greater than past, and existing, sea use. Installed offshore
wind capacity will increase by 600% in the next decade
(Figure 1A), requiring an extra ∼20, 000 10 MW+ turbines. Fixed, and floating, offshore wind infrastructure will penetrate
the thermocline, adding ‘anthropogenic’ mixing on top of natural
mixing. The impact of such infrastructure will be fundamentally
different from existing sea use. For example, even large surface Coarse numerics, i.e., Large Eddy Simulation (LES) or other
more sweeping turbulence closure schemes, may be desirable
modeling strategies for high Reynolds number flow (Rennau
et al., 2012; Carpenter et al., 2016). Nested or adaptive meshing, in
particular around known boundaries such as OWT foundations
(see, e.g., Rennau et al., 2012; Cazenave et al., 2016), can be used
to further reduce numerical cost or improve simulation accuracy. March 2022 | Volume 9 | Article 830927 Frontiers in Marine Science | www.frontiersin.org 17 Infrastructure Mixing in Shelf Seas Dorrell et al. However, without a firm understanding of the underlying
physics, e.g., from fully-resolved simulations, or robust datasets
for validation, low resolution numerical models should be
adopted with care, particularly for stratified (and therefore
highly anisotropic) flows (Hewitt et al., 2005; Khani and Waite,
2015). Integration of modeling disciplines ranging from Direct
Numerical Simulation (DNS) of idealized flows to regional
scale, operational forecast, ocean circulation models is therefore
required to determine the impact of offshore wind farms, which
scale to natural features and processes inherent to seasonally
stratified shelf seas (Figures 1, 4, 6). Field surveys are essential
to support and validate physical and numerical studies. However,
the limited work to date (Floeter et al., 2017; Schultze et al.,
2020a) has suffered from uncertainties, where either effects of
infrastructure were difficult to discern from topographical effects
or where wakes were difficult to separate from background
variability. While both studies concluded effects of infrastructure
on mixing may be large, further surveys are required to support
this. In addition there is a clear need for repeat and ‘before and
after’ surveys, as noted by van Berkel et al. (2020). cycle (Section 3.3). TKE production, and thus mixing, arising
from stratified flow past infrastructure is also dependent on the
drag coefficient cD, yet little is known about the dependence
of cD, particularly at high Reynolds numbers or where mixing
length scales are large in comparison to density length scales. 4.2. Shelf Sea Dynamics y
Offshore wind farms are anticipated to have large local
impacts on shelf sea dynamics, in a similar fashion to natural
topographically controlled mixing, e.g., driven by internal waves
and flow over seafloor sand banks. It is anticipated that a broad
and more diffuse thermocline would develop as a result of
enhanced mixing, weakening it as a barrier to vertical mixing
and transport (Figure 14). Subsequent change to surface water
characteristics would likely alter exchange across the ocean-
atmosphere interface, with impacts on heat storage, atmospheric
CO2 uptake and benthic resupply of O2. The scale of this
response will be site and infrastructure specific. At regional scales
the water column should re-stratify subject to natural buoyancy
forcing. In a more extreme scenario, strongly enhanced mixing
could prevent stratification from forming around wind farms. This effect is observed around islands, where reduced surface
temperatures and well-mixed waters result from enhanced
mixing due to flow acceleration and seabed shoaling (Simpson
et al., 1982). Small, but significant residual currents sweep this
well-mixed water into an observable downstream wake. There are also key research questions regarding the geometry
of offshore wind infrastructure, and how this impacts mixing. OWTs shed vortices with horizontal scales that are comparable
to the thickness of the thermocline. Such mixing contrasts with
previous work studying mixing by thin vertical cylinders, several
orders of magnitude smaller in diameter compared to density
length scales. It is vital to understand how large scale vortex
structures interact with a relatively thin thermocline; they could
produce Langmuir-like interactions with the thermocline (Polton
et al., 2008), or mixing processes may well be fundamentally
different to those studied before, with complex spatio-temporally
variable vortex, turbulence and mixing dynamics. In addition,
future floating technology raises further questions regarding
the impact of geometry on mixing. It is expected that spar-
buoy designs (Figure 2) will act in a similar way to monopiles,
given they penetrate through the thermocline and into the
well-mixed deep water. However, semi-submersible (or any
other small-draft designs) will introduce non-trivial effects by
intersecting the thermocline. Research of stratified flow over
finite topography, similar to small draft floating OWT, has
demonstrated that baroclinic effects can enhance drag by up to
two orders of magnitude (Section 3.2). Here, floating structures
will introduce infrastructure mixing via shed lee waves, internal
waves, blockage effects, and wake-wake interactions in the case of
semi-submersible designs. Frontiers in Marine Science | www.frontiersin.org 4. DISCUSSION Meanwhile, research on drag past other obstacle forms suggest
estimates for drag may be incorrect by orders of magnitude. Advancing our understanding of each of these processes is
vital for assessing their impact on shelf sea oceanography and
ecosystem functioning. 4.3. Shelf-Sea Ecosystem Functioning Enhanced mixing rates due to infrastructure, would not only lead
to temporal and spatial variations in vertical density structure,
but also impact biogeochemical function at a fundamental level
that would cascade up the ecosystem. The changes in the timing of stratification expected from
increases in mixing also need to be considered. Late development
of stratification and the spring bloom has the potential to impact
wildlife, which has evolved to take advantage of this abundance,
with seabird breeding populations and fish stocks a notable risk. At the other end of the stratified period, additional mixing will
lead to earlier re-mixing of the whole water column, shortening
the total productive time of the area and also altering the timing
of the autumn bloom. Similarly, some seabird colonies are located
to take advantage of the enhanced productivity at tidal mixing
fronts (Trevail et al., 2019), the location of which may be affected
by additional mixing from offshore wind farm infrastructure. Stronger mixing in the thermocline will drive more nutrients
from the bottom water up into the subsurface chlorophyll
maximum and, if the mixing is strong enough, up into the surface
layer where it could support additional phytoplankton growth. The rate of turbulent mixing strongly affects the simulation of
ecosystem behavior, as demonstrated by Luneva et al. (2019), who
found that different mixing schemes caused a shift in the spring
bloom by 1 month, and change regional chlorophyll differences
by order 100%. Mixing also alters the light experienced by the phytoplankton,
with stronger mixing potentially disrupting light sufficiently to
hinder photosynthesis. The net effect, i.e., whether or not the
extra mixing aids net phytoplankton production, will depend
on some balance between the nutrient and light effects. The
summer reduction in bottom water oxygen concentrations will
also respond to the increased mixing. Oxygen will be supplied
from the surface water downward, potentially offsetting some
of the normal bacterial demand for oxygen as they recycle the
sinking organic detritus from phytoplankton growth. At the same
time, however, there may be an increased supply of organic
detrital material if the net effect of mixing on primary production
was positive, in which case the bacterial demand for oxygen in
the bottom water will increase. 4.3. Shelf-Sea Ecosystem Functioning Understanding the net effect of
mixing on bottom water is important, as there are large areas
of stratified shelf seas that are currently viewed as being close
to experiencing oxygen depletion in late summer (Ciavatta et al.,
2016). 4.2. Shelf Sea Dynamics Enhanced mixing from infrastructure may also impact
seasonal, and shorter timescale, cycles. The first order response
of the vertical density structure to enhanced mixing in a wind
farm region would likely be delayed onset and early breakdown
of seasonal stratification with weaker stratification throughout
the summer season. Development of near surface stratification
during periods of low wind stress would no longer be expected
to occur as the enhanced mixing would act to persistently stir the
normally episodically mixed surface layer. With infrastructure development from local scale of a
single turbine to regional scales of multiple developments
offshore wind farms in shelf seas have the potential to have
regional impact. Horizontal variation in density may arise from
wind farm scale mixing, and these variations may enhance
submesoscale processes driving additional vertical transport
and mixing. Advanced regional oceanographic models, which
integrate physics based closures of offshore wind farm driven
mixing, will be required to understand impacts of such large scale
processes. The density of offshore wind farms and the regional
distribution of mixing and wake-wake interactions between wind
farms will be of critical importance in determining shelf sea
response to offshore wind development. Crucially, a lack of insight into key multi-scale mixing
processes adds uncertainty to current attempts to quantify
the impact of infrastructure mixing on shelf sea dynamics
and ecosystem functioning. For example, the turbulent flux
coefficient is often assumed constant, Ŵ
= 0.2 (Section 2.3
and 3.3), but it is unknown if this holds in the wake of OWTs
where fine scale density structures and strong spatial variability
are present. Further, production of TKE due to infrastructure
has been assumed constant in time and evenly distributed over
the area ‘occupied’ by the monopile. In reality production is
focused in the narrow wake of individual monopiles, and could
vary by two orders of magnitude during the spring-neap tidal As
has
been
discussed
herein,
deployment
of
OWT
foundations will lead to significant enhancement of natural sub
surface shear-production of turbulence, mixing the thermocline. However, OWT also generate significant atmospheric wakes, March 2022 | Volume 9 | Article 830927 18 Infrastructure Mixing in Shelf Seas Dorrell et al. which can extend downstream over multiple array scales, and
upstream wind blocking effects (Howland et al., 2019; Nygaard
et al., 2020). Frontiers in Marine Science | www.frontiersin.org 4.2. Shelf Sea Dynamics Changes in atmospheric wind speed or location of
stably stratified atmospheric boundary layers, due to atmospheric
wakes and wind blocking, will impact air-sea exchange of heat,
momentum and trace gases. Sea surface wind shear mediates
the rates of air-sea exchange of momentum, heat and trace
gases, e.g., CO2 (Komori et al., 1993; Lincoln et al., 2016). Wind
speeds are further directly coupled to surface wave dynamics. Change in wind speeds, decreasing turbulence production by
wave breaking, would result in changes to wave-driven Langmuir
turbulence which may extend across the near surface well
mixed layer (see, e.g., Lucas et al., 2019). Indeed the interaction
between the Stokes shear and the vertical vortices shed from
the infrastructure might energize the Langmuir turbulence to
establish a modified equilibrium mixed layer depth (Pearson
et al., 2015). In consequence the complex interplay of oceanic
surface mixed layer and atmospheric dynamics arising from
development of OWT in stratified shelf-seas will be a key area
for future research. column inventory of nutrients used for that year will have
increased. The shelf is not restocked with fresh nutrients from the
open ocean every year (Ruiz-Castillo et al., 2019), so the ability of
the shelf system to maintain the increased production will depend
on how efficient the system is at recycling organic nutrients back
to inorganic nutrients particularly over the winter. A natural analog to the effects of renewable energy
infrastructure in a stratified environment may be found in
the central Celtic Sea, where a number of seabed banks
interact with tidal flows to inject significant internal wave-
driven mixing into the thermocline (Palmer et al., 2013). While the exact biogeochemical and biological reasons are
not yet clear, it does appear that the bank-driven mixing
increases the overall biological activity of the region, ultimately
resulting in commercial fishing (see, e.g., Figures 1C, 4
Sharples et al., 2013). The island mixing effect too (Simpson
et al., 1982), produces significantly enhanced nutrient fluxes,
with a corresponding increase in plankton production and
fishing activity. Thus, combined with reduced fishing pressure,
appropriate infrastructure in suitable locations could prove a
positive impact on regional ecology, with benefits to wildlife
and fisheries. 4.4. Management and Mitigation
Requirements Environmental Impact Assessments (EIA) are required for all
new offshore wind farms to mitigate negative impacts resulting
from construction, operation and decommissioning across the
lifecycle of use. Surveying also covers geophysical site assessment,
focus on seabed composition, current assessments and the
potential for seabed scour around anchors and cables. Current
EIAs have been developed for mixed coastal environments to
ensure sustainable growth of the offshore wind sector. Beyond
the construction phase, impact has been focused on the benthic
habitat and individual species interactions with infrastructure. Impact on marine mammals, seabird colonies, and fisheries in
the region are individually assessed, but without consideration
of potential alteration of the primary production through
enhanced mixing. To ensure the continued growth of the sector (The Crown
Estate, 2020), impacts of the new generation of developments
in deeper seasonally stratified regimes will likely require a
more fundamental assessment. Baseline surveys must include
the natural cycle of water column stratification, biogeochemical Changes to biogeochemical functioning would need to be
assessed over several years. For instance, an immediate positive
impact on net phytoplankton production because of the extra
nutrient supply will mean that the total amount of the water March 2022 | Volume 9 | Article 830927 19 Infrastructure Mixing in Shelf Seas Dorrell et al. FIGURE 14 | Offshore wind infrastructure adds wake turbulence throughout the upper water column, including directly at the thermocline. Here swirl size indicates
turbulence intensity arising from near bed, near surface and flow-structure induced shear. Wake turbulence mixes cold nutrient rich bottom water with warm nutrient
poor surface water, reducing the strength of stratification and potentially enhancing plankton growth in the subsurface chlorophyll layer. Changes in the subsurface
chlorophyll layer would have further impacts on nutrient pathways, ecosystem functioning and oceanic carbon sequestration. FIGURE 14 | Offshore wind infrastructure adds wake turbulence throughout the upper water column, including directly at the thermocline. Here swirl size indicates
turbulence intensity arising from near bed, near surface and flow-structure induced shear. Wake turbulence mixes cold nutrient rich bottom water with warm nutrient
poor surface water, reducing the strength of stratification and potentially enhancing plankton growth in the subsurface chlorophyll layer. Changes in the subsurface
chlorophyll layer would have further impacts on nutrient pathways, ecosystem functioning and oceanic carbon sequestration. fluxes, and primary production. Accurately forecasting the
interaction between the flow, infrastructure and stratification
will require site, array and design specific observations and
model scenarios. 4.4. Management and Mitigation
Requirements Only with a comprehensive understanding of
this physical modification to the biogeochemical functioning
of shelf seas, can impact throughout the marine web be
adequately assessed. dissipates at least 140% more energy than exists naturally
at
the
thermocline
(Section
3.3). However,
in
stratified
waters, the additional baroclinic (wave) drag can exceed
the barotropic drag. For example, the baroclinic drag for a
stratified flow past topography can be 1–2 magnitudes larger
than the barotropic drag (Smith, 1978). Baroclinic drag from
vertical infrastructure in stratified flow, such as OWT, is
as yet unquantified. Moreover, the role of horizontal shear
on vertical mixing, here produced by obstacles that scale
with density gradient length scales, is poorly understood,
particularly in complex stratifications like the thermocline. It is of great importance to extend our understanding of the
fundamental fluid dynamics of flows past vertical structures
in ocean-realistic stratifications, in particular, the onset of
turbulence and ensuing mixing associated with the breakdown
of induced vortices. Frontiers in Marine Science | www.frontiersin.org REFERENCES Butterfield, S., Musial, W., Jonkman, J., and Sclavounos, P. (2007). Engineering
Challenges for Floating Offshore Wind Turbines. Technical report, National
Renewable Energy Lab. (NREL), Golden, CO. 4C Offshore (2021). Global Offshore Renewable Map. Available online at: https://
www.4coffshore.com/offshorewind (accessed September 30, 2021). gy
Carpenter, J. R., Merckelbach, L., Callies, U., Clark, S., Gaslikova, L., and Baschek, Carpenter, J. R., Merckelbach, L., Callies, U., Clark, S., Gaslikova, L., and Baschek,
B. (2016). Potential impacts of offshore wind farms on north sea stratification. p
, J
,
,
,
,
,
,
,
,
,
,
B. (2016). Potential impacts of offshore wind farms on north sea stratification. Arntsen, Ø. A. (1996). Disturbances, lift and drag forces due to the translation
of a horizontal circular cylinder in stratified water. Exp. Fluids 21, 387–400. doi: 10.1007/BF00189060 B. (2016). Potential impacts of offshore wind farms on north PLoS ONE 11, e0160830. doi: 10.1371/journal.pone.0160830 j
p
Castro, I., Snyder, W., and Baines, P. (1990). Obstacle drag in stratified flow. Proc Soc. Lond. A Math. Phys. Sci. 429, 119–140. doi: 10.1098/rspa.199 Augier, P., Billant, P., and Chomaz, J.-M. (2015). Stratified turbulence forced
with columnar dipoles: numerical study. J. Fluid Mech. 769, 403–443. doi: 10.1017/jfm.2015.76 R. Soc. Lond. A Math. Phys. Sci. 429, 119–140. doi: 10.1098/rspa.1990.0054
Castro-Santos, L., Silva, D., Bento, A. R., Salvacao, N., and Soares, C. G. (2020). y
,
p
Castro-Santos, L., Silva, D., Bento, A. R., Salvacao, N., and Soares, C. G. (2020). Economic feasibility of floating offshore wind farms in portugal. Ocean Eng. 207, 107393. doi: 10.1016/j.oceaneng.2020.107393 Austin, J. A., and Barth, J. A. (2002). Variation in the position of the
upwelling front on the oregon shelf. J. Geophys. Res. Oceans 107, 1–1. doi: 10.1029/2001JC000858 Caulfield,
C. (2021). Layering,
instabilities,
and
mixing
in
turbulent
stratified
flows. Annu. Rev. Fluid
Mech. 53,
113–145. doi: 10.1146/annurev-fluid-042320-100458 Basak, S., and Sarkar, S. (2006). Dynamics of a stratified shear layer with horizontal
shear. J. Fluid Mech. 568, 19–54. doi: 10.1017/S0022112006001686 Cazenave, P. W., Torres, R., and Allen, J. I. (2016). Unstructured grid modelling of
offshore wind farm impacts on seasonally stratified shelf seas. Prog Oceanogr. 145, 25–41. doi: 10.1016/j.pocean.2016.04.004 Bauer, J. E., Cai, W.-J., Raymond, P. A., Bianchi, T. S., Hopkinson, C. S., and
Regnier, P. A. (2013). The changing carbon cycle of the coastal ocean. Nature
504, 61–70. doi: 10.1038/nature12857 Ciavatta, S., Kay, S., Saux-Picart, S., Butenschön, M., and Allen, J. (2016). 5. CONCLUSIONS Previous work has considered the environmental impact
that offshore wind energy has on well-mixed shallow-water
marine ecosystems, including from benthic habits, fisheries
to seabirds. Much of this work remains relevant to enable
sector growth. However, sector expansion from well-mixed
shallow water to seasonally stratified deeper water represents a
fundamental change, where physical and environmental impacts
are not understood. Regionally, the first order paradigm for seasonally stratified
shelf seas is the balance between the stratifying influence
of surface heating, and the input of mechanical energy to
mix the water column at the upper and lower boundaries
(due
to
wind
stress
and
the
tidal
shear,
respectively). High resolution shelf sea models have some success in
reproducing this (Luneva et al., 2019). However, the addition
of mixing from large scale offshore wind farm development,
limits our ability to understand the trajectory of shelf sea
ecosystems. To address this, research is urgently needed
that scales processes from: a single turbine; an array of
turbines composing a wind farm; to an entire shelf sea
region with multiple farms. Advances in regional ecosystem For the first time planned developments of both fixed
and floating offshore wind infrastructure will add large
scale anthropogenic mixing to seasonally stratified shelf seas. Large scale mixing may force shelf sea physics, establishing
a ‘new normal’ for biogeochemical cycles and shelf sea
ecosystem functioning. The potential benefits and risks posed
by infrastructure mixing of stratified shelf seas, on top of
climate change, represents a combined hazard that has not
been considered. Locally, flow past offshore wind energy infrastructure
results in (barotropic) drag and turbulence that, by itself, March 2022 | Volume 9 | Article 830927 20 Infrastructure Mixing in Shelf Seas Dorrell et al. Oceanography of Stratification in Shelf Seas, Flow Structure
Interactions, Discussion, and Conclusions. C-cC contributed
to
sections
Flow
Structure
Interactions,
Discussion,
and
Conclusions. JS contributed to sections Oceanography of
Stratification in Shelf Seas and Discussion. BS contributed
to
section
Oceanography
of
Stratification
in
Shelf
Seas. DG contributed to sections Introduction and Discussion. RH
contributed
to
sections
Introduction,
Oceanography
of Stratification in Shelf Seas, Flow Structure Interactions,
Discussion,
and
Conclusions. All
authors
reviewed
the
entire manuscript. modeling
must
then
be
validated
against
direct
before-
and-after observations to skillfully assess the direct and
indirect impacts of anthropogenic mixing, and so guide
sustainable development. Growth
of
the
offshore
wind
energy
industry
must
be
accelerated
to
meet
global
2050
Net
Zero
commitments. AUTHOR CONTRIBUTIONS RD acknowledges the support of the UK Natural Environment
Research Council NE/S014535/1. CL acknowledges the support
of the Offshore Renewable Energy Catapult. BL acknowledges
the support of the Smart Efficient Energy Centre, Bangor
University, part funded by the European Regional Development
Fund. DG acknowledges the support of the UK Engineering
and Physical Sciences Research Council EP/S000747/1. JP
acknowledges
the
support
of
the
Natural
Environment
Research Council Climate Linked Atlantic Sector Science
(CLASS) programme. RD wrote and coordinated contributions to the article. CL
wrote section Flow Structure Interactions and contributed
to sections Introduction, Oceanography of Stratification in
Shelf Seas, Discussion, and Conclusions. BL wrote section
Oceanography of Stratification in Shelf Seas and contributed to
sections Discussion and Conclusions. TR and JHS contributed
to sections Oceanography of Stratification in Shelf Seas
Discussion, and Conclusions. JT and JP contributed to sections 5. CONCLUSIONS Risks posed by offshore wind development
in stratified shelf seas should be mitigated against, but
potential
benefits
should
be
identified
and
maximized. Research
should
consider
how
nature
based
solutions,
foundation
design
and
array
layout
can
enable
sector
growth,
mitigation
of
risks
and
maximization
of benefits. REFERENCES Decadal reanalysis of biogeochemical indicators and fluxes in the north
west european shelf-sea ecosystem. J. Geophys. Res. Oceans 121, 1824–1845. doi: 10.1002/2015JC011496 Billant, P., and Chomaz, J.-M. (2000a). Experimental evidence for a new instability
of a vertical columnar vortex pair in a strongly stratified fluid. J. Fluid Mech. 418, 167–188. doi: 10.1017/S0022112000001154 Cimbala, J. M., and Krein, M. V. (1990). Effect of freestream conditions on
the far wake of a circular cylinder. AIAA J. 28, 1369–1373. doi: 10.2514/3. 25227 Billant, P., and Chomaz, J.-M. (2000b). Theoretical analysis of the zigzag instability
of a vertical columnar vortex pair in a strongly stratified fluid. J. Fluid Mech. 419, 29–63. doi: 10.1017/S0022112000001166 Cocetta, F., Gillard, M., Szmelter, J., and Smolarkiewicz, P. K. (2021). Stratified
flow past a sphere at moderate reynolds numbers. Comput. Fluids 226, 104998. doi: 10.1016/j.compfluid.2021.104998 Bluteau, C., Jones, N., and Ivey, G. (2013). Turbulent mixing efficiency
at an energetic ocean site. J. Geophys. Res. Oceans 118, 4662–4672. doi: 10.1002/jgrc.20292 Cottura, L., Caradonna, R., Ghigo, A., Novo, R., Bracco, G., and Mattiazzo, G. (2021). Dynamic modeling of an offshore floating wind turbine for application
in the mediterranean sea. Energies 14, 248. doi: 10.3390/en14010248 Bosco, M., and Meunier, P. (2014). Three-dimensional instabilities of a stratified
cylinder wake. J. Fluid Mech. 759, 149–180. doi: 10.1017/jfm.2014.517 Boyer, D., Davies, P., Fernando, H., and Zhang, X. (1989). Linearly stratified flow
past a horizontal circular cylinder. Philos. Trans. R. Soc. Lond. A Math. Phys. Sci. 328, 501–528. doi: 10.1098/rsta.1989.0049 Cummins, P. F., Topham, D. R., and Pite, H. D. (1994). Simulated and
experimental two-layer flows past isolated two-dimensional obstacles. Fluid
Dyn. Res. 14, 105. doi: 10.1016/0169-5983(94)90055-8 Broström, G. (2008). On the influence of large wind farms on the upper ocean
circulation. J. Mar. Syst. 74, 585–591. doi: 10.1016/j.jmarsys.2008.05.001 Dalziel, S. B., Patterson, M. D., Caulfield, C., and Le Brun, S. (2011). The structure
of low-froude-number lee waves over an isolated obstacle. J. Fluid Mech. 689,
3–31. doi: 10.1017/jfm.2011.384 Burchard, H., and Rippeth, T. P. (2009). Generation of bulk shear spikes in shallow
stratified tidal seas. J. Phys. Oceanogr. 39, 969–985. doi: 10.1175/2008JPO4074.1 March 2022 | Volume 9 | Article 830927 Frontiers in Marine Science | www.frontiersin.org 21 Infrastructure Mixing in Shelf Seas Dorrell et al. spatial and temporal variability of internal tides. J. Geophys. Res. Oceans 123,
688–707. doi: 10.1002/2017JC012960 spatial and temporal variability of internal tides. J. Geophys. Res. Oceans 123,
688–707. REFERENCES Renew. Sustain. Energy Rev. 139, 110576. doi: 10.1016/j.rser.2020.110576 Global Wind Energy Council. (2020). Global Offshore Wind Report. Available
online
at:
https://gwec.net/global-offshore-wind-report-2020/
(accessed
August 8, 2021). Khani, S., and Waite, M. L. (2015). Large eddy simulations of stratified
turbulence: the dynamic smagorinsky model. J. Fluid Mech. 773, 327–344. doi: 10.1017/jfm.2015.249 Glorioso, P. D., and Flather, R. A. (1995). A barotropic model of the
currents offse south america. J. Geophys. Res. Oceans 100, 13427–13440. doi: 10.1029/95JC00942 Komori, S., Nagaosa, R., and Murakami, Y. (1993). Turbulence structure and mass
transfer across a sheared air-water interface in wind-driven turbulence. J. Fluid
Mech. 249, 161–183. doi: 10.1017/S0022112093001120 Golbraikh, E., and Beegle-Krause, C. (2020). A model for the estimation of the
mixing zone behind large sea vessels. Environ. Sci. Pollut. Res. 27, 37911–37919. doi: 10.1007/s11356-020-09890-y Kröger, S., Parker, R., Cripps, G., and Williamson, P. (2018). Shelf Seas: The
Engine of Productivity, Policy Report on NERC-Defra Shelf Sea Biogeochemistry
Programme. Lowesoft: Cefas. doi: 10.14465/2018.ssb18.pbd Lévy, M., Ferrari, R., Franks, P. J., Martin, A. P., and Rivière, P. (2012). Bringing physics to life at the submesoscale. Geophys. Res. Lett. 39. doi: 10.1029/2012GL052756 Gowen, R., Stewart, B., Mills, D., and Elliott, P. (1995). Regional differences
in
stratification
and
its
effect
on
phytoplankton
production
and
biomass in the northwestern irish sea. J. Plankton Res. 17, 753–769. doi: 10.1093/plankt/17.4.753 Lévy, M., Klein, P., and Treguier, A.-M. (2001). Impact of sub-mesoscale physics
on production and subduction of phytoplankton in an oligotrophic regime. J. Mar. Res. 59, 535–565. doi: 10.1357/002224001762842181 Graham, J. A., O’Dea, E., Holt, J., Polton, J., Hewitt, H. T., Furner, R., et al. (2018). Amm15: a new high-resolution nemo configuration for operational
simulation of the european north-west shelf. Geosci. Model Dev. 11, 681–696. doi: 10.5194/gmd-11-681-2018 Lie,
H.-J. (1989). Tidal
fronts
in
the
southeastern
hwanghae
(yellow
sea). Cont. Shelf
Res. 9,
527–546. doi:
10.1016/0278-4343(89)9
0019-8 Grashorn, S., and Stanev, E. V. (2016). Kármán vortex and turbulent
wake
generation
by
wind
park
piles. Ocean
Dyn. 66,
1543–1557. doi: 10.1007/s10236-016-0995-2 Lienhard, J. H., et al. (1966). Synopsis of lift, drag, and vortex frequency data for
rigid circular cylinders, Vol. 300. Technical Extension Service, Washington State
University Pullman, WA. Gray, T., Haggett, C., and Bell, D. (2005). Offshore wind farms and commercial
fisheries in the uk: a study in stakeholder consultation. Ethics Place Environ. 8,
127–140. doi: 10.1080/13668790500237013 Lincoln, B., Rippeth, T., and Simpson, J. (2016). REFERENCES doi: 10.1002/2017JC012960 Dannheim, J., Bergström, L., Birchenough, S. N., Brzana, R., Boon, A. R., Coolen,
J. W., et al. (2020). Benthic effects of offshore renewables: identification of
knowledge gaps and urgently needed research. ICES J. Mar. Sci. 77, 1092–1108. doi: 10.1093/icesjms/fsz018 Gula, J., Molemaker, M., and McWilliams, J. (2015). Topographic vorticity
generation, submesoscale instability and vortex street formation in the gulf
stream. Geophys. Res. Lett. 42, 4054–4062. doi: 10.1002/2015GL063731 Deloncle, A., Billant, P., and Chomaz, J.-M. (2008). Nonlinear evolution of the
zigzag instability in stratified fluids: a shortcut on the route to dissipation. J. Fluid Mech. 599, 229–239. doi: 10.1017/S0022112007000109 Gula, J., Taylor, J., Shcherbina, A., and Mahadevan, A. (2022). “Submesoscale
processes
and
mixing,”
in
Ocean
Mixing:
Drivers,
Mechanisms
and
Impacts, eds M. Meredith and A. N. Garabato (Elsevier), 181–214. doi: 10.1016/B978-0-12-821512-8.00015-3 Díaz, H., and Soares, C. G. (2020). Review of the current status, technology
and future trends of offshore wind farms. Ocean Eng. 209, 107381. doi: 10.1016/j.oceaneng.2020.107381 Hewitt, G., Vassilicos, C., et al. (2005). Prediction of Turbulent Flows. Cambridge:
Cambridge University Press. Eames, I., Jonsson, C., and Johnson, P. (2011). The growth of a cylinder wake in
turbulent flow. J. Turbulence 12:N39. doi: 10.1080/14685248.2011.619985 Holt, J., and Proctor, R. (2008). The seasonal circulation and volume transport
on the northwest european continental shelf: a fine-resolution model study. J. Holt, J., and Proctor, R. (2008). The seasonal circulation and volume transport
on the northwest european continental shelf: a fine-resolution model study. J. Geophys. Res. Oceans 113:C6. doi: 10.1029/2006JC004034 Egbert, G., and Ray, R. (2000). Significant dissipation of tidal energy in the
deep ocean inferred from satellite altimeter data. Nature 405, 775–778. doi: 10.1038/35015531 Howard, L. N. (1961). Note on a paper of john w. miles. J. Fluid Mech. 10, 509–512. doi: 10.1017/S0022112061000317 Egbert, G. D., and Erofeeva, S. Y. (2002). Efficient inverse modeling of
barotropic ocean tides. J. Atmosphere. Oceanic Technol. 19, 183–204. doi: 10.1175/1520-0426(2002)019andlt;0183:EIMOBOandgt;2.0.CO;2 Howland, M. F., Lele, S. K., and Dabiri, J. O. (2019). Wind farm power optimization
through wake steering. Proc. Natl. Acad. Sci. U.S.A. 116, 14495–14500. doi: 10.1073/pnas.1903680116 Esteban, M. D., Diez, J. J., López, J. S., and Negro, V. (2011). Why offshore wind
energy? Renew Energy 36, 444–450. doi: 10.1016/j.renene.2010.07.009 Inall, M. E., Rippeth, T. P., and Sherwin, T. J. (2000). Impact of nonlinear waves on
the dissipation of internal tidal energy at a shelf break. J. Geophys. Res. Oceans
105, 8687–8705. doi: 10.1029/1999JC900299 Esteban, M. REFERENCES D., López-Gutiérrez, J.-S., and Negro, V. (2019). Gravity-
based foundations in the offshore wind sector. J. Mar. Sci. Eng. 7, 64. doi: 10.3390/jmse7030064 Inall, M. E., Toberman, M., Polton, J. A., Palmer, M. R., Green, J. M., and Rippeth,
T. P. (2021). Shelf seas baroclinic energy loss: pycnocline mixing and bottom
boundary layer dissipation. J. Geophys. Res. Oceans 126, e2020JC016528. doi: 10.1029/2020JC016528 Exo, K.-M., Huppop, O., and Garthe, S. (2003). Birds and offshore wind farms:
a hot topic in marine ecology. Bull. Wader Study Group 100, 50–53. Available
online at: https://sora.unm.edu/node/121708 Floeter, J., van Beusekom, J. E., Auch, D., Callies, U., Carpenter, J., Dudeck, T., et al. (2017). Pelagic effects of offshore wind farm foundations in the stratified north
sea. Prog. Oceanogr. 156, 154–173. doi: 10.1016/j.pocean.2017.07.003 Jacobsen, A., and Godvik, M. (2021). Influence of wakes and atmospheric stability
on the floater responses of the hywind scotland wind turbines. Wind Energy 24,
149–161. doi: 10.1002/we.2563 Jacobsen, H. K., Hevia-Koch, P., and Wolter, C. (2019). Nearshore and
offshore wind development: costs and competitive advantage exemplified
by
nearshore
wind
in
denmark. Energy
Sustain. Dev. 50,
91–100. doi: 10.1016/j.esd.2019.03.006 Forster, R. M. (2018). The effect of monopile-induced turbulence on local suspended
sediment patterns around uk wind farms: field survey report. Technical report,
An IECS report to The Crown Estate, IECS, University of Hull, Hull, HU6 7RX. ISBN=978-1-906410-77-3. Frederiksen, M., Edwards, M., Richardson, A. J., Halliday, N. C., and Wanless,
S. (2006). From plankton to top predators: bottom-up control of a
marine food web across four trophic levels. J. Anim. Ecol. 75, 1259–1268. doi: 10.1111/j.1365-2656.2006.01148.x Jansen, M., Staffell, I., Kitzing, L., Quoilin, S., Wiggelinkhuizen, E., Bulder, B., et al. (2020). Offshore wind competitiveness in mature markets without subsidy. Nat. Energy 5, 614–622. doi: 10.1038/s41560-020-0661-2 Jiang, H. (2021a). Formation mechanism of a secondary vortex street in a cylinder
wake. J. Fluid Mech. 915,A127. doi: 10.1017/jfm.2021.195 Garrett, C., Keeley, J., and Greenberg, D. (1978). Tidal mixing versus thermal
stratification in the bay of fundy and gulf of maine. Atmosphere Ocean 16,
403–423. doi: 10.1080/07055900.1978.9649046 Jiang,
H.,
and
Cheng,
L. (2019). Transition
to
the
secondary
vortex
street in the wake of a circular cylinder. J. Fluid Mech. 867, 691–722. doi: 10.1017/jfm.2019.167 Ghosal, S., and Rogers, M. M. (1997). A numerical study of self-similarity in a
turbulent plane wake using large-eddy simulation. Phys. Fluids 9, 1729–1739. doi: 10.1063/1.869289 Jiang, Z. (2021b). Installation of offshore wind turbines: a technical review. REFERENCES doi: 10.1017/jfm.2013.612 Pauly, D., Christensen, V., Guénette, S., Pitcher, T. J., Sumaila, U. R., Walters, C. J., et al. (2002). Towards sustainability in world fisheries. Nature 418, 689–695. doi: 10.1038/nature01017 Mahadevan,
A. (2016). The
impact
of
submesoscale
physics
on
primary
productivity
of
plankton. Ann. Re.v
Mar. Sci. 8,
161–184. doi: 10.1146/annurev-marine-010814-015912 Pearson, B. C., Grant, A. L. M., Polton, J. A., and Belcher, S. E. (2015). Langmuir
turbulence and surface heating in the ocean surface boundary layer. J. Phys. Oceanogr. 45, 2897–2911. doi: 10.1175/JPO-D-15-0018.1 Mahadevan, A., and Archer, D. (2000). Modeling the impact of fronts and
mesoscale circulation on the nutrient supply and biogeochemistry of the upper
ocean. J. Geophys. Res. Oceans 105, 1209–1225. doi: 10.1029/1999JC900216 Pingree, R., and Griffiths, D. (1978). Tidal fronts on the shelf seas around the british
isles. J. Geophys. Res. Oceans 83, 4615–4622. doi: 10.1029/JC083iC09p04615 Mahadevan, A., D’asaro, E., Lee, C., and Perry, M. J. (2012). Eddy-driven
stratification initiates north atlantic spring phytoplankton blooms. Science 337,
54–58. doi: 10.1126/science.1218740 Pingree,
R.,
Mardell,
G.,
Holligan,
P.,
Griffiths,
D.,
and
Smithers,
J. (1982). Celtic sea and armorican current structure and the vertical
distributions of temperature and chlorophyll. Cont. Shelf Res. 1, 99–116. doi: 10.1016/0278-4343(82)90033-4 Mahadevan, A., and Tandon, A. (2006). An analysis of mechanisms for
submesoscale vertical motion at ocean fronts. Ocean Model. 14, 241–256. doi: 10.1016/j.ocemod.2006.05.006 Platt, T., Fuentes-Yaco, C., and Frank, K. T. (2003). Spring algal bloom and larval
fish survival. Nature 423, 398–399. doi: 10.1038/423398b j
Mahaffey, C., Palmer, M., Greenwood, N., and Sharples, J. (2020). Impacts of
climate change on dissolved oxygen concentration relevant to the coastal
and marine environment around the uk. MCCIP Sci. Rev. 2002, 31–53. doi: 10.14465/2020.arc02.oxy Polton, J., Smith, J., MacKinnon, J., and Tejada-Martínez, A. (2008). Rapid
generation of high-frequency internal waves beneath a wind and wave
forced
oceanic
surface
mixed-layer. Geophys. Res. Lett. 35,
L13602. doi: 10.1029/2008GL033856 Marmorino, G. O., Smith, G. B., North, R. P., and Baschek, B. (2018). Application
of airborne infrared remote sensing to the study of ocean submesoscale eddies. Front. Mech. Eng. 4, 10. doi: 10.3389/fmech.2018.00010 Ren, S., Xie, J., and Zhu, J. (2014). The roles of different mechanisms related to
the tide-induced fronts in the yellow sea in summer. Adv. Atmosphere. Sci. 31,
1079–1089. doi: 10.1007/s00376-014-3236-y Matutano, C., Negro, V., López-Gutiérrez, J.-S., and Esteban, M. D. (2013). Scour
prediction and scour protections in offshore wind farms. Renew. Energy 57,
358–365. REFERENCES doi: 10.1016/j.renene.2013.01.048 Rennau, H., Schimmels, S., and Burchard, H. (2012). On the effect of
structure-induced resistance and mixing on inflows into the baltic sea: a
numerical model study. Coastal Eng. 60, 53–68. doi: 10.1016/j.coastaleng.2011. 08.002 McWilliams, J. C. (2016). Submesoscale currents in the ocean. Proc. R. Soc. A
Math. Phys. Eng. Sci. 472,
20160117. doi:
10.1098/rspa.201
6.0117 Richardson, K., Visser, A., and Pedersen, F. B. (2000). Subsurface phytoplankton
blooms fuel pelagic production in the north sea. J. Plankton Res. 22, 1663–1671. doi: 10.1093/plankt/22.9.1663 Meunier, P. (2012). Stratified wake of a tilted cylinder. part 1. suppression of a von
kármán vortex street. J. Fluid Mech. 699, 174–197. doi: 10.1017/jfm.2012.92 Rippeth, T. P. (2005). Mixing in seasonally stratified shelf seas: a shifting
paradigm. Philos. Trans. R. Soc. A Math. Phys. Eng. Sci. 363, 2837–2854. doi: 10.1098/rsta.2005.1662 Miles, J. W. (1961). On the stability of heterogeneous shear flows. J. Fluid Mech. 10, 496–508. doi: 10.1017/S0022112061000305 Molemaker, M. J., McWilliams, J. C., and Dewar, W. K. (2015). Submesoscale
instability and generation of mesoscale anticyclones near a separation
of
the
california
undercurrent. J. Phys. Oceanogr. 45,
613–629. doi: 10.1175/JPO-D-13-0225.1 Rippeth, T. P., Palmer, M. R., Simpson, J. H., Fisher, N. R., and Sharples, J. (2005). Thermocline mixing in summer stratified continental shelf seas. Geophys. Res. Lett. 32, 4. doi: 10.1029/2004GL022104 Monismith, S. G., Koseff, J. R., and White, B. L. (2018). Mixing efficiency in the
presence of stratification: when is it constant? Geophys. Res. Lett. 45, 5627–5634. doi: 10.1029/2018GL077229 Rippeth, T. P., Wiles, P., Palmer, M. R., Sharples, J., and Tweddle, J. (2009). The diapcynal nutrient flux and shear-induced diapcynal mixing in the
seasonally stratified western irish sea. Cont. Shelf Res. 29, 1580–1587. doi: 10.1016/j.csr.2009.04.009 Nakamura, N. (1996). Two-dimensional mixing, edge formation, and permeability
diagnosed in an area coordinate. J. Atmospher. Sci. 53, 1524–1537. doi: 10.1175/1520-0469(1996)053andlt;1524:TDMEFAandgt;2.0.CO;2 Roobaert, A., Laruelle, G. G., Landschützer, P., Gruber, N., Chou, L., and
Regnier, P. (2019). The spatiotemporal dynamics of the sources and sinks
of co2 in the global coastal ocean. Global Biogeochem. Cycles 33, 1693–1714. doi: 10.1029/2019GB006239 Nygaard, N. G., Steen, S. T., Poulsen, L., and Pedersen, J. G. (2020). Modelling cluster wakes and wind farm blockage. J. Phys. 1618, 062072. doi: 10.1088/1742-6596/1618/6/062072 Ruiz-Castillo,
E., Sharples,
J.,
Hopkins,
J.,
and
Woodward,
M. (2019). Seasonality in the cross-shelf physical structure of a temperate shelf sea
and the implications for nitrate supply. Prog. Oceanogr. 177, 101985. doi: 10.1016/j.pocean.2018.07.006 Nylund, A. REFERENCES Surface mixed layer
deepening
through
wind
shear
alignment
in
a
seasonally
stratified
shallow sea. J. Geophys. Res. Oceans 121, 6021–6034. doi: 10.1002/2015JC0
11382 Guihou, K., Polton, J., Harle, J., Wakelin, S., O’Dea, E., and Holt, J. (2018). Kilometric scale modeling of the north west european shelf seas: Exploring the March 2022 | Volume 9 | Article 830927 Frontiers in Marine Science | www.frontiersin.org 22 Infrastructure Mixing in Shelf Seas Dorrell et al. Loder, J. W., and Greenberg, D. A. (1986). Predicted positions of tidal
fronts
in
the
gulf
of
maine
region. Cont. Shelf
Res. 6,
397–414. doi: 10.1016/0278-4343(86)90080-4 Omand, M. M., D’Asaro, E. A., Lee, C. M., Perry, M. J., Briggs, N., Cetini´c, I., et al. (2015). Eddy-driven subduction exports particulate organic carbon from the
spring bloom. Science 348, 222–225. doi: 10.1126/science.1260062 Lucas, D., Caulfield, C., and Kerswell, R. (2017). Layer formation in horizontally
forced stratified turbulence: connecting exact coherent structures to linear
instabilities. J. Fluid Mech. 832, 409–437. doi: 10.1017/jfm.2017.661 Ouellet, P., Fuentes-Yaco, C., Savard, L., Platt, T., Sathyendranath, S., Koeller, P.,
et al. (2011). Ocean surface characteristics influence recruitment variability of
populations of northern shrimp (pandalus borealis) in the northwest atlantic. ICES J. Mar. Sci. 68, 737–744. doi: 10.1093/icesjms/fsq174 Lucas, N. S., Grant, A. L., Rippeth, T. P., Polton, J. A., Palmer, M. R.,
Brannigan, L., et al. (2019). Evolution of oceanic near-surface stratification
in response to an autumn storm. J. Phys. Oceanogr. 49, 2961–2978. doi: 10.1175/JPO-D-19-0007.1 Palmer, M. R., Polton, J. A., Inall, M. E., Rippeth, T. P., Green, J. A. M., Sharples, J.,
et al. (2013). Variable behavior in pycnocline mixing over shelf seas. Geophys. Res. Lett. 40, 1–6. doi: 10.1029/2012GL054638 Luneva, M. V., Wakelin, S., Holt, J. T., Inall, M. E., Kozlov, I. E., Palmer, M. R., et al. (2019). Challenging vertical turbulence mixing schemes in a tidally energetic
environment: 1. 3-d shelf-sea model assessment. J. Geophys. Res. Oceans 124,
6360–6387. doi: 10.1029/2018JC014307 Palmer, M. R., Rippeth, T. P., and Simpson, J. H. (2008). An investigation of
internal mixing in a seasonally stratified shelf sea. J. Geophys. Res. Oceans 113,
1–14. doi: 10.1029/2007JC004531 Paskyabi, M. B., and Fer, I. (2012). Upper ocean response to large wind farm
effect in the presence of surface gravity waves. Energy Procedia 24, 245–254. doi: 10.1016/j.egypro.2012.06.106 Maffioli, A., Davidson, P., Dalziel, S., and Swaminathan, N. (2014). The
evolution of a stratified turbulent cloud. J. Fluid Mech. 739, 229–253. REFERENCES doi: 10.1175/1520-0485(1979)009andlt;0079:ASFOTCandgt;2.0.CO;2 Taylor, J. R., Smith, K. M., and Vreugdenhil, C. A. (2020). The influence of
submesoscales and vertical mixing on the export of sinking tracers in large-eddy
simulations. J. Phys. Oceanogr. 50, 1319–1339. doi: 10.1175/JPO-D-19-0267.1 g
p
g
g
y
simulations. J. Phys. Oceanogr. 50, 1319–1339. doi: 10.1175/JPO-D-19-0267.1 Sharples, J., Ellis, J. R., Nolan, G., and Scott, B. E. (2013). Fishing and
the oceanography of a stratified shelf sea. Progr. Oceanogr. 117, 130–139. doi: 10.1016/j.pocean.2013.06.014 The Crown Estate (2020). BROAD HORIZONS: Key Resource Areas for Offshore
Wind Summary Report. Available online at: https://www.thecrownestate.co. uk/media/3642/broad-horizons-offshore-wind-key-resource-area-summary-
report.pdf (accessed December 5, 2021). Sharples, J., Moore, C. M., and Abraham, E. R. (2001a). Internal tide dissipation,
mixing, and vertical nitrate flux at the shelf edge of ne new zealand. J. Geophys. Res. Oceans 106, 14069–14081. doi: 10.1029/2000JC000604 Thomas, H., Bozec, Y., Elkalay, K., and De Baar, H. J. (2004). Enhanced
open ocean storage of co2 from shelf sea pumping. Science 304, 1005–1008. doi: 10.1126/science.1095491 Sharples, J., Moore, M. C., Rippeth, T. P., Holligan, P. M., Hydes, D. J., Fisher, N. R., et al. (2001b). Phytoplankton distribution and survival in the thermocline. Limnol. Oceanogr. 46, 486–496. doi: 10.4319/lo.2001.46.3.0486 Thomas, L. N., Tandon, A., and Mahadevan, A. (2008). Submesoscale
processes
and
dynamics. Ocean
Model. Eddying
Regime
177,
17–38. doi: 10.1029/177GM04 Sharples, J., and Tett, P. (1994). Modelling the effect of physical variability
on
the
midwater
chlorophyll
maximum. J
Mar. Res. 52,
219–238. doi: 10.1357/0022240943077109 Thorpe, S. (2016). Layers and internal waves in uniformly stratified fluids stirred
by vertical grids. J. Fluid Mech. 793, 380–413. doi: 10.1017/jfm.2016.121 Sharples, J., Tweddle, J. F., Mattias Green, J., Palmer, M. R., Kim, Y.-N., Hickman,
A. E., et al. (2007). Spring-neap modulation of internal tide mixing and vertical
nitrate fluxes at a shelf edge in summer. Limnol. Oceanogr. 52, 1735–1747. doi: 10.4319/lo.2007.52.5.1735 Tonani, M., Sykes, P., King, R. R., McConnell, N., Pquignet, A.-C., O’Dea, E.,
et al. (2019). The impact of a new high-resolution ocean model on the met
office north-west european shelf forecasting system. Ocean Sci. 15, 1133–1158. doi: 10.5194/os-15-1133-2019 Sheehan, P. M., Berx, B., Gallego, A., Hall, R. A., Heywood, K. J., Hughes, S. L.,
et al. (2018). Shelf sea tidal currents and mixing fronts determined from ocean
glider observations. Ocean Sci. 14, 225–236. doi: 10.5194/os-14-225-2018 Tong, J., Gan, Z., Qi, Y., and Mao, Q. (2010). Predicted positions of tidal
fronts in continental shelf of south china sea. REFERENCES J. Mar. Syst. 82, 145–153. doi: 10.1016/j.jmarsys.2010.04.011 Shen, W., Chen, X., Qiu, J., Hayward, J. A., Sayeef, S., Osman, P., et al. (2020). A
comprehensive review of variable renewable energy levelized cost of electricity. Renew. Sustain. Energy Rev. 133, 110301. doi: 10.1016/j.rser.2020.110301 Trevail, A. M., Green, J. A., Sharples, J., Polton, J. A., Arnould, J. P. Y., and Patrick,
S. C. (2019). Environmental heterogeneity amplifies behavioural response to a
temporal cycle. Oikos 128, 517–528. doi: 10.1111/oik.05579 Umlauf, L., and Burchard, H. (2005). Second-order turbulence closure models for
geophysical boundary layers. a review of recent work. Continental Shelf Res. 25,
795–827. doi: 10.1016/j.csr.2004.08.004 Silvester, J. M., Lenn, Y.-D., Polton, J. A., Rippeth, T. P., and Maqueda, M. M. (2014). Observations of a diapycnal shortcut to adiabatic upwelling
of antarctic circumpolar deep water. Geophys. Res. Lett. 41, 7950–7956. doi: 10.1002/2014GL061538 United Nations (2015). Sustainable Development Goals. Available online at:
https://sdgs.un.org/goals (accessed August 15, 2021). Simpson, J., Allen, C., and Morris, N. (1978). Fronts on the continental shelf. J. Geophys. Res. Oceans 83, 4607–4614. doi: 10.1029/JC083iC09p04607 Uzunoglu, E., and Soares, C. G. (2020). Hydrodynamic design of a free-float
capable tension leg platform for a 10 mw wind turbine. Ocean Eng. 197, 106888. doi: 10.1016/j.oceaneng.2019.106888 Simpson, J., and Bowers, D. (1981). Models of stratification and frontal
movement in shelf seas. Deep Sea Res. A Oceanogr. Res. Pap. 28, 727–738. doi: 10.1016/0198-0149(81)90132-1 van Berkel, J., Burchard, H., Christensen, A., Mortensen, L. O., Petersen,
O. S., and Thomsen, F. (2020). The effects of offshore wind farms on
hydrodynamics and implications for fishes. Oceanography 33, 108–117. doi: 10.5670/oceanog.2020.410 Simpson, J., and Hunter, J. (1974). Fronts in the irish sea. Nature 250, 404–406. doi: 10.1038/250404a0 Simpson, J., Tett, P., Argote-Espinoza, M., Edwards, A., Jones, K., and Savidge, G. (1982). Mixing and phytoplankton growth around an island in a stratified sea. Cont. Shelf Res. 1, 15–31. doi: 10.1016/0278-4343(82)90030-9 van Haren, H., Maas, L., Zimmerman, J., Ridderinkhof, H., and Malschaert, H. (1999). Strong inertial currents and marginal internal wave stability in the
central north sea. Geophys Res Lett. 26, 2993–2996. doi: 10.1029/1999GL002352 Simpson,
J. H.,
Crawford,
W. R.,
Rippeth,
T. P.,
Campbell,
A. R.,
and
Cheok,
J. V. (1996). The
vertical
structure
of
turbulent
dissipation
in
shelf
seas. J. Phys. Oceanogr. 26,
1579–1590. doi: 10.1175/1520-0485(1996)026andlt;1579:TVSOTDandgt;2.0.CO;2 Vindenes, H., Orvik, K. A., Søiland, H., and Wehde, H. (2018). REFERENCES T., Arneborg, L., Tengberg, A., Mallast, U., and Hassellöv, I.-M. (2020). In situ observations of turbulent ship wakes and their potential implications for
vertical mixing. Ocean Sci. Discuss. 17, 1285–1302. doi: 10.5194/os-2020-59 Oakey, N. S., and Greenan, B. J. (2004). Mixing in a coastal environment: 2. a view from microstructure measurements. J. Geophys. Res. Oceans 109:C10. doi: 10.1029/2003JC002193 Scannell, B. (2020). Developing the Structure Function Method for Evaluating
Turbulent Kinetic Energy Dissipation Rate (Ph.D. thesis), Bangor, Wales, UK. Scannell, B. D., Lenn, Y.-D., and Rippeth, T. P. (2021). Impact of adcp motion on
structure function estimates of turbulent kinetic energy dissipation rate. Ocean
Sci. Discuss. 2021, 1–38. doi: 10.5194/os-2021-71 Offshore Renewable Energy Action Coalition (2020). The Power of Our Ocean
Report. Available online at: https://gwec.net/oreac/ (accessed August 15, 2021). March 2022 | Volume 9 | Article 830927 Frontiers in Marine Science | www.frontiersin.org 23 Infrastructure Mixing in Shelf Seas Dorrell et al. Schneider, J. A., and Senders, M. (2010). Foundation design: a comparison of oil
and gas platforms with offshore wind turbines. Mar. Technol. Soc. J. 44, 32–51. doi: 10.4031/MTSJ.44.1.5 Sun,
X.,
Huang,
D.,
and
Wu,
G. (2012). The
current
state
of
offshore
wind
energy
technology
development. Energy
41,
298–312. doi: 10.1016/j.energy.2012.02.054 Schultze, L., Merckelbach, L., Horstmann, J., Raasch, S., and Carpenter,
J. (2020a). Increased mixing and turbulence in the wake of offshore
wind farm foundations. J. Geophys. Res. Oceans 125, e2019JC015858. doi: 10.1029/2019JC015858 Taylor, J., and Zhou, Q. (2017). A multi-parameter criterion for layer formation in
a stratified shear flow using sorted buoyancy coordinates. J. Fluid Mech. 823,
R5. doi: 10.1017/jfm.2017.375 Taylor, J. R. (2016). Turbulent mixing, restratification, and phytoplankton
growth at a submesoscale eddy. Geophys. Res. Lett. 43, 5784–5792. doi: 10.1002/2016GL069106 Schultze, L., Merckelbach, L. M., and Carpenter, J. R. (2020b). “Towed chain
datasets and input files for simulations used in the manuscript “Increased
mixing and turbulence in the wake of offshore wind farm foundations”. J. Geophys. Res. Oceans 125, e2019JC015858. Taylor, J. R., and Ferrari, R. (2011). Ocean fronts trigger high latitude
phytoplankton blooms. Geophys. Res. Lett. 38, 5. doi: 10.1029/2011GL049312 Taylor, J. R., Smith, K. M., and Vreugdenhil, C. A. (2020). The influence of
submesoscales and vertical mixing on the export of sinking tracers in large-eddy Schumacher, J., Kinder, T., Pashinski, D., and Charnell, R. (1979). A structural front
over the continental shelf of the eastern bering sea. J. Phys. Oceanogr. 9, 79–87. REFERENCES Analysis of tidal
currents in the north sea from shipboard acoustic doppler current profiler data. Cont. Shelf Res. 162, 1–12. doi: 10.1016/j.csr.2018.04.001 Cont. Shelf Res. 162, 1–12. doi: 10.1016/j.csr.2018.04.001 Simpson, J. H., and Sharples, J. (2012). Introduction to the Physical and Biological
Oceanography of Shelf Seas. Cambridge: Cambridge University Press. Waite, M., and Smolarkiewicz, P. (2008). Instability and breakdown of a
vertical vortex pair in a strongly stratified fluid. J. Fluid Mech. 606, 1–35. doi: 10.1017/S0022112008001912 Smith, R. B. (1978). A measurement of mountain drag. J. Atmosphere. Sci. 35,
1644–1654. doi: 10.1175/1520-0469(1978)035<1644:AMOMD>2.0.CO;2 Wen, C., Dallimer, M., Carver, S., and Ziv, G. (2018). Valuing the visual impact
of wind farms: A calculus method for synthesizing choice experiments
studies. Sci. Total
Environ. 637,
58–68. doi:
10.1016/j.scitotenv.2018. 04.430 Soares-Ramos, E. P., de Oliveira-Assis, L., Sarrias-Mena, R., and Fernández-
Ramírez, L. M. (2020). Current status and future trends of offshore
wind power in europe. Energy 202, 117787. doi: 10.1016/j.energy.2020.1
17787 Weston, K., Fernand, L., Mills, D., Delahunty, R., and Brown, J. (2005). Primary
production in the deep chlorophyll maximum of the central north sea. J. Plankton Res. 27, 909–922. doi: 10.1093/plankt/fbi064 Srinivasan, K., McWilliams, J. C., Molemaker, M. J., and Barkan, R. (2019). Submesoscale vortical wakes in the lee of topography. J. Phys. Oceanogr. 49,
1949–1971. doi: 10.1175/JPO-D-18-0042.1 Williams, C., Sharples, J., Green, M., Mahaffey, C., and Rippeth, T. (2013a). The
maintenance
of
the
subsurface
chlorophyll
maximum
in
the
stratified western irish sea. Limnol. Oceanogr. Fluids Environ. 3, 61–73. doi: 10.1215/21573689-2285100 Stigebrandt,
A.,
and
Aure,
J. (1989). Vertical
mixing
in
basin
waters
of
fjords. J. Physical
Oceanogr. 19,
917–926. doi: 10.1175/1520-0485(1989)019andlt;0917:VMIBWOandgt;2.0.CO;2 March 2022 | Volume 9 | Article 830927 Frontiers in Marine Science | www.frontiersin.org 24 Dorrell et al. Infrastructure Mixing in Shelf Seas Williams, C., Sharples, J., Mahaffey, C., and Rippeth, T. (2013b). Wind-driven
nutrient pulses to the subsurface chlorophyll maximum in seasonally stratified
shelf seas. Geophys. Res. Lett. 40, 5467–5472. doi: 10.1002/2013GL058171 Conflict of Interest: The authors declare that the research was conducted in the
absence of any commercial or financial relationships that could be construed as a
potential conflict of interest. Williamson, C., and Prasad, A. (1993). A new mechanism for oblique
wave resonance in the natural far wake. J. Fluid Mech. 256, 269–313. Frontiers in Marine Science | www.frontiersin.org March 2022 | Volume 9 | Article 830927 REFERENCES doi: 10.1017/S0022112093002794 Publisher’s Note: All claims expressed in this article are solely those of the authors
and do not necessarily represent those of their affiliated organizations, or those of
the publisher, the editors and the reviewers. Any product that may be evaluated in
this article, or claim that may be made by its manufacturer, is not guaranteed or
endorsed by the publisher. Williamson, C. H. (1996). Vortex dynamics in the cylinder wake. Annu. Rev. Fluid
Mech. 28, 477–539. doi: 10.1146/annurev.fl.28.010196.002401 Winters, K. B., and D’Asaro, E. A. (1996). Diascalar flux and the rate of fluid
mixing. J. Fluid Mech. 317, 179–193. doi: 10.1017/S0022112096000717 Wollast, R. (1998). Evaluation and comparison of the global carbon cycle in the
coastal zone and in the open ocean. Sea 10, 213–252. Copyright © 2022 Dorrell, Lloyd, Lincoln, Rippeth, Taylor, Caulfield, Sharples,
Polton, Scannell, Greaves, Hall and Simpson. This is an open-access article
distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the
original author(s) and the copyright owner(s) are credited and that the original
publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these
terms. Xu, W., Liu, Y., Wu, W., Dong, Y., Lu, W., Liu, Y., et al. (2020). Proliferation
of offshore wind farms in the north sea and surrounding waters revealed
by satellite image time series. Renew. Sustain. Energy Rev. 133, 110167. doi: 10.1016/j.rser.2020.110167 Zhou, Q., Taylor, J., Caulfield, C., and Linden, P. (2017). Diapycnal mixing in
layered stratified plane couette flow quantified in a tracer-based coordinate. J. Fluid Mech. 823, 198–229. doi: 10.1017/jfm.2017.261 March 2022 | Volume 9 | Article 830927 Frontiers in Marine Science | www.frontiersin.org 25
|
https://openalex.org/W2106334526
|
https://link.springer.com/content/pdf/10.1007%2FJHEP01%282016%29029.pdf
|
English
| null |
Strong-coupling scales and the graph structure of multi-gravity theories
|
The Journal of high energy physics/The journal of high energy physics
| 2,016
|
cc-by
| 11,940
|
Received: November 13, 2015
Accepted: December 24, 2015
Published: January 7, 2016 Open Access, c⃝The Authors.
Article funded by SCOAP3. Keywords: Classical Theories of Gravity, Field Theories in Higher Dimensions 5
Conclusions 5
Conclusions A The spectrum of the generalised barbell graph
A.1 Spectrum of L(Md,∆) 3
Theory graphs 3
Theory graphs 3.1
Relation between the theory graph and the strong coupling scale 3.3
Bounds on the strong coupling scale
3.3.1
Higher order interaction terms 3.4
Mass diagonalisation 4
Relation to graviton masses 4
Relation to graviton masses 4.1
Relation to K(G) 4.2
Dimensional deconstruction 4.3
Coupling to matter Strong-coupling scales and the graph structure of
multi-gravity theories doi:10.1007/JHEP01(2016)029 doi:10.1007/JHEP01(2016)029 Contents
1
Introduction
1
2
Strong coupling scales in massive, bi-, and multi-gravity
2
2.1
Massive gravity
3
2.2
Bi-gravity
3
2.3
Tri-gravity
4
3
Theory graphs
6
3.1
Relation between the theory graph and the strong coupling scale
7
3.2
Placing bounds on the smallest eigenvalue of A(LT )
9
3.2.1
Upper bound
9
3.2.2
Lower bound
10
3.3
Bounds on the strong coupling scale
11
3.3.1
Higher order interaction terms
12
3.4
Mass diagonalisation
13
4
Relation to graviton masses
14
4.1
Relation to K(G)
15
4.2
Dimensional deconstruction
15
4.3
Coupling to matter
17
5
Conclusions
17
A The spectrum of the generalised barbell graph
18
A.1 Spectrum of L(Md,∆)
20 Contents
1
Introduction
1
2
Strong coupling scales in massive, bi-, and multi-gravity
2
2.1
Massive gravity
3
2.2
Bi-gravity
3
2.3
Tri-gravity
4
3
Theory graphs
6
3.1
Relation between the theory graph and the strong coupling scale
7
3.2
Placing bounds on the smallest eigenvalue of A(LT )
9
3.2.1
Upper bound
9
3.2.2
Lower bound
10
3.3
Bounds on the strong coupling scale
11
3.3.1
Higher order interaction terms
12
3.4
Mass diagonalisation
13
4
Relation to graviton masses
14
4.1
Relation to K(G)
15
4.2
Dimensional deconstruction
15
4.3
Coupling to matter
17
5
Conclusions
17
A The spectrum of the generalised barbell graph
18
A.1 Spectrum of L(Md,∆)
20 2
Strong coupling scales in massive, bi-, and multi-gravity JHEP01(2016)029 Strong-coupling scales and the graph structure of
multi-gravity theories JHEP01(2016)029 James H.C. Scargilla and Johannes Nollerb
aTheoretical Physics, University of Oxford, DWB,
Keble Road, Oxford, OX1 3NP, U.K. bAstrophysics, University of Oxford, DWB,
Keble Road, Oxford, OX1 3RH, U.K. E-mail: james.scargill@physics.ox.ac.uk, noller@physics.ox.ac.uk James H.C. Scargilla and Johannes Nollerb
aTheoretical Physics, University of Oxford, DWB,
Keble Road, Oxford, OX1 3NP, U.K. bAstrophysics, University of Oxford, DWB,
Keble Road, Oxford, OX1 3RH, U.K. Abstract: In this paper we consider how the strong-coupling scale, or perturbative cutoff,
in a multi-gravity theory depends upon the presence and structure of interactions between
the different fields. This can elegantly be rephrased in terms of the size and structure
of the ‘theory graph’ which depicts the interactions in a given theory. We show that the
question can be answered in terms of the properties of various graph-theoretical matrices,
affording an efficient way to estimate and place bounds on the strong-coupling scale of a
given theory. In light of this we also consider the problem of relating a given theory graph
to a discretised higher dimensional theory, `a la dimensional deconstruction. Keywords: Classical Theories of Gravity, Field Theories in Higher Dimensions ArXiv ePrint: 1511.02877 Open Access, c⃝The Authors. Article funded by SCOAP3. Open Access, c⃝The Authors. Article funded by SCOAP3. 1
Introduction Multi-gravity theories, or theories of multiple interacting spin-2 fields, have been extensively
studied in recent years. For consistency, all but one of the fields must be massive [1], leading
to an intimate connection with massive gravity (for reviews see [2, 3]). For a long time
it was thought to be impossible to construct a consistent non-linear theory of massive
gravity [4], yet this proved to be false [5–7]. To write a mass term for the metric requires
the introduction of a new, fiducial rank 2 tensor field [6, 8, 9], which can be rendered
dynamical, leading to bi-gravity [10, 11]; multi-gravity is then a generalisation of this,
involving further dynamical rank 2 tensor fields, and interactions among them [12]. Such theories are necessarily effective field theories which have some cutoff— an
energy scale above which they are not valid — along with some energy scale at which – 1 – perturbation theory breaks down, and higher order interaction terms become relevant (the
so-called ‘strong coupling’ scale); these need not be the same, and the latter only indicates
the energy scale above which the theory is strongly coupled. Whether, in the case of multi-
gravity theories, the strong coupling scale should also be considered as the cutoffof the
effective field theory is a matter of some debate (for a general effective field theory argument
to that effect see [13]) since these theories are known to be ghost free up to the Planck
scale. This is especially relevant in the context of local (e.g. solar system) gravitational
tests if such a modification of gravity is invoked to generate self-accelerating cosmological
solutions (and thereby be a step towards solving the cosmological constant problem), since
is such cases the strong coupling scale is usually very low, typically on the order of 103 km.1 JHEP01(2016)029 We do not dwell on this question here, as in any case, above the strong coupling scale
the nature of the theory clearly must change. Instead we focus on how the strong coupling
scale itself depends on the nature of the multi-gravity theory, and in particular on the size
and structure of the ‘theory graph’ (explained in section 3). 1
Introduction We will find that the structure
of the graph does have a large impact, since for example the strong coupling scale of a
theory based on a path graph decreases with the number of gravitons as ∼1/
√
N, whereas
that based on a star graph does not change. A further aspect which provides motivation for studying these theories is dimensional
deconstruction, which seeks to approximate a Kaluza-Klein reduced theory by discretising
the extra dimension [15]; multi-gravity theories naturally appear out of such a frame-
work [16]. This paper has the following structure. In section 2 we recapitulate the results for
the strong coupling scale in dRGT massive gravity and HR bi-gravity, as well as giving a
simple tri-gravity example to show how, and why, this can change as more spin-2 fields
are introduced. Section 3 discusses the graph representation of multi-gravity theories, and
derives bounds on the strong coupling scale which depend only on properties of the graph. In section 4 this is related to dimensional deconstruction, possible interpretation of general
multi-gravity theories in terms of discretisation of higher-dimensional theories is discussed,
and we briefly touch upon the coupling of these theories to matter. Finally, section 5
concludes. Throughout this paper we will work in four spacetime dimensions, though an
extension to arbitrary dimensions does not essentially change the conclusions. 1It is worth noting that this value is environment dependent [3, 14]. 2.1
Massive gravity In massive gravity there is one dynamical metric, g, which transforms under general co-
ordinate transformations (GC), and one fiducial non-dynamical one, f, which does not
transform under GC. The action consists of a kinetic term (Ricci scalar) for g, and the
following dRGT interaction terms: In massive gravity there is one dynamical metric, g, which transforms under general co-
ordinate transformations (GC), and one fiducial non-dynamical one, f, which does not
transform under GC. The action consists of a kinetic term (Ricci scalar) for g, and the
following dRGT interaction terms: L = Lkin + Lint = M2√−gR[g] + m2M2√−g
4
X
i=0
βiei
p
g−1f
,
(2.1) (2.1) JHEP01(2016)029 where in massive gravity M would be the Planck mass for g, but we leave it general here as
this is not necessarily the case in multi-gravity, and m is a mass parameter which controls
the size of the graviton mass. The interaction terms break the GC symmetry, which however can be restored by
introducing the St¨uckelberg fields Y µ, in a manner patterned after the gauge symmetry we
wish to (re-)introduce: fµν(x) →Fµν(x) = ∂µY α∂νY βfαβ(Y (x)). (2.2) (2.2) The transformation properties of Y µ are such that Fµν transforms as a tensor under GC. We then expand the metrics around flat space, and the St¨uckelberg field around the identity, gµν = ηµν + hµν,
fµν = ηµν,
Y µ(x) = xµ + Aµ,
(2.3) (2.3) ntroducing a U(1) St¨uckelberg field for Aµ: Aµ →Aµ + ∂µπ. before finally introducing a U(1) St¨uckelberg field for Aµ: Aµ →Aµ + ∂µπ. Note that the St¨uckelberg scalar field, π, always appears with two derivatives acting
on it, and so cannot at this stage have a canonical kinetic term. It gains one through
mixing with the spin-2 field hµν: at the quadratic order these fields are mixed, which can
be removed via a field redefinition of the form ∼hµν →hµν −πηµν. As well as removing the hπ mixing and introducing a canonical kinetic term for π, this
field redefinition also introduces terms of the form π(∂2π)n, which will turn out to be those
suppressed by the lowest scale. 2
Strong coupling scales in massive, bi-, and multi-gravity To find the strong coupling scale for the theory we employ the St¨uckelberg trick to restore
diffeomorphism invariance to all the spin-2 fields and expand around Minkowski space, in
order to make manifest the different degrees of freedom (for a review see [2]). There will
exist operators with dimension greater than four, suppressed by certain scales, the smallest
of which gives the strong coupling scale. We will recapitulate this for massive- and bi-
gravity, before giving a simple tri-gravity example, which contains a lot of the qualitatively
different features of multi-gravity. (See [17, 18], and [19] respectively for more details.) – 2 – 2.1
Massive gravity Once all the fields are canonically normalised one has L ⊃
1
4(β2 + β3)
1
Λ3
3
πXµ
(2),µ(π) +
1
6β3
1
Λ6
3
πXµ
(3),µ(π),
(2.4) (2.4) and hence the strong coupling scale of dRGT massive gravity is Λ3 =
m2M
1/3. 2.2
Bi-gravity In bi-gravity the action is as in massive-gravity, with the addition of a kinetic term for f,
which is now dynamical. The two dynamical fields each transform under a different copy of
GC; the interaction term breaks the symmetry GCg ×GCf down to the diagonal subgroup. L = M2√−gR[g] + M2p
−fR[f] + 1
2Lint,
(2.5) (2.5) – 3 – where Lint is as in (2.1),2 the factor of 1/2 comes from replacing M in the interaction term
by Meff= (M−2
g
+ M−2
f )−1/2 and we have chosen the ‘Planck masses’ Mg and Mf of the
two fields to be the same. Compared with massive gravity there is now a choice as to how we introduce the
St¨uckelberg fields: we can again make f transform as a tensor under GCg (and be in-
variant under GCf) via fµν(x) →Fµν(x) = ∂µY α
f→g∂νY β
f→gfαβ(Yf→g(x)), or we can do the
opposite, and change the transformation properties of g (making it a tensor under GCf and
invariant under GCg) via gµν(x) →Gµν(x) = ∂µY α
g→f∂νY β
g→fgαβ(Yg→f(x)). Comparing the
two cases, one finds Yg→f = Y −1
f→g, and that the St¨uckelberg scalar fields in each case, JHEP01(2016)029 Y µ
f→g(x) = xµ + ∂µπ,
Y µ
g→f(x) = xµ + ∂µφ,
(2.6) (2.6) are related to that in the other case by the so-called Galileon duality [20], which in it-
self is a direct consequence of the different, yet physically equivalent, ways of introducing
St¨uckelberg fields. The terms suppressed by Λ3 are now The terms suppressed by Λ3 are now 1
12(5β2 + 4β3 −β4)
1
Λ3
3
πXµ
(2),µ(π) +
1
24(7β3 −β4)
1
Λ6
3
πXµ
(3),µ(π),
(2.7) (2.7) where terms involving φ have been rewritten using the Galileon duality relation φ = −π +
1
2 (∂π)2 + · · · [18, 21]. Thus the strong coupling scale is still effectively Λ3. 2.3
Tri-gravity One can de-mix this through the following
field redefinition: h(1)
µν →h(1)
µν −(a1φ(1,2) + a2φ(1,3))ηµν,
h(i)
µν →h(i)
µν −ai−1π(1,i)ηµν
for i = 2,3. (2.1 To examine the quadratic and cubic order terms we must pick one from each pair of a
St¨uckelberg field and its dual, and rewrite the other in terms of that via To examine the quadratic and cubic order terms we must pick one from each pair of a
St¨uckelberg field and its dual, and rewrite the other in terms of that via φ = −π + 1
2 (∂π)2 + · · · ,
(2.11) (2.11) JHEP01(2016)029 (or equivalently for π in terms of φ). Calling the so-chosen fields ρ(i), i = 1, 2, and writing
˜ρ(i) = ai−1ρ(i), one has the following kinetic terms (or equivalently for π in terms of φ). Calling the so-chosen fields ρ(i), i = 1, 2, and writing
˜ρ(i) = ai−1ρ(i), one has the following kinetic terms Lρ□ρ ∝
˜ρ(1) ˜ρ(2)
2 σ
σ 2
! □˜ρ(1)
□˜ρ(2)
! = ˜ρT K□˜ρ,
(2.12) (2.12) where σ = 1 if {ρ(i)} = {π(1,2), π(1,3)} or {φ(1,2), φ(1,3)}, i.e. either g(1) is being always
mapped to the other sites, or g(2),(3) are both being mapped to site 1; and σ = −1 otherwise,
i.e. g(1) is mapped to another site in one interaction term, but not in the other, which
corresponds to {ρ(i)} = {π(1,2), φ(1,3)} or {φ(1,2), π(1,3)}. See figure 4. At cubic order there are two types of terms: those arising from ρ(i)Xµ
(2)µ(ρ(j)), as
appear in massive gravity, and those arising from the second order expansion of one of the
fields in a nominally quadratic term, e.g. φ(1,2)□φ(1,3) →1
2(∂ρ(1))2□ρ(2). We call these
respectively, A-, and B-type terms, and write LA ∝
˜ρ(1) ˜ρ(2)
CA
Xµ
(2)µ(˜ρ(1))
Xµ
(2)µ(˜ρ(2))
and
LB ∝
□˜ρ(1) □˜ρ(2)
CB
(∂˜ρ(1))2
(∂˜ρ(2))2
! , (2.13) (2.13) where the coefficient matrices CA,B depend on how the links are oriented, and are displayed
in table 1. As the scalars are mixed at the quadratic level, in order to ask questions concerning
when interactions become strongly coupled we must find the actual propagating modes, i.e. 2.3
Tri-gravity Consider now a tri-metric theory in which two of the fields (labelled 2 and 3) interact only
with a third (labelled 1), and whose interaction Lagrangian is (also see figure 2) Consider now a tri-metric theory in which two of the fields (labelled 2 and 3) interact only
with a third (labelled 1), and whose interaction Lagrangian is (also see figure 2) Lint = m2M2
"
p−g(2)
X
i
β(1,2)
i
ei
q
g−1
(2)g(1)
+ p−g(3)
X
i
β(1,3)
i
ei
q
g−1
(3)g(1)
#
, (2.8) and the kinetic sector for this and all subsequent N-metric theories consists of a sum of
the corresponding N Einstein-Hilbert terms, just as in the bigravity example above. We now introduce St¨uckelberg fields in each term mapping g(1) to site 2 or 3 as ap-
propriate.3 For the remainder of this subsection we set m = M = 1 in order to focus on
the interaction structure of the theory, since the explicit mass scales are the same as in
bigravity. To cubic order in the fields the scalar-tensor mixing then looks like Lhπ = h(1)
µν
h
a1Xµν
(1)(φ(1,2)) + b1,RXµν
(2)(φ(1,2)) + a2Xµν
(1)(φ(1,3)) + b2,RXµν
(2)(φ(1,3))
i
(2.9)
+ h(2)
µν
h
a1Xµν
(1)(π(1,2)) + b1,LXµν
(2)(π(1,2))
i
+ h(3)
µν
h
a2Xµν
(1)(π(1,3)) + b2,LXµν
(2)(π(1,3))
i
,
2Note that since √−gem
p
g−1f
= √−feD−m
p
f −1g
, Lint does not preference either of the two
fields. (2.9) 2Note that since √−gem
p
g−1f
= √−feD−m
p
f −1g
, Lint does not preference either of the two
fields. 2Note that since √−gem
p
g−1f
= √−feD−m
p
f −1g
, Lint does not preference either of the two
fields. 3As explained in more detail in [19], when applying the St¨uckelberg trick to a multi-gravity theory, one
can treat each interaction term independently, and for a given interaction term f(g(1), g(2)) one can choose
to make g(2) transform under the gauge symmetry of g(1), or vice versa. The two possibilities are equivalent
and related by the so-called Galileon duality [20, 21]. – 4 – where the coefficients ai, bi,R, bi,L are functions of the β(j,k)
i
parameters (and the L, R labels
refer to the direction of the St¨uckelberg fields). 4We also need to diagonalise their mass matrix, however in practice this does not have a significant
impact on the strong coupling scale, and so we ignore it here, but comment further in section 3.4. 5Though see [19] for examples of how this is no longer true when the assumption is invalid. 2.3
Tri-gravity a−1
1
−a−1
2
0
a−1
2
! Table 1. The matrices of coefficients of cubic terms (before quadratic demixing and canonical
normalisation of the modes), for the different possible orientations of the links. JHEP01(2016)029 Table 1. The matrices of coefficients of cubic terms (before quadratic demixing and canonical
normalisation of the modes), for the different possible orientations of the links. where where ˜Cijk =
1
p
λiλjλk
X
l,m
ClmUliUmjUmk,
(2.15) (2.15) and the explicit mass scale dependence via Λ3
3 has been reintroduced. The least suppressed
cubic interaction is the one with the largest (in magnitude) element of ˜C, ˜Cmax and hence
the effective strong coupling scale is Λ3/
˜Cmax
1/3
. and the explicit mass scale dependence via Λ3
3 has been reintroduced. The least suppressed
cubic interaction is the one with the largest (in magnitude) element of ˜C, ˜Cmax and hence
the effective strong coupling scale is Λ3/
˜Cmax
1/3
.
The strong coupling scale is thus shifted from the massive-/bi-gravity value. Whilst in
the tri-metric case, assuming that the coefficients in front of the initial interaction terms are
of the same order, this shift is not terribly large,5 when the number of fields involved is large
the effect can be significant, as we show below. The reader may wonder whether higher
order terms (quartic in the scalar and above) could lead to an effective strong coupling
scale which is lower than this. For tri-gravity this may be possible, since the numbers are
still O(1), however as we shall explain in section 3.3.1, one would expect higher order terms
to be less suppressed than cubic for larger numbers of interacting spin-2 fields. We emphasise that this requirement to de-mix the scalar fields, and the resulting shift
in the strong coupling scale, is a qualitatively different feature of multi-gravity, which does
not appear in bi- or massive- gravity. 2.3
Tri-gravity we must diagonalise the kinetic matrix K.4 Transforming to the eigenbasis of K involves
˜ρ →Uχ, where U is the matrix whose columns are the normalised eigenvectors of K;
the kinetic terms will then look like P2
i=1 λiχ(i)□χ(i), and so finally we must canonically
normalise the fields, χ(i) →χ(i)/√λi. Once this is done the cubic terms will look like Once this is done the cubic terms will look like LA =
X
ijk
˜CA
ijk
Λ3
3
χiXµ
(2),µ
χj, χk
,
and
LB =
X
ijk
˜CB
ijk
Λ3
3
□χi∂µχj∂µχk,
(2.14) (2.14) 4We also need to diagonalise their mass matrix, however in practice this does not have a significant
impact on the strong coupling scale, and so we ignore it here, but comment further in section 3.4. 4We also need to diagonalise their mass matrix, however in practice this does not have a significant
impact on the strong coupling scale, and so we ignore it here, but comment further in section 3.4. – 5 – {ρ(1), ρ(2)}
CA
CB
{π(1,2), π(1,3)}
a−2
1 (b1,L −b1,R)
−a−2
2 b2,R
−a−2
1 b1,R
a−2
2 (b2,L −b2,R)
! a−1
1
a−1
2
a−1
1
a−1
2
! {φ(1,2), φ(1,3)}
a−2
1 (b1,R −b1,L)
a−2
2 b2,R
a−2
1 b1,R
a−2
2 (b2,R −b2,L)
! a−1
1
0
0
a−1
2
! {π(1,2), φ(1,3)}
a−2
1 (b1,L −b1,R)
−a−2
2 b2,R
a−2
1 b1,R
a−2
2 (b2,R −b2,L)
! a−1
1
0
−a−1
1
a−1
2
! {φ(1,2), π(1,3)}
a−2
1 (b1,R −b1,L)
a−2
2 b2,R
−a−2
1 b1,R
a−2
2 (b2,L −b2,R)
! a−1
1
−a−1
2
0
a−1
2
! Table 1. The matrices of coefficients of cubic terms (before quadratic demixing and canonical
normalisation of the modes), for the different possible orientations of the links. {ρ(1), ρ(2)}
CA
CB
{π(1,2), π(1,3)}
a−2
1 (b1,L −b1,R)
−a−2
2 b2,R
−a−2
1 b1,R
a−2
2 (b2,L −b2,R)
! a−1
1
a−1
2
a−1
1
a−1
2
! {φ(1,2), φ(1,3)}
a−2
1 (b1,R −b1,L)
a−2
2 b2,R
a−2
1 b1,R
a−2
2 (b2,R −b2,L)
! a−1
1
0
0
a−1
2
! {π(1,2), φ(1,3)}
a−2
1 (b1,L −b1,R)
−a−2
2 b2,R
a−2
1 b1,R
a−2
2 (b2,R −b2,L)
! a−1
1
0
−a−1
1
a−1
2
! {φ(1,2), π(1,3)}
a−2
1 (b1,R −b1,L)
a−2
2 b2,R
−a−2
1 b1,R
a−2
2 (b2,L −b2,R)
! 3
Theory graphs The structure of a multi-gravity theory can be conveniently encoded in a so-called theory
graph [12, 19, 22]. Each spin-2 field in the theory is represented by a vertex, and where
two fields appear in an interaction term in the action the corresponding vertices in the
theory graph are connected. Examples are given in figure 1. Note that we are only
considering superpositions of bigravity-like interaction terms, but in general one may also
write interaction terms which involve more than two fields [12]. Also, certain aspects – 6 – 2
4
1
2
3
nt = f(g(1)
µν , g(2)
µν )
Lint = f1(g(1)
µν , g(4)
µν ) + f2(g(2)
µν , g(4)
µν ) + f3(g(3)
µν , g(4)
µν ) 1
2
4
1
2
3
Lint = f(g(1)
µν , g(2)
µν )
Lint = f1(g(1)
µν , g(4)
µν ) + f2(g(2)
µν , g(4)
µν ) + f3(g(3)
µν , g(4)
µν )
Figure 1. Examples of different types of theory graphs and their corresponding interaction
grangians; vertices correspond to metrics, and when two metrics are present in the same interact
term the corresponding vertices are connected by an edge. 1
2
Lint = f(g(1)
µν , g(2)
µν ) 4 2 3 Lint = f1(g(1)
µν , g(4)
µν ) + f2(g(2)
µν , g(4)
µν ) + f3(g(3)
µν , g(4)
µν ) Lint = f(g(1)
µν , g(2)
µν ) Figure 1. Examples of different types of theory graphs and their corresponding interaction La-
grangians; vertices correspond to metrics, and when two metrics are present in the same interaction
term the corresponding vertices are connected by an edge. Figure 1. Examples of different types of theory graphs and their corresponding interaction La-
grangians; vertices correspond to metrics, and when two metrics are present in the same interaction
term the corresponding vertices are connected by an edge. JHEP01(2016)029 of multi-gravity theories, in particular the cosmology, have been studied from a graph
theoretical perspective in [23]. of multi-gravity theories, in particular the cosmology, have been studied from a graph
theoretical perspective in [23]. Already some structural properties of the theory graph are known to have a bearing on
the theory. 3
Theory graphs In particular, if the theory graph contains cycles, then the equivalence between
the metric and vielbein versions of the theory breaks down [12], and in [24] it was shown that
a cycle causes a ghost to appear at a scale lower than Λ3 in the metric version of the theory,
which lowers the strong coupling scale and the cutoff. Furthermore, the vielbein version
of the theory, when examined in the decoupling limit, which makes explicit the strong
coupling scale, becomes exceedingly complicated when there is a cycle in the graph [24]. For these reasons, in what follows we will focus on theory graphs which are acyclic
(i.e. they are tree graphs), and it does not matter which version (metric or vielbein) of the
theory we use; for concreteness we use the metric version. Also note that to encode the full information about the theory this would be a weighted
graph, with the weight of each edge depending on the coefficient of the corresponding
interaction term, however in order to focus on the effect of the structure of the graph we
will choose these coefficients so that the we can consider simply an unweighted graph. 6Note that as we are considering only acyclic graphs, this is the only possibility with three vertices (up
to vertex relabelling). 7And since this is a tree graph it will have N −1 edges. 3.1
Relation between the theory graph and the strong coupling scale As explained in section 2, the strong coupling scale of a multi-gravity theory (in which all
the interaction mass scales are identical) is not simply Λ3 because one must de-mix and
canonically normalise the St¨uckelberg scalar fields, which leads to a hierarchy in the values
of the coefficients in front of the non-linear interaction terms (i.e. the Wilson coefficients). From (2.15) we see that the eigenvalues of the St¨uckelberg scalar kinetic matrix for the de-
mixed modes in question are a key ingredient in the size of the coefficient. Thus a first step
in finding the strong coupling scale is to find the smallest eigenvalue of the kinetic matrix. In order to see how this is related to the theory graph let us return to the tri-metric
theory considered in the previous section, whose theory graph is simply the path graph
on three vertices,6 P3, as shown in figure 2. If we choose the direction of the St¨uckelberg – 7 – 2
1
3 2
1
3 3 Figure 2. The path graph on three vertices, P3. Figure 2. The path graph on three vertices, P3. Figure 2. The path graph on three vertices, P3. 2 Figure 3. The operation of the line graph operator, L. Figure 3. The operation of the line graph operator, L. JHEP01(2016)029 2
1
3
2
1
3 2
1
3
2
1
3
Figure 4. St¨uckelberg mapping directions allowed by σ = 1 in (2.12). 2
1
3
2
1
3
Figure 4. St¨uckelberg mapping directions allowed by σ = 1 in (2.12). Figure 4. St¨uckelberg mapping directions allowed by σ = 1 in (2.12). Figure 4. St¨uckelberg mapping directions allowed by σ = 1 in (2.12). mapping so that σ = 1, we see from (2.12) that (in matrix notation) K = 2I2 + A(P2),
where IN is the N-dimensional identity matrix and A(G) is the adjacency matrix of a
graph G, which is defined as A(G)ij =
(
1
if vertices i and j are joined by an edge,
0
else;
(3.1) (3.1) we also note that P2 is the line graph of P3, where the line graph, LG, is the graph whose
vertices are the edges of G, and two vertices in LG are joined by an edge if the corresponding
edges in G are connected to the same vertex. This is illustrated by an example in figure 3. 9Extending this to graphs which contain cycles one has that, unless G is a tree or contains at least one
odd cycle, λmin(LG) = −2 (and λmin(LG) ≥−2 in general) [26], which means that not all the St¨uckelberg
scalar modes have a kinetic term — leading to all sorts of problems (for details see [24]).
One might
be tempted to conclude that this means that theory graphs containing at least one odd cycle can avoid
the problems, but note that our prescription concerning the directions of St¨uckelberg mapping fails in the
presence of an odd cycle and we no longer have K ∝2I + A(LG). 3.2
Placing bounds on the smallest eigenvalue of A(LT ) Some bounds are already known about the smallest eigenvalue of the line graph of a tree;
in particular from [25] we have −2 cos
π
N
≤λmin(LT ) ≤−1,
(3.3) (3.3) where the lower bound is saturated iffT = PN, and the upper bound iffT = SN, i.e. the ‘star graph’, which consists of one vertex connected to N −1 further vertices which
are not connected amongst themselves (the right graph in figure 1 is S4, for example).8
There is quite a gulf between these two graphs and their corresponding bounds, and in
fact the former results in a strong coupling scale which goes ∼1/
√
N whereas the latter
in one which doesn’t scale with N. This makes it difficult to draw conclusions about the
strong coupling scale for a general tree. Thus we attempt to find bounds which are tighter,
though note that they must necessarily depend on further properties of the graph, as the
bounds (3.3) are saturated in certain circumstances. JHEP01(2016)029 Incidentally we point out that from (3.3) we have λmin > −2 (for any finite N) and
hence K is positive definite (cf. equation (3.2)), which agrees with expectations since we
know that in the absence of cycles all the St¨uckelberg scalar modes have a kinetic term,
all of which are of the right sign.9 3.1
Relation between the theory graph and the strong coupling scale h
h
d
h
h
d
f h
S
k lb we also note that P2 is the line graph of P3, where the line graph, LG, is the graph whose
vertices are the edges of G, and two vertices in LG are joined by an edge if the corresponding
edges in G are connected to the same vertex. This is illustrated by an example in figure 3. The choice σ = 1 corresponds to choosing the direction of the St¨uckelberg mapping
such that either g(2) and g(3) are both mapped to the transformation properties of g(1) in
their respective interaction terms, or g(1) is mapped to the transformation properties of
g(2) or g(3) in the respective interaction terms. Or in other words, that each vertex in the
theory graph either only has fields being mapped to it, or away from it, as shown in figure 4. For a general multi-gravity theory we can consider all St¨uckelberg scalars in this way,
and furthermore for a tree graph we can always orient the edges so that each vertex has
only inward or only outward edges (corresponding to σ = 1; on the other hand, choosing
σ = −1 cannot be consistently extended to graphs with vertices of degree higher than two). Thus for a general theory graph T with N vertices7 we have K(T) ∝2IN−1 + A(LT ). (3.2) (3.2) Thus the problem of finding the smallest eigenvalue of the kinetic matrix is the same
as finding the smallest eigenvalue of the line graph. As we shall later see, λmin(K(T)) is
equal to the algebraic connectivity of the graph T. The following subsection is devoted to
placing bounds on the smallest eigenvalue of the line graph of a tree graph. – 8 – 8In standard notation the star graph would be K1,N−1, but we call it SN here to avoid confusion with
the kinetic matrix K.
9 3.2.1
Upper bound Consider adding a vertex (and edge) to T to form a new tree T ′. By suitably relabelling
vertices we have
! A(LT ′) =
A(LT ) a
aT
0
! ,
(3.4) (3.4) where aT = (0, . . . , 0,
z }| {
1, . . . , 1), and m is the degree of the vertex in T to which the new
vertex is attached. Let v be the eigenvector of A(LT ) with the smallest eigenvalue λmin(LT ),
and let v′T = (vT , 0), then by Rayleigh’s theorem we have λmin(LT ′) ≤v′T A(LT ′)v′
v′T v′
= λmin(LT ). (3.5) (3.5) Thus we see that as a tree grows, the minimum eigenvalue of its line graph can never
increase. As we shall see later, this means that we can never raise the strong coupling scale
of a theory by adding new vertices and/or edges in this way. Thus we see that as a tree grows, the minimum eigenvalue of its line graph can never
increase. As we shall see later, this means that we can never raise the strong coupling scale
of a theory by adding new vertices and/or edges in this way. 8In standard notation the star graph would be K1,N−1, but we call it SN here to avoid confusion with
the kinetic matrix K. – 9 – 1
2
d −3
d −2
⌊N+1−d
2
⌋
⌈N+1−d
2
⌉
Figure 5. The tree which minimises λmin(LT ) for a given number of edges N −1 and diameter d. ⌊X⌋denotes the largest integer less than or equal to X and vice versa for ⌈X⌉. Figure 5. The tree which minimises λmin(LT ) for a given number of edges N −1 and diameter d. ⌊X⌋denotes the largest integer less than or equal to X and vice versa for ⌈X⌉. Figure 5. The tree which minimises λmin(LT ) for a given number of edges N −1 and diameter d. ⌊X⌋denotes the largest integer less than or equal to X and vice versa for ⌈X⌉. JHEP01(2016)029 The diameter of a graph, d(G), is defined as the maximum of the set of minimum
distances between all pairs of vertices; thus for example d(PN) = N −1 and d(SN) = 2. 3.2.1
Upper bound Any tree can then be considered to be Pd(T)+1 with certain ‘extrusions’ attached to it, and
thus we immediately have λmin(LT ) ≤λmin(LPd(T )+1) = −2 cos
π
d(T) + 1
. (3.6)
d (3.6) 11These bounds do require knowledge of the diameter of the tree, however in general this is much eas
to compute than the spectrum (O(N) vs. O(N 3)). 10When d = 2 the tree must be a star, and we have λmin(K(SN)) = 1. 3.2.2
Lower bound The lower bound of (3.3) can also be improved. Given N −1 edges and a diameter d, the
tree whose line graph has the smallest eigenvalue is shown in figure 5. Thus the smallest eigenvalue of the associated line graph will give a lower bound for a
given diameter and number of edges. The derivation of the characteristic polynomial of the
adjacency matrix of the line graph of figure 5 can be found in appendix A, and the result
is that the smallest eigenvalue is −2 cos θ∗
N,d, where θ∗
N,d is the smallest non-zero root of fN,d(θ) = sin((d + 1)θ) + (N −1 −d) sin(dθ) +
1
4(N −3 −d)2 −1
sin((d −1)θ) fN,d(θ) = sin((d + 1)θ) + (N −1 −d) sin(dθ) +
1
4(N −3 −d)2 −1
sin((d −1)θ)
−1
2(N −1 −d)2 sin((d −2)θ) + 1
4(N −1 −d)2 sin((d −3)θ). (3.7) −1
2(N −1 −d)2 sin((d −2)θ) + 1
4(N −1 −d)2 sin((d −3)θ). (3.7) (3.7) Expanding this about θ = 0 (λ = −2) we can get a lower bound (valid for d > 2):10 this about θ = 0 (λ = −2) we can get a lower bound (valid for d > 2):10 θ∗
N,d =
2
p
N(d −2)
"
1 + O
1
Nd,
d
N
2!#
. (3.8) (3.8) We now have an improved set of bounds for the smallest eigenvalue of the line graph
of any tree of given N, d:11 We now have an improved set of bounds for the smallest eigenvalue of the line graph
of any tree of given N, d:11 −2 cos
π
d + 1
≥λmin(LT ) ≥−2 cos θ∗
N,d,
(3.9) (3.9) where the lower bound is saturated by the graph in figure 5. For large N, d this leads to
the following bounds on the smallest eigenvalue of the kinetic matrix 4
N(d −2) < λmin(K(T)) <
π2
(d + 1)2 . (3.10) (3.10) 10When d = 2 the tree must be a star, and we have λmin(K(SN)) = 1. 11 11These bounds do require knowledge of the diameter of the tree, however in general this is much easier
to compute than the spectrum (O(N) vs. O(N 3)). – 10 – 1
2
3
4
1
2
3
4
1
2
3
4
Li
Lo
Figure 6. 3.3
Bounds on the strong coupling scale To get to the strong coupling scale however requires more than just the eigenvalues of the
kinetic matrix, we also need to deal with the sum in (2.15). From table 1 we see that the
form of CA,B depends not just on the relative orientation of the links (related to the choice
of St¨uckelberg fields) connected to a given vertex (as for K), but also whether those links
are oriented towards or away from the vertex. Also relevant are the coefficients in front of
the different interaction terms, and thus to avoid unnecessary complication, and to focus
on the effect of the structure of the graph, in the sequel we demand that the interaction
Lagrangian of each link is of identical form, with the direction of the links in mind. Let us illustrate what we mean in this last point by an example: in the trimetric
theory considered previously, in (2.8), since the directions of the links to site 1 are the
same, β(1,2)
i
= β(1,3)
i
; if we were to add another site, connected to site 3, the interaction
term would have to be p−g(3)
P
i βiei
q
g−1
(3)g(4)
, whereas if it was connected to site 1,
the interaction term would have to be p−g(4)
P
i βiei
q
g−1
(4)g(1)
. q
In order to write expressions for CA,B for a general tree, we need to introduce two
more line graph operators. For a directed graph, the standard definition of its line graph
is that two edges are connected if they share a vertex and one edge is oriented towards the
vertex, and the other away from it. We introduce Li which connects edges if they share a
vertex and are both directed towards it, and Lo which connects edges if they share a vertex
and are both directed away from it. This is illustrated by an example in figure 6. With these in hand we can write CB ∝Co = IN−1 + A(Lo
G),
(3.11)
CA ∝bLCi −bRCo,
(3.12) (3.11)
(3.12) (3.11) (3.12) and given a theory graph, one can now use these expressions and (2.15) to determine the
largest Wilson coefficient, and hence the effective strong coupling scale. and given a theory graph, one can now use these expressions and (2.15) to determine the
largest Wilson coefficient, and hence the effective strong coupling scale. 3.2.2
Lower bound The operation of the additional line graph operators, Li and Lo. JHEP01(2016)029 Figure 6. The operation of the additional line graph operators, Li and Lo. 3.3
Bounds on the strong coupling scale From (2.15) we see that due to the inverse powers of λ(K(G)), the largest coefficient
will likely come from a term in which two of the fields are that with the minimal value of – 11 – λ(K(G)), and so we isolate this dependence by writing λ(K(G)), and so we isolate this dependence by writing λ(K(G)), and so we isolate this dependence by writing ˜Cmax =
f(G)
λmin(K(G)),
(3.13) (3.13) where where f(G) = λmin(K(G)) max
1
p
λiλjλk
X
lm
UliUljClmUmk
! . (3.14) (3.14) One may wonder why a third factor of λmin(K(G))−1/2 is not extracted, and the reason
is twofold: whilst CA,B depends on Li,o
G , and not just LG, the similarities are still such
that P
m ClmUmk ∼λk. This means that ˜C will not in general scale inversely with three
powers of λ(K(G)) and so the maximum coefficient will not generally have three fields with
minimal λ; secondly the form (3.13) will prove to have a nice interpretation in terms of
graviton mass eigenstates, as explained in section 4. JHEP01(2016)029 Having placed bounds on λmin(K(G)) it remains to examine the behaviour of f(G). Unfortunately, for given diameter, it is more difficult to determine which graphs extremise
this, and in fact the dependence of those extreme values on the diameter is much weaker
than for λmin(K(G)). (The same is true if one considers its dependence on the maximum
degree of the graph, for example.) None-the-less we can say something about its expected value (taking the prior distri-
bution to be uniform on the set of unlabelled trees of a given number of vertices); we find ⟨f(T)⟩≈
1
√
d −1. (3.15) (3.15) Thus, given (3.10) one has for d > 2 (note that d = 2 must be star graph, for which Λ = Λ3) Thus, given (3.10) one has for d > 2 (note that d = 2 must be star graph, for which Λ = Λ3) 1.5
√
d −1
N(d −2)
1/3
≲Λ
Λ3
≲2.1
√
d −1
(d + 1)2
1/3
,
(3.16) (3.16) or for large d: 1.5
√
d
d
N
1/3
≲Λ
Λ3
≲2.1
√
d
. (3.17) (3.17) Although these bounds are not strict (since we have only taken the average value of f(T))
they are fairly inclusive. Only 2.5% (resp. 2.0%) of trees with 15 (resp. 3.3
Bounds on the strong coupling scale 20) vertices exceed
the upper bound, and 0.4% (resp. 0.5%) fall below the lower bound. Although these bounds are not strict (since we have only taken the average value of f(T))
they are fairly inclusive. Only 2.5% (resp. 2.0%) of trees with 15 (resp. 20) vertices exceed
the upper bound, and 0.4% (resp. 0.5%) fall below the lower bound. 3.3.1
Higher order interaction terms Having focussed only on the cubic scalar interactions we now briefly consider higher order
(in the fields) interactions and explain why we would not expect them to lead to a lower
effective cutoffthan for cubic terms. An extra scalar has to come in with two extra derivatives (because of the multi-gravity
Λ3 decoupling limit structure [21]), and hence an (n + 2)th order term would look like 1
Λ3n
3
1
pλi1 · · · λin+2
X
lm
Ulin+2 · · · Uli2ClmUmi1
! L(χi1, · · · , χin+2). (3.18) (3.18) – 12 – As before extract n + 1 factors of √λmin, and writing the sum that is left as fn we have
that the effective cutoffis
+1 As before extract n + 1 factors of √λmin, and writing the sum that is left as fn we have
that the effective cutoffis
n+1 Λ
Λ3
= λ
n+1
6n
min
f
1
3n
n
. (3.19) (3.19) Since λmin is raised to an increasingly small power, fn would have to increase with n to
compensate (e.g. for a path graph, Λ ∼1/
√
N requires fn ∼N
n
2 −1). This seems unlikely,
especially as naive power counting in N would suggest opposite behaviour, and so it is
reasonable to conclude that the cubic order terms give the lowest effective cutoff. Since λmin is raised to an increasingly small power, fn would have to increase with n to
compensate (e.g. for a path graph, Λ ∼1/
√
N requires fn ∼N
n
2 −1). This seems unlikely,
especially as naive power counting in N would suggest opposite behaviour, and so it is
reasonable to conclude that the cubic order terms give the lowest effective cutoff. JHEP01(2016)029 3.4
Mass diagonalisation Thus far we have ignored the fact that the St¨uckelberg scalars are mixed not only in their
kinetic terms, but also in their mass terms, and really to find the propagating modes we
must find the mass eigenstates. In practice we find that not including mass diagonalisation
does not greatly affect the value of the strong coupling scale, and in this section we briefly
investigate why this is. Like their kinetic terms, the St¨uckelberg scalars do not initially have mass terms, but
gain them from the spin-2 mass terms when the field redefinition ∼h →h −πη is applied. It turns out that their mass terms mix St¨uckelberg scalars which are two links apart, rather
than just adjacent, and so the simplest example is a tetra-metric path theory, P4, since
this requires three St¨uckelberg scalars. Taking the interaction terms to be the same, as
described in the previous section, the spin-2 mass terms can be written, hT
µν
m2
RR m2
RL
0
0
m2
RL 2m2
LL m2
RL
0
0
m2
RL 2m2
RR m2
RL
0
0
m2
RL m2
LL
hαβMµναβ,
(3.20) (3.20) where Mµναβ contains the information about the tensor structure of these terms and would
be ηµναβ for Fierz-Pauli structure, whereas the other matrix of constant coefficients details
the ‘flavour’ structure. Once the appropriate field redefinitions to remove the quadratic
scalar-tensor mixing are applied, the scalar mass terms look like Lρ2 ∝ρT
m2
RR + 2m2
LL −2m2
RL
2(m2
LL −m2
RL)
−m2
RL
2(m2
LL −m2
RL)
2(m2
RR + m2
LL −m2
RL)
2(m2
RR −m2
RL)
−m2
RL
2(m2
RR −m2
RL)
2m2
RR + m2
LL −2m2
RL
ρ. (3 21)
(3.21) If we have m2
RR = m2
LL = −m2
RL,12 then the mass matrix in (3.21) becomes propor- 12For example, this is the case if the interaction terms are those of the so-called ‘minimal model’ for which
Lint(g, f) = √−g
3 −tr
p
g−1f + det
p
g−1f
. 13When m2
LL = m2
RR = −m2
RL is not the case, one would not expect that all the conclusions of this
section would remain exactly correct, but again, due to the similarity, even in the general case, of the
structure of M(G) to L(G), one would expect the general conclusions to remain accurate. 3.4
Mass diagonalisation In general the requirement is β0 + 4β1 + 6β2 + 4β3 + β4 =
0 = β1 + 3β2 + 3β3 + β4, or equivalently to the absence at the perturbative level of constant and tadpole
terms in the Lagrangian, whose presence at any rate would be equivalent to considering expanding around
a different background. – 13 – tional to
5 4 1
4 6 4
1 4 5
= 4 + 4A(LP4) + A(LP4)2,
(3.22) tional to
5 4 1
4 6 4
1 4 5
= 4 + 4A(LP4) + A(LP4)2,
(3.22) (3.22) where we recall that A is the adjacency matrix. It is easy to see that this property gener-
alises to other trees, i.e. Lρ2 ∝ρT 4 + 4A(LT ) + A(LT )2
ρ. (3.23) (3.23) JHEP01(2016)029 Now since the eigenvectors of K(T) are the eigenvectors of A(LT ), and hence of A(LT )2,
the mass diagonalistion becomes trivial in this case, being already performed by the kinetic
diagonalisation. Now since the eigenvectors of K(T) are the eigenvectors of A(LT ), and hence of A(LT )2,
the mass diagonalistion becomes trivial in this case, being already performed by the kinetic
diagonalisation. When m2
RR = m2
LL = −m2
RL is not satisfied by the interaction terms, then it is no
longer possible to write the scalar mass matrix in terms of A(LT ) and powers thereof, but,
especially if m2
LL ∼m2
RR, its structure is still very similar to that of A(LT ) and its powers,
and so it is perhaps unsurprising that it is safe to neglect its diagonalisation, since it will
already be approximately diagonalised by the kinetic term diagonalisation procedure. 14For graphs with multiple separate connected components, there is a corresponding number of zero
eigenvalues. 4.1
Relation to K(G) Given a graph G, and an arbitrary orientation of its edges, the incidence matrix of a graph,
is the |E| × |V | matrix with elements B(G)ij =
1
if edge i leaves vertex j
−1
if edge i arrives at vertex j
0
otherwise
,
(4.3) B(G)ij =
1
if edge i leaves vertex j
−1
if edge i arrives at vertex j
0
otherwise
,
(4.3) (4.3) then one has then one has L(G) = BT B,
and
K(G) = B+BT
+,
(4.4) (4.4) where B+,ij = |Bij|. Now for a tree (or, indeed, any graph which is odd-cycle-free) one
can consistently orient the edges such that each vertex has edges either all leaving, or all
arriving, in which case one has Bij = (−1)σjB+,ij, where σj = 0, 1 appropriately. Thus for a tree one has K(T) = BBT , and hence the non-zero eigenvalues of L(T) and
K(T) are identical, i.e. spec(L(T)) = {0}∪spec(K(T)), where spec(X) is the spectrum of a
matrix X. In particular this means that λmin(K(T)) = m2
1, and hence from (3.13) one has Λ3
eff= m2
1
M
f(T),
(4.5) (4.5) and (4.1) is recovered if f(T) = M/MPl. Thus in order to consider whether (4.1) gives an
accurate expression for the strong coupling scale one needs an expression for the physical
Planck mass to compare with M/f(G); we now consider two approaches to this. 4
Relation to graviton masses more “connected” theories have a
larger mass gap. JHEP01(2016)029 1
Relation to K(G) 4
Relation to graviton masses In the case of a path graph, one can write the strong coupling scale as [16] Λeff=
m2
1MPl
1/3 ,
(4.1) (4.1) where m1 is the mass of the lightest massive graviton state, and MPl is the physical Planck
mass, which are respectively related to the mass parameters m and M entering the action. Intuitively this makes sense, since if there is a gap between the lightest massive graviton
and the next, then at energies below the mass of the next to lightest graviton the theory
is effectively bi-gravity and so the strong coupling scale should be Λ3 built out of mass
parameters m1 and MPl. In this section we investigate to what extent this is true in general. First we need to find the graviton mass spectrum for a generic theory; from (3.20) one
can see that if m2
LL = m2
RR = −m2
RL = m2, extending to a general graph the mass matrix
satisfies M(G) = m2L(G) = m2 (D(G) −A(G)) ,
(4.2) (4.2) where D(G) is the diagonal matrix whose ii entry is the degree of vertex i. The combination
L(G), of this and the adjacency matrix, is called the Laplacian (or Kirchhoff) matrix of
the graph and its properties have been widely studied.13 where D(G) is the diagonal matrix whose ii entry is the degree of vertex i. The combination
L(G), of this and the adjacency matrix, is called the Laplacian (or Kirchhoff) matrix of
the graph and its properties have been widely studied.13 We note some basic facts about L(G) [27] and their interpretation in terms of multi-
gravity theories. We note some basic facts about L(G) [27] and their interpretation in terms of multi-
gravity theories – 14 – • L(G) has one zero eigenvalue, i.e. there is one massless graviton, since there is one
unbroken copy of GC.14 • The eigenvector corresponding to the zero eigenvalue is (1, · · · , 1)T , i.e. the sum mode
P
i h(i)
µν is always massless. • L(G) is positive semi-definite, i.e. none of the gravitons are tachyonic. • The smallest non-zero eigenvalue of L(G) is called the algebraic connectivity, and is
larger for graphs which are more “connected,” i.e. more “connected” theories have a
larger mass gap. • The smallest non-zero eigenvalue of L(G) is called the algebraic connectivity, and is
larger for graphs which are more “connected,” i.e. 4.2
Dimensional deconstruction A path or cycle graph has a natural interpretation as a discretised extra dimension [15, 16,
28–31], which leads to a relation between M and MPl. The number of sites in the discretised
theory is N = mR, where R is the size of the extra dimension, and hence the most massive
graviton has mass mN = 2m ∼N/R. One also has M2
Pl = M3R, and hence requiring the
validity of the effective theory via mN = M, one has the relation M2
Pl ∼NM2 [16], which
indeed agrees with MPl = M/f(PN). – 15 – 1
2
d −1
∆−1
Figure 7. The ‘mace graph,’ Md,∆. Figure 7. The ‘mace graph,’ Md,∆. Figure 7. The ‘mace graph,’ Md,∆. JHEP01(2016)029 More complicated graphs do not have so obvious an interpretation in terms of higher
dimensional theories, however using that M2
Pl = M3R is still expected to hold for general
dimensionally reduced theories and mN = M, one has f(T) = M
MPl
=
1
√mNR,
(4.6) (4.6) and there are several ways to proceed: one could postulate a relation between R and the
graph structure, and see whether (4.6) is satisfied, or alternatively one could take (4.6) as
a definition of R. As an example, consider the star graph: as already mentioned, f(SN) = 1, and we also
have m2
N = Nm2, and hence R = m−1/
√
N. On its face this is a little peculiar: keeping
the ‘discretisation scale,’ m−1, fixed, as we increase the number of sites, the size of the
extra dimension being described actually decreases. In reality though we probably should
not expect (4.1) to give a good expression for the strong-coupling scale in this case, because
the graviton mass spectrum is highly degenerate: spec(L(SN)) = {0, 1N−2, N}. (4.7) (4.7) i.e. there are N −2 gravitons with mass m1 = m. Thus it is no longer true that at
low enough energies one can treat the theory as simply bi-gravity, and we have no reason
to expect the cutoffto be (4.1), with the attendant interpretation as resulting from a
higher dimensional theory. This is not to say that a star graph theory has no place within
dimensional deconstruction, merely that its interpretation is a little more tricky. 15Of course, this relies on starting offwith order one parameters, and in particular coupling to Eeff=
P
i E(i), rather than, say Eeff=
1
√
N
P
i E(i) — though if the latter were chosen one still would not have a
result which was valid for all graphs. 4.3
Coupling to matter Outside of the framework of dimensional deconstruction an expression for the physical
Planck mass can be derived by coupling the theory to matter, since this coupling strength
is what determines the strength of gravity, and hence MPl. Considering this question in
the detail it deserves is unfortunately beyond the scope of this work, except to make a few
elementary observations. JHEP01(2016)029 Obviously in general this will depend on how one couples to matter (the known con-
sistent non-derivative such couplings are those of [34–36]), for instance if one couples to
just one site, then this will clearly depend on at least the graph structure in the vicinity
of that vertex. A more realistic scenario might be that one wants to couple to the (perturbative)
massless mode. Regardless of graph structure, the massless mode is h(m=0)
µν
= P
i h(i)
µν; this
leads to a kinetic term NM2h(m=0)Eh(m=0), and hence canonical normalisation involves
h(m=0) →
1
√
NM h(m=0). At the non-linear level coupling to this could arise from coupling
to an effective vielbein [35] Eeff= P
i E(i); linearising, and then canonically normalising
one has
1
Z 1
√
NM
Z
d4x h(m=0)
µν
T µν,
(4.10) (4.10) and thus the coupling strength is MPl =
√
NM, i.e. the same as in dimensional decon-
struction using a path graph. This gives f = M/MPl ∼1/
√
N, which is satisfied for trees
with d ∼N (see section 3.3). However there are many trees which violate this, and hence
violate (4.1); furthermore not all of these are ones which have a degenerate lightest graviton
mass (for which we already do not expect (4.1) to hold). This would seem to advise against
using (4.1) in general to estimate the strong coupling (at least when one is coupling matter
to the massless mode).15 4.2
Dimensional deconstruction Denoting by ∆the largest degree of any vertex in T, one has the following bounds [32,
33]
2
√
(
) Denoting by ∆the largest degree of any vertex in T, one has the following bounds [32,
] ∆+ 1 ≤m2
N < ∆+ 2
√
∆−1,
(4.8) (4.8) and hence from (3.15), R ∼m−1d/
√
∆. Intuitively this makes sense since as one increases
∆(resp. d) whilst keeping d (resp. ∆) fixed, it is increasingly similar to K1,∆−1 (resp. Pd+1). Thus a theory will avoid the counter-intuitive behaviour described above as its
theory graph ‘grows’ if d grows more quickly than does
√
∆. As a further example, consider the ‘mace graph,’ Md,∆, which we define in figure 7;
this indeed has R ∼d/
√
∆. Curiously however, the fact that R decreases when ∆grows
quickly enough compared to d cannot in this case be blamed on there being more than one – 16 – graviton at, or close to, mass m1, as in the case of the star graph. We show in appendix A
that for d > 4,
3 m2
m1
> 3
2
(4.9) (4.9) (and in fact numerically we find m2 > 2m1), i.e. the separation between the first two
massive gravitons is always of order the mass gap of the theory, which is precisely the case
for the path graph theory (which has m2 ≈2m1.) 5
Conclusions In this paper we have considered the question of how the strong coupling scale of a multi-
gravity theory is related to its topology, i.e. to the presence and structure of interactions
between the different fields. We have shown that this can elegantly be rephrased in terms
of various properties of the corresponding theory graph. In particular we have placed – 17 – bounds (3.17) on the strong coupling scale. The parameter to which these bounds are
most sensitive is the diameter of the graph. One would expect that tighter bounds could
in principle be derived by taking into account further graph properties (of course at the
expense of complexity and/or generality). These bounds give us a sense of how the range
of validity of the theory behaves as one changes the graph and hence the theory. One
important finding is that the strong coupling scale of a multi-gravity theory can never be
raised simply by adding extra spin-2 fields and/or interactions between them. We have also considered when the expression (4.1) ((m2
1MPl)1/3) for the strong coupling
scale, which one gets by just considering the lightest massive graviton, is valid. The lightest
graviton mass turns out to be the algebraic connectivity of the graph, and fits easily with
the previously computed bounds. One also has to determine the physical Planck mass, and
we have considered doing this via appealing to a higher-dimensional interpretation of the
theory, and by considering the coupling of the theory to matter. JHEP01(2016)029 The latter approach implies that (4.1) is not universally valid, whilst the former yields
curious behaviour in which, for a fixed discretisation scale, the size of the nominal extra
dimension decreases as the size of the graph increases. Whilst we would not expect theories
which have more than one graviton at mass scale m1 to have a cutoffdescribed by (4.1),
interestingly it also does not hold in some theories which do have a separation between the
first two massive gravitons. It would be useful to further investigate the question of whether and how general
theory graphs might be related to higher dimensional theories, in the way that path and
cycle graph theories can. Our results here seem to indicate that the usual approach to
interpretation is not appropriate. Finally, we have only considered bi-metric interaction
terms, whilst using vierbeins one may write interaction terms which involve more than two
fields. Acknowledgments We would like to thank Claudia de Rham and Andrew Tolley for several stimulating dis-
cussions regarding this work. JHCS is supported by STFC. JN acknowledges support from
the Royal Commission for the Exhibition of 1851 and BIPAC. 5
Conclusions It would be interesting to extend this work to those theories. 16In this notation the ordinary barbell graph would be Bn,0,n. A
The spectrum of the generalised barbell graph The line graph of the graph depicted in figure 5 consists of two complete graphs of ⌊N+3−d
2
⌋
and ⌈N+3−d
2
⌉vertices connected by a path graph with d −3 edges. This is an example of
a ‘generalised barbell graph’, Ba,b,c, which we define to consist of a path Pb+2 which has
each endpoint in common with each of two complete graphs, Ka and Kc.16 – 18 – The adjacency matrix is The adjacency matrix is The adjacency matrix is A(Ba,b,c) =
Ja,a
0a−1,1 0a−1,b−1
1
01,b−1
0a,c
01,a−1
1
0b−1,a−1 0b−1,1
Tb
0b−1,1 0b−1,c−1
1
01,c−1
0c,a
01,b−1
1
0c−1,b−1 0c−1,1
Jc,c
−Ia+b+c,
(A.1) (A.1) where Jn,n is the n×n matrix of 1’s, and Tn is a tri-diagonal matrix of 1’s. The characteristic
polynomial ∆a,b,c obeys the following recurrence relation in a: where Jn,n is the n×n matrix of 1’s, and Tn is a tri-diagonal matrix of 1’s. The characteristic
polynomial ∆a,b,c obeys the following recurrence relation in a: JHEP01(2016)029 ∆a,b,c = |λIa+b+c −A(Ba,b,c)| = λ∆a−1,b,c + (a −2)∆(1)
a−2,b,c + ∆(2)
a−2,b,c,
(A.2) ∆a,b,c = |λIa+b+c −A(Ba,b,c)| = λ∆a−1,b,c + (a −2)∆(1)
a−2,b,c + ∆(2)
a−2,b,c,
(A.2)
where (A.2) where where ∆(1)
a,b,c =
λIa+b+c+1−A(Ba+1,b,c)−
λ + 1
01,a+b+c
0a+b+c,1 0a+b+c,a+b+c
! =−∆a,b,c +(a−1)∆(1)
a−1,b,c +∆(2)
a−1,b,c,
(A.3)
d and and
∆(2)
a,b,c =
λIa+b+c+1−A(Ba+1,b,c)−
λ+1
01,a
1
01,b−1+c
0a+b+c,1 0a+b+c,a 0a+b+c,1 0a+b+c,b−1+c
! =−(λ + 1)a∆0,b,c. (A 4) These recurrence relations can be solved, noting that ∆1,b,c = ∆0,b+1,c, to get 6)
∆a,b,c = −(1 + λ)a−2 ((a −1)∆0,b,c −(λ −(a −2))∆0,b+1,c) . (A.5) (A.5) (A.5) ∆a,b,c = −(1 + λ)a−2 ((a −1)∆0,b,c −(λ −(a −2))∆0,b+1,c) . (A.5) Now ∆0,b,c obeys the following recurrence relation in b: Now ∆0,b,c obeys the following recurrence relation in b: ∆0,b,c = λ∆0,b−1,c −∆0,b−2,c,
(A.6)
which, by using ∆0,0,c = (λ+1)c−1(λ−(c−1)) and ∆0,1,c = λ∆0,0,c−(λ+1)c−2(λ−(c−2)),
can be solved to give (A.6) (A.6) ∆0,b,c =
1
2b+3 (λ + 1)c−2
"
(c −1)
λ −
p
λ2 −4
b+2
+
λ +
p
λ2 −4
b+2
! + (λ(c −3) + 2(c −2))
λ −
√
λ2 −4
b+2
−
λ +
√
λ2 −4
b+2
√
λ2 −4
#
. A.1
Spectrum of L(Md,∆) One has LMd,∆= B∆,d−4,2,17 and hence One has LMd,∆= B∆,d−4,2,17 and hence spec(L(Md,∆)) = {0, 2(1 −cos θ∗
i ), 1d+∆−8},
(A.10) (A.10) where θ∗
i is a root of f∆,d−4,2(θ) = sin((d + 1)θ) + (∆−2)(sin(dθ) −sin((d −1)θ))
(A.11)
= (∆−1) sin(dθ) sin(θ)
cot(dθ)−
∆−3
∆−1cot(θ)−∆−2
∆−1cosec(θ)
. (A.12) (A.11) JHEP01(2016)029 From the square bracket we see that (n −1
2) π
d < θ∗
n < n π
d. Hence we have for d > 4 From the square bracket we see that (n −1
2) π
d < θ∗
n < n π
d. Hence we have for d > 4
m2
m1
≈θ∗
2
θ∗
1
> 3
2. (A. From the square bracket we see that (n −1
2) π
d < θ∗
n < n π
d. Hence we have for d > 4
m2
m1
≈θ∗
2
θ∗
1
> 3
2. (A m2
m1
≈θ∗
2
θ∗
1
> 3
2. (A.13) (A.13) Open Access. This article is distributed under the terms of the Creative Commons
Attribution License (CC-BY 4.0), which permits any use, distribution and reproduction in
any medium, provided the original author(s) and source are credited. Open Access. This article is distributed under the terms of the Creative Commons
Attribution License (CC-BY 4.0), which permits any use, distribution and reproduction in
any medium, provided the original author(s) and source are credited. 17Note there is a degeneracy in the final two parameters: B∆,d−4,2 = B∆,d−3,1 = B∆,d−2,0. A
The spectrum of the generalised barbell graph (A.7) (A.7) We finally change variables: λ = −2 cos θ, and substitute (A.7) in (A.5), and a little
algebra yields We finally change variables: λ = −2 cos θ, and substitute (A.7) in (A.5), and a little
algebra yields ∆a,b,c = (−1)b(1 −2 cos θ)a+b−4cosec θ fa,b,c(θ),
(A.8) (A.8) where where fa,b,c(θ) = sin((b + 5)θ) + (a + c −4) sin((b + 4)θ) + ((a −3)(c −3) −1) sin((b + 3)θ)
−2(a −2)(c −2) sin((b + 2)θ) + (a −2)(c −2) sin((b + 1)θ). (A.9)
fN,d in (3.7) is then just (A.9) with a = c = N+1−d
2
and b = d −4. – 19 – References [1] N. Boulanger, T. Damour, L. Gualtieri and M. Henneaux, Inconsistency of interacting,
multigraviton theories, Nucl. Phys. B 597 (2001) 127 [hep-th/0007220] [INSPIRE]. [1] N. Boulanger, T. Damour, L. Gualtieri and M. Henneaux, Inconsistency of interacting,
multigraviton theories, Nucl. Phys. B 597 (2001) 127 [hep-th/0007220] [INSPIRE]. [2] K. Hinterbichler, Theoretical aspects of massive gravity, Rev. Mod. Phys. 84 (2012) 671
[arXiv:1105.3735] [INSPIRE]. [2] K. Hinterbichler, Theoretical aspects of massive gravity, Rev. Mod. Phys. 84 (2012) 671
[arXiv:1105.3735] [INSPIRE]. [3] C. de Rham, Massive gravity, Living Rev. Rel. 17 (2014) 7 [arXiv:1401.4173] [INSPIRE]. [4] D.G. Boulware and S. Deser, Can gravitation have a finite range?,
Phys. Rev. D 6 (1972) 3368 [INSPIRE]. [4] D.G. Boulware and S. Deser, Can gravitation have a finite range?,
Phys. Rev. D 6 (1972) 3368 [INSPIRE]. [5] C. de Rham and G. Gabadadze, Generalization of the Fierz-Pauli action,
Phys. Rev. D 82 (2010) 044020 [arXiv:1007.0443] [INSPIRE]. [5] C. de Rham and G. Gabadadze, Generalization of the Fierz-Pauli action,
Phys. Rev. D 82 (2010) 044020 [arXiv:1007.0443] [INSPIRE]. [6] C. de Rham, G. Gabadadze and A.J. Tolley, Resummation of massive gravity,
Phys. Rev. Lett. 106 (2011) 231101 [arXiv:1011.1232] [INSPIRE]. [6] C. de Rham, G. Gabadadze and A.J. Tolley, Resummation of massive gravity,
Phys. Rev. Lett. 106 (2011) 231101 [arXiv:1011.1232] [INSPIRE]. [7] S.F. Hassan and R.A. Rosen, On non-linear actions for massive gravity,
JHEP 07 (2011) 009 [arXiv:1103.6055] [INSPIRE]. [7] S.F. Hassan and R.A. Rosen, On non-linear actions for massive gravity,
JHEP 07 (2011) 009 [arXiv:1103.6055] [INSPIRE]. [8] S.F. Hassan and R.A. Rosen, Resolving the ghost problem in non-linear massive gravity,
Phys. Rev. Lett. 108 (2012) 041101 [arXiv:1106.3344] [INSPIRE]. [8] S.F. Hassan and R.A. Rosen, Resolving the ghost problem in non-linear massive gravity,
Phys. Rev. Lett. 108 (2012) 041101 [arXiv:1106.3344] [INSPIRE]. [9] S.F. Hassan, R.A. Rosen and A. Schmidt-May, Ghost-free massive gravity with a general
reference metric, JHEP 02 (2012) 026 [arXiv:1109.3230] [INSPIRE]. [9] S.F. Hassan, R.A. Rosen and A. Schmidt-May, Ghost-free massive gravity with a general
reference metric, JHEP 02 (2012) 026 [arXiv:1109.3230] [INSPIRE]. [10] S.F. Hassan and R.A. Rosen, Bimetric gravity from ghost-free massive gravity,
JHEP 02 (2012) 126 [arXiv:1109.3515] [INSPIRE]. [11] S.F. Hassan and R.A. Rosen, Confirmation of the secondary constraint and absence of ghost
in massive gravity and bimetric gravity, JHEP 04 (2012) 123 [arXiv:1111.2070] [INSPIRE]. [11] S.F. Hassan and R.A. References Rosen, Confirmation of the secondary constraint and absence of ghost
in massive gravity and bimetric gravity, JHEP 04 (2012) 123 [arXiv:1111.2070] [INSPIRE]. [11] S.F. Hassan and R.A. Rosen, Confirmation of the secondary constraint and absence of ghost
in massive gravity and bimetric gravity, JHEP 04 (2012) 123 [arXiv:1111.2070] [INSPIRE]. 17Note there is a degeneracy in the final two parameters: B∆,d−4,2 = B∆,d−3,1 = B∆,d−2,0 – 20 – [12] K. Hinterbichler and R.A. Rosen, Interacting spin-2 fields, JHEP 07 (2012) 047
[arXiv:1203.5783] [INSPIRE]. [13] U. Aydemir, M.M. Anber and J.F. Donoghue, Self-healing of unitarity in effective field
theories and the onset of new physics, Phys. Rev. D 86 (2012) 014025 [arXiv:1203.5153]
[INSPIRE]. [14] C. Burrage, N. Kaloper and A. Padilla, Strong coupling and bounds on the spin-2 mass in
massive gravity, Phys. Rev. Lett. 111 (2013) 021802 [arXiv:1211.6001] [INSPIRE]. [15] N. Arkani-Hamed, A.G. Cohen and H. Georgi, (De)constructing dimensions,
Phys. Rev. Lett. 86 (2001) 4757 [hep-th/0104005] [INSPIRE]. JHEP01(2016)029 JHEP01(2016)029 [16] C. de Rham, A. Matas and A.J. Tolley, Deconstructing dimensions and massive gravity,
Class. Quant. Grav. 31 (2014) 025004 [arXiv:1308.4136] [INSPIRE]. [17] C. de Rham, G. Gabadadze and A.J. Tolley, Ghost free massive gravity in the St¨uckelberg
language, Phys. Lett. B 711 (2012) 190 [arXiv:1107.3820] [INSPIRE]. [18] M. Fasiello and A.J. Tolley, Cosmological stability bound in massive gravity and bigravity,
JCAP 12 (2013) 002 [arXiv:1308.1647] [INSPIRE]. [19] J. Noller, J.H.C. Scargill and P.G. Ferreira, Interacting spin-2 fields in the St¨uckelberg
picture, JCAP 02 (2014) 007 [arXiv:1311.7009] [INSPIRE]. [20] C. de Rham, M. Fasiello and A.J. Tolley, Galileon duality, Phys. Lett. B 733 (2014) 46
[arXiv:1308.2702] [INSPIRE]. [21] J. Noller and J.H.C. Scargill, The decoupling limit of multi-gravity: multi-Galileons, dualities
and more, JHEP 05 (2015) 034 [arXiv:1503.02700] [INSPIRE]. [22] N. Arkani-Hamed, H. Georgi and M.D. Schwartz, Effective field theory for massive gravitons
and gravity in theory space, Annals Phys. 305 (2003) 96 [hep-th/0210184] [INSPIRE]. [23] T. Hanada, K. Kobayashi, K. Shinoda and K. Shiraishi, Classical and quantum cosmology of
multigravity, Class. Quant. Grav. 27 (2010) 225010 [arXiv:1004.5435] [INSPIRE]. [23] T. Hanada, K. Kobayashi, K. Shinoda and K. Shiraishi, Classical and quantum cosmology of
multigravity, Class. Quant. Grav. 27 (2010) 225010 [arXiv:1004.5435] [INSPIRE]. [24] J.H.C. Scargill, J. Noller and P.G. Ferreira, Cycles of interactions in multi-gravity theories
JHEP 12 (2014) 160 [arXiv:1410.7774] [INSPIRE]. [25] C. Yan, Properties of spectra of graphs and line graphs,
Appl. Math. J. Chin. Univ. B 17 (2002) 371. References [26] M. Doob, An interrelation between line graphs, eigenvalues, and matroids,
J. Combinat. Theor. B 15 (1973) 40. [27] R. Brualdi, The mutually beneficial relationship of graphs and matrices, in Regional
conference series in mathematics, American Mathematical Soc., U.S.A. (2011). [27] R. Brualdi, The mutually beneficial relationship of graphs and matrices, in Regional
conference series in mathematics, American Mathematical Soc., U.S.A. (2011). [28] N. Arkani-Hamed and M.D. Schwartz, Discrete gravitational dimensions,
Phys. Rev. D 69 (2004) 104001 [hep-th/0302110] [INSPIRE]. [28] N. Arkani-Hamed and M.D. Schwartz, Discrete gravitational dimensions,
Phys. Rev. D 69 (2004) 104001 [hep-th/0302110] [INSPIRE]. [29] M.D. Schwartz, Constructing gravitational dimensions, Phys. Rev. D 68 (2003) 024029
[hep-th/0303114] [INSPIRE]. [30] C. Deffayet and J. Mourad, Multigravity from a discrete extra dimension,
Phys. Lett. B 589 (2004) 48 [hep-th/0311124] [INSPIRE]. [31] N. Kan and K. Shiraishi, Multigraviton theory: a latticized dimension and the cosmological
constant, Class. Quant. Grav. 20 (2003) 4965 [gr-qc/0212113] [INSPIRE]. – 21 – [32] R. Grone and R. Merris, The Laplacian spectrum of a graph II,
SIAM J. Discrete Math. 7 (1994) 221. [33] D. Stevanovi´c, Bounding the largest eigenvalue of trees in terms of the largest vertex degree,
Lin. Alg. Appl. 360 (2003) 35. [33] D. Stevanovi´c, Bounding the largest eigenvalue of trees in terms of the largest vertex degree,
Lin. Alg. Appl. 360 (2003) 35. [33] D. Stevanovi´c, Bounding the largest eigenvalue of trees in terms of the largest vertex degree,
Lin. Alg. Appl. 360 (2003) 35. [34] C. de Rham, L. Heisenberg and R.H. Ribeiro, On couplings to matter in massive (bi-)gravity,
Class. Quant. Grav. 32 (2015) 035022 [arXiv:1408.1678] [INSPIRE]. [35] J. Noller and S. Melville, The coupling to matter in massive, bi- and multi-gravity,
JCAP 01 (2015) 003 [arXiv:1408.5131] [INSPIRE]. [36] S. Melville and J. Noller, Generalised matter couplings in massive bigravity,
arXiv:1511.01485 [INSPIRE]. JHEP01(2016)029 – 22 –
|
W4388557732.txt
|
https://journal.uokufa.edu.iq/index.php/jall/article/download/6648/6234
|
ar
|
بديع الترتيب في القرآن الكريم دراسة دلالية جمالية
|
مجلة اللغة العربية وادابها
| 2,013
|
cc-by
| 8,641
|
( ٣٢٣ ) ..............................................................................ﺑﺪﻳﻊ ﺍﻟﱰﺗﻴﺐ ﰲ ﺍﻟﻘﺮﺁﻥ ﺍﻟﻜﺮﻳﻢ
ﺑﺪﻳﻊ ﺍﻟﱰﺗﻴﺐ ﰲ ﺍﻟﻘﺮﺁﻥ ﺍﻟﻜﺮﻳﻢ ﺩﺭﺍﺳﺔ ﺩﻻﻟﻴﺔ ﲨﺎﻟﻴﺔ
ﺍﳌﺪﺭﺱ ﺍﻟﺪﻛﺘﻮﺭ
ﺧﺎﻟﺪ ﻛﺎﻇﻢ ﲪﻴﺪﻱ
ﺍﻟﻨﺠﻒ ﺍﻷﺷﺮﻑ-ﻛﻠﻴﺔ ﺍﻟﺸﻴﺦ ﺍﻟﻄﻮﺳﻲ ﺍﳉﺎﻣﻌﺔ
:ﺍﳌﻘﺪﻣﺔ
و،ب ا دب ا( من وا وام)ا: اّ
آن ا ر وأا ااص ا ا درا ا أداة إا
وا ج ا إم ة دا ء اء ا ل و.ا
ن وآمز اه ا و وا اذ و، وا
. ا وا اماره وأ
و،ر ة أ اا أنّ اإ
. ا أداة ا و ا ي أ ارا
إ ومّ :ول ا: اث ا ا ا ن اأ
م ؛ أو ا أو ا ّ ه و و ا
م وت دز و إذ ار، م اثا ا
اا وام ام: وا. وا ا ا رزةا
و ص ار ت وراح وت و م
ات و اث ب اا ا ُ أن أو
. ن
ن :ول ،ابدة ا ا أن ظو
ا و،وق وادب وا ا و،ة اا ير اا
ر ا ا ةات ا ة و ا
ا ا اوان ا : وا، وو
عم إذ إنّ ا،ا ا د ا ا إ أن م وإم،
. ا ا ة د ر
( ٣٢٤ ) ..............................................................................ﺑﺪﻳﻊ ﺍﻟﱰﺗﻴﺐ ﰲ ﺍﻟﻘﺮﺁﻥ ﺍﻟﻜﺮﻳﻢ
ءد ا دة وذ، ا ا وا
ا ، مدت إ أو و وروا
ر ء ا ا ورة ّ وء أو اه ا ن أن
ة د و،
ّ ا أن ط ّ ر
.اث ا ا أ ا، ا ا
را ا ا ات ا ا وا
ة اض م ، م ذب ا ،ء واا
رات ا إ،ا ا فن ا ،
روا ن ا أ.ف واء ا وء ا
،نا ا ل ذا ا ،ا و ءا
.ا
د درا،آن ا ا ا):ا ا أ و
اث ا ض ا ي :ول ا ،(
ص م : وا،ة ا ا و واا
. اآم اار ا أ ف آم
ﺍﳌﺒﺤﺚ ﺍﻷﻭﻝ
ﻣﻔﻬﻮﻡ ﺍﻟﱰﺗﻴﺐ ﰲ ﺍﻟﱰﺍﺙ ﺍﻟﻔﻠﺴﻔﻲ ﻭﺍﻟﺒﻼﻏﻲ
،ك أي:ّ و، رّ ء اّ ر (١) ا ا
لد (١٢.ي)ت قم ا ي اا ا ر و، أي أ ّور
ء
ّ ء و ا ))ا: ادء ا ا ّ ا
ا أمرة إ إ و ا ا ، ا
ن و و ، ا ر ا ا
ت ء ا ا أن ا .(٢)((
، إذا ت ا م ،ا
.إرادة وو
ةء ا ا)) يح ا ا اأ
( ٣٢٥ ) ..............................................................................ﺑﺪﻳﻊ ﺍﻟﱰﺗﻴﺐ ﰲ ﺍﻟﻘﺮﺁﻥ ﺍﻟﻜﺮﻳﻢ
ء ا ذا ... وا إ م ن و،ا اا
ّء ان ا
ّ أ .(٣)(( ن ذ ء وا إم م ا يا
أم اء ا وا إ أو رة ا ة
اء ء ا ة ا و ت ا اة ما
.اء اا
ا ، ء ا ا ا ّ أن ا
ا )) را م اClaparede ر ا
ا ن إذا ، ّب ا ء ا
.(٤)((ص نم يق اس ا أ
م ا إذ ا ّ أوا
أن م ا ء ا ا ، وا اء إا
ا(؛ و٥) د و إ دون ا ا ا ا
. و، ، ء واا
و،(٦)(٦ )قت اا ا ح ا ا أ
رد ف وا ف أو إ ا أن)): ّ
و ا ا ن و، ا ه و أو وا
ل ن ا أو ا ا زا و ظ ا
: ا
(٧)
اٍ ا ٍ
ء
.(٨)(( ا إ ا نم ا رف اون ا
، ةّه ظ و ( )ا (٩١١)تطد اوأ
فرد أو ))أن: ا ّو
و،(٩)((ا زا و و، ا فا
ٍَ ْ ِ
ٍُُمْ ُِب
ٍ َاُْ ِ
ْ َ ِياَُ: و. ذ (١٠)يما
َ اأََُُِْوََُِْْْ َ
ْ َِْ وً ُُ اُمَِ
ُ
ْ ُاأَُِْ
ُ
ِ ط
ْ ُ ُْ
ُ
( ٣٢٦ ) ..............................................................................ﺑﺪﻳﻊ ﺍﻟﱰﺗﻴﺐ ﰲ ﺍﻟﻘﺮﺁﻥ ﺍﻟﻜﺮﻳﻢ
.(١١)َنَََِْْ وَُ
إذ. ل ا ا و ر د ، ما ا و
، ا أ (Minesis)ة ا ، ا عا ا
ا ان ا ب أن و إ ر ة أن ا ام
، ا ا م دون (١٢)ة ا إ بأ
، ذي ا د أو،ة ار د )) اع
ّ أن ،(١٣)(( ر ة و فا
ا؛ ودب ا ا رودة ا دت أ ن و،
.ة ا ا ي ا ا ا أرادوا أمّم
ا ا( م٦٨١ن)ت ا ذو
ح ي م ا ا ّ ،ا مم ا
: ا ا ا ا
ا ا
وا ا ّ
(؛١٤)ر د وأا ا ن ر ا
. ء أ أ ي أم ظة ا ادا
ا ح ا ي ا اا ا وأ
:
ْ ا
ا
ْ ا وا
ا
(١٥)
اأ
ْ تأ
،ة ا يت ا أ ا أو
ا و ا ،عّع ا ي اوء ا ل ا
ا م ا ا و، ا
أ و ذ ( إا/و)ا،(/):اا
( ٣٢٧ ) ..............................................................................ﺑﺪﻳﻊ ﺍﻟﱰﺗﻴﺐ ﰲ ﺍﻟﻘﺮﺁﻥ ﺍﻟﻜﺮﻳﻢ
ا ا وح اد م ة اة ا وط
.ر ة
أن ،ة ا( م. ق٣٢٢٣٨٤) Aristotleّر أر طا او
ا ،اة ا ا ا دة ا ا ع اا
ا ا ، ا ا
مَِْْاِ
ٍ َْ رِ
ْ ُْ ْنسإ
ُ اَأَ: .(١٦)ء ا
ُِِْ وَمٍٍََََُُِْ وٍٍَََُُُِْْ ٍُُِْْمٍَُِابََُِْْ
ْ
َِْوَََُْْ
َِْْو
ُ اأَُُِْْ ِط
ُُْمُ
َ َُأإُءََمََ ْرا
ذ م ّ ا إذ.(١٧)ًَْ ِ َََُُِِِْْْْ ارْذَلأإَدَُْ
ة ا ا نمر ا ا
ء وارء وا ا ء اورة ا
رة ا ا إّ ا ا و، ا ت ا ،تا
ء ر أ ،(نم)ا ّ و و را اا
، ا إ ة ا إ ا ة ا ة ر ،(اب)ا
ا ،ف ا ا زوال ، ا إ وب ا او
ا م و ار ا ارد اآن اف ا ن
. ذ إ و ور و ن ا و،(... ر )إن:ا
ا ، و وظ ب أ أن ا او
م :(١٨٣١١٧٧٠) Hegel ل ي ادب ا ا إ
ّ اه؛ ّ وم ا ن ا ّرهي م ا))ا
أى ى ح ا إ ء ا
ّ ا ا واف اأو
.(١٨)(( اظ
إ، ن وا ا ا م مو
آنف ا أو و، ا و ّر دا ر ااا
ا وي وع م ءا ا
( ٣٢٨ ) ..............................................................................ﺑﺪﻳﻊ ﺍﻟﱰﺗﻴﺐ ﰲ ﺍﻟﻘﺮﺁﻥ ﺍﻟﻜﺮﻳﻢ
ء ن أو
ّ أ ا و، ز اّ إ
ا أن ااط واف او وم م و را
ر ط ن ، اى ا ا
:ﺍﳌﺒﺤﺚ ﺍﻟﺜﺎﻧﻲ
.ا ا اا
ﺗﻄﺒﻴﻘﺎﺕ ﻇﺎﻫﺮﺓ ﺍﻟﱰﺗﻴﺐ ﰲ ﻧﺼﻮﺹ ﻗﺮﺁﻧﻴﺔ
ِ َِ*س اَبُِذُقلْأ
ُ : ا ظت ات ا ا
.(س ا إ،س ا ،س)رب ا: ذ ا م أي،(١٩ )سِ ا*إسا
،(ب)ا ت ث ذ ا ءا ا
ا ت ا ا و،( )ا ت و،()ا و
ة ، ء ة د اة ا ّ أ
ي اة ا لء ا ،
ا ا و.(٢٠)ت ا اا
وأ،نم ا ب أم؛ب أو ان))ذ :(١٤٠٢)تطه اذ
، و و ه و وأ، أم ا ،و
.(٢١)((دي ا ط إ نمه ا وم؛ا
ام، س ا رج ا إ ا ا او
ّ )) ا ، اة إ ا رجن ا و، اط
و إ ،اء ّ ب و ا
ن
ّ ؛ ّ ا وأ. ا و، وم َ
إ ا ن وإن، إ أن ج م أ اه وموا
.(٢٢ )((اه
%١٠٠ ا س ا رج اد م أيا او
ّ أن ذ، و ا ام وإذا،% ا ا
ت آدا إ ا، ود داد ان ا ،ورة ب ق
( ٣٢٩ ) ..............................................................................ﺑﺪﻳﻊ ﺍﻟﱰﺗﻴﺐ ﰲ ﺍﻟﻘﺮﺁﻥ ﺍﻟﻜﺮﻳﻢ
ِيبِ وَاَِتُ اَ آِا: ،ن س ا أن أ إ ة
س
ا
َ َأَِو: و،(٢٣)َنُُِْ سَ اَأَِ وَ اَْ رِْلَإمأ
.(٢٥ )ََُُُِِْْنَأ ََ وَ َِ ذِنإ: و،(٢٤)َنُُِْ
ا س ا ب م
ّ إ (س )اف إ ا ر إ ا
ّ اه ا و،ا
،(٢٦)د بر أن ا ذ،ف إ ا اة إ ا رج وا
ّ ذا ل ، و ب
ّ ر
ّ و،ّ ربب و
ّ ار ر ن
د د ل و،(٢٧)ِ اْ دُونًَِْرْأَُمَُْْ وَرَُرَْ ُواأَا:َبرد ا
ْجعْإ
ِ ار: ل و،(٢٨)َنُُِ اُِ ُمَايَإَََْْ أَُ رمإ:
ن
ون ا ا ا ن ا و.(٢٩)ََُِْأَِْةِ اْلُ اَََُْر
ك وا، وب و
ّ أي، ّ رب وا
ّ ،نّن وو
و، واّد ا ، وا ةا او ن م ّ؛أ
وا وا،
ّ ،دون م ا
.(٣٠) اة إ ا رج د ا او،كا
ة إ ن أن َز ا ،س ا ؤه ا م و
ا و، دت ن
ْ وإ،ةا ا ا ا إ
،(س ا وإ،س ا و،س)رب ا: ،او ا ازا
.(٣١) واب وا ا ، دة ذوات أم
أن ، ا ا م ا إ، ا إ وإذا ام
ا إ ا ظ ا ا أ ا ا ا
،ر وا))ا:م اطف ا ا ا و،Coherence
اه ا و،(٣٢)(( را إت و ا و
و، و ن))ر (٧٥١)ت ال ا
و، ، ا ا، ا ا،ب ا ا ، وإ
( ٣٣٠ ) ..............................................................................ﺑﺪﻳﻊ ﺍﻟﱰﺗﻴﺐ ﰲ ﺍﻟﻘﺮﺁﻥ ﺍﻟﻜﺮﻳﻢ
ه اه ا و،ه ا ، وا،
.(٣٣)((ق وأع م أ ظ اا
ل ا ّ ا م ، اا ا و
ا و،ّب ا أ)): (٦٠٤ازي)ت ار إ أ وذ،ّوا
ٍ ه اه وأّ أن ر إ م أوا و،ه وإ
دة أن ا ، ا ّ ،ُِ َ و،ك أم ف
ا ، إف أم دة ا ده ف أن و، وا ز
.(٣٤)
((
إ آن ا ط د اا ا ّ إن ا ل او
.(٣٥)دة وا ا ار س ا ا
ة ّ إذ،ة أدّى وظا ا أن ل او
ا ا،ظ ام أ ا د امم اا
و. ا ا أ إ ة ت ّ ي ا،صا
،ده و و ا ن أو ا ات إه ا
ا ت ام ظ ا اط دة ز رج ا
لم ل ا دا وط دة طرج ا إذ،ا
:(٣٦) م ، ر وذ، ا إ اا
. : رب أ
. + : + َ : ب
. ط + :ه + إ : إ ج
و دة وظ ا سا ا و
و ا ي ا ا د توام
.(٣٧) ت اا وا ا ا
ء أن م ات ا ا و
دا ء ُ وأن ِاض أن ا د؛ اب اأ
( ٣٣١ ) ..............................................................................ﺑﺪﻳﻊ ﺍﻟﱰﺗﻴﺐ ﰲ ﺍﻟﻘﺮﺁﻥ ﺍﻟﻜﺮﻳﻢ
وَا: ا و، ا ر ا ظ
.(٣٨)ِِ اََا اَََ*وِِ ر*وَطنُْوَا
ا ا و،رة آم ا ا ون أنول ا
ةر اف ا أ ن( إم وا )ا (٣٩)ا ا اا
ا و، ا د أ ا دوا وا
ا د د و اا ا و،و ا ا
ا ي ا ا( و ر)وط: وام ا ا
؛k ا، ا ي( ا اا ا)و،
ونل ا ، ا ا ا اان ا نن او
ء أمان إ نم إم:ا ر ا ن وا ا
ّ
م "؛ر ز وط ر"ط:م ل رض ا ا ن))
ل "ن و"ا.انان و ل ""ا: و،ن وا ا
،(٤٠)((...ن وا او: م و، م ؛ا
اه د( و٥٣٨ي)ت ا ا و، ا( و ر)ط
و،م ز ا ر ا ا و، ا
ل ،ه ا ا ا أا
ءم ا ز
ظ و،رع اف ا مء اه ا ا))و:يا
ن وا ا ،ء وام ا وا ا
ن : و، ديي من ار ا وا،ه و واإ
ي ا و.(٤١)((...k ل ا ر و ى ي اا
و واإ: ءم ا وا ا و
.ا ا غ د إذ، k و
ا اب ا ه(و٥٤٦)تم ا ول ا و
))ا: م ا ا ه رأ( و٦٨س)ت ا
( ٣٣٢ ) ..............................................................................ﺑﺪﻳﻊ ﺍﻟﱰﺗﻴﺐ ﰲ ﺍﻟﻘﺮﺁﻥ ﺍﻟﻜﺮﻳﻢ
" رن وط وا"ا: و،ا إ ن وا،ح م
،ب ا أ ""ا:ط ا ل و، ا
أم " ر "ط وأ،[س ا] إ "نو"ا
.(٤٢)(( ا
ا ا ،ر ا ع اب ا ّ أن مو
ه س ا ا ى آراء أ راء ا ا أا
ءة "ا ر إ ا ا ا ا ،ا
ا ا))وا:(٧٤٥)تمن ا ل ا "نوا
.(٤٣)((..." ر"ط ُ و ذ ،ا اا
)ا: رت إ أراء ا ا ا وإنو
و وا إ: ء أوم (نوا
(ن وا)ا: وس ا أن ا ،ا إ ءأم
ا ال اط س ا ا
و دا ا،ء ا إ يع ارض وا ا ا ا
ٍَْأََةٍ طََ َ طَِ َََُبَ اَ َْ ََْأ: ة اا
أدم أم إُ وإنه ا و،(٤٤)ِءَ اٌََُِْ وَِ
ى أ رو إ ( ا ر )ط إذا وازم،ا
،(٤٥) ا أ ي ا ا ،ء رضا
و.( اا ا)و: آ ء رض ا أ
ء اب ا ل
ّ يا( ا)رة ا ا قط ا
رة إ ا ا () انّ ا و، ا
.(٤٦) رب ا الرة ا ا
ر زات ا ه عه ا ار ء ام ا او
( ٣٣٣ ) ..............................................................................ﺑﺪﻳﻊ ﺍﻟﱰﺗﻴﺐ ﰲ ﺍﻟﻘﺮﺁﻥ ﺍﻟﻜﺮﻳﻢ
ر ، k زع ا وأ،ى ا ا
ن اء دون ذ ، ظراط إ ا أدم
اع أ اع اثد أ و ، ا أ أا إ أ
ر أي، ن ا رهم رة ام ى
. آ أي مرة إع اا
تء ا ا ي ا ات ا ا و
ره ا ا ورا اب ا اءن امر ا ت را
و، ا ه ا آ و،دة ا
.ت ا إ
ء وا،(ّ و،ء)ا ا ا ا أدوات ا او
ْنِأ َِْ آ َِو: (٤٧)ا ا () و،اخ أو ا
ُ*ِ طْ ِنَم اَ َ أَََو: و، (٤٨)َُونََِ
ٌ ََ
ْ ُْمذَاأإ
ُب
ٍ َاُْ ِ
ْ َ
آن ا ا .(٤٩)ِِ ْ رُوَِِِ َاهُ ومَ ُ*ٍَءٍَُِْ ََُِْْمََ
مَِْْ اٍَِْ رُِْْ ْنسُإ اَ أَ: دة ا ااا
ُِِْ وَمٍٍََََُُِْ وٍٍَََُُُِْْ ٍُُِْْمٍَُِابََُِْْ
ْ
َِْوَََُْْ
َِْْو
ُ اأَُُِْْ ِط
ُُْمُ
َ َُأإُءََمََ ْرا
.(٥٠)
ًَْ ِ َََُُِِِْْْْ ارْذَلأإَدَُْ
،ا وا ا ا( ا) ث ا أنّ ام
ّغ و،ى وأ ة ات ز د ادط او
، م إا تم ا ة ا فا ا لا
م () ل ،( و،أ): ل اء ا و
: ا اا وا
( ٣٣٤ ) ..............................................................................ﺑﺪﻳﻊ ﺍﻟﱰﺗﻴﺐ ﰲ ﺍﻟﻘﺮﺁﻥ ﺍﻟﻜﺮﻳﻢ
اا
ف
اء أو اا
ا
ا
اا
ون إذا أم
ّ
اب
ٍ
ُ ِْ ُ
ْ ََ م َ
َ
ّ
نم ا أ -٧:ةا
َ ٍءَ
ْ ِ
ْ ِ ِ ِ
َ ا ُه َوَمَ
٢٠:وا
ط
ّ
ء ا
م
ّ
اب
و
ّ
م
ّ
و
٩
ِِرُو
٥:ا
ط
ذ ا غ ا ا او
ا،نمر ا (ُْأ اُ َْ ِ ُ):اءا
؛ ق إرادة ا ة ا و:وا
ّ وم، و،): ا اه ا ا ا
م و ا و ام ا ا إ و،( و،ر
د،ًُُ اُمَ ِ ُ
ْ ُأ اُ َْ ِ ُ : وذ،ن مما
ء م ، اام د إ ام
ي اوف ا ء ا و.ة ا نمء ا إي و
و،ا أ أي،(٥١)((ام و، )):
(ام) (ا)ال ا إ م ه ا و. ام
:
ن ام إرادة ا ، اام و:غ ا ١
م ةرب ا ب ا وا
( ٣٣٥ ) ..............................................................................ﺑﺪﻳﻊ ﺍﻟﱰﺗﻴﺐ ﰲ ﺍﻟﻘﺮﺁﻥ ﺍﻟﻜﺮﻳﻢ
ا ن طمّ ا ، واا
.ةرب ا رادة إذاوا
ا( دم)م ا ّ و:غ ا ٢
أن ا ،رادة ا رجّل ا ورة واا
، () و،)واو( ا (ن)ّم
ا اّل إث ا
إ ر و م ن؛م ا
)واون ا ،ب ا إ عوا
ل
ّ و،ء ا ّ أي،دا (ا
ن ، د أم
،ا ود (ا أ):
ا و، ا غ ا ّن ذ
اق ا (ل++):اب اإ
ة ا ، اظ ا دةا
ري إرادة ا إ ا دورة اا
ا ا إم وم
أو،اب ا: و ا أ اء اا ا
و مم ا ا و،(ء)ااب واا
ر إ أن ا ا
ة ا إ وأاط و وو
إدم ون إم ا أن
ل ، م م ا اوراء ا
ا ا ن إذام)):(٦٧١)تطا
ال أة و رآ م ا
وأوانل م أم ذ...
ى م و، أ ذ
( ٣٣٦ ) ..............................................................................ﺑﺪﻳﻊ ﺍﻟﱰﺗﻴﺐ ﰲ ﺍﻟﻘﺮﺁﻥ ﺍﻟﻜﺮﻳﻢ
ل ة إب وال ا م و
ي ا أم ... واا
م وم م وأن، ل ا
و م ال أ ذ و،ل ل إ
.(٥٢ )((
ى وا أ آ م اب اع إت وا ة دورة او
ُهَََ ُ*ِ طْ ٍُِ ْ َِنَم اََ ْ َو: أ ا اءورة اا
ً
ْ َِاَمْ ًَِ َُْ اََََُْ اَََ اََ
ُ*َِر
َاُِم
َََِِ اَْْمإُ*َنَََُِ ذَْْمإُ*ََُِِ اَْ ُأ ارَكََََ آَ ُهَمَمأُ
.(٥٣)نَُُْ
،ت ا ،ة)ا:د ا ّ آ ث ةل دورة او
ر اس وا ا س اْ ( م ةا
.(٥٤ )
((اعء وام ا ار ا)) د،را
،(ء)ا ىة أ( و) ة ال م ة ا ا وم
(١٠٦٩)ت اب ا ا ار أ ذو
(ّ) اطف ا))وا: (٧٩١وي)ت ا
ة ن
ّ أ(٥٥) ا ارد أنّ ا ء ا وا ا اا
ة أو ا ( إن م) ا أن ن أر ّ ا
.(٥٦)(( ء إن م و،و
وا، () ات ا ه إذا وازم او
ا ا وأ.()اءف ا أ( أو) اء ا
و ّ ))أنت ووت ا لا
ن ؛ ا واا ا أو ر د ا () فا
ء د ا ا ا و،ا ا اء أ ل ا
و ا و،رةن وا ا ا ف .أ
( ٣٣٧ ) ..............................................................................ﺑﺪﻳﻊ ﺍﻟﱰﺗﻴﺐ ﰲ ﺍﻟﻘﺮﺁﻥ ﺍﻟﻜﺮﻳﻢ
إُ إذا م أم ،(٥٧)((ه ا
ّ ا و...
وج): ا و،ء ،و م (٥٨)ل اا
ى أل إ ّلة ا ول ا ط إُ إذا م أ.( ُ
ة (ََ آَ ُهَم َم أُ ):() (١٤) ا ام ()
.( ء ا ):ت اا
ُ َْْمُأَمَُْْْمأنَ*أَُُُْْْأَأ: ار ا و
ََُِِْْْ وَمَْلَأَُنْمأَ*َََُُِِْْمََْتَ وَْ اَََْرْمَُْنَ*مَِا
َُْْمُأَمَُْرَُْْمأنَ*أَََُُُْْْأُونَ*أَْو اةْْ اَُِْ َْنَ*وَُْ
ِيءَ اَْ اُْأَأ*َنُُوَُْ
ْ َْمَ*َنُ
َ ُْمنَ*إََُ
ْ ًُُ ُهَََُءََمْ *َنُارا
ر
َ ْ اُْأَأ*َُونَْْ ًَُهأَََ ُءََمْ *َنُاُ َْمْ نأُْاْ ُِهَُُمأ
ْ ُْمأنَ*أََُْ
.(٥٩)ًََََُِ وَةَََََِْ َُْنَ*مُُُِ اَْْمأَََََ َُْمْأُْمأرُونَ*أُِا
ء ذ و،نم ا دة ا ا م ذُ ترت ا
وا،اء ا ن
ّ ؛ ا م أ ،(٦٠)د و
ّ ر د أ زقن ا و،ء ا و ا
ن
ّ وب؛ ا ،اء ا م ؛ل أو ا: ر أ و نما
اَُْا وَا: ل ،ل ا ب ا ا و،ار ا
نج وام ا ح ا ر ا ا ذ ،(٦١)َنَْ َ ُْ َِ ًَِ
و وإذ، ب وا و ،ءةوا
.ء ا ذ ا ءاء واا
ا أم
ّ ظ نمد ا ا رق ا ء ّ ذ
و،ب ا ّ م ب أ ّ ده م و
ن ار وت ا ا إرع و ا م ورر او
وام،(ا) :ن وب ا نما
، ء ا ،ءد ا ب ذات ،(٦٢)()دوا
( ٣٣٨ ) ..............................................................................ﺑﺪﻳﻊ ﺍﻟﱰﺗﻴﺐ ﰲ ﺍﻟﻘﺮﺁﻥ ﺍﻟﻜﺮﻳﻢ
ا ا و،ء ا وف ا ا ،
أراد ا ا ا د أو و و،ء ا
م ؛ ا ا ا ا
إنّ ا: ل ،(ت ز ،واء أو ام ا ا) :ل ل
ا و.ء ( ظ) أو،بواء أ وا ا إ،( ) ا
ن ا
ّ أ ا نم .ارة ا تت آّ را
ب أو ا ُ إذا إ، ا ءهن أ
ّ أ ؛ ا
،ار ود ارع وإت ام ا أم و. ا أ
:، ن ا إاد ا ت ارت ا أو
ن ا
إدراك ا را
ا
َْتَْ ا
َََْرم
ْ ُ
ْ َم
َنَُِ ا
ْ َمْ أَُمَْ
ْ ُْأمأ
َنَُُُْْْأَأ
ًُ ُهَََُءََمْ
ُ
ْ َمْ أَُمُر
َ َْ
ْ ُْأمأ
َنُر
اا
َنََُُُْْْأَأ
ُ
ْ َمْ أن
ْ ُْ اُِهُُ
َ أمُْْأمأ
ِيءَ اَْ اُْأَأ
َن
ُا
َنََُْ
م ء م):
ُ
ْ َْمأََََ
َ َُْمْأُْأمأ
ِرَاْ اُْأَأ
.رعج ا م ا،(
َنُُِا
َرُونُ
ًَهُأَََ ُءََمْ
ه أن م ي أراد ا ا ا ة ا ا مو
ا ر اه ا ّن ؛ب ا و ،تا
ائ ا و، ا ا إ ا ا، اا اأم
،ت ا أو ا ودوا اا ة( أيرادة ا)ا
ل )إرادة د إ ا ا أنّ ان إذن إا
، م اق ا إء أن م ا ،(ا
ث إ ةن ا
ّ ؛ ا ن ا
ّ ؛ ل ذوم
( ٣٣٩ ) ..............................................................................ﺑﺪﻳﻊ ﺍﻟﱰﺗﻴﺐ ﰲ ﺍﻟﻘﺮﺁﻥ ﺍﻟﻜﺮﻳﻢ
.ق ا أ أ و،أ
س
َُِز:ان رة آل ورد أر ا و
ُعََِْثَِذَ وَاَْم
ِ وَاََُ اَِْ وَا
ِِ وَاَ اْ َِِةُْا
ِطَ وَاَ َِِء وَاَ اْ ِت
ِ َاَ ا
ُ
.(٦٣)ِبَُ اُْ َُهِْ ُ وَاَْمةِ اََا
د (َِاتَ اُ س
ِ
َ ُ)ز:ظ ا ر ا ا ّإن
، أء ا ا ام ن أنّ ا وا
. ا ام
اء ي( ا)ف ا ء ا أ ا ن ا
ّ أو
ماه ا و.ث ء وام ا أت ا ا و،(٦٤)ا
ن م ،اب إة ا وم ااه ا ا ا
وظه ا وع ا ُ أة ، أ ُ اوظ
إ ا أ. ب م و أظ
أ ُ ي ا ا ّم ؛ ا ، ة أةا
أة ا إذ، ظ ا ا ةح ا ،أةا
(٦٥)أةن ا ذاك و،ة ا ا و،ىع و
:
ّ ُا و و،ن أ ن و،ن
ُْ رَاوَدَو: و،(٦٦)ٌَِ َْ نإ ِْ ْ ُِمَلإ
ُُ دْ ِ َُِىرَأ
ِإَ ُأْ الَ رَب*ََِ اَِنََ وََُُُْهُ آَََََِْْْ وَََِِْْْمَ
ل ال ر و،(٦٧)ََِِْ اِْ وَأُْإْأَُْ َ ْفَِْ وَإْإَِمَُْ
.(٦٨)((ل ا ء ا أ ّ أ ي )):k
ازاء ا وامن ا اق ا ّو
ء ا رة أ ادات ا ن د
ّ َاتِ(؛َ اُ سَُِ)ز:
ءج ا من ا
ّ (؛٦٩) ا ذ ذ ، واا
ََِْ
ِزوَا
ْ ْأِْ
ََََوًزوَا
ْ ْأِمأْ ِ
ْ َََ ُوَا: ل أوا
و ا ا اء ا ءن ا م و،(٧٠)ًَةَ َو
( ٣٤٠ ) ..............................................................................ﺑﺪﻳﻊ ﺍﻟﱰﺗﻴﺐ ﰲ ﺍﻟﻘﺮﺁﻥ ﺍﻟﻜﺮﻳﻢ
َ أَ: ّ ا ّر ا و،م ة ا
ء ا .(٧١)َُْرُوْ ْاُوَ ِْدْوْ وَأِ زْوَاْأِناإََُ آِا
ةن ا م ن ،دوال وا ا ا ذ و.دوا
،(٧٢)ٌَِ ٌْ َهُأِْ َ ان وَأَُِْْدْوْ وَأَاْأَماأُْوَا:
ا أي، أ أ رج اق ا دو ا الا
ء اان ر رة آل آ ،ء ا ا
ان ا و،ء ا ط يء ا ن اما
ود إ أ (٧٣)ا ا إد داو
إ ور وا ا ظ ا وو ةن ام ارؤ
.(٧٤)ذ
ط وا، ا دوء وا (٧٥)(ة اط )او
() و. ق ة أو اة أي ا وو،ل اا
،(٧٦)(( ل ا وط ن)) أو ( وا ا):
،ء ا م ؛ م إ واوا
و، اَ م نم ا ا وا،(٧٧)ء ا
رةل ا ا وظُ ا و،ل ا ن م
أي ا: ا ووظ،(٧٨)ًََُِْ اََ ِْ اَاْءَأَا اَُُْو:
ا نم ا ة ت د و، ا
اد أ ت ا ا و ا وإم، ا ا
. ذ إ ا وا
ّ ا اا
و ، ن أن ا ال ا دونو
رغف ا واان ا م :ك ، ط
د و،ّ اء ا د و ا إ،ا
. رون ف أ ُ م ه أُ ور
،حر ا ر وا ا را م روا
ا ط وا أ ،م ذ و
( ٣٤١ ) ..............................................................................ﺑﺪﻳﻊ ﺍﻟﱰﺗﻴﺐ ﰲ ﺍﻟﻘﺮﺁﻥ ﺍﻟﻜﺮﻳﻢ
ط د ا ط ،سب ا ر ور
.ء واه وا ا ، او إ ا ة
ن و،(ث)ا
َ ّ ارااء ا م نوا
.وم وارب ا ر
إة ا اطي ا وا ا ال ا
ّ و
د وا وا ا ،د اا ف ا أ
ض ا آء إ ا ير دم إط ذ ،اد اأ
.ت ا
ء ا ي م( ا٧٩)ط رأي ا ا ا موأ
،م وا، ا وا، وا)ا:ل ا ف أر
ا و.( وا،ادي ا وأ،ك وا،ر)ا: ّل،(ثوا
ا م ّ و،ا ور ة ا ا ل
دل ا ا ه و، ف ا إو
رف اه ا ا أ و، ا ا ء ا
ة وإم اط ا ه و ف أ و
ّ ا،ال و ك ا : و، دم مو
. ل واا وا، ا: و.(٨٠) ا ا
ء ا ا اك و ا ل ّ و ا او
ء م وا، اا ّ ء فت ا ا
. ا
ّ ت ا واا
)) ي إ ا و،م ا،ل وإم ا
م ا ن ا ،(٨١)((:دهن أو ّل َمو...ح و وط
رة ل أن م و، ا وط هل و ا
.و ا
ورا اراك اّ ا ث او
، روح ا وإد ا ادي إ و ا و
.اد ا أ و وا ا
( ٣٤٢ ) ..............................................................................ﺑﺪﻳﻊ ﺍﻟﱰﺗﻴﺐ ﰲ ﺍﻟﻘﺮﺁﻥ ﺍﻟﻜﺮﻳﻢ
ا ا م اا ا و
دي ا ا ّ ز ،ن ام و اأ:(َ ُل)زا
ن
ّ و،ن ا ن ا ا ا
ّ و، وظ
س ّ ز ،نم إرادة ا رن ا ن طا
ة ا واه ا وأ، اتّ ا
ن ءه ا ة ا ، وظ وا افا
ءه ا ا اف ا ، ل ا و،ا ب ا
اط اّ ،اه ده إ نم ا ا ى ا
ا ل ا د إ د واو ا ء ومّ ا
وط ّ ، و ل ا و
ذا ،ور ا ر هُ ء زا ه او
ع ار إ ّ نمّل ا ا و. ّ
ّف ّ إ، اع ا ام ا ظ
د ء اء ا ا ،ءه ا
و ا ر ا م م ( و٨٢)ا
ي اات ا ،ءاب و ءد ا أنّ و
ازم ا إ ا ا .ت ا ده و ا و، ذا
ا ا وي ا واي ام ا:د ا
ن ات ا و.(٨٣)ِبَاُ ُْ َُهِْ ُ وَاَْمِة اََُعاََ
َِذ:
k لن ا ة ود ا را
ََِِ َِ ُرَْم
اََِِْْيَْت
ٌ َ
ْ َ رَ ِْ ْا اَ ِِ
ْ َِ ذْ ِ
َِْ
ْ َُؤُمأْ :
ا ة ا ا و.(٨٤)ِدَِِ
ٌ َُِ وَاِ اْ َِاٌنْة وَرَُزْوَاٌجوَأ
ي اع ا إ طاه و ّ ين ام د ا ا
.ي ا
:ﺍﳋﺎﲤﺔ ﻭﺍﻟﺘﻮﺻﻴﺎﺕ
: أ م إ ا
( ٣٤٣ ) ..............................................................................ﺑﺪﻳﻊ ﺍﻟﱰﺗﻴﺐ ﰲ ﺍﻟﻘﺮﺁﻥ ﺍﻟﻜﺮﻳﻢ
د ي او أنّ ا و،ء ا ة َ ا ١
فن ا ، ة اض م ذإ
و، أ أن:ل ءه أو اا
و نا ا و،اء وا ا ا ا ر
.ا و ا
ة أن ا ام، ا أ ة ا ا ٢
إ ب أا ان ا ب أن و إ ر أ
ّ أن ، ا ا م دون ،ةا
.ة ا ا ي اك ا
ِ ،
م ا وا اا ا ة اؤت ا أ٣
ء ا ر ا ، ُ ا اا
ا اب ات و اا ، وا
.ة اه ا
ء إذ،طتِ ا آن ا ا أن ا ارا ا ٤
د أمو ا ا وإن،آم ا ا ن م ادا و
م وإ ار ا إظ ي ا ا و، ت اد
.ا
ا آن ا ا اام ا ام ا -٥
او ر درارس ا ان ا ؤوا
. وا اظ و ن راه دواط
ﺍﳌﻠﺨﺺ
را ا ا ات ا ا ا
ن ، ة اض م ،ء واا
ء ا وء ا رات ا إ،ا ا فا
ا ،ا و ء ا روا ن ا أ.فوا
. ا ،نا ا ل ذا
ض ا ي :ول ا ا ا أ و
( ٣٤٤ ) ..............................................................................ﺑﺪﻳﻊ ﺍﻟﱰﺗﻴﺐ ﰲ ﺍﻟﻘﺮﺁﻥ ﺍﻟﻜﺮﻳﻢ
ار ا أ ف آم ص م : وا
. اآما
: أ م إ ا و
د ي او أنّ ا و،ء ا ة َ ا ١
فن ا ، ة اض م ذإ
.ا و و ا نا ا و، ها
ة أن ا ام، ا أ ة ا ا ٢
إ ب أا ان ا ب أن و إ ر أ
ّ أن ، ا ا م دون ،ةا
.ة ا ا ي اكِ ا ،
م ا وا اا ا ة اؤت ا أ٣
ء ا ر ا ، ُ ا اا
ا اب ات و اا ، وا
.ة اه ا
ء إذ،طت ا
ِ آن ا ا أن ا ارا ا ٤
د أمو ا ا وإن،آم ا ا ن م ادا و
م وإ ار ا إظ ي ا ا و، ت اد
ا
Abstract:
The arrangement is considered one of the rhetorical terms that did
not go through enough study by the ancient and modern studiers
because of the ambiguity within the theory of simulation by many of
them. This made them ignore this term except for some of the signs by
the interpretation scholars and the scholars of similarity and
difference. Concerning the modernists, they followed the steps of the
ancient ones in theory and application, therefore, they did not add
anything important for not having modern research methods.
Here lies the significance of the study which came into two parts,
the first is theoretical for the purpose of identifying the arrangement
and the other is an application to the Quran texts to reveal some of the
secrets of its miraculous arrangement.
( ٣٤٥ ) ..............................................................................ﺑﺪﻳﻊ ﺍﻟﱰﺗﻴﺐ ﰲ ﺍﻟﻘﺮﺁﻥ ﺍﻟﻜﺮﻳﻢ
The research reached some results, the most important of which
are:
1- The arrangement did not receive enough study by the scholars of
the Arabic Language. The reason that led to this is the ambiguity of
simulation theory for many of the scholars which made them ignore
many of its signs. The modernists followed the steps of the ancient
ones in theory and application.
2- The rhetoric of the arrangement is related to the simulation theory
concerning its aesthetic aspect based on the fact that the simulation
in its simplest images necessitate that the creative work should be
more like life and that it is a simple relation depicting reality as it
is. In other words, not every arrangement is a creative work if it
does not simulates the essence that arouses meditation in life.
3- The new vision of arrangement helped deciding the analytical rules
and uncovering the profound inspirational meanings that has never
been revealed before because of the interrelation of the terms and
the absence of the methods followed in revealing the meaning of
this rhetorical phenomena.
:ﻫﻮﺍﻣﺶ ﺍﻟﺒﺤﺚ
.١٢٨/٥:(دة)ر ،ر ا،بن ا :( ظ١)
.٤١١/١:يم ا،ن وات اف ا (٢)
.٤١٢/١:ن. (٣)
، ا ت : ظ،٢٦٧/١: ، ا( ا٤)
.١٠٥:زيزاق ا ا
، ا؛ ا٤١٢/١:يم ا،ن وات اف ا :( ظ٥)
.٢٦٧/١:
در ا ل إ،ادر ا م (٦)
(٥٢٦) برة ا إ ي ا، ا ي ا ا
د و إ أه و ب ا ا و،(٥٥٨) و
ن ا و، ه ا و ما
أ أ ب ورد م و،ر آم ا ا اا
ا ن ن ج ون أ :ا
دسن ا ا أواه إ ع إر ل و. اا
( ٣٤٦ ) ..............................................................................ﺑﺪﻳﻊ ﺍﻟﱰﺗﻴﺐ ﰲ ﺍﻟﻘﺮﺁﻥ ﺍﻟﻜﺮﻳﻢ
رت إ أ اش ا إ ا م أ.يا
ن و،ش ا ء و ا م در اا
و، ر ة و وَل
ر آ و و،ارعو
ب ا ء و،ب وا أ ذ ب ا:ظ.ان ا نا
ور ا : ظ،٥٣:١٨٥٧ ،ا ا،(٢٨٧ي)تة ا ،وا
ر ا،٥: ا س أ ا أ،اس اارك ا
.١٠٢/٢:ي ا س أ ا أ،ب واا
: و،٣٢٥:ام ا انح د (٧)
ا ا د اء
.٦٠/٤:ي ا ا،رب ادب و اام (٨)
.٤٤٤/١:ط ا،آن ا ن( ا٩)
.٤١١/١:يم ا،ن وات اف ا :( ظ١٠)
.٦٧: رة (١١)
ا، و١٥٦: و ، و درا، اا:( ظ١٢ )
.٥٧: ،ل ا
.٣٣٨:ا إ ز. د، ا ا ا (١٣)
.٢٣٧/٣:ن ا،نت ا و:( ظ١٤ )
.١٩٩٩/٣:يان ا( د١٥)
ة وا ا،١٦٨: و ، و درا، اا:( ظ١٦)
.٥٨-٥٧:ه ا . د، وااث اا
.٥:رة ا (١٧)
.١٨٣: ،مو واي ا ا( ا١٨)
.٣-١:سرة ا (١٩)
. و١٧٢: و ، درا، ا ا:( ظ٢٠)
.٤٥٩/٢٠:ط ا،آن ا ان( ا٢١ )
.٦١٢/٨: ا،رت وا ا رر ا( م٢٢ )
.١:رة ا (٢٣)
.١٧:د رة (٢٤ )
.١٩٠ و،١٧٤ و،١٥٨ و،١٣٩ و،١٢١ و،١٠٣ و،٦٧ و،٨:اءرة ا (٢٥)
( ٣٤٧ ) ..............................................................................ﺑﺪﻳﻊ ﺍﻟﱰﺗﻴﺐ ﰲ ﺍﻟﻘﺮﺁﻥ ﺍﻟﻜﺮﻳﻢ
.٨١٧/٤:ي ا،ف ا:( ظ٢٦)
.٣١:رة ا (٢٧)
.٢٣: رة (٢٨)
.٥٠: رة (٢٩)
.٤٨/١:ا ا .د،م ا ا ط :( ظ٣٠)
،٤٦٦:ز ا ا، ا ا،٥٣٤/٨:من ا ا، ا ا:( ظ٣١)
.٦١٣/٨: ا،رت وا ا رر ام
.٤٧١:ط ا،آن ا ن( ا٣٢)
.٤٦٧-٤٦٦:ز ا ا، ا( ا٣٣)
.١٩٦/١٦:ازي ا، ا (٣٤)
،ري وا ا رر ا م،٤٦٦:ز ا ا، ا ا:( ظ٣٥)
.٦١٣/٨:ا
، ا ا ،٥٠-٤٨:م ن ،ات ا ا دروس:( ظ٣٦)
ا.د،،ت ارا ا ا ا،١٧٣:ن .د
.٢٧٩-٢٧٨:ا
.٥٨-٥٧:ر رو،تم ا :( ظ٣٧)
.٣-١:رة ا (٣٨)
ا،ب ا ا ر ا ا،٧٦٣/٤:ي ا،ف ا:( ظ٣٩)
ن ا، ا ا،٨/٣٢:ازي ا، ا ،٤٩٩/٥:
.٤٧٠/٨: ا،رت وا ا رر ا م،٤٨٥/٨:ما
ا،ب ا ا ر ا ا: ظ،٧٦٣/٤:ي ا،ف( ا٤٠)
ب اا إ اد ا إر،٨/٣٢:ازي ا، ا ،٤٩٩/٥:
.٦٤٥/٦:د ا أ،ا
.٧٦٣/٤:ي ا،ف( ا٤١ )
، ا : ظ،٤٤٩/٥: ا،ب ا ا ر ا( ا٤٢ )
،٤٤٦/٦:د ا أ،ب ا اا إ اد ا إر،٩/٥:ازيا
.٣٩٤/١٥: ا،مروح ا
.٤٨٥/٨:من ا ا، ا( ا٤٣)
.٢٤:ارة إ (٤٤ )
( ٣٤٨ ) ..............................................................................ﺑﺪﻳﻊ ﺍﻟﱰﺗﻴﺐ ﰲ ﺍﻟﻘﺮﺁﻥ ﺍﻟﻜﺮﻳﻢ
.٨٣-٨٢:زي ن .د،آن ا ا ع وا ا :( ظ٤٥)
.٨٣-٨٢:ن. :( ظ٤٦)
،موف ا ام ا ا،٢٥٠:م ا ا ،ز ا د:( ظ٤٧)
.٤٢٦ و،٦١:اديا
.٢٠:ورة ا (٤٨)
.٩-٧:ةرة ا (٤٩)
.٥:رة ا (٥٠)
.١٨٢/٤:ي ا،ف( ا٥١)
.٢٠٢/٢:ط ا،آن ا ( ا٥٢)
.١٦-١٢:نرة ا (٥٣)
.١٨٢/٤:ي ا،ف( ا٥٤)
ن أر أ ن أ
ّ ))إ: k ل ا ر ( ورد٥٥)
،وح ا ا إ ذ ن ذ ذ
ه إ يا . أو و و وأ رز :ت ر و
ذراع إ و ن ا أ إن أ
ر ا أ وإن أ، ر ا أ با
.( ا أ ب ا ذراع إ و ن
.(٢٦٤٣) ح،١٢٥/١:ري ا ،
ب ،ا ا و ا ة ا،وي ا ب ا (٥٦)
.٥٦٣/٦: اا
.١٩١/١٨:رو ا ا ،مآن واب ا إ:ظ،٥٦٣/٦:ن. (٥٧)
.١٩٠/١٨:رو ا ا ،مآن واب ا إ:( ظ٥٨)
.٧٣-٥٨:ارة ا (٥٩)
ارك ا ،٤٦٦:طم ا،وك ا ،١٨٠/٢٩:ازي ا، ا :( ظ٦٠)
.١٥٠/٢٧: ا،م روح ا،٢١١/٤: ا،و او
.١٩:ررة ا (٦١)
.٦٤٨/١ : . د، اا:( ظ٦٢)
.١٤:ان رة آل (٦٣)
.٥٩٧:م ا،وف ام ،٣٤٧/٤: ،ب ا:( ظ٦٤)
( ٣٤٩ ) ..............................................................................ﺑﺪﻳﻊ ﺍﻟﱰﺗﻴﺐ ﰲ ﺍﻟﻘﺮﺁﻥ ﺍﻟﻜﺮﻳﻢ
ا،٢١٠/٧:ازي ا، ا ،٢٨٣/١:ي ا، ا :( ظ٦٥)
.٩٩/٣: ا،م روح ا،٢٩/٤:ط ا،آنا
.٢٨: رة (٦٦)
.٣٣-٣٢: رة (٦٧)
.٢٠٠/٥: أ، أ (٦٨)
ا ّن ؛ رد او ا أم إ ا ( ذ٦٩)
ء ان ذ
ّ أ إ ذ ،م ا أا
.٩٩/٣: ا،م روح ا،٢١٠/٧:ازي ا، ا : ظ.ا
.٧٢:رة ا (٧٠)
.١٤:رة ا (٧١)
.٢٨:لمرة ا (٧٢)
.٢٥٢/٢: ا،ن ا :( ظ٧٣)
.٢١٠/٧:ازي ا، ا :( ظ٧٤)
ن وأ: ا و،را أو أون د وأ: رار ا ( ا٧٥)
أو،را ا ن أم: و،ر د أو أ،رف دة أ: و،ل
:ظ.ل ا ا ر أن ا د ره ا و ا رط
، ا ،٤١١/٢: ا،نا،٢٧٤/٣:ي ا،نا
.٢٥٣/٢: ا،ن ا ،٤١٦/١:ي ا،فا،٢٨٤/١:يا
.١٠٠/٣: ا،م( روح ا٧٦)
.٢١١/٧:ازي ا، ا :( ظ٧٧)
.٥:ءرة ا (٧٨)
،س ا ّل ل ا ع م ،ل ا ف أ ))أرط ا( ذ٧٩)
وأ،ك ا ّل ا ا وأ،ر ا ّل وا اأ
ا.(( ا أ ّل ث ا وأ،ادي ا أ ّل ما
.٣٦/٧:ط ا،آن ا
، ا ،٤١٢/٢: ا،ن ا،٢٧٦/٣:ي ا،ن ا :( ظ٨٠)
٢٥٣/٢: ا،ن ا ،٤١٦/١:ي ا،فا،٢١١/٧:ازيا
.٤٤٦/٥:(دة)م ،رس ا، ا (٨١)
.١٨-١٧: ر ، وا ا :( ظ٨٢)
( ٣٥٠ ) ..............................................................................ﺑﺪﻳﻊ ﺍﻟﱰﺗﻴﺐ ﰲ ﺍﻟﻘﺮﺁﻥ ﺍﻟﻜﺮﻳﻢ
.١٤:ان رة آل (٨٣)
:ﻗﺎﺋﻤﺔ ﺍﳌﺼﺎﺩﺭ ﻭﺍﳌﺮﺍﺟﻊ
.١٥:ان رة آل (٨٤)
.آن ا ا
دار ا،(٩١١)تط ا ا ل ا ،آن ا ن ا.١
.(٢٠٠٧/١٤٢٨)١ ،ن ،وت،ا
د ا أ،ب ا اا إ اد ا إر.٢
دار ا، ا ا ا و،(٩٨٢)تدي اا
.(١٩٩٩/١٤١٩)،١،ن ،وت ،ا
و ،ي ا س أ ا أ،ب وار ا ا.٣
.١٩٥٤ ،ءار ا ا،ب دار ا،ي اي و ا
دار ا، وا ا،رو ا ا ،مآن واب ا إ.٤
.(١٩٩٩/١٤٢٠)،٧،وت د، وا
و درا،(٧٤٥)تمن ا ا ، ا ا.٥
،وت ، ا دار ا، وآ،د ا دل أ ا،و
.(٢٠٠٧/١٤٢٨)٢،ن
أ و ،(٤٦٠)ت ا ا أ،()ن ا.٦
.(١٩٦٣/١٣٨٣)ف ا ا،ن ا ، ا
ل دار ا،ه ا . د، وا ا ا ة وا ا.٧
.١٩٨٠ ،١ ،ن ،ا ط،ز وا وا
ب أ أ ا أ ا ، ا ا.٨
دار ا، ا ،وي ا أو ،(٧٥١)تزا
.(٢٠٠٧/١٤٢٨) ١،ن ،وت،ا
،(٣١٠ي)ت ا أ،آن آي او ن ا .٩
.(١٩٥٤/١٣٧٣)٢ ، ،ده وأو ا ا
( ٣٥١ ) ..............................................................................ﺑﺪﻳﻊ ﺍﻟﱰﺗﻴﺐ ﰲ ﺍﻟﻘﺮﺁﻥ ﺍﻟﻜﺮﻳﻢ
،(٦٧١)تطري ام ا أ ،آن ا ا.١٠
.(٢٠٠٤/١٤٢٤) ،٢،ن ،وت ، ا دار ا،ريا
،(٧٤٩ادي)ت ا ا ،موف ا ام ا ا.١١
ن ،وت ، ا دار ا، م ذا،وة ا
.(١٩٩٢/١٤١٣)١،
ب ،ا ا و ا ة ا،وي ا ب ا .١٢
اد وأج آ و،(١٠٦٩)ت ا أا
.(١٩٩٧/١٤١٧)١ ،ن،وت ، دار ا،يزاق ا ا
وف ا، أ،رب ادب و اام
.١٣
،در دار،ب د . د و درا،(٨٣٧ي)تا
.(٢٠٠٥/١٤٢٥)١،ن،وت
ا،دي ا ،م ن ،،ات ا ا دروس.١٤
.١٩٦٦، وادث ات وارا ا م،ما
م ا ا ا أ،ز ا د.١٥
،ة ،م دار ا، د : أه و،(٤٧١)ت
.(١٩٩٢/١٤١٣)٣
، ،رف دار ا، ا و و ،يان ا د.١٦
.(ت. )د،٢
ا ا أ،م ا واآن ا ا مروح ا
.١٧
ءدار إ، ا ا و، أ أ ،(١٢٧٠ادي)تا
.(١٩٩٩/١٤٢٠)١ ،ن ،وت ،اث اا
، ا س أ ا أ،اس اارك ا ور ا
.١٨
ا،س ن إ. د ،(ر )اّ ال ا
.(١٩٨٠/١٤٠٠)١ ،ت وارا ا
( ٣٥٢ ) ..............................................................................ﺑﺪﻳﻊ ﺍﻟﱰﺗﻴﺐ ﰲ ﺍﻟﻘﺮﺁﻥ ﺍﻟﻜﺮﻳﻢ
رد ا ،زي ن .د،آن ا ا ع وا ا .١٩
.(٢٠١١/١٤٣٢) ،١ ،اق ا،ف اا
،(٢٠٨ري)تم ا ا ،ام ا انح د .٢٠
.(ت. )د،٢ ، رفدار ا،ن ا . د وا
،(٢٦١ري )تي اج ا ا ا أ ، .٢١
اد : ود وأا وأ و ور و م
.(ت. )د اء ا دار إ ،ة ا، دار ا،ا
،داب وا ا ،ا ا . د،م ا ا ط
.٢٢
.(٢٠٠٢/١٤٢٣) ،را
، دار،ا إ ز. د، ا ا ا
.٢٣
.١٩٦٦،ةا
دار،ارج ط:،،موا،ا،ي ا ا.٢٤
.١٩٨٦ ،٢ ،وت،ا
ا وو ،(١٨٠()ت) ن و ،با
.٢٥
،ن،وت ، ا دار ا،ب أ. د،رو
.(١٩٩٩/١٤٢٠)١
،(١٢ ق.ي)تم ا ، ن وات اف ا
.٢٦
إر ا ا م،وج د . د ، ا . دااف ووإ
،ن ،رج ز . د، ا ا،ي ا ا . د،ا
.١٩٩٦ ،١ ،ن ،وت
د ا،ون ا و ا فا
.٢٧
دار،يزاق ا ا: م ،(٥٣٨)تارزي اا
.(٢٠٠١/١٤٢١) ،ن،وت، ار ا،اث اء اإ
( ٣٥٣ ) ..............................................................................ﺑﺪﻳﻊ ﺍﻟﱰﺗﻴﺐ ﰲ ﺍﻟﻘﺮﺁﻥ ﺍﻟﻜﺮﻳﻢ
ريمر ا ا أل ا ،بن ا
.٢٨
و، أ ا وو و ،(٧١١ي)ت اا
.(٢٠٠٦/١٤٢٦)،١ ،ن،وت ، ا دار ا،ا إ ا
أ) ا ا ا أ،آن ا ن ا
.٢٩
ا ، اء وا ا و ،(٦ق
.(١٩٩٥/١٤١٥)١،ن ،وت ،ت ا ، اا
ا أ،ب ا ا ر اا
.٣٠
، ا دار ا، ا ا ،(٥٤٦ )تما
.(٢٠٠١/١٤٢٢) ،١،ن ،وت
،(٧١٠)تد ا أ ا،و ا وارك ا
.٣١
، ا دار ا،ات ز اد وأج آ و
.(١٩٩٥/١٤١٥)١،ن،وت
.١٩٩٣ ،ردن ا،ن ، ،ل ا ا
.٣٢
ا،،يدر ا ا . د ،نر رو،تم ا
.٣٣
٢٠٠٧ ،١ ،وت ، ا
.(ت. )د،وت ،در (دار٢٤١)ت أ، أ
ا ا (٥٤٠ي)تد ا ا، ا
.٣٤
.٣٥
.(١٩٩٧/١٤١٧) ،٤ ،ز وا دار ط،وآ
،(٣٨٤ي )ت ام ا ا ،وف ام
.٣٦
وق دار ا، ح إ ا . د، ه وا جو
.(١٩٨٤/١٤٠٤) ،٣ ،ة ، وازوا
.١٣٨٥ ،١ ، ،ادهم ، . د، اا
دار ا،زيزاق ا ا ، ا ت
.٣٧
.٣٨
.٢٠١٢ ،١ ،ن،وت ،ا
( ٣٥٤ ) ..............................................................................ﺑﺪﻳﻊ ﺍﻟﱰﺗﻴﺐ ﰲ ﺍﻟﻘﺮﺁﻥ ﺍﻟﻜﺮﻳﻢ
دار،(٦٠٤ازي)ت اء ا ا ، ا
.٣٩
.(١٩٨٥/١٤٠٥)،٣،ن ،وت،ا
ا : ،(٣٩٥ )ت ز رس ، ا .٤٠
.(ت. )دز وا وا دار ا،رون
أ، آي ا ا ا د واوي ا ط اوك ا .٤١
ا و،(٧٠٨)تطم ا ا ا ا إ أ
.(٢٠٠٦/١٤٢٧)١ ،ن،وت ، ادار ا، ا ا
،ءار ا ا، ر رع ، دار ا،ن . د، ا ا .٤٢
.(١٩٧٩/١٤٠٠)با
ر ا . د،ف ا ا ة رؤ، ا اا
.٤٣
.(١٩٨٠/١٤٠٠) ا ،
وت ،ت ا ،ط ا ،آن ا ان ا.٤٤
.(١٩٩٧/١٤١٧)ن ،
ا إ ا أن ا،رت وا ا رر ا م.٤٥
،ي ا زاق ا ا وود وأج آ ،(٨٨٥)تا
.(٢٠٠٣/١٤٢٤)٢،ن،وت ، ادار ا
،اد ز . د، و ، و درا، اا
.٤٦
.١٩٨١ ،٢،وت ،ت وارا اا
ن إ. د ،(٦٨١ن)ت ا،نء اء أن وأمت او
.٤٧
.(ت.)د،ن،وت ، دار ا،س
|
|
https://openalex.org/W4386636387
|
https://ard.bmj.com/content/annrheumdis/early/2023/09/12/ard-2023-223955.full.pdf
|
English
| null |
HLA-DRB1 and HLA-DQA1 associated with immunogenicity to adalimumab therapy in patients with rheumatoid arthritis
|
Annals of the rheumatic diseases
| 2,023
|
cc-by
| 2,696
|
HLA-DRB1 and HLA-DQA1 associated with
immunogenicity to adalimumab therapy in
patients with rheumatoid arthritis RA diagnosis (HR=1.02, p=0.01, table 1). Compared with
TNFi monotherapy, combination therapy with csDMARD
reduced the rate of ADA development by more than twofold
(HR=0.379, p=1.27e−07). Importantly, a statistically signif-
icant difference in the rate of immunogenicity was observed
when MTX cotherapy was compared with cotherapy with
alternative csDMARDs; MTX conferring higher protec-
tion from immunogenicity (HR=0.425, p=1.27e−05). However, non-MTX csDMARD use also trended towards a
reduced rate of immunogenicity (HR=0.66; 95% CI 0.429
to 1.012, p=0.056). RA diagnosis (HR=1.02, p=0.01, table 1). Compared with
TNFi monotherapy, combination therapy with csDMARD
reduced the rate of ADA development by more than twofold
(HR=0.379, p=1.27e−07). Importantly, a statistically signif-
icant difference in the rate of immunogenicity was observed
when MTX cotherapy was compared with cotherapy with
alternative csDMARDs; MTX conferring higher protec-
tion from immunogenicity (HR=0.425, p=1.27e−05). However, non-MTX csDMARD use also trended towards a
reduced rate of immunogenicity (HR=0.66; 95% CI 0.429
to 1.012, p=0.056). Advanced targeted therapies including tumour necrosis factor
inhibitors (TNFis) have transformed the clinical management
of rheumatoid arthritis (RA). However, monoclonal antibody
(MAb)-derived TNFis are associated with development of immu-
nogenicity resulting in low circulating drug levels (online supple-
mental figure S5).1 A genetic predictor of immunogenicity would
have clinical utility by providing a pretreatment biomarker that
could be used to inform therapy selection. Previous genetic
studies of TNFi immunogenicity have focused on alleles within
the HLA locus on chromosome 6.2–4 on October 23, 2024 by guest. Protected by copyright. http://ard.bmj.com/
ublished as 10.1136/ard-2023-223955 on 12 September 2023. Downloaded from p
Following quality control of the genetic data, 166 HLA
alleles were available for analysis in 435 patients with non-
missing covariate data. The most statistically significant
association with immunogenicity was observed for HLA-
DQA1*03 (HR 0.6; 95% CI 0.474 to 0.775, p=6.4e−05)
and HLA-DRB1*04 (HR 0.6; 95% CI 0.476 to 0.775,
p=6.3e−05) (4-digit and amino-acid results are reported
in online supplemental material S1). In the Kaplan-Meier
analysis, carriage of HLA-DQA1*03 and HLA-DRB1*04
alleles under an additive model was associated with reduced
rate of immunogenicity (figure 1A–C). The two HLA alleles
were in LD (R2: 0.94),5 suggesting a single protective
effect. In carriers of at least one copy of HLA-DQA1*03
or HLA-DRB1*04, MTX was observed to provide stronger
protection against ADA development compared with other
csDMARDs (HR 0.44; 95% CI 0.24 to 0.78, p=5.7e−03,
figure 1B–D). HLA-DRB1 and HLA-DQA1 associated with
immunogenicity to adalimumab therapy in
patients with rheumatoid arthritis We also investigated HLA alleles that have
previously been reported on in RA and Crohn’s disease
and provide support for alleles at HLA-DQA1*05, HLA-
DRB1*11 and HLA-DRB1*03 (online supplemental figure
S4). Patients were followed for 12 months with serum samples
collected at 3 months, 6 months and 12 months following
commencement on adalimumab (TNFi) therapy. Neutral-
ising antidrug antibodies (ADAs) were detected using a
drug-sensitive/drug-tolerant radioimmunoassay (Sanquin,
NL). The presence of ADAs was determined by radioimmu-
noassay. A positive ADA titre was defined as >12 arbitrary
units/mL. If a patient developed ADA at any time in the
study, they were classed as ADA positive. Genotyping was
carried out using the Illumina array, and HLA alleles were
imputed using SNP2HLA and the T1DGC reference panel
following standard data quality control (full details in online
supplemental S1). Drug immunogenicity rates were deter-
mined using Kaplan-Meier analysis, and Cox proportional
hazards regression, which was used to adjust genetic models
for biological sex, age, concurrent conventional synthetic
disease-modifying antirheumatic drug (csDMARD) use,
disease duration and first within-sample principal compo-
nent from the genetic dataset. In conclusion, in the largest study of its type in RA to
date, carriage of HLA-DQA1*03 and HLA-DRB1*04
reduced the rate of drug immunogenicity to adalim-
umab. The strongest protection from immunogenicity was
conferred by csDMARD cotherapy, particularly in combina-
tion with MTX. Our results suggest that the use of alterna-
tive csDMARDs should be encouraged for patients treated
with MAb TNFi who are MTX intolerant. Larger studies are
now needed to determine if genetic testing could optimise In total 445 patients were studied, of whom 96 (21.6%)
became ADA positive during treatment. A total of 377
(85.3%) patients received cotherapy with csDMARDs of
which 302 (81.4%) patients received methotrexate (MTX,
online supplemental table S1). *Comparison within patients with complete MTX information, those with missing information were not included in this analysis. mokers and current smokers.
ide antibody; ADA, antidrug-antibody; BMI, body mass index; csDMARD, conventional synthetic disease-modifying antirheumatic drug ; DAS28, disease act MTX information, those with missing information were not included in this analysis.
ving known combination therapy, as well as complete MTX information.
k ent smokers.
DA, antidrug-antibody; BMI, body mass index; csDMARD, conventional synthetic disease-modifying antirheumatic drug ; DAS28, disease activity score in CPA, anti-citrullinated peptide antibody; ADA, antidrug-antibody; BMI, body mass index; csDMARD, conventional synthetic disease-modifying antirheumatic
joints †Comparison within recorded patients of having known combination therapy, as well as complete MTX information.
‡Ever smoker refers to ex smokers and current smokers. p
,
g
y
d patients of having known combination therapy, as well as complete MTX information.
mokers and current smokers s with complete MTX information, those with missing information were not included in this analysis.
d patients of having known combination therapy as well as complete MTX information Letter Letter MTX information, those with missing information were not included in this analysis. †Comparison within recorded patients of having known combination therapy, as well as complete MTX information.
‡Ever smoker refers to ex smokers and current smokers.
ACPA anti-citrullinated peptide antibody;ADA antidrug-antibody; BMI body mass index; csDMARD conventional synthetic disease-modifying antirheumatic drug ; DAS28 disease activity score in ‡Ever smoker refers to ex smokers and current smokers.
ACPA, anti-citrullinated peptide antibody; ADA, antidrug-antibody; BMI, body mass index; csDMARD, conventional synthetic disease-modifying antirheumatic drug ; DAS28, disease activity score in HLA-DRB1 and HLA-DQA1 associated with
immunogenicity to adalimumab therapy in
patients with rheumatoid arthritis Disease duration modestly
increased the rate of immunogenicity for every year since Table 1 Cox regression output for the clinical attributes, where N is the number of samples available within each variable
N
P value
HR
ADA negative
ADA positive
Concurrent csDMARD usage
442
1.27e−07
0.38 (0.26–0.54)
354 (80%)
88 (20%)
Methotrexate (MTX) usage*
371
1.93e−05
0.41 (0.28–0.62)
312 (84%)
59 (16%)
MTX versus other csDMARD†
377
1.27e−05
0.43 (0.29–0.62)
315 (84%)
62 (16%)
Concurrent csDMARD (excluding MTX)
143
0.06
0.66 (0.43–1.01)
95 (66%)
48 (34%)
First biologic
444
0.88
0.95 (0.53–1.73)
356 (80%)
88 (20%)
Age
445
0.18
0.99 (0.97–1.00)
357 (80%)
88 (20%)
Sex
445
0.29
1.21 (0.85–1.73)
357 (80%)
88 (20%)
BMI
364
0.87
1.00 (0.97–1.03)
296 (81%)
68 (19%)
ACPA status
239
0.83
1.06 (0.65–1.72)
192 (80%)
47 (20%)
Never versus current smoker
151
0.27
0.66 (0.32–1.37)
125 (83%)
26 (17%)
Never versus ever smoker‡
254
0.47
0.85 (0.54–1.33)
207 (82%)
47 (18%)
Disease duration
438
0.01
1.02 (1.00–1.04)
351 (80%)
87 (20%)
Baseline DAS28 score
439
0.59
0.95 (0.78–1.15)
353 (80%)
86 (20%)
f
f 1 Ann Rheum Dis Month 2023 Vol 0 No 0 Letter
Figure 1 (A, C) Kaplan-Meier (KM) plot showing rate of drug antidrug antibody development, stratified by the number of HLA alleles carried (A,
HLA-DQA1*03; C, HLA-DRB1*04). The tables presented underneath the KM plots represents the number of participants at risk over time. Blue, orange
and green indicate 0, 1 and 2 copies of the alleles respectively. (B, D) Kaplan-Meier plot of drug immunogenicity rate for carriers of at least one
copy of HLA-DQA1*03 and HLA-DRB1*04, respectively, for different types of csDMARD cotherapy. Solid line and darkest shade of colour represent
cotherapy with MTX, dashed line and middle shade represents non-MTX csDMARD, dotted line with the lightest shade represents monotherapy with
only adalimumab. csDMARD, conventional synthetic disease modifying antirheumatic drug; MTX, methotrexate. on October 23, 2024 by guest. Protected by copyrigh
http://ard.bmj.com/
Ann Rheum Dis: first published as 10.1136/ard-2023-223955 on 12 September 2023. Downloaded from on O
http://ard.bmj.com/
Ann Rheum Dis: first published as 10.1136/ard-2023-223955 on 12 September 2023. Downloaded from Ann Rheum Dis: first published as 10.1136/ard-2023-223955 on 12 September Figure 1 (A, C) Kaplan-Meier (KM) plot showing rate of drug antidrug antibody development, stratified by the number of HLA alleles carried (A,
HLA-DQA1*03; C, HLA-DRB1*04). ORCID iDs Chuan Fu Yap http://orcid.org/0000-0001-5256-5642
Ann W Morgan http://orcid.org/0000-0003-1109-624X
John D Isaacs http://orcid.org/0000-0002-6103-7056
Anthony G Wilson http://orcid.org/0000-0003-4855-3926
Anne Barton http://orcid.org/0000-0003-3316-2527
Darren Plant http://orcid.org/0000-0003-1395-9344 Disclaimer The views expressed are those of the authors and not necessarily those
of the NHS, the NIHR or the Department of Health. Competing interests AWM is supported National Institute for Health and
Care Research (NIHR) and Medical Research Council (MRC). AWM has acted as
consultant for Roche/Chugai, Vifor and AstraZeneca. AWM is member of speakers’
bureaus for Roche/Chugai. AWM is on Data Safety Monitoring Board for GSK and
Regeneron/Sanofi. AWM is on the board for MRC and Vasculitis UK. KH has received
grant/research support from Pfizer, Bristol Myers Squibb (BMS). KH has received
honoraria for speaking at educational meeting by Abbvie. AB is supported by the
NIHR Manchester Biomedical Research Centre. AB has received grant/research
support from Pfizer, BMS, Scipher Medicine and Galapagos (paid to host institution). AB is member of speakers’ bureaus for Galapagos (paid to host institution). DP
has received grant/research support from BMS, Versus Arthritis and European
Commission. Received 30 January 2023
Accepted 22 July 2023 Received 30 January 2023
Accepted 22 July 2023 Received 30 January 2023
Accepted 22 July 2023 Funding This research was supported by the NIHR Manchester Biomedical
Research Centre. We thank Versus Arthritis (grant number 21173, grant number
21754 and grant number 21755) for support. This project has received funding from
the Innovative Medicines Initiative 2 Joint Undertaking (JU) under grant agreement
No. 831434 (3TR). The JU receives support from the European Union’s Horizon 2020
research and innovation programme and EFPIA. Dublin, University College Dublin, Dublin, Ireland
10 Open access This is an open access article distributed in accordance with the
Creative Commons Attribution 4.0 Unported (CC BY 4.0) license, which permits
others to copy, redistribute, remix, transform and build upon this work for any
purpose, provided the original work is properly cited, a link to the licence is given,
and indication of whether changes were made. See: https://creativecommons.org/
licenses/by/4.0/. University of Manchester, Manchester, UK Correspondence to Dr Darren Plant, Centre for Musculoskeletal Research, The
University of Manchester Centre for Genetics and Genomics Versus Arthritis,
Manchester, UK; Darren.Plant@manchester.ac.uk Manchester, UK; Darren.Plant@manchester.ac.uk © Author(s) (or their employer(s)) 2023. Re-use permitted under CC BY. Published
by BMJ. Handling editor Josef S Smolen selection of treatment and to quantify effects of non-MTX
csDMARDs on immunogenicity. selection of treatment and to quantify effects of non-MTX
csDMARDs on immunogenicity. Ethics approval This study involves human participants and ethics was approved
by the North West 6 Central Manchester South Research Ethics Committee (COREC
04/Q1403/37) and all patients provided written consent. Participants gave informed
consent to participate in the study before taking part. Chuan Fu Yap ,1 Nisha Nair,1,2 Annick de Vries,3 Floris C Loeff,3
Ann W Morgan ,4,5,6 John D Isaacs ,7,8 Anthony G Wilson ,9
Kimme L Hyrich,2,10 Anne Barton ,1,2 Darren Plant 1 Provenance and peer review Not commissioned; externally peer reviewed. Provenance and peer review Not commissioned; externally peer reviewed. Supplemental material This content has been supplied by the author(s). It
has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have
been peer-reviewed. Any opinions or recommendations discussed are solely those
of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and
responsibility arising from any reliance placed on the content. Where the content
includes any translated material, BMJ does not warrant the accuracy and reliability
of the translations (including but not limited to local regulations, clinical guidelines,
terminology, drug names and drug dosages), and is not responsible for any error
and/or omissions arising from translation and adaptation or otherwise. 1Centre for Genetics and Genomics Versus Arthritis, Centre for Musculoskeletal Research, The University of Manchester, Manchester, UK
2 Research, The University of Manchester, Manchester, UK y
IHR Biomedical Research Centre, Manchester University NHS Founda y
2NIHR Biomedical Research Centre, Manchester University NHS Found Manchester Academic Health Science Centre, Manchester, UK
3 3Diagnostic Services, Sanquin, Amsterdam, The Netherlands 4School of Medicine, University of Leeds, Leeds, UK
5 5NIHR Leeds Biomedical Research Centre, Leeds Teaching Hospitals NHS Trust, Leeds,
UK
6 6NIHR In Vitro Diagnostic Co-operative, Leeds Teaching Hospitals NHS Trust, Leeds,
UK 7Translational and Clinical Research Institute, Newcastle University, Newcastle upon
Tyne, UK HLA-DRB1 and HLA-DQA1 associated with
immunogenicity to adalimumab therapy in
patients with rheumatoid arthritis The tables presented underneath the KM plots represents the number of participants at risk over time. Blue, orange
and green indicate 0, 1 and 2 copies of the alleles respectively. (B, D) Kaplan-Meier plot of drug immunogenicity rate for carriers of at least one
copy of HLA-DQA1*03 and HLA-DRB1*04, respectively, for different types of csDMARD cotherapy. Solid line and darkest shade of colour represent
cotherapy with MTX, dashed line and middle shade represents non-MTX csDMARD, dotted line with the lightest shade represents monotherapy with
only adalimumab. csDMARD, conventional synthetic disease modifying antirheumatic drug; MTX, methotrexate. y
py g Figure 1 (A, C) Kaplan-Meier (KM) plot showing rate of drug antidrug antibody development, stratified by the number of HLA alleles carried (A,
HLA-DQA1*03; C, HLA-DRB1*04). The tables presented underneath the KM plots represents the number of participants at risk over time. Blue, orange
and green indicate 0, 1 and 2 copies of the alleles respectively. (B, D) Kaplan-Meier plot of drug immunogenicity rate for carriers of at least one
copy of HLA-DQA1*03 and HLA-DRB1*04, respectively, for different types of csDMARD cotherapy. Solid line and darkest shade of colour represent
cotherapy with MTX, dashed line and middle shade represents non-MTX csDMARD, dotted line with the lightest shade represents monotherapy with
only adalimumab. csDMARD, conventional synthetic disease modifying antirheumatic drug; MTX, methotrexate. Ann Rheum Dis Month 2023 Vol 0 No 0 2 on October 23, 2024 by guest. Protected by copyrig
http://ard.bmj.com/
Ann Rheum Dis: first published as 10.1136/ard-2023-223955 on 12 September 2023. Downloaded from Letter Contributors DP and CFY conceived or designed the study and data analyses. NN,
AdV and FCL acquired the data. CFY analysed the data. CFY, NN, AB and DP had
access to the data. All authors were involved in interpretation of data and reviewed
and approved the manuscript’s content before submission. CFY accepts final
responsibility for this work and controlled the decision to publish. To cite Yap CF, Nair N, de Vries A, et al. Ann Rheum Dis Epub ahead of print: [please
include Day Month Year]. doi:10.1136/ard-2023-223955 Handling editor Josef S Smolen ►Additional supplemental material is published online only. To view, please visit
the journal online (http://dx.doi.org/10.1136/ard-2023-223955). Twitter Chuan Fu Yap @chuanfuyap and John D Isaacs @ProfJohnIsaacs Twitter Chuan Fu Yap @chuanfuyap and John D Isaacs @ProfJohnIsaacs Acknowledgements We thank Asma Kalei for carefully coordinating sample
analysis at Sanquin Diagnostic Services. REFERENCES 1 Jani M, Chinoy H, Warren RB, et al. Clinical utility of random anti-tumor necrosis factor
drug-level testing and measurement of antidrug antibodies on the long-term treatment
response in rheumatoid arthritis. Arthritis Rheumatol 2015;67:2011–9. p
2 Billiet T, Vande Casteele N, Van Stappen T, et al. Immunogenicity to Infliximab is
associated with HLA-Drb1. Gut 2015;64:1344–5. 3 Sazonovs A, Kennedy NA, Moutsianas L, et al. HLA-Dqa1*05 carriage associated with
development of anti-drug antibodies to Infliximab and Adalimumab in patients with
Crohn’s disease. Gastroenterology 2020;158:189–99. 4 Liu M, Degner J, Davis JW, et al. Identification of HLA-Drb1 Association to Adalimumab
Immunogenicity. PLoS One 2018;13:e0195325. Patient and public involvement Patients and/or the public were not involved in
the design, or conduct, or reporting, or dissemination plans of this research. Patient and public involvement Patients and/or the public were not involved in
the design, or conduct, or reporting, or dissemination plans of this research. 5 Rogers AR, Huff C. Linkage disequilibrium between Loci with unknown phase. Genetics
2009;182:839–44. Patient consent for publication Consent obtained directly from patient(s). Patient consent for publication Consent obtained directly from patient(s). Ann Rheum Dis Month 2023 Vol 0 No 0 3
|
https://openalex.org/W2806456319
|
https://zenodo.org/record/1287279/files/Embedding%20Power%20Line%20Communication.pdf
|
English
| null |
Embedding Power Line Communication in Photovoltaic Optimizer by Modulating Data in Power Control Loop
|
IEEE transactions on industrial electronics
| 2,019
|
cc-by
| 11,165
|
Embedding Power Line Communication in
Photovoltaic Optimizer by Modulating Data
in Power Control Loop Yue Zhu, Jiande Wu, Member, IEEE, Ruichi Wang, Student Member, IEEE,
Zhengyu Lin, Senior Member, IEEE, and Xiangning He, Fellow, IEEE induced by manufacture difference, nonuniform aging, etc. d by manufacture difference, nonuniform aging, et Abstract—In Photovoltaic (PV) system, dc-dc power op-
timizer (DCPO) is an option to maximize output power. At
the same time, data links among DCPOs are often required
for system monitoring and controlling. This paper pro-
poses a novel power line communication (PLC) method for
the DCPOs, in which the data of a DCPO is modulated into
the control loop of power converter, and then transmitted
through the series-connected dc power line to other
DCPOs. In the process of communication, differential
phase shift keying (DPSK) modulation and discrete Fourier
transformation (DFT) demodulation are employed. To ana-
lyze the quality of communication, the communication
model of the system is built, based on small-signal model. Furthermore, the noises of the system, including switch-
ing, maximum power point tracking (MPPT) and additive
white Gaussian noise (AWGN), are discussed and meas-
ured to evaluate the signal-to-noise ratio (SNR). At last, an
experimental system including 6 DCPOs is established and
tested, which verifies the feasibility and effectiveness of
the proposed method. To mitigate the aforementioned problem, dc-dc power op-
timizer (DCPO), which is a dc-dc converter with input con-
nected to the PV panel and output connected to other DCPOs in
series, has been proposed and widely studied [1]-[9]. By
equipping each PV panel with a DCPO, the panel can work
independently at its MPP, thus distributed maximum power
point tracking can be achieved. p
g
In [1]-[4], different topologies of DCPO have been compared
and analyzed. Buck, boost, buck-boost, etc., are candidate
topologies for DCPO. Buck and boost converters are the most
efficient and low-cost topologies [2][3], while buck is only
suited for long string and boost is for short strings. Buck-boost,
cuk and sepic converters are flexible in voltage ranges but
always at an efficiency and alternatively cost shortcoming
[1][2][4]. In [5]-[7], modeling of a single DCPO and of a PV
string equipped with DCPOs have been studied in detail, which
is very helpful for system designing. In [8], a novel maximum
power point tracking (MPPT) technique based on DCPO is
proposed, which shows the effectiveness of DCPO in panel
MPPT. IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS Embedding Power Line Communication in
Photovoltaic Optimizer by Modulating Data
in Power Control Loop [9] proposes an intelligent PV module concept based on
DCPO and PLC, and it shows great potentials of DCPO in
intelligent PV system development. Index Terms—Photovoltaic (PV) system, Power line
communication, DC power optimizer. Manuscript received February 4, 2018; revised April 24, 2018; accepted
May 4, 2018. This work is sponsored by the National Nature Science Founda-
tion of China under Grants 51577170, National Key Technologies R&D
Program of China No. 2017YFE0112400, and the European Union’s Horizon
2020 research and innovation programme under grant agreement No. 734796.
Y. Zhu, J. Wu, R. Wang and X. He are with the College of Electrical En-
gineering, Zhejiang University, Hangzhou; 310027, China. (e-mail:
y_zhu@foxmail.com; eewjd@zju.edu.cn; rui_chi@163.com;hxn@zju.edu.cn).
Z. Lin is with the Department of Electrical, Electronic and Power Engi-
neering, Aston University, Birmingham, B4 7ET, U.K. (email: z.lin@ieee.org). Manuscript received February 4, 2018; revised April 24, 2018; accepted
May 4, 2018. This work is sponsored by the National Nature Science Founda-
tion of China under Grants 51577170, National Key Technologies R&D
Program of China No. 2017YFE0112400, and the European Union’s Horizon
2020 research and innovation programme under grant agreement No. 734796. Y. Zhu, J. Wu, R. Wang and X. He are with the College of Electrical En-
gineering, Zhejiang University, Hangzhou; 310027, China. (e-mail:
y_zhu@foxmail.com; eewjd@zju.edu.cn; rui_chi@163.com;hxn@zju.edu.cn). y
j
j
j
Z. Lin is with the Department of Electrical, Electronic and Power Engi-
neering, Aston University, Birmingham, B4 7ET, U.K. (email: z.lin@ieee.org). I. INTRODUCTION In a PV system equipped with DCPOs, PV status monitoring,
intelligent control algorithm, and outage management can be
implemented only if communication links among DCPOs are
provided. Due to the cost-sensitive characteristic of PV system,
low-cost communication method is preferred. For conventional
PV status monitoring systems, wireless communication is
commonly used [10]-[14], in which ZigBee technique gets
more attention. In [14], ZigBee is employed in a distributed PV
generator due to the advantages of low power consumption and
cost-effectiveness. However, if the Wi-Fi access point is de-
ployed in the same area, it may corrupt ZigBee packet and
cause severe interference [19]. Meanwhile, the vulnerable
characteristic of wireless communication may damage the
security operation of PV system. Thus, wireless communica-
tion is not the best scheme for PV system. P HOTOVOLTAIC (PV) generation has been one of the most
popular renewable energy technologies in the world. Con-
ventionally, dozens of panels are series-connected to increase
the output voltage, and then feeds to a grid-tied inverter. The
output currents of all panels are equal due to the series con-
nection. However, the current generation of PV panel is de-
pendent on solar intensity and its condition. If some panels
generate less current than the string current due to partial
shading, the unshaded panels have to lower their output current
to match with the shaded ones. Consequently, the unshaded
panels will not work at their maximum power point (MPP),
which leads to a waste of power [1]. Such problem can also be
P DC PLC technology, which utilizes the dc power line as an
alternative communication channel, is considered as a viable
communication method in distribution power systems. Due to
its reliable and low-cost advantages [15], dc PLC has become
an attractive solution for PV monitoring [16]-[21]. In [16], a
low-cost low-frequency PLC for PV monitoring was developed,
using baseband transmission and direct spread code division
multiple access (DS-CDMA). However, the achieved bit rate in IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS vPV
+-
MPPT
iPV
+
Carrier
vref
PID
+
-
fs
vc
δ(t)
Driver
01
Digital controller
dc-dc converter
Input
Output
Power
Info
Power
Data
PSK
(a)
DC
DC
PV1
vPV
iPV
Tx
Digital
Controller
DC
DC
PVn
Rx
Digital
Controller
Sampling
&filtering
Transmitting
data 010... Vload
Receiving
data 010... 1
1
ˆ
o
o
V
v
ˆ
on
on
V
v
δ(t)
DCPO1
DCPOn
(b)
Fig. 1. I. INTRODUCTION (a) Structure of signal modulation in a single DCPO’s power control
loop. (b) Structure of the proposed communication method in a PV string. this literature is 5 bps, which can only be employed in
low-demand PV monitoring occasions. In [20], A capacitive
coupling PLC scheme is proposed. A coupling transformer is
utilized to transmit signal carrier wave. However, due to the
parallel connection, the transmission efficiency is low, which
results in high transmitting power. In [21], a parallel resonant
coupling unit is designed to increase the transmission effi-
ciency of dc PLC, and 10kbps bit rate and low error rate are
achieved. However, since the coupling unit is series-connected
with the string and the generating current of PV panels flows
through it, the efficiency of the power conversion would de-
crease slightly. [9] introduces an approach which integrates
PLC with DCPO. However, in this scheme, the DCPO and PLC
circuits are designed independently, hence the total cost and
volume are increased compared with a mere DCPO. In general,
conventional PLC design consists of three components, in-
cluding a signal source, coupling circuits, and a demodulation
circuit. To reduce cost, it would be ideal if DCPO and PLC can
share a common circuit. Power (a)
DC
DC
PV1
vPV
iPV
Tx
Digital
Controller
DC
DC
PVn
Rx
Digital
Controller
Sampling
&filtering
Transmitting
data 010... Vload
Receiving
data 010... 1
1
ˆ
o
o
V
v
ˆ
on
on
V
v
δ(t)
DCPO1
DCPOn
(b) Data communication can be integrated with power converter,
and transmitted with the power either sequentially [22]-[23] or
simultaneously [24]-[27]. [22] proposed a novel PLC scheme in
which power electronic circuits transmit power in discontinues
mode, at the period of power off, data signal can be transmitted
at the same lines. [23] optimizes this sequential transmitting
scheme and achieves a maximum bit rate of 4.68Mbps. How-
ever, in sequential transmission strategy, power is transmitted
as pulse in the power line. In long distance situation, e.g. PV
system, transmission line effect will significantly interfere the
signal and lead to data corruption [22]. Therefore, sequential
transmission strategy is unsuitable for PV system. In the aspect
of simultaneous transmission strategy, [24] proposed a PLC
scheme among dc-dc converters sharing a common input dc bus,
in which the communication signal is generated by switching
frequency modulation of power converters. I. INTRODUCTION [25]-[27] investi-
gate the common method that embedding data signal into pulse
width modulation (PWM), and term it power and signal dual
modulation (PSDM) technique. The systems adopting PSDM
technique can reduce cost and simplify circuitry, which corre-
sponds to the requirement of PV system. According to the
studies, switching frequency and the pulse phase of the con-
verter are the common elements to carry information. [25]
realizes communication using frequency shift keying (FSK)
modulation on PWM carrier, while [26] employs phase shift
keying (PSK) modulation and direct sequence spread spectrum
(DSSS) technique on PWM carrier. The carrier-modulated
converter can send out data signal, but the amplitude of the
signal is small and cannot be regulated, thus it is restricted in
long-range communication applications, including PV system. [27] proposes a PSDM method that can regulate the amplitude
of the carrier, but this method is based on phase shift full bridge
(PSFB) converter and is hard to apply in other topologies. (b) Fig. 1. (a) Structure of signal modulation in a single DCPO’s power control
loop. (b) Structure of the proposed communication method in a PV string. 1) A novel PLC method for PV system is presented, in which
the data signal is modulated through the power control loop of
the DCPO and sent out from the output of DCPO. 2)The model of the communication system is built, and the
noise influences from switching process, MPPT and additive
white Gaussian noise (AWGN) in the PV system are analyzed. 3) An experimental system is built, and a reliable communica-
tion with 2kbps bit rate has been achieved, which verifies the
correctness of this method. B. Modeling
hi
l This paper selects buck converter for further study due to its
low cost and high efficiency advantages. A series string con-
nection of buck DCPOs allows total independence of output
voltage and power. Besides, the internal diode helps the con-
verter get bypassed automatically while the module is inactive
[2]. Nevertheless, a minimum number of PV panels is required
to guarantee the dc bus voltage while under extreme
non-homogeneous irradiance conditions, which is regarded as a
constraint of buck DCPO application. value vref, which is continuously updated by MPPT algorithm. However, in a view of communication system, it can also be
exploited as a data transmitter. If data carrier is added to the
reference voltage vref of the power control loop, a perturbation
carrying data information will be introduced in the duty-cycle d. Thus, the data will be modulated at the output voltage of DCPO
and transmitted via the cascaded power line. From the view of
communication, the DCPO with data modulation is regarded as
a data transmitter (Tx) and the power line of the PV string is a
communication channel. Meanwhile, the other DCPOs in the
string are the receivers (Rxs). In the following part, a buck converter with input voltage
feedback control is employed. MPPT process is not concerned
for modeling. For DCPOs, the circuit parameters and feedback network are
designed for power conversion primarily. In a well-designed
DCPO, data modulation should be designed according to the
following principles. Fig. 1(b) shows the system structure which applies the pro-
posed PLC scheme. A string composed of n panels with asso-
ciated DCPOs is presented in the system. As a first-order
approximation, it is possible to model the PV inverter as a
voltage source Vload with a series resistance [5].The resistance is
ignored in this paper because the value is small comparing with
the impedance of DCPOs in the string. 1) The signal gain from reference to output voltage should be
large to acquire large signal amplitude, whilst the gain from
signal reference to input voltage should be small to reduce the
influence on PV panel. The following discussion is based on the scenario that
DCPO1 is sending out data, and the other DCPOs are receiving. When a sinusoidal wave is added to vref of DCPO1, a voltage
perturbation
1
ˆov
with the same frequency will be produced at
the output. A. Principle For a PV system equipped with DCPOs, a data link among
DCPOs is necessary for PV status monitoring, including input
and output voltages and currents, temperature, and other pa-
rameters. Besides, innovative control algorithm may be de-
veloped if data is exchanged among DCPOs. For industrial application, considering the conversion effi-
ciency and cost issue, buck topology is identified as the most
cost-efficient topology for DCPO. This paper proposes a new PSDM method for PV system
equipped with DCPOs: modulating data in power control loop
of DCPOs by digital controller and demodulating the data from
the outputs of other DCPOs. Fig. 1(a) shows the structure of a single DCPO with data
modulation. The DCPO is originally utilized to maximize the
output power of PV panel. A common DCPO control strategy is
employed, in which MPPT is implemented by regulating PV
output voltage and measuring PV output power. The controller
of the DCPO regulates the PV voltage to follow a time-variant The contributions of this paper are as follows. IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS iPV
+
-
vPV
Co
Cin
+
-
io
L
Cin
RPV
1:D
L
Co
2 ˆ
I d
vo
oˆv
ˆPV
v
i2
ˆ
PV
V d
PV
ˆ
PV
i
2ˆi
Gc(s)
1/VM
Kf
ˆcv
ˆd
+
-
vref
ˆPV
v
vo
Fig. 2. Topology and small-signal model of a single buck DCPO Second, the impact to PV power system should be evaluated. Since the PV module is not a voltage source, the data signal
modulated in control loop not only generates a perturbation at
the output of the DCPO, but also introduces a disturbance at the
input of the DCPO, which may result in performance degrada-
tion of the PV panel. Thus, it is significant that the disturbance
at the input should be small while the perturbation at the output
should be large. Third, multiple noises, including switching process and
MPPT algorithm, influence the communication reliability and
increases bit error rate (BER). Therefore, noise analysis is
necessary to give a reference for selecting proper signal am-
plitude. Further study focuses on these problems and gives analysis
in detail. ig. 2. Topology and small-signal model of a single buck DCP B. Modeling
hi
l Meanwhile, the voltage perturbation will transmit to
the outputs of other DCPOs and can be extracted by voltage
sampling circuits. Assuming that the DCPOs have the same
output impedance, the relationship between
1
ˆov
and output
perturbation of receivers ˆ i
v
is 2) The carrier frequency should be high to achieve a high bit
rate. And it would be better if switching frequency is integral
multiple of the carrier wave frequency, which will be explained
in Section III-B. Small-signal model of a DCPO is essential for analysis. The
schematic of a single DCPO under study and the small-signal
model are shown in Fig. 2, where the equivalent series re-
sistance (ESR) of the inductor and capacitor are ignored. RPV is
the dynamic resistance of PV module at MPP and can be ex-
pressed as perturbation of receivers ˆoi
v is 1
ˆ
ˆ-
/ (
1)
{2,3..., }
oi
o
v
v
n
i
n
(1) (1) 1
PV
PV
PV
MPP
i
R
v
(2) where symbols with hats represent small-signal variations
around the steady state values. 1
ˆov
is the output voltage per-
turbation of transmitter, and ˆoi
v is the voltage perturbation at
the output of ith DCPO. (2) To simplify the power control loop design of DCPO, the
output voltage is assumed as a constant, thus ˆov is considered as
zero. The open-loop transfer functions of a single DCPO from
ˆd to ˆpv
v
is derived by [7] Three major problems should be analyzed before applying
this scheme.
_
2
2
ˆ
0
ˆ
( )
( )
ˆ
2
( )
PV
o
o
o
PV
v
d
OL
o
o
o
v
K
s
a
v
s
G
s
s
s
d s
(3) First, to keep the communication reliable under different
situations, the signal amplitude at output voltage should be
regulated by the digital controller. Therefore, the gain of signal
carrier amplitude from digital controller to output voltage must
be calculated. (3) where where IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS Gc(s)
GVpvd_OL(s)
1/VM
Kf
G1(s)
G2(s)
G3(s)
ˆcv
ˆd
ˆPV
v
+ -
+ +
++
1
ˆov
-1/n
ˆoi
v
vsig
Tx
Rx
Fig. 4. The block diagram of the communication system. B. Modeling
hi
l (4) 1
2
o
o
in
PV
o
o
o
in
o
PV
in
I
K
C
V
D
a
LI
D
LC
L
DR
C
. (4) 2
3
2
1
n
sL
Z
Z
Z
s LC
. (10) (10) The simplified communication system is shown in Fig. 3(b). Using superposition principle, the Tx output perturbation
1
ˆov
is
derived as (11), and the Rx output perturbation ˆoi
v is derived
as (12) (4)
1
2
2
o
o
//(
1)
ˆ
ˆ
ˆ
//(
1)
1
1
ˆ
ˆ
1
1
o
n
o
pv
pv
o
n
pv
pv
sC
n
Z
v
Dv
V d sL
sC
n
Z
V
n
n
D
d
v
n
s LC
n
s LC
(11)
1
1
ˆ
ˆ
2,3,
1
oi
o
v
v
i
n
n
. (12) From the poles and zeros of
( )
PV
v
dOL
G
s , a lead-lag compensator
function Gc(s) is designed as (11)
2
c
2
( )
2
o
c
o
s
K
s
G s
s s
a
s
f
. (5) (5)
1
1
ˆ
ˆ
2,3,
1
oi
o
v
v
i
n
n
. (12) (12) In a feedback control system, the loop transfer function is
_
( )
( )
pv
c
f
c
v d
OL
M
T s
K G s G
s V
(6) In a feedback control system, the loop transfer function is
_
( )
( )
pv
c
f
c
v d
OL
M
T s
K G s G
s V
(6) Fig. 4 shows a block diagram of the proposed communication
system, in which a signal carrier vsig is modulated into the
power control loop. G1(s), G2(s), and G3(s) are the directly
transfer functions from ˆpv
v
to ˆov , from ˆd to ˆov , and from ˆov
to ˆpv
v
, respectively. From (8) and (11), G1(s), G2(s), and G3(s)
are calculated as Fig. B. Modeling
hi
l 4 shows a block diagram of the proposed communication
system, in which a signal carrier vsig is modulated into the
power control loop. G1(s), G2(s), and G3(s) are the directly
transfer functions from ˆpv
v
to ˆov , from ˆd to ˆov , and from ˆov
to ˆpv
v
, respectively. From (8) and (11), G1(s), G2(s), and G3(s)
are calculated as where Kf is the feedback factor, and VM is the amplitude of
PWM sawtooth wave. The value of KC decides the
gain-crossover frequency of Tc(s). Thus, for a single buck
DCPO, the close-loop transfer function from ˆd to ˆpv
v
is
_
_
( )=
1
PV
c
vpvd
OL
M
v
d
CL
c
G
s G
s
V
G
s
T
s
. (7) (7) 1
2
1
( )
1
o
n
D
G s
n
s LC
(13)
2
2
1
( )
1
PV
o
n
V
G s
n
s LC
(14)
3
2
2
( )
in
PV
D
G s
L
s LC
s
D
R
. (15) (13) In the communication system, since the DCPO is a data
transmitter, and the carrier wave is modulated on the output of
the DCPO, ˆov should not be considered as zero as before. In the communication system, since the DCPO is a data
transmitter, and the carrier wave is modulated on the output of
the DCPO, ˆov should not be considered as zero as before. Therefore, both ˆpv
v
and ˆov are state variables, which are in-
fluenced by input variable ˆd . The transfer functions from ˆd to
ˆpv
v
and ˆd to ˆov are of high order, but it can be simplified by
superposition principle. Assuming ˆd is zero, the transfer func-
tion from ˆov to ˆpv
v
can be derived from Fig. 2 and written as (14) (15) For the case study, the PV module parameters and the buck
converter parameters are listed in TABLE I. With the help of PSIM ac sweep, bode diagrams of ˆ
ˆ
PV
sig
v
v
and ˆ
ˆ
o
sig
v v
are shown in Fig. 5(a). Since phase shifting from
input to output does not affect signal transmitting, only ampli-
tude response is depicted. B. Modeling
hi
l L
Co
oˆv
ˆ
PV
V d
ˆPV
Dv
ˆ
ˆ
PV
PV
V d
Dv
Tx
L
Co
Rx
L
Co
buck DCPO
small-signal model
(a)
ˆ
ˆ
PV
PV
V d
Dv
L2
Co2
Ln
Con
Inverter
Tx
Z2
Zn
Rx-2
Rx-n
L1
Co1
1
ˆov
2
ˆov
ˆon
v
(b)
Fig. 3. (a) The simplified model of Tx and Rx. (b) The model of the com-
munication system. Gc(s)
GVpvd_OL(s)
1/VM
Kf
G1(s)
G2(s)
G3(s)
ˆcv
ˆd
ˆPV
v
+ -
+ +
++
1
ˆov
-1/n
ˆoi
v
vsig
Tx
Rx
Fig. 4. The block diagram of the communication system. _
o
ˆ 0
ˆ
ˆ
ˆ
( )
( )
PV
PV
o
PV
v
d
CL
v
v
d
v
G
s d
G
s
v
(9) _
o
ˆ 0
ˆ
ˆ
ˆ
( )
( )
PV
PV
o
PV
v
d
CL
v
v
d
v
G
s d
G
s
v
(9) To calculate ˆo
v precisely, the close-loop impedance of every
DCPO should be deduced, which is complicated, and the de-
tailed derivation is given by [5]. However, in the communica-
tion model, signal carrier frequency is much higher than the
crossover frequency fg of Tc(s). Feedback network of Rx will
not response to such signal. Therefore, for Rx, ˆd will be zero,
and the dependent voltage source
ˆ
PV
V d can be regarded as
shorted. Meanwhile, if the input capacitor Cin is large, the
perturbation at the input ˆPV
v
of Rx will be small, and
ˆPV
Dv
is
approximately equal to zero. Thus, Rx can be regarded as a
parallel LC branch, which greatly simplifies the calculation, as
shown in Fig. 3(a). The impedance of each Rx is (b) Fig. 3. (a) The simplified model of Tx and Rx. (b) The model of the com-
munication system. 1
2
o
o
in
PV
o
o
o
in
o
PV
in
I
K
C
V
D
a
LI
D
LC
L
DR
C
. B. Modeling
hi
l 19.3
20.2
21.1
21.9
22.8
24.0
25.0
26.1
27.0
18.0
20.0
22.0
24.0
26.0
28.0
200
300
400
500
600
700
800
900
1000
Solar Intensity (W/m2 )
(dB)
10
ˆ
ˆ
o
PV
kHz
v v
(c) (c) The gain from ˆPV
v
to ˆo
v at different frequency is shown in
Fig. 5(b) based on the simplified model. According to above
analysis, 10 kHz is selected as carrier wave frequency in this
paper. The ratio from ˆo
v to ˆPV
v
when modulating the signal in
the reference voltage is ( )
Fig. 5. (a) Bode diagram of transfer functions from vsig to vPV and from vsig to vo. (b) Gain under different carrier frequency. (c) Gain at 10 kHz under different
solar intensity. Fig. 5. (a) Bode diagram of transfer functions from vsig to vPV and from vsig to vo. (b) Gain under different carrier frequency. (c) Gain at 10 kHz under different
solar intensity.
1
ˆ
cos
PV
PV
c
v
A
t
(17)
1
2
ˆ
cos
o
o
c
v
A
t
. (18) (17) (17)
(18)
ˆ
20lg
6.1
13.2
19.3dB
ˆ
o
PV
v
v
. (16) (18) (16) For the transmitter, due to the swing of PV working point
caused by communication carrier, the additional loss is calcu-
lated by [6] Assuming in a string with 6 DCPOs, 0.5Vamp sinusoidal wave
vsig is modulated into the power control loop of Tx. Then, 1.03V
amplitude sinusoidal wave will be generated at the output
voltage of Tx. At the same time, the amplitude of input voltage
perturbation will be 0.11V, which only produce a slight influ-
ence on PV panel. Assuming in a string with 6 DCPOs, 0.5Vamp sinusoidal wave
vsig is modulated into the power control loop of Tx. Then, 1.03V
amplitude sinusoidal wave will be generated at the output
voltage of Tx. At the same time, the amplitude of input voltage
perturbation will be 0.11V, which only produce a slight influ-
ence on PV panel. 2ˆ
ˆ
PV
i
MPP
v
P
R
. (19) 2ˆ
ˆ
PV
i
MPP
v
P
R
. (19) (19) Thus, the power loss for the DCPO can be calculated a
2
2
2
1
0
cos
1
. B. Modeling
hi
l Switching model bode diagrams are
also depicted, which show that the simplified communication With the help of PSIM ac sweep, bode diagrams of ˆ
ˆ
PV
sig
v
v
and ˆ
ˆ
o
sig
v v
are shown in Fig. 5(a). Since phase shifting from ˆ
0
2
2
( )
PV
o
v
v
d
in
PV
D
G
s
L
s LC
s
D
R
. (8) (8) input to output does not affect signal transmitting, only ampli-
tude response is depicted. Switching model bode diagrams are
also depicted, which show that the simplified communication In such way, from (7) and (8), ˆPV
v
is derived as IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS 1000
5000
10000
Frequency (Hz)
Frequency (Hz)
20000
2530
Gian-crossover
frequency fg
0
-5
-10
-15
-20
-25
5
0
-5
-10
5
10
15
Magnitude (dB)
20
25
Signal carrier
frequency fc
The simplified model
Switching model
From:vsig
To:vPV
To:vo
6.1 dB
-13.2 dB
(a)
19.5
23.0
23.8
23.1
22.4
21.9
21.4
21.2 20.8
20.6
20.3 20.1 19.9 19.7
19.6 19.4
19.3
19.0
20.0
21.0
22.0
23.0
24.0
2
3
4
5
6
7
8
9
10
Frequency (103 Hz)
(dB)
ˆ ˆ
o
PV
v v
(b)
19.3
20.2
21.1
21.9
22.8
24.0
25.0
26.1
27.0
18.0
20.0
22.0
24.0
26.0
28.0
200
300
400
500
600
700
800
900
1000
Solar Intensity (W/m2 )
(dB)
10
ˆ
ˆ
o
PV
kHz
v v
(c)
Fig. 5. (a) Bode diagram of transfer functions from vsig to vPV and from vsig to vo. (b) Gain under different carrier frequency. B. Modeling
hi
l 2
cT
PV
c
PV
i
c
PV
PV
A
t
A
P
dt
T
R
R
(20) Different solar intensities and mismatch situations will not
influence signal receiving, since the Rx is taken as a parallel LC
branch. For Tx, using the same system model and linear anal-
ysis tool,
ˆ ˆ
20lg
o
PV
v v
under different solar intensities are de-
picted in Fig. 5(c), which shows that the gain is even larger in
shading condition. (20) For the communication system, comparing with the imped-
ance of DCPOs, the resistance of DC power line Rl is much
small, thereby the output current of the carrier is deduced as
2
2
1
2
ˆ
4
1
ˆ =
sin
1 2
1
1
//
o
c
o
o
o
c
c
o
v
f LC
i
A
t
n
f L
n
sL
sC
. (21) B. Modeling
hi
l (c) Gain at 10 kHz under different
l
i t
it 1000
5000
10000
Frequency (Hz)
Frequency (Hz)
20000
2530
Gian-crossover
frequency fg
0
-5
-10
-15
-20
-25
5
0
-5
-10
5
10
15
Magnitude (dB)
20
25
Signal carrier
frequency fc
The simplified model
Switching model
From:vsig
To:vPV
To:vo
6.1 dB
-13.2 dB
(a) TABLE I
PARAMETERS OF THE SYSTEM
Symbol
Parameter
Value
S
Solar intensity
1000W/m2
VMPP
Maximum power point voltage
21.8V
IMPP
Maximum power point current
3.56A
RPV
Dynamic resistance at MPP
6.70Ω
fs
Converter switching frequency
50kHz
n
Number of PV panels in the string
6
L
Converter inductance
35H
Cin
Input capacitance
200F
Co
Output capacitance
22F
D
Steady duty cycle
0.80
Vo
Output voltage
17.4V
Io
Output current
4.45A
Kf
Feedback factor
1
KC
Coefficient of the compensator network
1.32×105
VM
Amplitude of PWM sawtooth wave
1V
f
G i
f
f
( )
2 53 103H
Switching model
The simplified model
Input: 10kHz 1.0VP-P Sine wave
vPV (V)
vo1 (V)
vo2 (V)
0.22V
2.0V
0.40V
Fig. 6. Simulation result of vPV, vo of DCPO1(Tx) and vo of DCPO2(Rx) in
small-signal model and switching model. 19.5
23.0
23.8
23.1
22.4
21.9
21.4
21.2 20.8
20.6
20.3 20.1 19.9 19.7
19.6 19.4
19.3
19.0
20.0
21.0
22.0
23.0
24.0
2
3
4
5
6
7
8
9
10
Frequency (103 Hz)
(dB)
ˆ ˆ
o
PV
v v
(b) Kf
Feedback factor
1
KC
Coefficient of the compensator network
1.32×105
VM
Amplitude of PWM sawtooth wave
1V
Fig. 6. Simulation result of vPV, vo of DCPO1(Tx) and vo of DCPO2(Rx) in
small-signal model and switching model. model is corresponding with switching model around carrier
frequency. In the diagram, the proper range of carrier frequency
is from 2 kHz to 10 kHz, in which the amplitude gain, from ˆPV
v
to ˆo
v , is about 20dB. It means that when the signal is modu-
lated into the power control loop of the DCPO, a large pertur-
bation will be produced at the output, while the input pertur-
bation will be small. Consequently, the influence on PV panel
will be slight. If carrier frequency is above 10kHz, communi-
cation may still work, but small-signal model is imprecise
because perturbation frequency is more than 1/5 switching
frequency. A. Data Modulation and Demodulation A. Data Modulation and Demodulation To modulate baseband data in the PV string, three basic
modulation methods can be exploited, including ASK, FSK and
PSK. This paper employs binary differential phase shift keying
(2DPSK) method due to the features of narrow bandwidth, high
anti-noise performance and easy implementation, as shown in
Fig. 7(a).
0
cos
cos
cos
2
b
T
l
c
c
b
x
A
t
t dt
AT
(28)
0
cos
sin
-
sin
2 . b
T
q
c
c
b
x
A
t
t dt
AT
(29) (28) (29) Considering sequence s0(t) and s1(t), which represents a bit in
a continuous data stream, and has a differential phase Δthe
reflection values are (xl0,xq0), (xl1,xq1) respectively. Using (30),
the result Ar is calculated, and then tested by the decision
device according to (31). For bit 0, the phase stays the same,
which means Δthus, Ar>0. Conversely, for bit 1, Ar<0. Therefore, decisions of bit data can be made by the value of Ar . Non-coherent demodulation and coherent demodulation are
generally used in DPSK communication. However, coherent
demodulation requires a phase-synchronized clock signal
between the transmitter and receivers, which is complex for
implementation. Therefore, this paper employs non-coherent
orthogonal demodulation method. Still, bit-synchronization is
required, which can be easily achieved by program [28].
2
2
0
1
0
1
cos
/ 4
r
l
l
q
q
b
A
x x
x x
A T
(30)
0
(
0)
data= 1
(
0)
r
r
A
A
. (31) (30) q
y
y p
g
For Rx, the received signal s(t) after band pass filter (BPF) is
( )
cos(
)
0,2
c
s t
A
t
(25) q
y
y p
g
For Rx, the received signal s(t) after band pass filter (BPF) is
( )
cos(
)
0,2
c
s t
A
t
(25) (31) where A, ωc, θ are the amplitude, angular frequency and relative
phase of the carrier wave after BPF, respectively. A symbol is
modulated into 5 cycles of carrier wave. A. Data Modulation and Demodulation For 10kHz carrier
wave, the symbol rate RB will be 2kBaud. As binary, the bit rate
of this communication system could reach 2kbps. The period of
one-bit Tb is The demodulation processes, including A/D conversion,
discrete Fourier transform(DFT) calculation and bit decision,
are all implemented in MCU by software. For the hardware of
demodulator, only a voltage sampling circuit and a BPF circuit
are needed, hence the cost of the receiver is very low. D. Modeling Verification (a)
s(t)
Decision
device
BPF
Delay
Tb
Delay
Tb
+
+
0
b
T dt
0
b
T dt
Acos(ωct+θ)
cos(ωct)
sin(ωct)
Acos(ωct+θ+Δθ )
e2DPSK(t)
xl1
xq1
Ar
xl0
xq0
data
(b) System simulation is implemented to check the proposed
small-signal model. The simulation software used is PSIM 9.0,
and the parameters are as shown in TABLE I. Six DCPOs are
connected in series, and a perturbation is added to the reference
voltage of DCPO1. The simulation result is shown in Fig. 6. The red line shows the output of switching model, which is the
result using switching circuit. The blue line shows the output of
small-signal model. It proves the accuracy of the model. (b) Fig. 7. (a) Illustrative waveform for 2DPSK. (b) Block diagram of 2DPSK
receiver. g
C. Power Loss Analysis (21) The embedded communication signal would introduce addi-
tional power loss for the DCPO, and the transmitting signal
would consume power, thus, it is necessary to analyze the
power loss of the system. For expression simplification, let For expression simplification, let p
2
2
4
1
1 2
c
o
Io
o
c
f LC
A
A
n
f L
. (22) (22) To calculate the power loss, the voltage-perturbation func-
tions of vPV and vo are defined as Thus, the transmitting power consumption on the power line is Thus, the transmitting power consumption on the power line is IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS 0
1
1
0
Binary Data
Modulated
Wave
Phase
Differential Δθ
0
0
Tb
π
π
(a)
s(t)
Decision
device
BPF
Delay
Tb
Delay
Tb
+
+
0
b
T dt
0
b
T dt
Acos(ωct+θ)
cos(ωct)
sin(ωct)
Acos(ωct+θ+Δθ )
e2DPSK(t)
xl1
xq1
Ar
xl0
xq0
data
(b)
Fig. 7. (a) Illustrative waveform for 2DPSK. (b) Block diagram of 2DPSK
receiver. 0
1
1
0
Binary Data
Modulated
Wave
Phase
Differential Δθ
0
0
Tb
π
π 2
2
0
1
ˆ
2
cT
Io
l
o
o
l
c
A R
P
i R dt
T
(23) (23) and the total power loss for data transmission is and the total power loss for data transmission is total
i
o
P
P
P
. (24) (24) total
i
o
P
P
P
. B. Noise Analysis 5
b
c
T
T
(26) (26) Noise analysis is vital for a communication system. In the
proposed system, many types of interferences are involved,
including switching noise, MPPT interference, and white noise. Noise analysis is vital for a communication system. In the
proposed system, many types of interferences are involved,
including switching noise, MPPT interference, and white noise. Grid influence, which introduces a 100Hz ripple on the dc bus,
is not concerned in this paper because it is far from the center
frequency of BPF. where
cT is the period of carrier wave. The bandwidth of the
carrier B is where
cT is the period of carrier wave. The bandwidth of the
carrier B is 2
b
B
T
. (27) (27) The procedure diagram of modulation is depicted in Fig. 7(b). Separating the signal by two orthometric vectors cos(ωct) and
sin(ωct), the reflection values xl and xq are calculated as IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS Voltage (V)
0.5
1.5
2.5
2.0
1.0
Time (ms)
0
2
4
6
8
10
(a)
0
0.5
1
1.5
2
2.5
3
3.5
4
0
200
400
600
800
1000
1200
1400
1600
1800
Distribution diagram
Gaussian curve
Voltage values (V)
(b)
Fig. 8. (a) Noise sampling at receiving device. (b) The distribution diagram and
density function curve of samples. TABLE Ⅱ TABLE Ⅱ
PERFORMANCE PARAMETERS OF COMMUNICATION PERFORMANCE PARAMETERS OF COMMUNICATION PERFORMANCE PARAMETERS OF COMMUNICATION
Symbol
Parameter
Value
Tc
Signal carrier wave period
10μs
Tb
One-bit period
50μs
RB
Symbol rate
2kBaud
B
Bandwidth
4kHz
A
Signal Amplitude after BPF
1.0V
σ2
Noise power
0.0123W
SNR
Signal-to-Noise Ratio
16.1dB 2
2
=10lg 2
S
A
N
(36) (36) The quantities related to the communication system are
listed in TABLE Ⅱ. The SNR is 16.1dB, which is a large value. Therefore, the operations of DCPOs in the string will not
seriously affect communication effectiveness. (b)
Fig. 8. (a) Noise sampling at receiving device. (b) The distribution diagram and
density function curve of samples. y
In applications, the imperfection of BPF, the inaccuracy of
ADC, and the deviation of bit-synchronization may cause
demodulation error. However, these influences are slight, thus
they are not discussed in this paper. For the receiver, the switching noise is mainly from the
converter itself, including fundamental wave, harmonics and
their sidebands, and the voltage spikes caused by power
switching. IV. EXPERIMENTAL VERIFICATION An experimental system, which consists of six DCPOs in
series and a constant voltage (CV) load at the terminal, is
designed to verify the proposed method, the circuit connection
of the system is shown in Fig. 9(a). In the DCPO, buck con-
verter is employed, and controlled by a STM32F407 MCU
operating in 168MHz. The picture of the system and the DCPO
board are shown in Fig. 9(b) and (c) respectively. To build a
stable experimental platform, the PV module is substituted by a
46.6V voltage source and a 6.7Ω resistance connected in series. Other parameters of the system have been shown in TABLE I. By programming, a sinusoidal wave, with 0.5V amplitude and
10kHz frequency, is added to the reference voltage of power
control loop of DCPO1. A dc electronic load (CHROMA 6314)
is employed as the CV load, which keeps the bus voltage at
104.4V. The total power of the system is 466W. To imitate the
MPPT process of a real PV system, a 500Hz/0.1Vamp square
wave is added to the reference voltage of each DCPO except
Tx.
cos 2
s
s
s
s
g
t
A
f t
(32)
0
( )cos
0
b
T
ls
s
c
x
g t
t dt
(33)
0
( )sin
0 . b
T
qs
s
c
x
g t
t dt
(34) Perturbation and observation (P&O) method is the most
popular MPPT method in applications. Consequently, a low
frequency perturbation around the MPP is introduced. In this
paper, to simulate the noise environment, each DCPO produces
a 0.1V perturbation at vPV every millisecond. At the output of
the DCPO, a mixture of MPPT perturbation from itself and
other DCPOs, together with white noise, makes the noise
randomly and complicated. However, after sampling the noise
voltage wave at BPF, as shown in Fig. 8(a), a distribution
diagram of the values shows that the mixed noise n(t) is
Gaussian, as shown in Fig. 8(b), and the density function is For Tx, data signal is directly modulated to the buck con-
verter by software in the MCU, and no hardware circuit is
added. B. Noise Analysis Most of them are filtered by the BPF, especially the
high frequency harmonics. Sidebands of switching frequency
will also be filtered due to the high frequency. Voltage spikes
can be avoided by separating sampling with switching actions
in a PWM period. Thus, switching fundamental wave gs(t) is
the major part, which can be written as (32), where φs and fs are
the relative angle and frequency of the converter respectively. When fs is integral multiple of the carrier frequency fc, the
demodulation results will be zero, as shown in (33) and (34). Therefore, the influence of switching fundamental can be
eliminated. IV. EXPERIMENTAL VERIFICATION Waveform of ˆPV
v
, ˆov of Tx, and ˆPV
v
, ˆov of Rx. Tx vo
Rx vo
Rx BPF
Rx demodulation
results
pilot-carrier
1 0 1 1 0 0 1 0
Fig. 12. Waveform of ˆov of Tx and Rx, waveform at BPF, and data demodu-
lation results at a standard state. (c)
Fig. 9. (a) Circuit connection of the system. (b) Picture of experimental
platform. (c) Picture of the DCPO. result agrees with the analysis. The ratio
ˆ
ˆ
20lg(
)
PV
o
v
v
is 19.4dB. result agrees with the analysis. The ratio 20lg(
)
PV
o
v
v
is 19.4dB. The error between experimental result and theoretical result is
0.1dB, which proves the accuracy of modeling. The amplitude
of voltage perturbation at output of Rx is 0.20V. The value is
less than ˆ
5
o Tx
v
, mainly because part of the signal voltage is
dropped on the CV load. The amplitude of voltage perturbation
at input of Rx is 0.01V, which is too small to affect the opera-
tions of Rx, thus it can be ignored. result agrees with the analysis. The ratio 20lg(
)
PV
o
v
v
is 19.4dB. The error between experimental result and theoretical result is
0.1dB, which proves the accuracy of modeling. The amplitude
of voltage perturbation at output of Rx is 0.20V. The value is
less than ˆ
5
o Tx
v
, mainly because part of the signal voltage is
dropped on the CV load. The amplitude of voltage perturbation
at input of Rx is 0.01V, which is too small to affect the opera-
tions of Rx, thus it can be ignored. Fig. 12 shows the output voltage of Tx and Rx, and the de-
modulation at Rx when Tx is sending a data frame with binary
sequence 10110010 at the beginning. The amplitude of carrier
wave after BPF reaches 1.0V, and correct sequence is demod-
ulated at Rx using DFT program. The first five cycles of carrier
wave, named as pilot-carrier, is used to provide a reference
phase for the coming sequence. Moreover, the pilot-carrier is
used by program for bit-synchronization. Fig. 12. Waveform of ˆov of Tx and Rx, waveform at BPF, and data demodu-
lation results at a standard state. voltage is changed to 131V. IV. EXPERIMENTAL VERIFICATION For Rx, a differential amplifier and a BPF are used to
extract the signal from the output voltage of the buck converter,
and the demodulation process is achieved by MCU program, as
shown in Fig. 10. When the data carrier is modulated into the power control
loop of Tx(DCPO1), a perturbation is generated upon the dc
voltage. Setting that Tx is sending out a continuous carrier, the
detailed waveforms of perturbation ˆPV
v
and ˆov of Tx, together
with ˆPV
v
and ˆov of Rx, are shown in Fig. 11. To measure the
perturbations, these waveforms are recorded by ac-coupling of
oscilloscope. The accurate amplitude of 10kHz sinusoidal wave
is acquired using FFT function of the oscilloscope. The am-
plitude of ˆov is 1.21V, while the amplitude of ˆPV
v
is 0.13V. The
2
2
1
( )
exp
2
2
Y
y
f
y
(35) (35) where Y is the random variable, y is the possible value of Y,
μ is the mean and σ2 is variance of the samples, and σ2 can be
regarded as noise power. In this situation, the receiving envi-
ronment is the same as a sine wave plus narrowband noise. As a
result, the signal-to-noise ratio (SNR) can be written as where Y is the random variable, y is the possible value of Y,
μ is the mean and σ2 is variance of the samples, and σ2 can be
regarded as noise power. In this situation, the receiving envi-
ronment is the same as a sine wave plus narrowband noise. As a
result, the signal-to-noise ratio (SNR) can be written as ˆ
RX
=21.8V
(10kHz)
0.01V
PV
PV
v
dc
v
amp
：
ˆ
RX
=17.4V
(10kHz)
0.20V
o
o
v dc
v amp
：
ˆ
TX
=21.8V
(10kHz)
0.13V
PV
PV
v
dc
v
amp
：
ˆ
TX
=17.4V
(10kHz)
1.21V
o
o
v dc
v amp
：
Fig. 11. Waveform of ˆPV
v
, ˆov of Tx, and ˆPV
v
, ˆov of Rx. Tx vo
Rx vo
Rx BPF
Rx demodulation
results
pilot-carrier
1 0 1 1 0 0 1 0
Fig. 12. Waveform of ˆov of Tx and Rx, waveform at BPF, and data demodu-
lation results at a standard state. IV. EXPERIMENTAL VERIFICATION Rx
ipv +
-
Vpv
Co
Cin
+
-
VCo
io
Spwm
D
S
L
-
+
V+
V-
-
+
V+
V-
Voltage Sampling
& Amplifier
Sallen-Key
topology BPF
MCU
STM32F407
Demodulation
& Judgement
R1
R3
R4
R2
R7
R8
R1
R6
C2
C1
R5
Results
Vo
-
Vo
+
Vo
+
Vo
-
10nF
10nF
3.3V
10μF
C3
R9
R10
DC Block
& Bias
10kΩ
10kΩ
1.6kΩ
1kΩ
1.93kΩ
420Ω
9kΩ
51kΩ
20kΩ
20kΩ
51kΩ
DFT
Program
ADC
Cin1
Cin2
100nF
100nF
Fig. 10. Demodulation process of Rx. IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS DC
DC
DCPO1
DC
DC
DCPO6
CV
Load
DC
DC
DCPO2
+
-
+
-
+
-
+
-
6.7Ω
6.7Ω
6.7Ω
46.6V
46.6V
46.6V
104.4V
17.4V
17.4V
17.4V
(a)
Scope
CV Load
DCPO1
Rx1
Rx2
Rx3
Rx4
Rx5
DC Power
Supply
Power
Resistance
Tx
(b)
Output
Input
Aux
Power
Digital
Controller
Power
Circuit
(c)
Fig. 9. (a) Circuit connection of the system. (b) Picture of experimental
platform. (c) Picture of the DCPO. Rx
ipv +
-
Vpv
Co
Cin
+
-
VCo
io
Spwm
D
S
L
-
+
V+
V-
-
+
V+
V-
Voltage Sampling
& Amplifier
Sallen-Key
topology BPF
MCU
STM32F407
Demodulation
& Judgement
R1
R3
R4
R2
R7
R8
R1
R6
C2
C1
R5
Results
Vo
-
Vo
+
Vo
+
Vo
-
10nF
10nF
3.3V
10μF
C3
R9
R10
DC Block
& Bias
10kΩ
10kΩ
1.6kΩ
1kΩ
1.93kΩ
420Ω
9kΩ
51kΩ
20kΩ
20kΩ
51kΩ
DFT
Program
ADC
Cin1
Cin2
100nF
100nF
Fig. 10. Demodulation process of Rx. DC
DC
DCPO1
DC
DC
DCPO6
CV
Load
DC
DC
DCPO2
+
-
+
-
+
-
+
-
6.7Ω
6.7Ω
6.7Ω
46.6V
46.6V
46.6V
104.4V
17.4V
17.4V
17.4V
(a) (a) Scope
CV Load
DCPO1
Rx1
Rx2
Rx3
Rx4
Rx5
DC Power
Supply
Power
Resistance
Tx Voltage Sampling
& Amplifier Fig. 10. Demodulation process of Rx. ˆ
RX
=21.8V
(10kHz)
0.01V
PV
PV
v
dc
v
amp
：
ˆ
RX
=17.4V
(10kHz)
0.20V
o
o
v dc
v amp
：
ˆ
TX
=21.8V
(10kHz)
0.13V
PV
PV
v
dc
v
amp
：
ˆ
TX
=17.4V
(10kHz)
1.21V
o
o
v dc
v amp
：
Fig. 11. Waveform of ˆPV
v
, ˆov of Tx, and ˆPV
v
, ˆov of Rx. (b) Digital
Controller Fig. 11. IV. EXPERIMENTAL VERIFICATION Han et al [20]
2ASK
0.7
Not mentioned Tx vo 21.3V
Rx vo 21.8V
Rx BPF
Rx demodulation
results
No switching ripple
Shoot-
Through
(a)
Tx vo 8.7V
Rx vo 19.1V
Rx BPF
Rx demodulation
results
Partial
Shading
(b)
Fig. 13. (a) Waveform of ˆov at Tx and Rx, waveform at BPF, and data de-
modulation results when RXs are working at shoot-through and bus voltage
h
i
131V (b) W
f
ˆ
f T
d R
f
BPF
d vpv 28.6 20.0V
Output power 38.9W 54.4W
Receiving signal
ipv1.36A 2.72A
MPPT
Process
0.1V/ms
MPPT Start
400us/div
Fig. 14. MPPT waveform of Rx while communication is linked. Shoot-
Through Process TABLE Ⅲ TABLE Ⅲ
COMPARISON AMONG THE PROPOSED METHOD AND RELATED WORKS
Authors [Reference]
Modem
Scheme
Bit Rate
(kbps)
Communication
Cost (USD)
The proposed method
2DPSK
2.0
0.14
W. Mao et al [21]
2ASK
10.0
5.00
E. Roman et al [9]
2FSK
2.4
12.20
H. Nosato et al [16]
Baseband
0.005
2.0
J. Han et al [20]
2ASK
0.7
Not mentioned COMPARISON AMONG THE PROPOSED METHOD AND RELATED WORKS Fig. 13. (a) Waveform of ˆov at Tx and Rx, waveform at BPF, and data de-
modulation results when RXs are working at shoot-through and bus voltage
changes into 131V. (b) Waveform at ˆov of Tx and Rx, waveform at BPF, and
data demodulation results when Tx’s input current changes into 1.60A. the dc bus voltage. In this experiment, the input current of Tx is
changed to 1.60A, and the duty-cycle is changed to 0.4. Fig. 13(b) shows experimental results of ˆov at Tx and Rx, and de-
modulation results at Rx. The waveform shows almost the same
as before, and the sequence is demodulated correctly. mercial MCU chips; thus, the software cost of this method is
acceptable in applications. Comparisons among the proposed
PLC method and some related conventional PLC methods are
presented in TABLE Ⅲ. It shows that the proposed one has
great economic competitiveness in PV industry. Therefore, in either shoot-through or partial shading situa-
tion, the amplitude of signal is large enough for demodulation. It shows that the communication system will be reliable in
different working conditions. IV. EXPERIMENTAL VERIFICATION For Tx, the output voltage couldn’t
be larger than 21.8V, hence the dc reference voltage of Tx
changes into 21.3V when transmitting signal. Experimental
results of ˆov at Tx and Rx, and demodulation results at Rx are
shown in Fig. 13(a). All the data are demodulated correctly. Because Rxs are working at shoot-through, no switching ripple
appears at the output of Rx except those from Tx. Thus, the
noise environment is even better at shoot-through state. BER is an important feature to the quality of communication. To estimate BER, Tx is set to transmit 2×105 bits in a form of
m-sequence, in which symbol 0 and symbol 1 show in equal
probability and look randomly. During the test, only one-bit
error occurs, which shows the BER is lower than 10-5. Experiments under different situations have been imple-
mented to verify the reliability. For buck-DCPO, shoot-through
is a common state, in which the duty-cycle of buck converter is
1 and the switching is always on. To simulate this situation, all
the five Rxs are set to work in shoot-though state. The dc bus In partial shading situation, the MPP of the PV panel will be
changed. In most cases, the current will decrease severely. For
the DCPO, to match its output current with the string current,
the converter will decrease its duty-cycle. Meanwhile, other
DCPOs in the string will increase their duty-cycle to balance IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS Tx vo 21.3V
Rx vo 21.8V
Rx BPF
Rx demodulation
results
No switching ripple
Shoot-
Through
(a)
Tx vo 8.7V
Rx vo 19.1V
Rx BPF
Rx demodulation
results
Partial
Shading
(b)
Fig. 13. (a) Waveform of ˆov at Tx and Rx, waveform at BPF, and data de-
modulation results when RXs are working at shoot-through and bus voltage
changes into 131V. (b) Waveform at ˆov of Tx and Rx, waveform at BPF, and
data demodulation results when Tx’s input current changes into 1.60A. vpv 28.6 20.0V
Output power 38.9W 54.4W
Receiving signal
ipv1.36A 2.72A
MPPT
Process
0.1V/ms
MPPT Start
400us/div
Fig. 14. MPPT waveform of Rx while communication is linked. TABLE Ⅲ
COMPARISON AMONG THE PROPOSED METHOD AND RELATED WORKS
Authors [Reference]
Modem
Scheme
Bit Rate
(kbps)
Communication
Cost (USD)
The proposed method
2DPSK
2.0
0.14
W. Mao et al [21]
2ASK
10.0
5.00
E. Roman et al [9]
2FSK
2.4
12.20
H. Nosato et al [16]
Baseband
0.005
2.0
J. REFERENCES [1] M. Kasper, D. Bortis and J. W. Kolar, “Classification and comparative
evaluation of PV panel-integrated DC–DC converter concepts,” IEEE
Trans. Power Electron., vol. 29, no. 5, pp. 2511-2526, May. 2014. [22] J. Wu, C. Li and X. He, "A novel power line communication technique
based on power electronics circuit topology, " in Proc. 25th Annu. IEEE
Applied Power Electronics Conf. Expo. (APEC), 2010, pp. 681-685. pp
y
[2] G. R. Walker and P. C. Sernia, “Cascaded DC-DC converter connection
of photovoltaic modules,” IEEE Trans. Power Electron., vol. 19, no. 4,
pp. 1130-1139, Jul. 2004. [23] A. Gupta and S. Mazumder, "Sequential Co-Transmission of
High-Frequency Power and Data Signals," IEEE Trans. Ind. Informat.,
doi: 10.1109/TII.2018.2794373 [3] W. Xiao, N. Ozog and W. G. Dunford, “Topology study of photovoltaic
interface for maximum power point tracking,” IEEE Trans. Ind. Elec-
tron., vol. 54, no. 3, pp. 1696-1704, Jun. 2007. [24] W. Stefanutti, S. Saggini, P. Mattavelli and M. Ghioni, “Power line
communication in digitally controlled DC–DC converters using switch-
ing frequency modulation,” IEEE Trans. Ind. Electron., vol. 55, no. 4, pp. 1509-1518, Apr. 2008. [4] E. Duran, M. Sidrach-de-Cardona, J. Galan, and J. M. Andujar, “Com-
parative analysis of buck–boost converters used to obtain I–V character-
istic curves of photovoltaic modules,” in Proc. PESC, 2008, pp. 2036–
2042. [25] J. Du, J. Wu, R. Wang, Z. Lin and X. He, “DC power-line communication
based on power/signal dual modulation in phase shift full-bridge con-
verters,” IEEE Trans. Power Electron., vol. 32, no. 1, pp. 693-702, Jan. 2017. [5] N. Femia, G. Lisi, G. Petrone, G. Spagnuolo and M. Vitelli, “Distributed
maximum power point tracking of photovoltaic arrays: Novel approach
and system analysis,” IEEE Trans. Ind. Electron., vol. 55, no. 7, pp. 2610-2621, Jul. 2008. [26] R. Wang, Z. Lin, J. Du, J. Wu and X. He, “Direct sequence spread
spectrum-based PWM strategy for harmonic reduction and communica-
tion,” IEEE Trans. Power Electron., vol. 32, no. 6, pp. 4455-4465, June
2017. [6] N. Femia, G. Petrone, G. Spagnuolo and M. Vitelli, “Optimization of
perturb and observe maximum power point tracking method,” IEEE
Trans. Power Electron., vol. 20, no. 4, pp. 963-973, Jul. 2005. [27] J. Wu, J. Du, Z. Lin, Y. Hu, C. Zhao and X. He, “Power conversion and
signal transmission integration method based on dual modulation of DC–
DC converters,” IEEE Trans. Ind. Electron., vol. 62, no. 2, pp. 1291-1300, Feb. 2015. V. CONCLUSIONS MPPT is the most important feature of DCPO, thus com-
munication should not interfere with the MPPT process. For
Tx, MPPT process and data transmission can be separated by
the program of controller, so that they do not affect each other. For Rx, the perturbation caused by communication at DCPO’s
input is small and can be filtered after sampling, hence it will
not impact MPPT process. The detailed waveform of MPPT
process is shown in Fig. 14, in which the output power is well
regulated by conventional P&O method, while communication
is linked at the same time. This paper presents a novel DCPO design, in which DCPO
can not only maximize the PV panel output power in mod-
ule-level, but also construct a data link among modules. The
proposed DCPO design integrates data communication with
power conversion. The data is modulated in the power control
loop of DCPOs and transmitted to other DCPOs. The technique
is easy for implementation, and only a few additional compo-
nents are required for data communication. A suit of experi-
mental tests has been conducted in a 466W experimental sys-
tem, where 2kbps communication among six DCPOs has been
achieved, hence the effectiveness of the technique is verified. Comparing with conventional PLC technique in PV system, the
proposed method has the advantages of low-cost and integra-
tion. Besides, by adjusting the signal carrier wave in MCU
program, different amplitudes and forms of the signal can be The above scheme is cost-efficient. For hardware cost, only a
dual-channel amplifier TI-LF353DR is used for voltage sam-
pling and BPF, and the price is 0.14USD. For MCU program,
the method needs an ADC interrupt routine operating in every
10us, in which the DFT and decode algorithm is processed. The
real processing time of the algorithm for STM32F407 is less
than 1us. Such demands can be easily satisfied by other com- IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS based photovoltaic plants parameters smart monitoring communications
and control module,” in Proc. 2nd Int. Conf. Power Eng., Energy Electr. Drives., 2011, pp. 1–6. modulated, thus the bit rate can be increased by using advanced
modulation methods, e.g. M-QAM. However, the proposed method has some constraints. First,
this PLC method can only link DCPOs in the same string. In a
PV array with several strings connected in parallel, communi-
cation among different strings may require extra hardware. V. CONCLUSIONS Second, due to the modulation mode, the carrier wave fre-
quency is limited to lower than half of switching frequency,
hence the communication rate is low comparing with some
conventional PLC methods. [18] P. Jonke, C. Eder, J. Stöckl and M. Schwark, “Development of a module
integrated photovoltaic monitoring system,” in Proc. 39th Annu. Conf. IEEE Ind. Electron. Soc. (IECON), 2013, pp. 8080-8084. [19] J. Han, C. s. Choi, W. k. Park, I. Lee and S. h. Kim, “PLC-based photo-
voltaic system management for smart home energy management system,”
IEEE Trans. Consum. Electron., vol. 60, no. 2, pp. 184-189, May. 2014. [20] J. Han, I. Lee and S. H. Kim, “User-friendly monitoring system for
residential PV system based on low-cost power line communication,”
IEEE Trans. Consum. Electron., vol. 61, no. 2, pp. 175-180, May. 2015. [21] W. Mao, X. Zhang, R. Cao, F. Wang, T. Zhao and L. Xu, “A research on
power line communication based on parallel resonant coupling technol-
ogy in PV module monitoring,” IEEE Trans. Ind. Electron., vol. 65, no. 3,
pp. 2653-2662, Mar. 2018. REFERENCES [7] O. Khan and W. Xiao, “An efficient modeling technique to simulate and
control submodule-integrated PV system for single-phase grid connec-
tion,” IEEE Trans. Sustain. Energy., vol. 7, no. 1, pp. 96-107, Jan. 2016. [28] S. Haykin, “Passband digital transmission,” in Communication System,
4th ed., New York, NY, USA: Wiley, 2001, pp. 407-409. gy
pp
[8] H. S. H. Chung, K. K. Tse, S. Y. R. Hui, C. M. Mok and M. T. Ho, “A
novel maximum power point tracking technique for solar panels using a
SEPIC or Cuk converter,” IEEE Trans. Power Electron., vol. 18, no. 3,
pp. 717-724, May. 2003. pp
y
[9] E. Roman, R. Alonso, P. Ibanez, S. Elorduizapatarietxe and D. Goitia,
“Intelligent PV module for grid-connected PV systems,” IEEE Trans. Ind. Electron., vol. 53, no. 4, pp. 1066-1073, Jun. 2006. Yue Zhu received the B.S. degree in electrical
engineering, Zhejiang University, Hangzhou, China,
in 2016, where he is currently a master student in
College of Electrical Engineering. Yue Zhu received the B.S. degree in electrical
engineering, Zhejiang University, Hangzhou, China,
in 2016, where he is currently a master student in
College of Electrical Engineering. [10] B. Andò, S. Baglio, A. Pistorio, G. M. Tina and C. Ventura, “Sentinella:
smart monitoring of photovoltaic systems at panel level,” IEEE Trans. Instrum. Meas., vol. 64, no. 8, pp. 2188-2199, Aug. 2015. pp
g
[11] P. Guerriero, F. Di Napoli, G. Vallone, V. d'Alessandro and S. Daliento,
“Monitoring and diagnostics of PV plants by a wireless self-powered
sensor for individual panels,” IEEE J. Photovolt., vol. 6, no. 1, pp. 286-294, Jan. 2016. During his study, he focuses on power optimiza-
tion of renewable generation and communication
technique applied in power electronics. During his study, he focuses on power optimiza-
tion of renewable generation and communication
technique applied in power electronics. [12] P. Guerriero, F. Di Napoli, V. d'Alessandro and S. Daliento, “A wireless
controlled circuit for PV panel disconnection in case of fire,” in Proc. IEEE Int. Symp. Power Electron. Elect. Drives, Autom. Motion, 2014, pp. 982-986. [13] S. Moon, S. G. Yoon and J. H. Park, “A new low-cost centralized MPPT
controller system for multiply distributed photovoltaic power condition-
ing modules,” IEEE Trans. Smart Grid., vol. 6, no. 6, pp. 2649-2658,
Nov. 2015. Jiande Wu (M’11) was born in Zhejiang, China, in
1973. He received the B.Sc., M.SC and Ph.D. REFERENCES degree from the College of Electrical Engineering,
Zhejiang University, Hangzhou, China, in 1994,
1997 and 2012, respectively. Since 1997, he has
been a faculty member at Zhejiang University,
where he is currently an associate professor. From
2013 to 2014, he was an academic visitor at the
University of Strathclyde, Glasgow, U.K. His
research interests include power electronics control,
distributed power electronics system and fieldbus Jiande Wu (M’11) was born in Zhejiang, China, in
1973. He received the B.Sc., M.SC and Ph.D. degree from the College of Electrical Engineering,
Zhejiang University, Hangzhou, China, in 1994,
1997 and 2012, respectively. Since 1997, he has
been a faculty member at Zhejiang University,
where he is currently an associate professor. From
2013 to 2014, he was an academic visitor at the
University of Strathclyde, Glasgow, U.K. His
research interests include power electronics control,
distributed power electronics system and fieldbus communication. [14] M. E. Andreoni López, F. J. Galdeano Mantiñan and M. G. Molina,
“Implementation of wireless remote monitoring and control of solar
photovoltaic (PV) system,” in Proc. 6th IEEE/PES Transm. Distrib.:
Latin Amer. Conf. Expo. (T&D-LA), 2012, pp. 1-6. [15] A. Haidine, B. Adebisi, A. Treytl, H. Pille, B. Honary and A. Portnoy,
“High-speed
narrowband
PLC
in
Smart
Grid
landscape
—
State-of-the-art,” in Proc. 15th Int. Symp. Power Line Commun. Its Appl. (ISPLC), 2011, pp. 468-473. pp
[16] H. Nosato, Y. Kasai, E. Takahashi and M. Murakawa, "A very low-cost
low-frequency PLC system based on DS-CDMA for DC power lines," in
Proc. 16th Int. Symp. Power Line Commun. Its Appl. (ISPLC), 2012, pp. 398-403. communication. [17] F. J. Sánchez-Pacheco, P. J. Sotorrio-Ruiz, J. R. Heredia-Larrubia, F. Pérez-Hidalgo and M. Sidrach-de-Cardona, “Low cost DC lines PLC IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS Ruichi Wang (S’14) received the B.S. degree in
electrical engineering from Zhejiang University,
Hangzhou, China, in 2013. Currently, she is working
toward the Ph.D. degree in the College of Electrical
Engineering,
Zhejiang
University,
Hangzhou,
China. She visited Aston University, Birmingham,
UK, from September, 2015 to March, 2016. Her
current research interests include communication
technique applied in power electronics and EMI
mitigation for SMPS. Ruichi Wang (S’14) received the B.S. degree in
electrical engineering from Zhejiang University,
Hangzhou, China, in 2013. Currently, she is working
toward the Ph.D. degree in the College of Electrical
Engineering,
Zhejiang
University,
Hangzhou,
China. She visited Aston University, Birmingham,
UK, from September, 2015 to March, 2016. REFERENCES Her
current research interests include communication
technique applied in power electronics and EMI
mitigation for SMPS. Zhengyu Lin (S’03–M’05–SM’10) received the
B.Sc. and M.Sc. degrees from the College of
Electrical
Engineering,
Zhejiang
University,
Hangzhou, China, in 1998 and 2001, respectively,
and the Ph.D. degree from Heriot-Watt University,
Edinburgh, U.K.,in 2005. He is currently a Lecturer
in Electrical, Electronic and Power Engineering with
Aston University, Birmingham, U.K. He was a
Research Associate with the University of Sheffield
from 2004 to 2006, an R&D Engineer with Emerson
Industrial Automation, Control Techniques PLC from 2006 to 2011, a Senior
Research Scientist with Sharp Laboratories of Europe Ltd. from 2011 to 2012,
and a Lecturer with Coventry University from 2013 to 2014. His research
interests include power electronics and its applications in renewable energy,
energy storage, motor drives and power systems. Xiangning He (M’95--SM’96--F’10) received the
B.Sc. and M.Sc. degree from Nanjing University of
Aeronautical and Astronautical, Nanjing, China, in
1982 and 1985, respectively, and Ph.D. degree from
Zhejiang University, Hangzhou, China, in 1989. In 1991, he obtained a Fellowship from the Royal
Society of U.K., and conducted research in
Heriot-Watt University, Edinburgh, U.K., as a
Post-Doctoral Research Fellow for two years. In
1994, he joined Zhejiang University as an Associate
Professor. Since 1996, he has been a Full Professor in the College of Electrical
Engineering, Zhejiang University. His research interests are power electronics
and their industrial applications. Dr. He is a Fellow of The Institute of Electrical and Electronics Engineers
(IEEE) and has been appointed as IEEE Distinguished Lecturer by the IEEE
Power Electronics Society in 2011. He is also a Fellow of the Institution of
Engineering and Technology (formerly IEE), U.K. In 1991, he obtained a Fellowship from the Royal
Society of U.K., and conducted research in
Heriot-Watt University, Edinburgh, U.K., as a
Post-Doctoral Research Fellow for two years. In
1994, he joined Zhejiang University as an Associate
Professor. Since 1996, he has been a Full Professor in the College of Electrical
Engineering, Zhejiang University. His research interests are power electronics
and their industrial applications. Dr. He is a Fellow of The Institute of Electrical and Electronics Engineers
(IEEE) and has been appointed as IEEE Distinguished Lecturer by the IEEE
Power Electronics Society in 2011. He is also a Fellow of the Institution of
Engineering and Technology (formerly IEE), U.K. Dr. REFERENCES He is a Fellow of The Institute of Electrical and Electronics Engineers
(IEEE) and has been appointed as IEEE Distinguished Lecturer by the IEEE
Power Electronics Society in 2011. He is also a Fellow of the Institution of
Engineering and Technology (formerly IEE), U.K.
|
https://openalex.org/W2907616410
|
https://www.ufrgs.br/det/index.php/det/article/download/490/252
|
Portuguese
| null |
Diretrizes para projeto de e-books com enfoque em crianças em fase de alfabetização
|
Design & Tecnologia
| 2,018
|
cc-by
| 16,151
|
Resumo Neste artigo, os autores propõem-se a explorar de que modo o projeto de e-books voltados a crianças em fase de
alfabetização pode ser aprimorado, de modo a fazer uso mais consciente dos recursos tecnológicos e, assim, melhorar a
experiência de leitura infantil nesse contexto de ensino-aprendizagem. Para isso, partem de uma pesquisa bibliográfica
dividida em duas partes: a primeira busca identificar características do público-alvo, suas capacidades e limitações no que diz
respeito à leitura em suporte digital; a segunda, por sua vez, procura elucidar os aspectos técnicos envolvidos no projeto de
livros digitais infantis, sob o enfoque do design, abordando tópicos como tipografia, ilustração e diagramação. Posteriormente, realizam uma análise de similares, em que observam e avaliam dez publicações digitais voltadas ao público
infantil. A partir dos dados obtidos, propõem, então, diretrizes preliminares de projeto de e-books infantis e destacam alguns
pontos de discussão, que são avaliados por seis especialistas, selecionados entre designers gráficos com experiência
teórico/prática e especialistas em educação e/ou pedagogos. As diretrizes preliminares e pontos de discussão passam, então,
por uma reavaliação, a partir dos apontamentos feitos pelo corpo de especialistas consultados (a consulta direta com o
público infantil não foi realizada até esta etapa da pesquisa). Os autores definem, por fim, como resultado da pesquisa, vinte
diretrizes de projeto de e-books infantis que permitam ao designer envolvido em um projeto de e-book infantil explorar de
forma adequada os recursos tecnológicos, aprimorando a qualidade técnica dos livros digitais em contexto de alfabetização –
pretende-se, dessa forma, contribuir com os esforços acadêmicos na área, elucidando formas de se trabalhar a leitura em
suporte digital como ferramenta efetiva em ambiente educacional. Palavras-chave: Design editorial digital, E-books, Livros infantis. www.pgdesign.ufrgs.br www.pgdesign.ufrgs.br Design & Tecnologia 16 (2018) Thais A. Fensterseifer, thaisarnold@gmail.com – Programa de Pós-Graduação em Design,
Departamento de Design e Exp. Gráfica, Universidade Federal do Rio
Grande do Sul, Porto Alegre, Brasil Fábio G. Teixeira, fabiogt@ufrgs.br
– Programa
de
Pós-Graduação
em
Design,
Departamento de Design e Exp. Gráfica, Universidade Federal do Rio
Grande do Sul, Porto Alegre, Brasil fabiogt@ufrgs.br
– Programa
de
Pós-Graduação
em
Design,
Departamento de Design e Exp. Gráfica, Universidade Federal do Rio
Grande do Sul, Porto Alegre, Brasil Keywords: Digital design, eBooks, Children’s eBooks. Palavras-chave: Design editorial digital, E-books, Livros infantis. Diretrizes para projeto de e-books com enfoque
em crianças em fase de alfabetização Thais A. Fensterseifer, 2.1 As crianças em fase de alfabetização e a leitura
em suporte digital O problema é que esses recursos podem,
eventualmente, ao invés de aumentar o engajamento e
facilitar a compreensão da criança com relação à história,
acabar por ser meras distrações, elementos que desviam o
foco da ação de ler. Por outro lado, é possível considerar-se
que, em um mundo cercado pela tecnologia digital, os e-
books façam justamente esta ‘ponte’ entre os atrativos
tecnológicos e a leitura do texto escrito, auxiliando, assim, o
processo de alfabetização e incentivando a leitura e o
interesse pelos livros – ainda que digitais. Para Roger Chartier
(2007), esse processo é inevitável: “o essencial da leitura,
hoje, passa pela tela”. Nota-se, assim, a relevância do tema, tanto no sentido
de entender as mudanças sociais ocorridas a partir da
mudança do suporte de leitura, quanto no sentido de
explorar as novas possibilidades de projeto decorrentes da
concretização desse novo cenário. O que se observa (e tal
percepção será demonstrada ao longo deste artigo) é que
grande parte dos e-books infantis disponíveis no mercado
para compra ou para download gratuito, não têm, de modo
geral, apresentado uso adequado da tecnologia. A simples
digitalização de livros infantis impressos, por exemplo, resulta
em e-books com problemas de proporção de página (com mal
aproveitamento da tela do e-reader), com subutilização de
recursos de áudio e de vídeo, e com composições tipográficas
que não favorecem a leitura em tela, oferecendo ao leitor
uma experiência de uso limitada. Nesse sentido, estudo de campo realizado por Colombo
& Landoni (2014) concluiu, justamente, que muitas crianças
têm expectativa de uma experiência de leitura melhor (ou,
pelo menos, diferente) com o livro digital, em comparação
com o livro impresso. As autoras configuraram sua pesquisa
com dois grupos de crianças: o primeiro foi colocado em
contato com um e-book ‘básico’ – simplesmente uma
digitalização de um livro infantil impresso –, e o segundo
grupo teve acesso a um e-book ‘aprimorado’ – contendo
diversas opções de interação, além de áudio e vídeo. O
primeiro grupo teve uma experiência negativa com o livro
digital, que falhou em atender suas expectativas. As crianças
desse grupo relataram preferir o livro impresso em
detrimento do digital: quase um terço das crianças
participantes desse grupo não finalizou a leitura do livro. Por
outro lado, as crianças do segundo grupo experimental
terminaram de ler o e-book e demonstraram maior intenção
e vontade de repetir a atividade. 1. INTRODUÇÃO Na primeira década do século XXI, quando se iniciou a
popularização dos e-books (abreviação de eletronic books,
livros digitais), publicações desse tipo voltadas ao público
infantil não eram usuais, conforme afirma Guernsey (2011). Para a autora, a atenção estava voltada a livros de texto
focados no público adulto e disponíveis para download e
leitura em e-readers como o Kindle (leitor de livros digitais
produzido pela empresa Amazon). No entanto, com o
lançamento de aparelhos portáteis com telas coloridas
sensíveis ao toque, esse cenário foi mudando. A chegada do
iPad, da empresa americana Apple, transformou revistas
cheias de fotos e livros totalmente coloridos em uma
realidade para os leitores digitais. Assim, foi surgindo um sem
número de produtos especializados no público infantil,
fazendo uso de atrativos como cor, ilustração e animação
(GUERNSEY, 2011). 2.1 As crianças em fase de alfabetização e a leitura
em suporte digital Para Dresang (1999, p.1123), “o ambiente digital online
proporciona uma ‘arena’ poderosa, sem precedentes, para os
jovens
obterem
informação
para
tomar
parte
nas
comunidades globais de aprendizado”. Estudos recentes, no
entanto, sugerem que boa parte das crianças fazem uso
apenas superficial das tecnologias, não ‘aproveitando’ as
mídias para exercer sua criatividade e sua capacidade de
inovação (Facer et al., 2003; Holloway; Valentine, 2003;
Livingstone; Bober, 2004 apud Buckingham, 2010). Nesse
cenário, o papel do designer é fundamental, já que cabe a
esse profissional projetar o melhor uso possível dos recursos
gráficos e tecnológicos disponíveis, desenvolvendo, assim, e-
books que sejam adequados ao público-alvo e que facilitem o
acesso ao conteúdo, permitindo que o livro digital assuma
funções relevantes no processo de ensino-aprendizagem. Para Roger Chartier (2007), o uso do meio digital pelas
crianças, quando acompanhado de métodos pedagógicos,
pode, inclusive, melhorar o processo de alfabetização,
tornando mais rápido e efetivo o aprendizado da leitura e
aprimorando o domínio da língua escrita. No Brasil, a grande
maioria das escolas, em especial da rede pública, trabalha
com o método pedagógico tradicional, em que o professor é
figura central, detentor do conhecimento. Em termos de
leitura, nas séries iniciais, é comum a prática da “Hora do
Conto”, observada por Klohn & Fensterseifer (2012), em que
o professor lê o livro em voz alta, mostrando as ilustrações
aos alunos, normalmente sentados no chão, em círculo. Quanto ao acesso e ao uso da tecnologia, Buckingham (2010,
p.44) afirma que “poucas escolas oferecem amplo ou
irrestrito acesso à Internet” – e, para o autor, quando o
contato com o computador e com as redes ocorre, no meio
escolar, acaba ficando limitado, por exemplo, à visitação de
sites específicos sugeridos pelo professor (experiência muito
menos multimídia do que algumas crianças têm fora da
escola). Um dos questionamentos que surgem, quando se fala
em público infantil e em leitura em suporte digital, é o
quanto a tecnologia auxiliará ou dificultará o processo de
alfabetização desta nova geração. A criança tem à sua
disposição, no suporte digital, múltiplas formas de interação
com o livro: é possível ouvir o narrador, assistir a ilustrações
em movimento, e alguns livros até dispõem de jogos
integrados. Abstract In this paper, the authors explore how the project of eBooks aimed at children learning to read can be improved in order to
take better advantage of technological resources and thus enhance children's reading experience in this educational context. In
order to do that, it begins with a literature research, divided in two main areas: the first aims to understand the target
audience, their abilities and limitations with regard to reading, and the second approaches children’s eBooks, involving
technical aspects under the focus of graphic design, such as typography, illustration and editorial design. Subsequently, an
analysis on children’s books available in the market is performed on ten digital publications focused at children, which are
observed and analyzed. From the data obtained so far, the authors propose some preliminary guidelines and discussion points,
which are evaluated by experts, selected among senior designers, with theoretical or market experience, and professionals of
education or pedagogy. After the interview with the experts, the preliminary guidelines and the discussion points are
reevaluated (direct consultation with children was not held until this stage of the research). The authors propose, finally, as a
result of this research, twenty design guidelines for children's eBooks, enabling the designer to explore adequately the
technological resources, and improving the technical quality of digital books in an educational context – the authors intend, by
presenting this research, to contribute with the academic efforts in this area, elucidating ways of making digital reading an
effective tool in the educational environment. Design & Tecnologia 16 (2018) 53 suporte digital; a segunda aborda o livro digital infantil, sob o
enfoque do projeto gráfico-visual. 2.1 As crianças em fase de alfabetização e a leitura
em suporte digital Para as autoras do estudo,
conclui-se que “e-books com uma melhor experiência de
leitura podem motivar as crianças – especialmente aquelas
relutantes com a atividade da leitura (...) – a lerem mais”
(COLOMBO & LANDONI, 2014, p.142). Elas perceberam,
ainda, que as crianças do segundo grupo que não haviam
finalizado a leitura do livro impresso acabaram por ler o livro
digital do início até o final, corroborando com as conclusões
da pesquisa – algumas crianças relataram, inclusive, não
terem gostado da história, mas, apesar disso, terem Este artigo procura analisar, assim, o projeto de e-books
infantis, buscando definir parâmetros que auxiliem os
designers a fazer melhor uso dos recursos gráficos e
tecnológicos disponíveis em ambiente digital, desenvolvendo,
assim, e-books que sejam adequados ao público infantil em
fase de alfabetização. 2. REVISÃO BIBLIOGRÁFICA As mesmas autoras
também concluíram que a disponibilidade de atividades de
ler + jogar ou de ler + ouvir existentes em alguns e-books
aprimora as competências de alfabetização e de letramento
infantil, quando se compara essa experiência à experiência
com um livro que disponibiliza apenas função ‘básica’ de
leitura. Tais pesquisas oferecem subsídios para crer que o
livro digital infantil tem potencialidade para assumir funções
relevantes no processo de ensino-aprendizagem, se os
recursos
possibilitados
pelo
suporte
digital
forem
devidamente explorados. Os e-books de layout fixo, por sua vez, permitem menos
customização por parte do usuário e, portanto, conferem
mais controle ao designer sobre o layout do livro. Esse tipo de
layout é frequentemente usado em projetos em que o layout
da página é mais complexo do que uma coluna única de
texto, ou, ainda, quando há necessidade de posicionar
determinado conteúdo em um ponto fixo na página. Considera-se que a configuração em layout fixo é, de modo
geral, mais adequada para o desenvolvimento de e-books
infantis, visto que permite explorar a relação entre texto e
ilustração/animação de maneira mais complexa do que a
possível nos e-books de layout fluido (em que há a restrição
do layout de coluna única, por exemplo), além de permitir
que o designer trabalhe a tipografia e a seleção das fontes de
forma mais cuidadosa e precisa. Analisando-se os tipos de e-
book mais comumente utilizados no mercado – Mobipocket,
KF8, ePub2, ePub3, epic, iBooks Author e PDF –, que foram
avaliados e comparados, concluiu-se que os formatos ePub2
e ePub3 são os que, de maneira geral, mostram-se mais
adequados ao projeto de e-books voltados ao público infantil,
visto serem compatíveis com ampla quantidade de e-readers,
oferecendo suporte ao uso de layout fixo e à inclusão de
mídia e áudio ao livro. Nos aspectos relacionados à seleção cromática e ao
estilo de ilustração, observou-se, por meio da bibliografia
consultada, as diversas possibilidades de exploração desses
recursos, sem que houvesse alguma recomendação de paleta
cromática mais adequada ao público-alvo da pesquisa ou de
estilo de ilustração que causasse maior apelo visual junto às
crianças. 2. REVISÃO BIBLIOGRÁFICA Nos aspectos relacionados à seleção cromática e a
estilo de ilustração, observou-se, por meio da bibliograf
consultada, as diversas possibilidades de exploração dess
recursos, sem que houvesse alguma recomendação de pale
cromática mais adequada ao público-alvo da pesquisa ou d
estilo de ilustração que causasse maior apelo visual junto
crianças. Apenas duas pesquisas de campo consultad
indicaram resultados nesse sentido: Brookshire, Scharff
Moses (2002) notaram uma preferência infantil p
ilustrações de estilo realista (em detrimento de ilustraçõ
mais abstratas), em cores vibrantes (em detrimento de cor
mais neutras ou sóbrias); no mesmo sentido, Klohn
Fensterseifer (2012) observaram em um grupo de crianç No mesmo sentido, também atribuindo melhoras no
processo de ensino-aprendizagem à utilização de livros
digitais, diversos autores têm observado que a combinação
da leitura visual com o áudio tem o potencial de fazer a
leitura infantil em suporte digital uma experiência mais rica e
atraente para as crianças, como se lê a seguir. Estudo
realizado por Roskos, Burstein, Shang e Gray (2014) compilou
diversificado aporte teórico para atestar os benefícios da
sincronia audiovisual na sustentação da atenção infantil ao
conteúdo do e-book, alcançando melhores resultados de
aprendizagem da leitura: "o realce das palavras e a
contiguidade temporal do áudio (narração e música) com
informação visual (ilustrações) parece chamar a atenção
visual das crianças para as figuras e as palavras, de maneira a
tornar o texto mais ‘concreto’, fazendo-o, assim, mais real e
mais memorável para elas" (ROSKOS, BURSTEIN, SHANG &
GRAY, 2014). Lewin (2000 apud AMANTE, 2007), no mesmo
sentido, assinala que os programas interativos multimídia
complementam o desenvolvimento do letramento infantil na
medida em que os recursos de áudio permitem trabalhar
junto às crianças a pronúncia correta das palavras, além de
proporcionar leitura fracionada em sílabas ou centrada em
segmentos fonêmicos, desenvolvendo, desse modo, a
consciência fonológica da criança. Cabe ressaltar que, quando
se trata, aqui, de “pronúncia correta”, está-se referindo à
pronúncia que privilegie a norma culta da língua,
respeitando-se, evidentemente, que há diferenças regionais
na fala. A mesma pesquisa de Roskos, Burstein, Shang e Gray
(2014) concluiu que “a sincronia espacial e temporal de olhar
(recursos visuais), ouvir (recursos de áudio) e tocar (recursos
hápticos) pode ser o ‘ponto central’ que atrai a atenção da
criança ao texto digital, de forma a respaldar a experiência de
alfabetização e de aprendizagem". 2. REVISÃO BIBLIOGRÁFICA Esta seção apresenta um breve resumo da investigação
bibliográfica realizada a fim de sustentar as fases
subsequentes do trabalho, oferecendo subsídios preliminares
para a geração de diretrizes de projeto de e-books infantis. A
revisão foi dividida em duas partes principais: a primeira
busca identificar características do público-alvo, suas
capacidades e limitações no que diz respeito à leitura em Design & Tecnologia 16 (2018) 54 finalizado a leitura por “gostarem de interagir com o livro
enquanto liam o texto”. Dentre as conclusões de sua pesquisa
de campo, as autoras citam, por fim, que “a experiência de
leitura infantil em suporte digital seria bastante beneficiada a
partir da inclusão de multimídia e de interatividade – desde
que tais características proporcionem uma contribuição
significativa para a história” (COLOMBO & LANDONI, 2014,
p.143). Há, basicamente, dois tipos principais de e-books no que
tange ao layout: o layout padrão/fluido e o layout fixo. Entender as características de cada um e as diferenças entre
eles é primordial, pois essa é uma das primeiras decisões de
projeto quando se fala em desenvolvimento de e-books –
essa decisão influencia, inclusive, na definição do formato
digital por meio do qual o livro será disponibilizado. Os e-
books de layout fluido (também chamado padrão ou
responsivo) são o tipo mais comum de publicação digital. Utilizado para conteúdos compostos por texto e poucos
recursos visuais, o layout fluido permite que o conteúdo do
livro se adapte à largura da tela do e-reader, de modo que o
leitor não precise fazer uso de ferramentas como zoom e
scroll para ler. O layout fluido permite que o leitor altere a
fonte que compõe o texto e que aumente ou diminua o corpo
do texto, customizando o livro para sua experiência de uso. finalizado a leitura por “gostarem de interagir com o livro
enquanto liam o texto”. Dentre as conclusões de sua pesquisa
de campo, as autoras citam, por fim, que “a experiência de
leitura infantil em suporte digital seria bastante beneficiada a
partir da inclusão de multimídia e de interatividade – desde
que tais características proporcionem uma contribuição
significativa para a história” (COLOMBO & LANDONI, 2014,
p.143). 2. REVISÃO BIBLIOGRÁFICA No mesmo sentido, também atribuindo melhoras no
processo de ensino-aprendizagem à utilização de livros
digitais, diversos autores têm observado que a combinação
da leitura visual com o áudio tem o potencial de fazer a
leitura infantil em suporte digital uma experiência mais rica e
atraente para as crianças, como se lê a seguir. Estudo
realizado por Roskos, Burstein, Shang e Gray (2014) compilou
diversificado aporte teórico para atestar os benefícios da
sincronia audiovisual na sustentação da atenção infantil ao
conteúdo do e-book, alcançando melhores resultados de
aprendizagem da leitura: "o realce das palavras e a
contiguidade temporal do áudio (narração e música) com
informação visual (ilustrações) parece chamar a atenção
visual das crianças para as figuras e as palavras, de maneira a
tornar o texto mais ‘concreto’, fazendo-o, assim, mais real e
mais memorável para elas" (ROSKOS, BURSTEIN, SHANG &
GRAY, 2014). Lewin (2000 apud AMANTE, 2007), no mesmo
sentido, assinala que os programas interativos multimídia
complementam o desenvolvimento do letramento infantil na
medida em que os recursos de áudio permitem trabalhar
junto às crianças a pronúncia correta das palavras, além de
proporcionar leitura fracionada em sílabas ou centrada em
segmentos fonêmicos, desenvolvendo, desse modo, a
consciência fonológica da criança. Cabe ressaltar que, quando
se trata, aqui, de “pronúncia correta”, está-se referindo à
pronúncia que privilegie a norma culta da língua,
respeitando-se, evidentemente, que há diferenças regionais
na fala. A mesma pesquisa de Roskos, Burstein, Shang e Gray
(2014) concluiu que “a sincronia espacial e temporal de olhar
(recursos visuais), ouvir (recursos de áudio) e tocar (recursos
hápticos) pode ser o ‘ponto central’ que atrai a atenção da
criança ao texto digital, de forma a respaldar a experiência de
alfabetização e de aprendizagem". As mesmas autoras
também concluíram que a disponibilidade de atividades de
ler + jogar ou de ler + ouvir existentes em alguns e-books
aprimora as competências de alfabetização e de letramento
infantil, quando se compara essa experiência à experiência
com um livro que disponibiliza apenas função ‘básica’ de
leitura. Tais pesquisas oferecem subsídios para crer que o
livro digital infantil tem potencialidade para assumir funções
relevantes no processo de ensino-aprendizagem, se os
recursos
possibilitados
pelo
suporte
digital
forem
devidamente explorados. 2. REVISÃO BIBLIOGRÁFICA Há, basicamente, dois tipos principais de e-books no qu
tange ao layout: o layout padrão/fluido e o layout fix
Entender as características de cada um e as diferenças ent
eles é primordial, pois essa é uma das primeiras decisões d
projeto quando se fala em desenvolvimento de e-books
essa decisão influencia, inclusive, na definição do forma
digital por meio do qual o livro será disponibilizado. Os
books de layout fluido (também chamado padrão o
responsivo) são o tipo mais comum de publicação digita
Utilizado para conteúdos compostos por texto e pouc
recursos visuais, o layout fluido permite que o conteúdo d
livro se adapte à largura da tela do e-reader, de modo que
leitor não precise fazer uso de ferramentas como zoom
scroll para ler. O layout fluido permite que o leitor altere
fonte que compõe o texto e que aumente ou diminua o corp
do texto, customizando o livro para sua experiência de uso. Os e-books de layout fixo, por sua vez, permitem men
customização por parte do usuário e, portanto, confere
mais controle ao designer sobre o layout do livro. Esse tipo d
layout é frequentemente usado em projetos em que o layo
da página é mais complexo do que uma coluna única d
texto, ou, ainda, quando há necessidade de posicion
determinado conteúdo em um ponto fixo na págin
Considera-se que a configuração em layout fixo é, de mod
geral, mais adequada para o desenvolvimento de e-boo
infantis, visto que permite explorar a relação entre texto
ilustração/animação de maneira mais complexa do que
possível nos e-books de layout fluido (em que há a restriçã
do layout de coluna única, por exemplo), além de permit
que o designer trabalhe a tipografia e a seleção das fontes d
forma mais cuidadosa e precisa. Analisando-se os tipos de
book mais comumente utilizados no mercado – Mobipocke
KF8, ePub2, ePub3, epic, iBooks Author e PDF –, que fora
avaliados e comparados, concluiu-se que os formatos ePub
e ePub3 são os que, de maneira geral, mostram-se ma
adequados ao projeto de e-books voltados ao público infant
visto serem compatíveis com ampla quantidade de e-reader
oferecendo suporte ao uso de layout fixo e à inclusão d
mídia e áudio ao livro. 2. REVISÃO BIBLIOGRÁFICA Há ainda, autores que defendem a
composição do texto infantil inteiramente em caixa alta, visto
que o repertório visual das crianças ainda é bastante limitado
e, por isso, a leitura infantil envolve menos reconhecimento
visual instantâneo e mais operações de análise e de síntese
(KATO, 1999 apud RUMJANEK, 2009). Dessa forma, as
palavras lidas por uma criança seriam decifradas a partir de
unidades menores (letras e, depois, sílabas) o que indicaria,
por consequência, que os caracteres maiúsculos são mais
apropriados, uma vez que o primeiro contato infantil com as
letras costuma ser com o alfabeto das maiúsculas. compartimentação e conjunção –, vinculados à disposição
dos elementos no espaço da página (Figura 1). A autora
defende que haja variação entre os tipos de diagramação de
página a página, o que garantiria maior fluidez no
encadeamento do livro. Figura 1: Imagens dos tipos de diagramação: LINDEN (2011). Esquemas: KLOHN & FENSTERSEIFER (2012). Ainda quanto à tipografia, tão importante quanto a
escolha tipográfica é a composição do texto – espaçamento
entrelinhas, comprimento da linha e corpo da fonte são
algumas características que têm influência direta na
facilitação do contato com o texto, quando se fala em leitores
iniciantes. Para Lourenço (2011), o recurso de hifenação do
texto deve ser evitado, e as quebras de linha devem ocorrer
sempre de acordo com a sintaxe do texto. O autor afirma,
ainda, que “para as crianças existe uma real necessidade de
um espaçamento maior entre letras e entre palavras do que
para um adulto” (LOURENÇO, 2011, p.123). Para Burt, Cooper
e Martin (1955), considerando-se um público-alvo de crianças
até 8 anos de idade, o tamanho mínimo do tipo deve ser de
18 pontos, com linhas compostas, preferencialmente, por até
38 caracteres e blocos de texto contendo até 18 linhas, no
máximo. Apesar de as recomendações dos autores terem
enfoque em publicações impressas, esses valores podem ser
transpostos para composições de texto em suporte digital. Quanto aos recursos tecnológicos, diversos estudos
relatam, como já se viu, bons resultados no processo de
alfabetização decorrentes da utilização de recursos de áudio
de maneira simultânea à leitura do texto escrito. Para
Colombo e Landoni (2014), a combinação da leitura visual
com o áudio tem o potencial de fazer a leitura infantil em
suporte digital uma experiência mais rica e atraente para as
crianças. 2. REVISÃO BIBLIOGRÁFICA Apenas duas pesquisas de campo consultadas
indicaram resultados nesse sentido: Brookshire, Scharff e
Moses (2002) notaram uma preferência infantil por
ilustrações de estilo realista (em detrimento de ilustrações
mais abstratas), em cores vibrantes (em detrimento de cores
mais neutras ou sóbrias); no mesmo sentido, Klohn &
Fensterseifer (2012) observaram, em um grupo de crianças
em fase de alfabetização, que a maioria tinha preferência por
livros com paletas de cores variadas, em detrimento de livros
monocromáticos ou compostos por cores menos saturadas. 2.2 Projeto de e-books voltados ao público infantil
O projeto de e-books envolve diversos aspectos, como
diagramação, tipografia, seleção cromática, ilustração,
seleção e projeto dos recursos de áudio, de animação e de
vídeo,
planejamento
da
inserção
de
recursos
de
interatividade e de hipertexto, consideração de aspectos
relacionados à usabilidade e à experiência do usuário, além
de definições técnicas que envolvem, por exemplo, a escolha
do formato de arquivo do e-book, tendo influência direta no
tipo de e-reader que poderá ser utilizado para leitura e na
plataforma de distribuição em que o e-book será
comercializado. Nesta seção, esses aspectos serão abordados
de forma resumida, destacando as principais decisões de
projeto que cabem ao profissional de design. Em termos de diagramação, é importante observar que
a já comentada opção pelo e-book de layout fixo também se
sustenta na ampliação de possibilidades de diagramação das
páginas, uma vez que, no livro infantil – seja impresso, seja
digital –, a diagramação é regida pela ilustração, e é por meio
das relações propostas entre ilustração, texto e demais
elementos de layout que é condicionado o ‘discurso’ do livro. Assim, a compreensão de opções diversas para diagramação
do e-book voltado a crianças viabiliza uma implementação de
técnicas de projeto editorial mais efetivas. A diagramação do
livro infantil, para Sophie Van der Linden (2011), pode ser
classificada em quatro tipos – dissociação, associação, Design & Tecnologia 16 (2018) 55 que se escreve. Para Walker, o primordial é a diferenciação
entre os caracteres infantis, principalmente no que tange às
letras ‘a’, ‘g’ e ‘o’ (motivo que leva outras correntes teóricas a
defenderem o uso de letras serifadas em publicações infantis,
visto que a diferenciação entre caracteres costuma ser maior
nesse tipo de fonte). 2. REVISÃO BIBLIOGRÁFICA Isso ocorre, para as autoras, porque a combinação
de texto escrito e som proporciona às crianças o acesso ao
conteúdo do livro por meio de dois sentidos diferentes –
visão e audição – o que as obriga a dedicar mais recursos de
atenção à atividade de leitura. Há, por outro lado, poucos
estudos realizados acerca das potencialidades educacionais
do uso de recursos de animação e de vídeo em e-books, no
sentido da facilitação do acesso infantil ao texto. Diversos
autores, no entanto, como Lisa Guernsey (2011), relatam
preocupações no sentido de que esse tipo de recurso seja,
realmente, fator de distração e de afastamento da criança em
relação ao texto. Figura 1: Imagens dos tipos de diagramação: LINDEN (2011). Esquemas: KLOHN & FENSTERSEIFER (2012). Quanto à tipografia mais adequada para crianças em
fase de alfabetização, não houve consenso entre os autores
consultados, razão pela qual a questão foi levada, como se
verá mais adiante neste artigo, à apreciação de especialistas. Autores como Roa (2013) e Lourenço (2011) defendem a
utilização de fontes com caracteres infantis, que são
estruturas de traço diferente para os caracteres ‘a’ e ‘g’, com
formas que correspondem de maneira aproximada ao modo
como as crianças escrevem” (Figura 2). Figura 2: Caracteres ‘adultos’ e caracteres ‘infantis’. Adaptado de Lourenço (2011). Para suprir as lacunas da pesquisa bibliográfica realizada
e para que fosse possível aos autores desta pesquisa
posicionarem-se com relação a aspectos em que não houve
consenso entre os autores consultados, optou-se por realizar
uma análise de similares, avaliando e-books infantis
disponíveis no mercado nacional e internacional e,
posteriormente, consultar especialistas, de modo a triangular
os dados a que se teve acesso – tais etapas são descritas mais
detalhadamente na seção seguinte, dedicada à metodologia
de pesquisa. Figura 2: Caracteres ‘adultos’ e caracteres ‘infantis’. Adaptado de Lourenço (2011). No entanto, há algumas evidências de que a utilização
desse tipo de caractere não traria tantas vantagens à
alfabetização infantil. A pesquisadora Sue Walker (2005)
notou, em pesquisa de campo, que a maioria das crianças
sabia que existiam diferentes desenhos dos caracteres ‘a’ e
‘g’, e algumas até formularam uma explicação para isso,
referindo que o ‘a adulto’ é o que se lê e que o ‘a infantil’ é o 3. METODOLOGIA A metodologia de trabalho proposta para a presente pesquisa
compreende cinco etapas: 1) estudar; 2) entender e observar; Design & Tecnologia 16 (2018) Design & Tecnologia 16 (2018) 56 3) construir; 4) avaliar e 5) concluir. As duas primeiras etapas
estão ligadas à compreensão e ao estudo dos fenômenos
relacionados à temática de pesquisa; a terceira e a quarta
etapas compreendem a construção teórica, que envolve
análise de dados e avaliação desses, em que possíveis ajustes
podem ser aplicados aos resultados do trabalho, expostos na quinta etapa, de conclusão. Em cada uma das etapas,
especificam-se os procedimentos que foram adotados, a fim
de alcançar os objetivos a que esta pesquisa se propõe
(Figura 3). Figura 3: Procedimentos metodológicos. Figura 3: Procedimentos metodológicos. Assim,
para
que
fosse
possível
identificar
as
características principais do público infantil em fase de
alfabetização, compreender as especificidades desse leitor na
era digital, observar e identificar as principais características e
possibilidades de e-books e e-readers, realizou-se pesquisa
bibliográfica em livros, artigos e websites relacionados às
áreas de interesse. que 1) não eram adequados à faixa etária que corresponde
ao público-alvo da presente pesquisa (crianças em fase de
alfabetização);
2)
não
tinham
conteúdo
narrativo
(enciclopédias, dicionários etc.); 3) não continham, no
mínimo, oito páginas. Foram descartados, ainda, livros com
autor e editora coincidentes com outro livro já previamente
selecionado. Considerou-se, ademais, imprescindível que o
livro estivesse disponível para compra imediata na data de
realização da análise, com conta brasileira na iBookstore (loja
online para compra de e-books no iPad e em outros produtos
da Apple). Para
viabilizar,
na
segunda
etapa
metodológica
(“Entender & Observar”), a identificação dos principais
recursos utilizados em e-books infantis disponíveis no
mercado –, foi realizada uma pesquisa de similares,
subdividida em três etapas principais: 1) selecionar amostra
de e-books infantis; 2) realizar um levantamento dos recursos
utilizados em cada livro digital; 3) analisar o projeto gráfico e
a relação entre texto, ilustração, animação e áudio adotada
em cada obra. Na terceira etapa metodológica (“Construir”), partindo-
se dos dados obtidos por meio das etapas anteriores, foi
realizada uma análise dos pontos de consenso e dos pontos
de discussão, chegando-se, assim, nas diretrizes preliminares
de projeto. A partir dessa listagem, foi feita uma consulta
com especialistas (etapa “Avaliar” dos procedimentos
metodológicos), a fim de extrair suas opiniões acerca das
diretrizes preliminares e de cada ponto de discussão listado. 4.1 Análise dos livros mais vendidos nos EUA Nesta seção, apresenta-se uma análise dos cinco e-books
selecionados dentre os mais vendidos da iTunes Store
americana (Quadro 1), conforme os critérios previamente
apresentados. Figura 4: Página interna do livro Peppa Meets the Queen, da
Editora Penguin Books Ltd., visualizada com o iPad
em modo horizontal. Quadro 1: Livros da iTunes Store americana analisados. Título
Autor
Ano
Olivia Loves Halloween
Maggie Testa
2014
Crankenstein
Samantha Berger
2013
The Giving Tree
Shel Silverstein
2014
Peppa Meets the Queen
Neville Astley e Mark
Baker
2012
The Day the Crayons Quit
Drew Daywalt
2013 Quadro 1: Livros da iTunes Store americana analisados. Em ‘Olivia Loves Halloween’, conforme o narrador faz a
leitura, as palavras são simultaneamente realçadas no texto –
o que auxilia a criança a reconhecer o conteúdo textual do
livro. Há que se observar, no entanto, que, no caso desse
livro, a passagem automática de página ocorre poucos
segundos após o narrador finalizar a leitura. Assim, resta
pouquíssimo tempo para que a criança possa observar e ‘ler’
cuidadosamente a informação visual da página. Tal
configuração é prejudicial no sentido de induzir a criança a
acompanhar apenas uma das leituras possíveis do livro por
vez (textual ou imagética). Ainda, considerando-se que há
páginas em que a leitura imagética complementa a leitura
textual (e não apenas a repete), tal problemática pode
incorrer em dificuldade de compreensão de determinados
trechos da narrativa (Figura 5). Primeiramente, observaram-se aspectos gerais quanto
ao funcionamento dos e-books e quanto ao aproveitamento
da tela durante a leitura. Todos os e-books analisados
adaptaram-se a ambas as orientações de leitura – tanto com
o iPad em sentido horizontal/paisagem, quanto em sentido
vertical/retrato. Além disso, o projeto dos cinco livros
considerou o espaço da tela com uma divisão vertical – a
página ‘espelhada’ ou página dupla simula um livro impresso
aberto, cuja face visível apresenta-se dividida pela dobra do
papel. Embora esse conceito tenha relação direta com a
natureza física do livro, muitos e-books ‘simulam’ essa divisão
da dobra do papel, como foi o caso dos e-books analisados. Esse fato parece ser resultado, no caso desses livros, da
adaptação da obra impressa para a versão digital, sem que
sejam feitas alterações na diagramação das páginas. 3. METODOLOGIA Para isso, foram consultados seis especialistas, entre
designers gráficos com
experiência
teórico/prática
e
especialistas em educação e/ou pedagogos. A consulta
realizada constituiu-se de uma entrevista aberta, com roteiro
previamente elaborado pelos pesquisadores. Foram expostos
aos participantes, individualmente, os pontos de divergência
bibliográfica e as diretrizes preliminares. Cada especialista
pôde responder de forma aberta e fornecer outras
contribuições, caso desejasse. O áudio das entrevistas foi
gravado apenas para futura consulta, e os participantes
assinaram um termo de consentimento livre e esclarecido –
eles não são identificados nominalmente em nenhum
momento. A partir dos dados obtidos na entrevista, que
foram triangulados com os dados da pesquisa bibliográfica e
da análise de similares, foram feitos os devidos ajustes nas
diretrizes preliminares, procurando-se, também, definir os
pontos de não consenso, chegando-se, enfim, às diretrizes A seleção das obras a serem analisadas foi feita a partir
do catálogo digital da iTunes Store, serviço online de venda
de música, filmes, aplicativos e livros da Apple. Selecionou-se
este catálogo por ser o mais popular quando se fala em
publicações digitais para público infantil e, ainda, pois os
formatos de arquivo de e-book compatíveis com o e-reader
iPad, vendidos na iTunes Store, mostraram-se mais
adequados a esse tipo de publicação, contendo mais recursos
interativos, conforme foi abordado na revisão bibliográfica. Foram selecionadas cinco publicações dentre a lista das mais
vendidas no Brasil e cinco publicações dentre as mais
vendidas nos Estados Unidos (selecionou-se esse país a partir
da constatação de que o mercado editorial digital norte-
americano é um dos mais desenvolvidos da atualidade, com
maior catálogo referente a publicações digitais para crianças;
as publicações norte-americanas do setor normalmente são
tomadas como referência para as publicações de outros
países, inclusive). Para a seleção das obras a serem
analisadas, seguiu-se a ordem da listagem de mais vendidos
na categoria infantil em cada país, descartando-se os livros Design & Tecnologia 16 (2018) 57 ofereceu recurso diferente, utilizando rolagem lateral, sem
simulação de dobra em página impressa). Dos livros
analisados, apenas dois (‘Olivia Loves Halloween’ e
‘Crankenstein’) possuíam recursos de leitura em voz alta –
‘read aloud’ –, em que o usuário pode optar por ouvir um
narrador lendo a história, tal qual é apresentada em texto. Por consequência, esses foram também os únicos em que o
usuário pode configurar o livro para que a passagem de
páginas seja feita automaticamente (a página vira quando o
narrador termina a leitura). 3. METODOLOGIA conclusivas de projeto de e-books voltados a crianças em fase
de alfabetização (etapa “Concluir” dos procedimentos
metodológicos). Neste artigo, são apresentadas, ainda que de
forma resumida, todas as etapas do procedimento
metodológico proposto. 4. ANÁLISE DE SIMILARES Conforme indicado na metodologia, para a análise de
similares, foram selecionados dez livros, sendo cinco
publicações da lista dos mais vendidos na iTunes Store nos
Estados Unidos e cinco publicações dentre os mais vendidos
na iTunes Store no Brasil. Figura 4: Página interna do livro Peppa Meets the Queen, da
Editora Penguin Books Ltd., visualizada com o iPad
em modo horizontal. Para cada publicação, foi preenchida uma ficha objetiva
com as principais características identificadas no e-book
quanto ao formato, à diagramação, aos elementos pré e pós-
textuais, ao esquema cromático, à tipografia, aos recursos de
áudio, de animação e de vídeo e à interatividade – neste
artigo, apresenta-se uma análise compilada a partir das fichas
preenchidas para cada livro. As publicações foram analisadas
dentro do software iBooks, tendo em vista ser esse o
aplicativo padrão para leitura de e-books no iPad. 4.1 Análise dos livros mais vendidos nos EUA Figura 8: Página interna de The Day the Crayons Quit, da
Editora Harper Collins, com texto simulando a escrita
infantil à mão livre. Figura 7: Página interna de The Giving Tree, da Editora Harper
Collins, com tipografia serifada. Figura 6: Página
interna
do
livro
Crankenstein,
da
EditoraLittle, Brown Books for Young Readers. Exemplo de animação em e-book infantil, a
onomatopeia “Mehhrrrr!” vibra quando o narrador
faz a leitura. Figura 7: Página interna de The Giving Tree, da Editora Harper
Collins, com tipografia serifada. Figura 8: Página interna de The Day the Crayons Quit, da
Editora Harper Collins, com texto simulando a escrita
infantil à mão livre. Figura 6: Página
interna
do
livro
Crankenstein,
da
EditoraLittle, Brown Books for Young Readers. Exemplo de animação em e-book infantil, a
onomatopeia “Mehhrrrr!” vibra quando o narrador
faz a leitura. Figura 8: Página interna de The Day the Crayons Quit, da
Editora Harper Collins, com texto simulando a escrita
infantil à mão livre. Ao final da leitura de ‘Crankenstein’, há um pequeno
jogo, do tipo ‘quiz’, com dez perguntas objetivas para que o
leitor possa ‘testar’ se é um ‘crankenstein’ (personagem
principal da história, rabugento e mal-humorado). A inclusão
do jogo ao livro parece, neste caso, válida, por ser
diretamente relacionada à temática da história e por testar as
capacidades de interpretação e de compreensão do texto. Pode-se observar, também, que, por estar posicionado após o
término da narrativa, o jogo parece interferir menos, no
sentido de não distrair a criança do foco no entendimento da
história. Já nas obras ‘Peppa Meets the Queen’, ‘The Giving
Tree’ e ‘The Day the Crayons Quit’, há pouquíssimos ajustes
de adaptação do livro do meio físico para o meio digital
(todos esses e-books têm uma versão impressa, lançada
antes). Não há recursos de áudio, passagem automática de
páginas, animações ou qualquer outro tipo de interatividade. Quanto à diagramação, notou-se pouca variação de
layout de página a página – a grande maioria das páginas dos
livros analisados fez uso de diagramação do tipo associação
(conforme a classificação de Linden, 2011). Nesse tipo de
layout, não há uma distinção entre página de texto e página
de imagem, reunindo pelo menos um enunciado verbal e um
enunciado visual no espaço da página. 4.2 Análise dos livros mais vendidos no Brasil
Nesta seção, apresenta-se uma análise dos cinco e-books
selecionados dentre os mais vendidos da iTunes Store
brasileira (Quadro 2). 4.1 Análise dos livros mais vendidos nos EUA Essa
característica pode ser considerada positiva, se se considera
válida a aproximação com a estética do livro impresso, mas
há que se considerar que, na maioria dos casos, a proporção
da página, resultado da opção pelo aproveitamento do
formato da página impressa, não é a mais adequada à
proporção de tela
do iPad, causando prejuízo ao
aproveitamento da área de tela do aparelho (pode-se
observar, na Figura 4, o fundo cinza que 'preenche' a área
não ocupada pelas páginas do livro, no e-book ‘Peppa Meets
the Queen’). Recurso que também faz referência direta ao
livro impresso, a simulação de dobra no canto da página para
indicar a passagem de páginas foi um recurso utilizado por
quatro dos cinco e-books analisados (apenas ‘The Giving Tree’ Figura 5: Página interna do livro Olivia Loves Halloween, da
editora Simon Spotlight. As palavras são grifadas
conforme o narrador faz a leitura em voz alta. Figura 5: Página interna do livro Olivia Loves Halloween, da
editora Simon Spotlight. As palavras são grifadas
conforme o narrador faz a leitura em voz alta. Já no caso do livro ‘Crankenstein’, o recurso de leitura
em voz alta não realça no texto as palavras que o narrador Design & Tecnologia 16 (2018) Design & Tecnologia 16 (2018) 58 está lendo. Dessa forma, as vantagens do uso do recurso de
áudio ficam limitadas, considerando-se que
diversas
correntes
teóricas
apontam
os
benefícios
para
a
aprendizagem infantil da marcação concomitante no texto do
conteúdo lido pelo narrador, conforme visto na revisão
bibliográfica. ‘Crankenstein’ foi a única publicação que
explorou o uso de animações: as letras da onomatopeia que
representa o grunhido do Crankenstein – “Mehhrrrr!” –
movem-se, vibrando visualmente, ao mesmo tempo em que
o narrador faz a leitura desta palavra. A animação – que
reaparece diversas vezes ao longo do livro – é apenas um
detalhe sutil, que não parece distrair ou afastar a criança da
história. Pelo contrário, a vibração da palavra reforça a
sensação de ouvir alguém gritando ou grunhindo alto, no
mundo real. Parece, assim, por estar diretamente relacionada
ao conteúdo, ser um bom exemplo de uso do recurso de
animação em livros digitais infantis (Figura 6). há dificuldade de entender determinados caracteres. Em
alguns trechos, a tipografia é serifada (não há caracteres
infantis). Figura 7: Página interna de The Giving Tree, da Editora Harper
Collins, com tipografia serifada. 4.1 Análise dos livros mais vendidos nos EUA Quadro 2: Livros da iTunes Store brasileira analisados. Título
Autor
Ano
Você deve tentar, tentar e tentar! Michael O’Reilly
2013
Duda e o Bichinho do Medo
Renato Carvalho
2013
The Fox on the Box
Barbara Gregorich
1996*
O Universo
Raquel Durães
2015
Chapeuzinho Vermelho
Cristina Marques
2015
*ano de publicação do livro homônimo, impresso. Quadro 2: Livros da iTunes Store brasileira analisados. Quanto à tipografia, notou-se bastante diversidade de
estilos dentre os livros analisados. ‘Olivia Loves Halloween’ e
‘The Giving Tree’ fizeram uso de fontes serifadas, bastante
convencionais. No caso do segundo livro, a quantidade de
texto por página parece inadequada ao público-alvo, levando-
se em conta as recomendações bibliográficas (Figura 7). O e-book ‘Crankenstein’ traz tipografia sem-serifas, com
texto composto inteiramente em caixa-alta. Dentre os
analisados, ‘Peppa Meets the Queen’ foi o único a fazer uso
de tipografia sem serifas e com caracteres infantis. No caso
de ‘The Day the Crayons Quit’, a tipografia predominante no
livro simula a escrita infantil à mão livre (Figura 8). Por vezes, Design & Tecnologia 16 (2018) 59 Diferentemente dos livros da iTunes Store americana
analisados, quatro das cinco publicações mais vendidas na
loja brasileira têm projeto desenvolvido especificamente para
o meio digital – e não adaptado de obra impressa (The Fox on
the Box é a exceção na lista). Apesar disso, nem todas fazem
uso de recursos tecnológicos. ‘Você deve tentar, tentar e
tentar!’, por exemplo, apesar de projetada originalmente
para ser um e-book, não dispõe de narração em áudio, nem
de passagem automática de páginas e tampouco faz uso de
recursos de vídeo ou de animação. A única vantagem advinda
do fato de o projeto ter sido pensado para o meio digital é a
proporção da tela (Figura 9), que se encaixa perfeitamente à
tela do iPad (apenas em formato de visualização horizontal,
no entanto). Já no e-book ‘Chapeuzinho...’, o realce é feito por
palavra, facilitando o acompanhamento da leitura pela
criança e favorecendo o reconhecimento fonético. Em
algumas páginas, há uma música instrumental, que funciona
como ‘pano de fundo’ para a leitura feita pelo narrador. Em
determinados momentos, o uso de tal recurso parece válido,
ao contextualizar o leitor sobre o conteúdo daquele trecho da
história – a música ajuda a despertar sentimentos como
tensão ou alegria, dependendo dos acontecimentos narrados
na página. 4.1 Análise dos livros mais vendidos nos EUA No entanto, mesmo com o volume diminuído
enquanto se lê o texto, a música parece, em alguns
momentos, influenciar negativamente, distraindo o leitor do
conteúdo do texto em si. Poderia, talvez, haver um modo de
optar por ouvir ou não a música simultaneamente à narração. Ao final da narrativa, é disponibilizado um vídeo, posicionado
em uma página à parte (Figura 11). É possível assistir ao vídeo
em miniatura, centralizado na página em que se encontra, ou
ampliá-lo para que ocupe toda a tela do iPad. O vídeo conta a
mesma história lida no livro, mas com muito maior riqueza de
detalhes – parece, por isso, ser bastante negativo no sentido
do incentivo ao gosto pela leitura (o texto acaba sendo mais
‘pobre’ que o vídeo, do ponto de vista do conteúdo). No
momento em que a narrativa em texto tem menos qualidade
do que a narrativa em vídeo, a criança provavelmente
preferirá assistir ao vídeo – justamente a hipótese levantada
por Guernsey (2011), descrita na pesquisa bibliográfica. Figura 9: Página interna de Você deve tentar, tentar e tentar!,
da Editora Trans Eye Connect LLC. A proporção da
página ajusta-se ao formato de tela do iPad. Figura 1: Página interna de Chapeuzinho Vermelho, da Editora
Todolivro. No lado direito da página, observa-se
miniatura de vídeo que reconta a história lida. Figura 9: Página interna de Você deve tentar, tentar e tentar!,
da Editora Trans Eye Connect LLC. A proporção da
página ajusta-se ao formato de tela do iPad. Já os e-books ‘Duda e o Bichinho do Medo’ e
‘Chapeuzinho Vermelho’ apresentam mais esforços no
sentido de aproveitamento dos recursos digitais. Ambos
possuem recurso de narração em voz alta, com passagem
automática de páginas. No caso de ‘Duda...’, conforme o
narrador faz a leitura, os períodos (da letra maiúscula ao
ponto final) são simultaneamente realçados no texto,
facilitando que a criança acompanhe a leitura (Figura 10). Cabe observar, no entanto, que o realce por palavra ou por
sílaba seria mais interessante em termos de percepção
fonética (conforme os estudos de LEWIN, 2000 apud
AMANTE, 2007, apresentados na revisão bibliográfica). Figura 1: Página interna de Chapeuzinho Vermelho, da Editora
Todolivro. No lado direito da página, observa-se
miniatura de vídeo que reconta a história lida. ‘O Universo’, por sua vez, traz recurso de narração em
voz alta, mas sem passagem automática de páginas e sem
realce contíguo no texto (Figura 12). 4.1 Análise dos livros mais vendidos nos EUA Esse é o único dos livros
analisados que possui um player de áudio próprio (não utiliza
o recurso padrão do iBooks/iPad). Em cada página, há um
ícone de som, no canto superior esquerdo, em que é possível
pausar/continuar ou repetir a leitura da página. Não há, no
entanto, possibilidade de pausar e reiniciar. Figura 10: Página interna de Duda e o Bichinho do Medo, de
editora independente. O destaque no texto é feito
por período, conforme o narrador lê. Também não há possibilidade de desligar a narração do
livro como um todo: será necessário pausá-la cada vez que se
passa para uma nova página. Desse modo, apesar de
personalizado, o player funciona com mais restrições e sem
nenhuma vantagem com relação ao player padrão. Um ponto
positivo do projeto, por outro lado, é que, assim como em
‘Você deve tentar...’, a opção pelo layout de página única (e
não dupla ou espelhada) garante ajuste perfeito à proporção
do iPad, com ótimo aproveitamento do espaço da tela. Figura 10: Página interna de Duda e o Bichinho do Medo, de
editora independente. O destaque no texto é feito
por período, conforme o narrador lê. Design & Tecnologia 16 (2018) 60 Figura 12: Página interna de O Universo, da Editora O Fiel
Carteiro. O layout de páginas únicas ajusta-se à
proporção de tela do iPad. Na mesma imagem, é
possível ver o ícone que permite pausar a narração. entre si, no entanto. No caso de ‘Você deve tentar...’, o texto
é composto com uma família tipográfica neutra – a maior
problemática da composição acaba sendo o espaço
entrelinhas, ligeiramente ‘apertado’, e a própria mancha
gráfica do livro (os textos são posicionados muito próximos às
margens, faltando ‘área de respiro’). Já em 'Duda...', a
tipografia escolhida para o livro, do tipo condensada, com
espaço entre letras justo, acaba prejudicando bastante a
legibilidade (principalmente nas páginas com bastante texto,
que são frequentes). Em ‘Chapeuzinho...’, a composição
inteiramente em caixa alta, formando blocos de texto,
prejudica bastante a legibilidade, também porque o espaço
entrelinhas não é suficiente. As escolhas tipográficas de ‘The
Fox...’ e de ‘O Universo’ parecem ter sido as mais acertadas –
no caso do segundo, a fonte, do tipo ‘rounded’, poderia ser
apenas um pouco maior, conforme as especificações
advindas da revisão bibliográfica. Figura 12: Página interna de O Universo, da Editora O Fiel
Carteiro. 4.1 Análise dos livros mais vendidos nos EUA O layout de páginas únicas ajusta-se à
proporção de tela do iPad. Na mesma imagem, é
possível ver o ícone que permite pausar a narração. Quanto à diagramação, notou-se, assim como no caso
dos livros da iTunes Store americana, pouquíssima variação
de layout, com a quase totalidade das páginas dos livros
fazendo uso de diagramação do tipo associação (conforme a
classificação de Linden, 2011). Dentre os e-books analisados, ‘The Fox on the Box’, cujo
projeto foi adaptado da versão impressa da obra, é o que
menos explora os recursos digitais (Figura 13). 5. PROPOSTA DE DIRETRIZES PRELIMINARES PARA
PROJETO DE E-BOOKS INFANTIS Nesse e-book, não há recurso de narração em voz alta,
nem de passagem automática de páginas – o e-book é, na
verdade, a simples digitalização do livro físico. Parece, por
isso, ser mera coincidência o fato de que, em modo de
visualização horizontal no iPad, a proporção da página
permita um bom aproveitamento da tela. A partir da pesquisa bibliográfica e da análise de similares, foi
possível compilar os principais dados referentes ao projeto de
e-books voltados ao público infantil em fase de alfabetização
em formato de diretrizes preliminares de projeto. Essas
diretrizes foram categorizadas conforme a temática que
abordam dentro do grande campo do projeto editorial de
livros digitais: formato de arquivo, tipo de layout,
diagramação, tipografia, paleta cromática, áudio, vídeo,
animação, interatividade e hipertexto. Em seguida, foram
elencados alguns pontos de não consenso – são aspectos em
que não se chegou a um resultado minimamente conclusivo a
partir da pesquisa bibliográfica e da análise de similares. Figura 13: Página interna de The Fox on the Box, da Editora
School Zone, simples digitalização do livro impresso,
o e-book não traz nenhum recurso tecnológico extra. Para que se chegasse a esse resultado preliminar, os
autores
listaram,
em
cada
uma
das
categorias,
recomendações em que se notou um consenso entre os
autores consultados e o que se observou na prática, por meio
da análise dos e-books – essas recomendações foram
nomeadas ‘diretrizes preliminares’. Quando não houve esse
consenso (seja entre os autores em si ou entre autores e
observação prática), a recomendação foi, então, enquadrada
como ‘ponto de discussão’ (Quadros 3 e 4). Tanto esses
pontos quanto as diretrizes preliminares em si foram levados,
em etapa subsequente, à avaliação por especialistas. Para
cada uma das diretrizes preliminares ou pontos de discussão
foi atribuído um código, para facilitar a referência a cada um
desses itens na etapa de consulta a especialistas. Figura 13: Página interna de The Fox on the Box, da Editora
School Zone, simples digitalização do livro impresso,
o e-book não traz nenhum recurso tecnológico extra. Os cinco livros analisados fizeram uso de tipografia sem
serifas e sem caracteres infantis, com algumas diferenças Quadro 3: Pontos de discussão. Categoria
Cód. Diretriz
Tipografia
PD1
Devem-se utilizar fontes serifadas ou não serifadas no projeto de e-books infantis? PD2
É importante a utilização de fontes com caracteres infantis? 5. PROPOSTA DE DIRETRIZES PRELIMINARES PARA
PROJETO DE E-BOOKS INFANTIS PD3
A composição do texto deve ser totalmente em caixa alta, ou em caixa alta e caixa baixa? Paleta cromática
PD4
A paleta cromática do e-book deve ser sempre em tons vibrantes e saturados ou deve ser definida a partir
da temática da história? Ilustração
PD5
Há algum estilo de ilustração mais recomendável ou indicado? Quadro 3: Pontos de discussão. Categoria
Cód. Diretriz
Tipografia
PD1
Devem-se utilizar fontes serifadas ou não serifadas no projeto de e-books infantis? PD2
É importante a utilização de fontes com caracteres infantis? PD3
A composição do texto deve ser totalmente em caixa alta, ou em caixa alta e caixa baixa? Paleta cromática
PD4
A paleta cromática do e-book deve ser sempre em tons vibrantes e saturados ou deve ser definida a partir
da temática da história? Ilustração
PD5
Há algum estilo de ilustração mais recomendável ou indicado? Design & Tecnologia 16 (2018) 61 Quadro 4: Diretrizes preliminares. Quadro 4: Diretrizes preliminares. Categoria
Cód. Diretriz
Formato de arquivo
D1
Preferencialmente, trabalhar com os formatos ePub2 ou ePub3. Tipo de layout
D2
Utilizar layout fixo para confecção do e-book. Diagramação
D3
Na diagramação do e-book, prever variação entre os tipos de diagramação de página a página. Tipografia
D4
Utilizar fonte com boa diferenciação entre caracteres. D5
Selecionar, para o projeto gráfico, família tipográfica que tenha boa renderização em tela. D6
Na composição tipográfica, priorizar colunas de texto estreitas (até 30 letras por linha) e curtas (até 12
linhas por bloco de texto) e espaço entrelinhas amplo. D7
Evitar a utilização de hifenação, e 'quebrar' as linhas conforme o conteúdo do texto. Paleta cromática
D8
Definir a paleta cromática do livro selecionando cores com brilho adequado à visualização em tela, de
modo que proporcionem conforto visual durante a leitura. As cores devem ter boa diferenciação entre si,
independentemente da configuração de brilho da tela do e-reader. Áudio
D9
Sempre que viável, incluir recurso de ‘read aloud’ (leitura em voz alta) ao e-book infantil. Preferencialmente, as palavras ou sílabas devem ser grifadas individualmente, em sincronia com a
narração (em detrimento do grifo por frase ou parágrafo). D10
Quando houver inclusão de músicas ao e-book, incluir um player de áudio em separado, para que o leitor
possa escolher quando – e se – deseja ouvir a música. Vídeo
D11
Quando for incluído conteúdo em vídeo, apresentá-lo em página à parte do conteúdo de texto. 5. PROPOSTA DE DIRETRIZES PRELIMINARES PARA
PROJETO DE E-BOOKS INFANTIS Ainda, se
possível, conferir o controle da reprodução do vídeo ao leitor. D12
Quando se optar pela inclusão de vídeo, observar que seu conteúdo seja relevante e que enriqueça a
experiência de leitura do livro, não apenas repetindo a narrativa textual
Animação
D13
As animações incluídas no e-book devem ter relação com o conteúdo e, preferencialmente, ser acionadas
apenas sob o comando do leitor. Interatividade e
Hipertexto
D14
Quando for feito uso de recursos de hipertexto, estruturar as informações de forma clara e deixá-las
acessíveis ao leitor. 6. ENTREVISTA COM ESPECIALISTAS Os entrevistados foram selecionados pelos autores
deste trabalho, observando-se que estivessem divididos entre
designers com experiência acadêmica e/ou prática e
pedagogos ou professores, que pudessem trazer seus
conhecimentos das áreas da pedagogia e da educação para
complementar e confirmar as análises realizadas (Quadro 5). No Quadro 6, apresentam-se as questões propostas pelos
pesquisadores aos especialistas e os resultados da consulta. De modo a validar e a esclarecer as diretrizes preliminares e
os pontos de discussão, listados na seção anterior, foi
realizada uma entrevista guiada com especialistas. As
perguntas foram feitas de forma aberta, com base em um
questionário previamente elaborado pelos pesquisadores,
permitindo-se que o especialista pudesse dar o máximo de
contribuições acerca de cada tópico temático. Quadro 5: Especialistas consultados. Sigla
Formação do especialista
Área de atuação do especialista
A
Mestrado em Educação pela Universidade Federal do Rio
Grande do Sul e Doutorado em Informática na Educação. Arquitetura e Design Gráfico (editorial e de identidade visual,
principalmente). B
Graduação em Design Gráfico. Experiência prática em desenvolvimento de e-books e de
objetos de aprendizagem. C
Graduação e Mestrado em Design Gráfico. Doutorado em
andamento, na mesma área. Design, produção de livros e tipografia. D
Graduação em Pedagogia (habilitação em Educação Infantil,
Séries Iniciais, Orientação Educacional e Matérias Pedagógicas
do Ensino Médio), Especialização em Gestão da Educação e
Mestrado em Educação. Experiência prática com crianças em fase de alfabetização. Atuação no desenvolvimento de objetos de aprendizagem em
meio digital, para várias faixas etárias. E
Graduação em Design Gráfico. Experiência prática em design editorial impresso e digital. Ampla experiência em tipografia, tendo ministrado diversos
cursos e mini palestras na área. F
Graduação em Letras – Português e Literaturas de Língua
Portuguesa. Experiência em docência no Ensino Fundamental. Quadro 5: Especialistas consultados. Design & Tecnologia 16 (2018) 62 Quadro 6: Consulta aos especialistas acerca das diretrizes preliminares e pontos de discussão. O código (segunda coluna) indica a
diretriz preliminar ou o ponto de discussão que deu origem à questão proposta aos especialistas. Quadro 6: Consulta aos especialistas acerca das diretrizes preliminares e pontos de discussão. O código (segunda coluna) indica a
diretriz preliminar ou o ponto de discussão que deu origem à questão proposta aos especialistas. Questão proposta
Código*
Opinião dos especialistas consultados**
Qual a sua percepção acerca da
estratégia de composição editorial em
que se prevê variação entre os tipos de
diagramação de página a página? 6. ENTREVISTA COM ESPECIALISTAS D3
Especialistas A, C, D, E, F consideraram a diretriz útil e aplicável no contexto. Especialista B fez ressalva relativa à redução da produtividade de mercado que seria
decorrente da aplicação desta diretriz, algo que se afasta dos objetivos desta
pesquisa. Qual a sua opinião acerca da utilização
de caracteres infantis na composição do
texto? PD2
Especialistas B, C, D, E, F mostraram-se de certa forma inclinados ao uso de
caracteres infantis nas publicações voltadas a crianças em fase de alfabetização. O
especialista A não teceu comentário específico sobre a utilização de caracteres
infantis. Devem-se utilizar fontes serifadas ou
não serifadas no projeto de e-books
infantis? Há alguma família tipográfica
específica que você considere adequada
ao público-alvo? PD1
Especialistas B, C e F posicionaram-se a favor do uso de fontes sem serifas, em
função da simplificação do desenho. O argumento do especialista F enriqueceu a
argumentação a favor desse tipo de tipografia, ao explicar uma das teorias sobre o
processo de decodificação dos caracteres pela criança. Nenhum dos especialistas
consultados pareceu condenar o uso de fontes com serifas. A composição do texto deve ser
totalmente em caixa alta, ou em caixa
alta e caixa baixa? PD3
Todos os especialistas que teceram comentários acerca desse tópico referiram que a
composição do texto em caixa alta pode ser interessante, mas não é uma
necessidade ou obrigatoriedade, uma vez que não há como comprovar que a criança
está sendo alfabetizada primeiramente apenas com o alfabeto das maiúsculas. Indique
alguma(s)
fonte(s)
que
considere adequada(s) ao público-alvo e
às publicações digitais. D5
Apenas dois entrevistados indicaram fontes que consideram adequadas ao público-
alvo, com boa renderização em tela: houve uma referência à Sassoon e duas
referências à Kindergarten. Qual a sua percepção acerca da seleção
cromática para obras infantis? PD4
Alguns especialistas confirmaram a hipótese da preferência infantil por paletas de
cores saturadas e variadas. No entanto, também referiram que essa preferência não
precisa, necessariamente, restringir ou limitar a seleção cromática para os livros
digitais infantis. O contraste entre as cores foi citado como característica decisiva
para um bom resultado de projeto. Há algum estilo de ilustração que você
considere, atualmente, mais adequado
ao público-alvo? PD5
A variedade nos estilos de ilustração para o livro digital infantil pareceu ser consenso
entre os entrevistados, como forma de que a criança crie repertório visual. 6. ENTREVISTA COM ESPECIALISTAS É importante a presença de comandos
para pausar o áudio da leitura ou para
retornar ao começo e repetir a leitura
de cada página? É relevante que haja
possibilidade de selecionar a velocidade
da leitura em voz alta? Qual a sua
opinião sobre a inclusão de músicas aos
livros
digitais
infantis? Em
que
momento da leitura seria possível ouvi-
las? D9, D10
Os especialistas, de modo geral, consideraram que a inclusão de algumas
funcionalidades no player de áudio poderia aprimorar a experiência de uso das
crianças: opção de velocidade de narração, opção de repetir a leitura da página e
opção de selecionar o ponto de reprodução do áudio foram confirmados como
possíveis melhorias nesse sentido. Os especialistas B e C referiram que a inclusão do
player de áudio na própria página do e-book facilitaria o uso, desde que mantida a
simplicidade de uso dos comandos. Quanto à música, parece ser consenso que, se o
e-book optar por incluir tal mídia, deve ser inserida à parte da narração e deve se
permitir que o leitor tenha controle sobre a reprodução desse conteúdo. O livro digital infantil deve conter
animações? As animações devem ser
acionadas automaticamente ou apenas
ao clicar/tocar? O livro digital infantil
deve conter vídeos? Se sim, em que
momento da leitura seria possível
assisti-los? Que conteúdo deveriam ter
esses vídeos? D11,
D12,
D13
Não pareceu haver consenso entre os entrevistados sobre a validade ou não da
inclusão de conteúdo em vídeo aos e-books infantis. Houve, no entanto, referência
ao conteúdo do vídeo, que não deve substituir a leitura do texto escrito. Também
referiu-se que o usuário tenha controle sobre a reprodução desse tipo de mídia. As
animações parecem ser mais bem aceitas, do ponto de vista dos especialistas
consultados – não houve declarações explícitas contra o uso de ilustrações animadas
(como houve no caso do vídeo). Dois especialistas (B e D) indicaram o uso de
animações que exijamalguma ação/interação do leitor, relacionada ao contexto da
narrativa. O livro digital infantil deve ser de leitura
linear ou deve ser hipertextual e
interativo, permitindo que a criança
transite pelo conteúdo de diversas
maneiras distintas? Por quê? D14
De modo geral, os especialistas mostraram-se favoráveis à utilização de recursos de
interatividade e de hipertexto. *A correspondência entre o código e a diretriz/ponto de discussão encontra-se nos quadros 3 e 4.
** Os especialistas são referenciados, neste quadro, de acordo com as siglas apresentadas no quadro 5 7. DIRETRIZES PARA PROJETO DE E-BOOKS INFANTIS Esta seção apresenta as diretrizes conclusivas para projeto de
e-books voltados a crianças em fase de alfabetização. Essas
diretrizes advêm das diretrizes preliminares e dos pontos de
discussão (apresentados na seção 5), que foram ajustados a
partir das considerações dos especialistas (seção 6). São
resultado, portanto, da triangulação entre os dados da
pesquisa bibliográfica, da análise de similares e da consulta
aos especialistas. Diretriz #3 // DIAGRAMAÇÃO - Na diagramação do e-book,
prever variação entre os tipos de diagramação de página a
página. Esses parâmetros, que podem ser entendidos, em
conjunto, como um guia para o projeto de e-books voltados
ao público infantil, são apresentados em quadros cujo título
corresponde ao tópico temático em que se inserem
(configuração do e-book, formato, diagramação, tipografia,
cor, ilustração, áudio, vídeo, animação, interatividade e
hipertexto). Abaixo de cada quadro, resgatam-se, de maneira
resumida, os dados que embasaram a proposição de cada
diretriz. A recomendação bibliográfica de Sophie Van der Linden
(2011) quanto à diagramação indica que cada página no e-
book infantil seja tratada quase que como um projeto único –
variar o posicionamento da ilustração, animação e conteúdo
textual garante, conforme a autora, fluidez à leitura. A cada
nova página, a criança tem uma surpresa, por meio da
variação na configuração visual, o que ajuda a gerar interesse
e a prender a atenção à leitura do livro. Nos e-books infantis
analisados, a recomendação bibliográfica não se confirmou:
em poucos deles, e às vezes, em apenas algumas páginas,
havia algum tipo de variação no layout. A entrevista com os
especialistas, no entanto, confirmou a opinião de Linden – a
quase totalidade dos que teceram comentários sobre esse
tópico consideraram de extrema importância a variação de
página a página, a fim de enriquecer o projeto visual,
especialmente considerando a grande quantidade de
elementos gráficos presentes em um livro digital infantil. Diretriz #1 // CONFIGURAÇÃO DO E-BOOK - Utilizar layout
fixo para confecção do e-book. Utilizar páginas inteiras (sem
divisão simulando dobra), cuja proporção seja adequada ao
suporte digital. A partir da pesquisa bibliográfica específica sobre e-
books infantis, foi possível concluir que a presença de
ilustração em cada página é uma característica imprescindível
dessas publicações. 7. DIRETRIZES PARA PROJETO DE E-BOOKS INFANTIS Ainda, a partir dos autores pesquisados,
foi possível notar que uma composição tipográfica cuidadosa
é importante para facilitar o acesso infantil ao texto – tais
cuidados envolvem a seleção da família tipográfica, a
definição do espaço entrelinhas e da largura da coluna de
texto, entre outros aspectos. Também da pesquisa
bibliográfica foi possível aferir que tal composição de layout,
com múltiplas colunas de texto, com ilustração e tipografia
ocupando espaços definidos na página, e com controle total
da tipografia pelo designer, só é possível quando se trabalha
com e-books de layout fixo. Diretriz #4 // TIPOGRAFIA - Utilizar fonte com boa
diferenciação entre caracteres. Dentre diversos aspectos que permeiam definições
tipográficas para crianças em fase de alfabetização em que
não há consenso entre os pesquisadores, a diferenciação
entre caracteres parece ser uma exceção. Amplo aporte
bibliográfico que versa sobre o tema concorda que a clareza
dos caracteres é fundamental quando se fala em iniciação à
leitura. Para citar dois exemplos, Walker (2005) considera
primordial
a
diferenciação
entre
os
caracteres,
principalmente no que tange às letras ‘a’, ‘g’ e ‘o’, e Lourenço
(2011) comenta que, independentemente de se trabalhar
com fontes serifadas ou sem serifas, a clareza entre
caracteres é fundamental. Observaram-se, na análise de
similares, algumas escolhas tipográficas ruins, em que a
diferenciação
entre
caracteres
ficava
prejudicada. A
recomendação bibliográfica foi, no entanto, endossada pelos
especialistas, que muitas vezes citaram espontaneamente
essa característica como primordial à tipografia para e-books
infantis. A análise de similares permitiu confirmar o que foi
levantado na pesquisa bibliográfica, com a observação de que
a totalidade dos e-books analisados trabalhavam com layout
fixo e conseguiam, por isso, ter bons resultados na
composição de páginas com texto e ilustração. Nas
entrevistas com os especialistas, essas características de
layout que levam à necessidade de optar pelo layout fixo
foram confirmadas, uma vez que diversos dos entrevistados
citaram a importância da composição tipográfica cuidadosa e
reiteraram a importância da diagramação composta por
ilustração, animação e texto ocupando lugares definidos e
coexistindo no espaço da página. 6. ENTREVISTA COM ESPECIALISTAS Alguns (B, E e F) fizeram a ressalva de que o
hipertexto seja utilizado para fornecer informações sem relação cronológica com a
narrativa, considerando a faixa etária do público-alvo. *A correspondência entre o código e a diretriz/ponto de discussão encontra-se nos quadros 3 e 4. ** Os especialistas são referenciados, neste quadro, de acordo com as siglas apresentadas no quadro 5 Design & Tecnologia 16 (2018) 63 suporte a e-books de layout fixo é nativo. O ePub3 oferece,
ainda, algumas vantagens, como a possibilidade de
incorporação de mídia ao livro – o que poderá ser
interessante, dependendo dos requisitos de projeto. Todos os
e-books avaliados na análise de similares foram executados
em formato ePub2 ou ePub3, corroborando com as assertivas
bibliográficas. Por ser uma questão puramente técnica, esta
diretriz não foi levada à análise dos especialistas. Diretriz #2 // FORMATO - Preferencialmente, trabalhar com
os formatos ePub2 ou ePub3. 7. DIRETRIZES PARA PROJETO DE E-BOOKS INFANTIS Diretriz #6 // TIPOGRAFIA - Na composição tipográfica,
priorizar colunas de texto estreitas (até 30 letras por linha) e
curtas (até 12 linhas por bloco de texto) e espaço entrelinhas
amplo. Evitar a utilização de hifenação, e ‘quebrar’ as linhas
considerando-se as construções sintáticas do texto. Diretriz #6 // TIPOGRAFIA - Na composição tipográfica,
priorizar colunas de texto estreitas (até 30 letras por linha) e
curtas (até 12 linhas por bloco de texto) e espaço entrelinhas
amplo. Evitar a utilização de hifenação, e ‘quebrar’ as linhas
considerando-se as construções sintáticas do texto. As recomendações bibliográficas de Burt, Cooper e
Martin (1955), que especificam parâmetros tipográficos de
acordo com a faixa etária do leitor, são úteis de modo a evitar
que linhas muito longas ou tamanhos de fonte muito
pequenos desestimulem ou dificultem o acesso ao texto pelo
público-alvo. Para Lourenço (2011, p.123), “existe uma real
necessidade de um espaçamento maior entre letras e entre
palavras do que para um adulto”. A literatura específica,
nesse sentido, ressalta que o espaçamento entrelinhas, o
comprimento da linha e o corpo da fonte são algumas
características que têm influência direta na facilitação do
contato com o texto, quando se fala em leitores iniciantes. Uma segunda alternativa de composição do texto, que
segue o mesmo argumento de apresentar um desenho de
letra familiar à criança é a composição do texto totalmente
em caixa alta. Defendida por alguns autores pesquisados,
essa diretriz baseia-se no fato de que crianças em fase de
alfabetização leem letra a letra e são, comumente, em um
primeiro momento apresentadas ao alfabeto das letras
maiúsculas. Alguns dos especialistas consultados também se
mostram favoráveis a esse tipo de composição, levando-se
em consideração o público-alvo. Ressalta-se, no entanto, que
essa estratégia só é válida para livros com quantidade
realmente pequena de texto por página – caso essa
quantidade exceda duas frases ou duas a três linhas, a
legibilidade fica prejudicada, e portanto, devem-se evitar
composições desse tipo. Também para Lourenço (2011), o recurso de hifenação
do texto deve ser evitado, e as quebras de linha devem
ocorrer sempre de acordo com o conteúdo do texto. Essas
recomendações parecem estar sendo seguidas, de modo
geral, pelos livros digitais disponíveis no mercado, em que, na
maioria das vezes, o projeto gráfico prima por linhas curtas,
com pouca quantidade de texto por página. O recurso de
hifenação também não é recorrente nos livros analisados. 7. DIRETRIZES PARA PROJETO DE E-BOOKS INFANTIS Alguns especialistas consultados referiram-se, de forma
direta, aos aspectos da quantidade de texto, do comprimento
da linha
e do tamanho da fonte, reforçando as
recomendações bibliográficas. Diretriz #9 // COR - Definir a paleta cromática do livro
selecionando cores com brilho adequado à visualização em
tela, de modo que proporcionem conforto visual durante a
leitura. Bom contraste é fundamental – as cores devem ter
boa
diferenciação
entre
si,
independentemente
da
configuração de brilho da tela do e-reader. Diretriz #7 // TIPOGRAFIA - Observar que o desenho da fonte
esteja alinhado ao conceito pensado para o livro. Fontes com
traçado mais orgânico e menos mecânico são, de modo geral,
mais bem aceitas pelo público-alvo. Dar preferência a fontes
não serifadas ou fontes com serifas retas (slab serif). Para Vanderschantz (2008), é imprescindível que o
designer tenha atenção com o contraste ao definir a paleta
de cores de um livro digital, visto que variações de cor são
mais propensas a ocorrer em suporte digital do que em
suporte impresso – pode haver uma grande escala de
variação de cor entre uma tela e outra, e também
dependendo da configuração de brilho do dispositivo
utilizado para leitura. É fundamental, ainda, preocupar-se
com o conforto visual, evitando matizes demasiado vibrantes
ou claras. A realização de testes em telas diversas e com
diversas configurações de brilho pode auxiliar o designer a
garantir que o contraste entre cores e que o conforto visual
da paleta definida para o projeto estejam adequados. Nos e-
books avaliados na análise de similares, não se observaram
problemas quanto ao conforto visual, mas a problemática do
contraste acabou por prejudicar inclusive a legibilidade, em
casos pontuais, o que reforça a importância dessa diretriz. A
questão do contraste, inclusive, foi lembrada pelos
especialistas, quando questionados sobre a paleta cromática
do livro digital infantil. Apesar de a escolha entre fontes serifadas ou sem
serifas para público infantil não ser um consenso na
bibliografia específica, observou-se uma inclinação da maior
parte dos autores pesquisados pelas famílias sans-serif. A
argumentação de que tais fontes são mais acessíveis ao
público-alvo, por terem menos detalhes e por se
aproximarem mais do modo como a criança escreve, foi
repetida pelos especialistas consultados, reforçando essa
tese. 7. DIRETRIZES PARA PROJETO DE E-BOOKS INFANTIS Ainda quanto à
configuração da página, é indicado que a proporção seja
adequada ao suporte digital em que a publicação será
distribuída, garantindo que a área da tela seja utilizada em
sua totalidade – divisões de página simulando dobra não são
indicadas, pois, além de não terem sentido em ambiente
digital (são mera simulação da configuração do livro
impresso), prejudicam a visualização da página no e-reader. Diretriz #5 // TIPOGRAFIA - Selecionar, para o projeto
gráfico, família tipográfica com boa renderização em tela. De acordo com Vanderschantz (2008), a tipografia para
tela e a forma de suas letras devem ser robustas, claras e
reconhecíveis, com considerável uso de espaço entre letras
para garantir legibilidade e leiturabilidade. As informações
bibliográficas foram referendadas pelo comentário de alguns
especialistas, que reafirmaram a importância de se considerar
esse aspecto na seleção tipográfica, visando à maior
‘definição’ dos caracteres em suporte digital, ainda que as
telas de grande parte dos e-readers tenha, hoje em dia,
resolução suficiente para exibir de forma adequada a maioria
das fontes disponíveis no mercado. Diretriz #2 // FORMATO - Preferencialmente, trabalhar com
os formatos ePub2 ou ePub3. Diretriz #2 // FORMATO - Preferencialmente, trabalhar com
os formatos ePub2 ou ePub3. Diretriz #2 // FORMATO - Preferencialmente, trabalhar com
os formatos ePub2 ou ePub3. Além da compatibilidade com o iPad e com grande parte
dos demais e-readers com tela em cores, os formatos ePub2
e ePub3 permitem a configuração do e-book em layout fixo,
seguindo a primeira diretriz de projeto. No caso do ePub2, o
suporte ao layout fixo depende da utilização de uma
extensão desenvolvida pela Apple. Já no caso do ePub3, o Design & Tecnologia 16 (2018) 64 das fontes não serifadas, observou-se uma tendência entre os
pesquisadores da área a favor da utilização de tais caracteres. O argumento principal é aproximar-se do desenho de letra a
que a criança já está acostumada e do caractere que ela
mesma ‘desenha’ quando escreve à mão livre. Dessa forma,
reduzem-se as barreiras de acesso ao texto neste momento
inicial, de alfabetização. O conhecimento da variedade
tipográfica poderia, nessa lógica, acontecer em um segundo
momento, em que a criança já tem mais maturidade de
leitura. Alguns dos entrevistados mostraram-se favoráveis,
também, à utilização dos caracteres infantis, confirmando a
tendência observada na bibliografia específica. 7. DIRETRIZES PARA PROJETO DE E-BOOKS INFANTIS Diretriz #14 // ÁUDIO - Se possível, desenvolver um player de
áudio próprio para o livro e oferecer o máximo de controle ao
usuário sobre o conteúdo. Opções de reiniciar a leitura da
página ou de selecionar um ponto específico de reprodução
do conteúdo são interessantes, desde que o player se
mantenha simples de utilizar. Por vezes, as opções oferecidas pelo player do iBooks,
software padrão para leitura de e-books no iPad, não
parecem suficientes quando se fala em público infantil. O
player oferece, simplesmente, opção de ativar/desativar a
narração em
voz alta, ativar/desativar a passagem
automática de páginas e controlar o volume do áudio. Parece
faltar controle ao usuário, conforme as heurísticas de
usabilidade de Nielsen (1995) indicam. Não há, por exemplo,
opção de repetir a narração da página, ou de retornar a um
ponto específico da narração (caso a criança não tenha
entendido alguma palavra, por exemplo). Da entrevista com especialistas, extraíram-se diversos
apontamentos de que não deve haver restrições quanto ao
estilo de ilustração no livro digital infantil – a variedade de
estilos ajudará a criança a criar repertório visual. Diretriz #12 // ÁUDIO - Sempre que possível, incluir recurso
de ‘read aloud’ (leitura em voz alta) ao e-book infantil. Preferencialmente, as palavras ou sílabas devem ser grifadas
individualmente, em sincronia com a narração (em
detrimento do grifo por frase ou parágrafo). Diretriz #12 // ÁUDIO - Sempre que possível, incluir recurso
de ‘read aloud’ (leitura em voz alta) ao e-book infantil. Preferencialmente, as palavras ou sílabas devem ser grifadas
individualmente, em sincronia com a narração (em
detrimento do grifo por frase ou parágrafo). Além disso, o acesso ao player não é suficientemente
intuitivo – é necessário tocar no centro da tela para que as
configurações de áudio sejam mostradas, no canto superior
direito. A partir dessas considerações, pondera-se que a
inclusão de um player próprio para o livro digital infantil, que
fique posicionado em local fixo, na própria página do e-book,
poderia solucionar essas problemáticas de usabilidade do
player do iBooks, oferecendo um recurso mais fácil de
acessar e com mais funcionalidades, considerando-se o
público-alvo. Alguns dos e-books avaliados exploraram a
inclusão de player próprio – e apesar de não incluírem
funções adicionais, o fato de o player estar posicionado na
própria página já oferece um ganho, por ser mais intuitivo
que o acesso ao player do iBooks. 7. DIRETRIZES PARA PROJETO DE E-BOOKS INFANTIS Essa segunda
tendência bibliográfica foi considerada mais adequada pelos
especialistas consultados, que defendem que a escolha
cromática seja adequada ao conceito definido para o livro
digital infantil, garantindo a coerência da obra como um todo,
e ampliando o repertório visual da criança. Diretriz #13 // ÁUDIO - Oferecer opções de velocidade de
leitura, na funcionalidade de ‘read aloud’, ajuda a contemplar
crianças em níveis diversos de aprendizado da leitura. Ao oferecer opções de velocidade na funcionalidade de
‘read aloud’, favorece-se a leitura fracionada e o aprendizado
fonêmico, além de contemplarem-se crianças em fases
distintas do processo de alfabetização. Para uma criança
dando os primeiros passos na leitura, por exemplo, a
velocidade padrão de narração oferecida poderá ser rápida
demais, dificultando que a criança acompanhe os grifos no
texto; assim, fica prejudicado o acompanhamento da leitura
com o áudio. Apesar de a opção de duas ou mais velocidades
de narração não estar disponível em nenhum dos e-books
observados na análise de similares, a pesquisa bibliográfica
pareceu indicar que essa seria uma boa solução no sentido de
potencializar o aprendizado da leitura por meio do
acompanhamento em áudio e por meio da inclusão do maior
número possível de crianças, com habilidades em fases de
aprimoramento distintas. A entrevista com os especialistas
confirmou
que
tal
possibilidade
seria
interessante,
especialmente do ponto de vista pedagógico. Em alguns livros observados na análise de similares, o
uso de paleta cromática mais sóbria, em alguns casos,
mostrou-se adequado ao conceito do livro, sem que se
abrisse mão de uma estética infantil, por vezes fazendo uso
de alguns pontos de cor mais vibrantes – o que corrobora
com a diretriz advinda da bibliografia e da consulta aos
especialistas. Diretriz #11 // ILUSTRAÇÃO - Definir o estilo de ilustração a
partir do conceito do livro, sem restringir-se a ilustrações
realistas ou de estilo tradicional. Da pesquisa bibliográfica, compreendeu-se que, a partir
do contato com livros ilustrados, a criança forma-se como
leitora de textos verbais e não verbais. (Zimmermann, 2009). Alguns autores observaram, também, uma preferência
infantil por ilustrações realistas, como foi o caso da pesquisa
de campo realizada por Brookshire, Scharff e Moses (2002). Ao observar-se, no entanto, os e-books infantis na análise de
similares, notou-se uma variedade maior de estilo de
ilustração – alguns, com ampla aceitação pelo público, como
é o caso da personagem Peppa Pig, apresentam ilustrações
mais conceituais e/ou abstratas. 7. DIRETRIZES PARA PROJETO DE E-BOOKS INFANTIS Também da consulta aos especialistas, destacou-se que
a escolha tipográfica esteja alinhada ao conceito do livro e
que as letras tenham traçado mais orgânico e menos
mecânico – nesse sentido, caso o designer opte pelas
serifadas, as fontes de serifas retas são mais indicadas, por
terem, de modo geral, desenho mais informal. As
recomendações dessa diretriz também foram referendadas
pela análise de similares, em que o uso de fontes sem serifas
mostrou-se predominante nas publicações avaliadas. Diretriz #8 // TIPOGRAFIA - Considerar o uso de fontes com
caracteres infantis, por serem letras cujo desenho se
aproxima ao que a criança já está familiarizada. Caso a
quantidade de texto por página se resuma a poucas frases
curtas, compor a tipografia apenas em caixa alta pode,
também, ser uma alternativa viável de projeto, visto que há
uma tendência de que o primeiro contato de muitas crianças
com as letras se dê por meio do alfabeto das maiúsculas. Diretriz #10 // COR - Fazer a seleção da paleta cromática do
livro de acordo com a temática da narrativa e com o conceito
definido para o livro - nem todos os livros infantis precisam
ser compostos em cores vibrantes e variadas (e nem devem
restringir-se às cores primárias). Algumas referências de pesquisa apontam a preferência
infantil por obras compostas em cores variadas e vibrantes. Outros autores, por oposição, observam que a paleta
cromática pode ser guiada pela temática da história – e é,
inclusive, importante que haja variação entre os tipos de A utilização de caracteres infantis na composição do
texto para crianças em fase de alfabetização também não é
consenso entre os pesquisadores, mas, assim como no caso Design & Tecnologia 16 (2018) 65 narrativas para o livro ilustrado infantil – ou seja,
dependendo da temática central do livro, poderá não ser
possível explorar uma paleta de cores vivas. Essa segunda
tendência bibliográfica foi considerada mais adequada pelos
especialistas consultados, que defendem que a escolha
cromática seja adequada ao conceito definido para o livro
digital infantil, garantindo a coerência da obra como um todo,
e ampliando o repertório visual da criança. narrativas para o livro ilustrado infantil – ou seja,
dependendo da temática central do livro, poderá não ser
possível explorar uma paleta de cores vivas. 7. DIRETRIZES PARA PROJETO DE E-BOOKS INFANTIS Preferencialmente, fazer uso do hipertexto para oferecer
informações extra – que podem ou não fazer parte da
narrativa –, mas que não deixem a narrativa incompleta caso
se siga apenas a leitura linear. Para as pesquisadoras Linda Reynolds e Sue Walker
(2000, p.233), “informações claramente estruturadas e
acessíveis são importantes para ajudar as crianças a se
sentirem no controle”. Dessa forma, garante-se que a criança
terá a possibilidade de ‘navegar’ pelo conteúdo, mas com
menos chance de ‘perder-se’ na leitura do livro – essa
possibilidade de ‘navegação’ pelos conteúdos, para Basso
(2005),
proporciona
“a
descoberta
e
a
invenção,
possibilitando a formação de alunos capazes de construir seu
próprio conhecimento. O recurso de hipertexto não foi explorado nos e-books
analisados, mas os dados advindos da bibliografia e a
consulta com especialistas indicou que a utilização de tal
recurso pode ser incorporada de forma positiva ao livro
digital infantil. Especialistas referiram, por exemplo, que o
hipertexto pode ser uma oportunidade de disponibilizar
conteúdo extra no e-book – ao utilizar-se o recurso, por
exemplo, para esclarecer de forma interativa e rápida
palavras que a criança desconheça. Informações adicionais
sobre a narrativa ou as personagens também podem vir sob a
forma de hipertexto, oferecendo à criança conteúdo ‘a
explorar’ no e-book e incentivando a curiosidade e a busca
por conhecimento. Nesse sentido, considerando-se o público-
alvo, é importante que a narrativa se sustente e tenha
sentido completo mesmo que a criança não explore todos os
conteúdos ‘escondidos’ na navegação não linear. Diretriz #17 // VÍDEO - Quando se optar pela inclusão de
vídeo, observar que seu conteúdo seja relevante e que
enriqueça a experiência de leitura do livro, não apenas
repetindo a narrativa textual. Diretriz #17 // VÍDEO - Quando se optar pela inclusão de
vídeo, observar que seu conteúdo seja relevante e que
enriqueça a experiência de leitura do livro, não apenas
repetindo a narrativa textual. Complementando a diretriz anterior, ressalta-se que é
necessário, quando o projeto do e-book prevê inclusão de
vídeo, selecionar cautelosamente o conteúdo dessa mídia –
se for mera repetição da narrativa escrita, poderá, como já
referido, competir com o conteúdo em texto pela atenção da
criança. Alguns especialistas consultados mostraram-se
contra a utilização desse tipo de mídia no e-book infantil,
referindo que a estética do vídeo ‘rompe’ em demasiado a
estética do livro. 7. DIRETRIZES PARA PROJETO DE E-BOOKS INFANTIS Tal solução pode,
eventualmente, vir a prejudicar o aprendizado fonêmico, que
ocorre com a sincronização de áudio e de grifo no texto,
como se viu anteriormente. Nas entrevistas com
especialistas, opinião no mesmo sentido corroborou com as
indicações bibliográficas: um dos especialistas comentou
acreditar que a utilização de música no e-book infantil pode
acabar resultando em um excesso de informação para a
criança – assim, o ideal seria apresentar esse conteúdo
musical separadamente, para que não haja dispersão da
atenção. apreensão do texto. Para isso, é necessário oferecer,
conforme indicam as heurísticas de Nielsen (1995), controle e
liberdade ao usuário. Essa heurística traduz-se, nesse caso,
em permitir ao leitor optar por ouvir ou não a música – e em
que momento. Em um dos e-books avaliados na análise de
similares, o áudio da música estava disponível junto ao
recurso de ‘read aloud’ – a música estava disponível como
‘pano de fundo’ da narração do texto. Tal solução pode,
eventualmente, vir a prejudicar o aprendizado fonêmico, que
ocorre com a sincronização de áudio e de grifo no texto,
como se viu anteriormente. Nas entrevistas com
especialistas, opinião no mesmo sentido corroborou com as
indicações bibliográficas: um dos especialistas comentou
acreditar que a utilização de música no e-book infantil pode
acabar resultando em um excesso de informação para a
criança – assim, o ideal seria apresentar esse conteúdo
musical separadamente, para que não haja dispersão da
atenção. mídia agregue informação ao livro, ao invés de dispersar ou
de competir com o restante do conteúdo. Deve-se considerar
que a atividade de assistir a um vídeo, apesar de ser vendida
como ‘interativa’, é mais passiva para a criança do que a
atividade de ler, por exemplo. Diretriz #18 // ANIMAÇÃO - As animações incluídas no e-
book devem ter relação com o conteúdo e, preferencialmente,
ser acionadas apenas sob o comando do leitor. Optar, sempre
que possível, por animações que envolvam ação do leitor. Apesar de algumas referências bibliográficas indicarem
cautela no uso de animações no e-book infantil, a maioria dos
especialistas consultados mostrou-se a favor da utilização
desse tipo de recurso, desde que a animação tenha relação
com o conteúdo do texto e que seja utilizada, assim, como
ferramenta de auxílio ao entendimento e à interpretação da
narrativa. 7. DIRETRIZES PARA PROJETO DE E-BOOKS INFANTIS Nos e-books analisados, uma das formas de utilização do
recurso da animação mostrou-se adequada justamente por
auxiliar na interpretação do texto: a onomatopeia que
representava um grito vibrava em sincronia com o áudio,
indicando a entonação da leitura. Para os especialistas, esse
tipo de funcionalidade é interessante porque se utiliza de um
recurso do meio digital para favorecer a aprendizagem. Especialistas referiram, ainda, que a animação pode exigir
alguma ação da criança, preferencialmente relacionada aos
verbos na narrativa – mover um objeto ou um personagem
na página, por exemplo, acompanhando o que é descrito no
texto. Diretriz #16 // VÍDEO - Quando houver inclusão de vídeo ao
e-book, apresentá-lo em página à parte do conteúdo de texto. Preferencialmente, conferir o controle da reprodução do vídeo
ao leitor. A pesquisa bibliográfica permitiu identificar que a
inclusão de conteúdo em vídeo no e-book infantil, quando
apresentado dividindo o espaço da página com o texto e com
a ilustração, poderá ser prejudicial, por ‘dividir’ a atenção da
criança – para Mayer & Moreno (2003, p.45-46), por
exemplo, “a atenção visual do aprendiz é dividida entre ver a
animação e ler o texto em tela”. Dos e-books analisados,
apenas um dispunha de conteúdo em vídeo – nesse, apesar
de a mídia estar inserida em página exclusiva, o conteúdo em
vídeo pareceu não favorecer a leitura do livro, por ser mera
repetição da narrativa escrita. Junto aos especialistas, houve
referência quanto à inclusão de vídeo ao e-book infantil ter
mais apelo comercial do que vantagens reais para a criança. Um dos especialistas também referiu que o e-book que optar
por incluir vídeo deverá ter atrativos de animação e de
interatividade suficientes de modo que o conteúdo do livro
como um todo seja tão atrativo e interessante quanto o
conteúdo em vídeo. Conclui-se que, além da inclusão dessa
mídia à parte do restante do conteúdo, e do oferecimento de
controle ao leitor sobre a reprodução do conteúdo, deve-se
avaliar cautelosamente a real necessidade de inclusão do
vídeo – se for mera repetição da narrativa escrita, poderá
resultar em ‘preguiça’ da criança para ler o texto, já que, para
ela, a leitura ainda é algo difícil, certamente mais difícil do
que simplesmente dar ‘play’ e assistir à história em vídeo. Diretriz #19 // INTERATIVIDADE E HIPERTEXTO - Quando for
feito uso de recursos de hipertexto, estruturar as informações
de
forma
clara
e
deixá-las
acessíveis
ao
leitor. 7. DIRETRIZES PARA PROJETO DE E-BOOKS INFANTIS Os especialistas, de modo
geral, concordaram com a disponibilização de player próprio
para o e-book infantil, desde que a inclusão de novas
funcionalidades não dificulte o uso – deve-se manter a
simplicidade, para garantir que seja suficientemente intuitivo
e fácil de usar. A pesquisa bibliográfica indicou a importância da
presença de recurso de ‘read aloud’ nos e-books infantis:
para Colombo e Landoni (2014), por exemplo, a combinação
da leitura visual com o áudio tem o potencial de fazer a
leitura infantil em suporte digital uma experiência mais rica e
atraente para as crianças. Além de atrair a atenção infantil
para o texto, outras pesquisas apontam ganhos em termo de
aprendizagem quando há tal recurso disponível, por meio da
sincronia audiovisual (ROSKOS, BURSTEIN E GRAY, 2014). Quanto à sincronia do grifo no texto com a narração, a
opção por grifar palavras ou sílabas, em detrimento de frases
ou parágrafos, sustenta-se na pesquisa de Lewin (2000 apud
AMANTE, 2007), que credita os bons resultados da função
‘read aloud’ à leitura fracionada em sílabas ou centrada em
segmentos fonêmicos, o que auxilia a criança a aprender, aos
poucos, a pronúncia correta das palavras. Para os especialistas consultados, o recurso de narração
em voz alta é interessante e é uma ótima forma de
diferenciar e de oferecer vantagem no uso do livro digital em
comparação ao impresso. Nos e-books avaliados, aqueles
cujo projeto foi desenvolvido para o digital (e não
simplesmente transposto da versão impressa), de modo
geral, continham recurso de ‘read aloud’, o que confirma a
opinião dos entrevistados nesse sentido. Diretriz #15 // ÁUDIO - Quando houver inclusão de músicas
ao e-book, incluir um player de áudio em separado, para que
o leitor possa escolher quando – e se – deseja ouvir a música. Ao incluir música ao e-book, deve-se estar atento para
que essa mídia não distraia a criança da leitura e da Design & Tecnologia 16 (2018) 66 apreensão do texto. Para isso, é necessário oferecer,
conforme indicam as heurísticas de Nielsen (1995), controle e
liberdade ao usuário. Essa heurística traduz-se, nesse caso,
em permitir ao leitor optar por ouvir ou não a música – e em
que momento. Em um dos e-books avaliados na análise de
similares, o áudio da música estava disponível junto ao
recurso de ‘read aloud’ – a música estava disponível como
‘pano de fundo’ da narração do texto. 8. CONSIDERAÇÕES FINAIS Com o que se procurou investigar no presente trabalho,
torna-se possível afirmar que, seguindo-se algumas diretrizes
de projeto, é possível aprimorar a qualidade técnica dos e-
books voltados a crianças em fase de alfabetização,
melhorando-se a experiência de uso desses objetos digitais
em contexto de ensino-aprendizagem infantil. Por meio
dessas diretrizes, torna-se viável a utilização adequada dos
recursos gráfico-digitais que o suporte digital disponibiliza,
possibilitando ao designer projetar livros em que a criança
utilize a visão, o toque e a audição, enriquecendo sua
experiência de acesso ao texto. Recursos como a animação,
por exemplo, podem ser utilizados para auxiliar a criança a
atribuir sentido no mundo real ao conteúdo textual contido
no livro – essa ‘leitura significativa’ é parte imprescindível do
processo de alfabetização e de letramento, conforme se pôde
verificar na bibliografia específica consultada. Da mesma
forma, a tipografia, se selecionada e composta seguindo-se as
diretrizes propostas, pode contribuir com a legibilidade da
publicação, influenciando diretamente (de forma positiva) na
experiência de leitura infantil. Seguindo a mesma lógica,
todas as demais decisões de projeto que envolvem o
desenvolvimento de um e-book infantil devem ser tomadas
considerando-se o público-alvo e suas necessidades –
passando pela seleção da paleta cromática, do estilo de
ilustração, do formato da página e das animações, vídeos e
outros recursos interativos que farão parte da obra. [9]. COLOMBO, L.; LANDONI, M. A diary study of children's
user experience with EBooks using flow theory as
framework. In: IDC '14 Proceedings of the 2014
conference on Interaction design and children. Aarhus,
Denmark. June 2014, p.135-144. [10]. DAYWALT, D.; JEFFERS, O. The Day the Crayons Quit. 2013. Disponível em: <https://itunes.apple.com/br/book
/the-day-the-crayons-quit/id578772370?mt=11>. Acesso
em: 25 out. 2015. [11]. DRESANG, E. T. More research needed: Informal
information-seeking behavior of youth on the Internet. In: Journal of the American Society for Information
Science. Vol. 50, p.1123–1124, 1999. [12]. DURÃES, R.; PASSOS, V. O Universo. 2015. Disponível em:
<https://itunes.apple.com/br/book/o-universo/id10277
62084?mt=13>. Acesso em: 25 out. 2015. [13]. GREGORICH, B.; MASHERIS, R. The fox on the Box. 1996. Disponível em: <https://itunes.apple.com/us/book/the-
fox-on-the-box/id414956890?mt=11>. Acesso em: 25
out. 2015. [14]. GUERNSEY, Lisa. Are e-books any good? 2011. Disponível
em: <http://www.schoollibraryjournal.com/slj/printissu
ecurrentissue/890540-427/are_ebooks_any_good.html
.csp>. Acesso em 21 jun. 2013. [15]. KLOHN, S. C.; FENSTERSEIFER, T. A. Contribuições do
Design Editorial para a Alfabetização Infantil. In:
Infodesign | Revista Brasileira de Design da Informação. São Paulo, v.9, n.1, p.45-51, 2012. 7. DIRETRIZES PARA PROJETO DE E-BOOKS INFANTIS Dadas essas restrições, conclui-se de fato
imprescindível que o roteiro do conteúdo em vídeo seja
cautelosamente projetado, de modo que a inclusão dessa Design & Tecnologia 16 (2018) 67 As_TIC_na_Escola_e_no_Jardim_de_Infancia_motivos_e
_factores_para_a_sua_integracao>. Acesso em: 27 set. 2014. Diretriz #20 // INTERATIVIDADE E HIPERTEXTO - Caso o
designer opte por uma navegação não linear no livro digital
infantil, preferencialmente oferecer, também, uma opção de
navegação linear, de modo a contemplar o maior número
possível de crianças, em fases distintas do desenvolvimento
de suas habilidades. [2]. ASTLEY, N.; BAKER, M. Peppa Meets the Queen. 2012. Disponível em: <https://itunes.apple.com/br/book/pepp
a-pig-peppa-meets-queen/id522357276?mt=11>. Acesso
em: 25 out. 2015. Quando se fala em público infantil e hipertexto, diversos
autores fazem a ressalva de que, apesar das vantagens do uso
desse tipo de recurso no que diz respeito à exploração infantil
do conteúdo, há que se considerar que o hipertexto pode
acabar por desviar o leitor de seu objetivo principal de
apreensão do conteúdo do texto (VIEIRA, 2007 apud LIMA,
2009). Além disso, considerando-se as características de
desenvolvimento diferenciadas de cada criança, a não
linearidade do e-book pode, eventualmente, ser um
empecilho para a leitura do livro, dependendo da criança. Dessa forma, considera-se adequado que o livro ofereça uma
opção de navegação linear, de forma a atender às
necessidades de crianças com capacidades distintas. Tal
assertiva foi corroborada na entrevista com especialistas, que
também referenciaram que se tenha cautela na utilização de
recursos de interatividade e de hipertexto no e-book voltado
a crianças em fase de alfabetização. [3]. BERGER, S.; SANTAT, D. Crankenstein. 2013. Disponível
em: <https://itunes.apple.com/us/book/crankenstein/id
64 9512518?mt=11>. Acesso em: 25 out. 2015. [4]. BROOKSHIRE, J.; SCHARFF, L. F.V; MOSES, L.E. The
influence of Illustrations on children's book preferences
and comprehension. In: Reading Psychology, 23:5, p.323-
339, 2002. [5]. BUCKINGHAM, D. Cultura Digital, Educação Midiática e o
Lugar da Escolarização. In: EDUCAÇÃO & REALIDADE. Porto Alegre: UFRGS / FACED, v. 35, n. 3, p.37-58,
set./dez. 2010. [6]. BURT, C.; COOPER, W. F.; MARTIN, J. L. A Psychological
Study of Typography. In: The British Journal of Statistical
Psychology. London, Vol. VIII, Part I, p.29-56, may/1955. [7]. CARVALHO, R.; MARTINS, R. Duda e o Bichinho do Medo. 2013. Disponível em: <https://itunes.apple.com/br/book
/duda-e-o-bichinho-do-medo/id856607585?mt=11>. Acesso em: 25 out. 2015. [8]. CHARTIER, R. Os Livros Resistirão às Tecnologias Digitais. 2007. Disponível em: <https://novaescola.org.br/conteu
do/938/roger-chartier-os-livros-resistirao-as-tecnologias
-digitais>. Acesso em: 10 set. 2017. 8. CONSIDERAÇÕES FINAIS Espera-se que os resultados desta pesquisa possam
contribuir para que o e-book consolide-se, por meio do
aprimoramento
de
seu
projeto
gráfico-visual,
como
ferramenta de suporte ao ensino e como objeto de auxílio e
de incentivo à leitura, especialmente junto ao público infantil. Espera-se, ainda, que designers envolvidos em projetos deste
tipo possam ter, nesta pesquisa, subsídios para auxiliar e
aprimorar o desenvolvimento de seus projetos. [16]. LINDEN, Sophie Van der. Para ler o livro ilustrado. São
Paulo, Cosac Naify, 2011. [17]. LOURENÇO, Daniel Alvares. Tipografia para livro de
literatura infantil: desenvolvimento de um guia com
recomendações tipográficas para designers. 2011,
Dissertação (Mestrado em Design), Programa de Pós-
graduação em Design, Universidade Federal do Paraná,
Curitiba, 2011. REFERÊNCIAS [1]. AMANTE, L. As TIC na Escola e no Jardim de Infância:
motivos e factores para a sua integração. 2007. Disponível em: <https://www.academia.edu/3561151/ [1]. AMANTE, L. As TIC na Escola e no Jardim de Infância:
motivos e factores para a sua integração. 2007. Disponível em: <https://www.academia.edu/3561151/ [18]. MARQUES, C.; PASSOS, V. Chapeuzinho Vermelho. 2014. Disponível em: <https://itunes.apple.com/br/book/chap Design & Tecnologia 16 (2018) 68 euzinho-vermelho/id877423993?mt=11>. Acesso em: 25
out. 2015. [22]. RUMJANEK, L. G. Tipografia para crianças: um estudo de
legibilidade. Dissertação (Mestrado em Design) - Escola
Superior de Desenho Industrial, Universidade do Estado
do Rio de Janeiro. 2009. [19]. O'REILLY, M.; WATANABE, M. Você deve tentar, tentar e
tentar! 2013. Disponível em: <https://itunes.apple.com
/br/book/voce-deve-tentar-tentar-e/id681331813?mt=1
1>. Acesso em: 25 out. 2015. [23]. SILVERSTEIN, S. The Giving Tree. 2014. Disponível em:
<https://itunes.apple.com/br/book/the-giving-tree/id66
0420754?mt=11>. Acesso em: 25 out. 2015. [20]. ROA, M. M.; ¿Puede el diseño tipográfico fomentar el
interés en la lectura y mejorar el rendimiento de los más
pequeños? In: Entre líneas – Ensayo sobre diseño,
tipografía y lectura. 1a. edição, p. 64-71, 2013. [24]. TESTA, M.; OSTERHOLD, J. Olivia Loves Halloween. 2014. Disponível em: <https://itunes.apple.com/br/book/olivia
-loves-halloween/id721184959?mt=11>. Acesso em: 25
out. 2015. [21]. ROSKOS, K., BURSTEIN, K., SHANG, Y., GRAY, E. Young
Children’s Engagement with E-Books at School. 2014. Disponível em: <http://sgo.sagepub.com/content/4/1/
2158244013517244>. Acesso em: 25 set. 2014 [25]. WALKER, S. Typography in Children's Books. 2005. Disponível em: <http://booksforkeeps.co.uk/issue/154/
childrens-books/articles/other-articles/typography-in-chi
ldrens-books>. Acesso em: 13 jul. 2015.
|
https://openalex.org/W1990769998
|
https://europepmc.org/articles/pmc4322544?pdf=render
|
English
| null |
Increased use of hypnotics in individuals with celiac disease: a nationwide case-control study
|
BMC gastroenterology
| 2,015
|
cc-by
| 6,910
|
© 2015 Mårild et al.; licensee BioMed Central. This is an Open Access article distributed under the terms of the Creative
Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and
reproduction in any medium, provided the original work is properly credited. The Creative Commons Public Domain
Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article,
unless otherwise stated. Abstract Background: Although poor sleep is common in numerous gastrointestinal diseases, data are scarce on the risk
of poor sleep in celiac disease. The objective of this study was to estimate the risk of repeated use of hypnotics
among individuals with celiac disease as a proxy measure for poor sleep. Methods: This is a nationwide case–control study including 2933 individuals with celiac disease and 14,571
matched controls from the general Swedish population. Poor sleep was defined as ≥2 prescriptions of hypnotics
using prospective data from the National Prescribed Drug Register (data capture: July 2005-January 2008). We
estimated odds ratios and hazard ratios for poor sleep before and after celiac disease diagnosis respectively. Results: In this study, poor sleep was seen in 129/2933 individuals (4.4%) with celiac disease, as compared with
487/14,571 controls (3.3%) (odds ratio = 1.33; 95% CI = 1.08-1.62). Data restricted to sleep complaints starting ≥1
year before celiac disease diagnosis revealed largely unchanged risk estimates (odds ratio = 1.23; 95% CI = 0.88-1.71)
as compared with the overall risk (odds ratio 1.33). The risk of poor sleep in celiac disease was essentially not
influenced by adjustment for concomitant psychiatric comorbidity (n = 1744, adjusted odds ratio =1.26;
95% CI = 1.02-1.54) or restless legs syndrome (n = 108, adjusted odds ratio = 1.33; 95% CI = 1.08-1.63). Poor sleep
was also more common after celiac disease diagnosis as compared with matched controls (hazard ratio = 1.36;
95% CI = 1.30-1.41). Conclusions: In conclusion, individuals with celiac disease suffer an increased risk of poor sleep, both before and after
diagnosis. Although we cannot rule out that surveillance bias has contributed to our findings, our results are consistent
with previous data suggesting that sleep complaints may be a manifestation of celiac disease. Keywords: Coeliac, Immune-mediated, Insomnia, Sleep disorders, Small intestine Poor sleep commonly includes difficulties in initiating
and maintaining sleep, or a non-restorative sleep with or
without daytime consequences [4]. According to Swedish
data, up to 20% of those with a self-reported sleeping
disorder use prescribed sleeping pills [5]. Although pre-
scription of hypnotics, such as benzodiazepines or
benzodiazepine-related drugs, is primarily recommended
for short-term treatment of acute sleep disorders, in
routine clinical practice pharmacological treatment is
also common for prolonged-periods (months to years)
of poor sleep [6]. Mårild et al. BMC Gastroenterology (2015) 15:10
DOI 10.1186/s12876-015-0236-z Mårild et al. BMC Gastroenterology (2015) 15:10
DOI 10.1186/s12876-015-0236-z Increased use of hypnotics in individuals with
celiac disease: a nationwide case-control study Karl Mårild1,2,8*, Timothy I Morgenthaler3, Virend K Somers4, Suresh Kotagal5, Joseph A Murray6
and Jonas F Ludvigsson1,7 Background Celiac disease (CD) is a chronic autoimmune disorder
affecting 1-2% of the Western population [1]. In CD,
dietary gluten not only leads to small-intestinal villous
atrophy, but also affects organs outside the gastrointestinal
tract [2]. Individuals with CD therefore present with a great
variety of symptoms and signs: classical symptoms of
malabsorption (e.g. diarrhea with abdominal pain) and
non-classical symptoms (e.g. fatigue, peripheral neuropathy
and restless legs syndrome (RLS) [3]). Poor sleep frequently occurs in a number of gastro-
intestinal diseases, e.g. inflammatory bowel disease (IBD)
[7]. Not only may gastrointestinal symptoms impair the
quality of sleep, but poor sleep may also have a detrimental * Correspondence: karlmarild@gmail.com
1Department of Medical Epidemiology and Biostatistics, Karolinska Institutet,
Stockholm, Sweden
2Division of Epidemiology, Norwegian Institute of Public Health, Oslo, Norway
Full list of author information is available at the end of the article * Correspondence: karlmarild@gmail.com
1Department of Medical Epidemiology and Biostatistics, Karolinska Institutet,
Stockholm, Sweden
2Division of Epidemiology, Norwegian Institute of Public Health, Oslo, Norway
Full list of author information is available at the end of the article Mårild et al. BMC Gastroenterology (2015) 15:10 Page 2 of 8 males aged 76 year in the part of Dalsland where this
man lived. effect on the course of both functional and inflammatory
gastrointestinal conditions [7]. Poor sleep may also add to
the burden of disease by reducing the quality of life [8] and
predispose to the development of comorbidities, such as
psychiatric diseases [9]. Although poor sleep has been
closely studied in many other gastrointestinal diseases [10],
to our knowledge, only one previous study has addressed
the risk of poor sleep in CD [11]. That cross-sectional
study, based on 60 patients identified at one tertiary center,
found poorer sleep quality among individuals with CD,
both before and after treatment with a gluten-free diet, as
compared with healthy volunteers [11]. We then linked data on individuals with CD and their
matched controls to the Prescribed Drug Register [12]
(established July 2005) in order to identify individuals
with repeated use of hypnotics before CD diagnosis (and
corresponding date in controls). Our main analysis was
therefore restricted to individuals diagnosed with CD be-
tween July 2005 and January 2008, including 2,933 indi-
viduals with CD and 14,571 matched controls. Study population Between 2006 and 2008, we searched the computerized
register from all Sweden’s pathology departments (n =
28) to identify individuals with biopsy-verified CD, de-
fined as small-intestinal villous atrophy (Marsh grade 3)
[13]. We have earlier shown that 95% of Swedish indi-
viduals with villous atrophy have CD [13]. The current
study was based on the same dataset as described in our
study on mortality in CD, [14] originally including
29,096 individuals with CD diagnosed between 1969 and
2008. Time of CD diagnosis was defined as time of first
biopsy showing villous atrophy. For each of these
individuals with CD the government agency Statistics
Sweden tried to identify five controls matched for age,
sex, calendar period of birth and county of residence
through the Total Population Register, which includes
demographic data on all Swedish residents [15]. Controls
were sampled from all Swedish residents in whom the
computerized pathology registers did not indicate prior
duodenal/jejunal biopsy. After exclusion of controls with
data irregularities and those that could not be matched
144,522 controls were identified. Due to lack of eligible
controls some individuals with celiac disease only had
four controls and the average numb of controls per indi-
vidual with celiac disease was slightly below 5. Where a
patient had celiac disease, especially if diagnosed in old
age, the responsible government agency sometimes failed
to identify 5 matching controls. E.g. 76-year old man di-
agnosed with celiac disease in 1989 in Dalsland county. This means that this person was born in 1913. It may
then have been that in 1989 there were only 4 other g
y
We collected data on use of hypnotics from July 1st
2005 through January 29th 2008 and up to the date of
CD diagnosis (and the corresponding date in matched
controls). In order to minimize the risk of false positive
cases we restricted our definition of poor sleep as ≥2
prescriptions of hypnotics before CD diagnosis. Hyp-
notics were classified according to the Anatomical
Therapeutic Chemical (ATC) pharmaceutical classifica-
tion system [16]: benzodiazepine-related drugs (i.e. non-
benzodiazepine GABA-A receptor agonists), benzodiaze-
pines, other hypnotics and melatonin receptor agonists
(see Additional file 1). A similar classification of hyp-
notics has previously been used [17]. Regrettably, the
National Patient Register is unsuitable for identification
of poor sleep (e.g. Background In order
to estimate the risk of poor sleep after CD diagnosis, we
performed a pre-planned subanalysis including 26,587
individuals with CD diagnosed between 1969 and 2008
and 133,465 matched controls with a follow-up until July
2005 (start of the Prescribed Drug Register) and with no
use of hypnotics prior to CD diagnosis. See Figures 1
and 2 for overview of study methodology. In the current nationwide study including some 2900
individuals with CD we examined the risk of poor sleep
defined as repeated use of hypnotics, both before and
after CD diagnosis. Poor sleep
Th
P The Prescribed Drug Register, established in July 2005,
contains prospectively recorded high-quality data on vir-
tually all prescribed drugs in Sweden [12]. In Sweden,
the use of hypnotics is strictly controlled and sleep med-
ications are generally not sold over the counter. Methods Data on individuals with CD were linked to the Swedish
Prescribed Drug Register [12] in order to examine the
risk of poor sleep in CD. Poor sleep was defined as re-
peated use (≥2 prescriptions) of hypnotics. Study population diagnosed insomnia) as this registry is
limited to International Classification of Disease (ICD)
codes from inpatient and non-primary outpatient care
clinics with low sensitivity to ascertain poor sleep. Celiac disease
(n=2933)
Matched controls
(n=14,571)
Prescribed Drug Register
July 2005 - January 2008
Repeated use of hypnotics
Time: July 2005-January 2008
Figure 1 Overview methodology of main analysis: risk of poor
sleep before celiac disease diagnosis. Figure 1 Overview methodology of main analysis: risk of poor
sleep before celiac disease diagnosis. Page 3 of 8 Mårild et al. BMC Gastroenterology (2015) 15:10 Celiac disease
(n=26,587)
Matched controls
(n=133,465)
Prescribed Drug Register
July 2005 - January 2008
Repeated use of hypnotics
Time: 1969-January 2008
Figure 2 Overview methodology of subanalysis: risk of poor
sleep after celiac disease diagnosis. We considered education level [4,23], as a measure
of socioeconomic position, psychiatric comorbidity [24]
and epilepsy [25] as potential confounders. In a subana-
lysis we therefore adjusted for education level using
seven predefined education categories determined by
Statistics Sweden. We also adjusted for epilepsy and
psychiatric comorbidity, including substance use and
mood disorders, using relevant ICD-codes recorded in
the National Patient Register [26] (see Additional file 3
for ICD-codes). Subanalyses We used Cox regression to estimate Hazard Ratios
(HRs) and 95% CIs for poor sleep (≥2 prescriptions of
hypnotics) after CD diagnosis. In this subanalysis we in-
cluded 26,587 individuals with CD diagnosed between
1969 and 2008 and 133,465 matched controls with a
follow-up until July 2005 (start of the Prescribed Drug
Register) and with no use of hypnotics prior to CD diag-
nosis (see Figure 2 for overview of methodology of suba-
nalysis). Follow-up started at time of CD diagnosis (and
corresponding date in controls) and ended at time of
first prescription of hypnotics (July 2005 or later), death,
emigration or on December 31st 2009, whichever oc-
curred first. In controls, follow-up could also end if the
individual was diagnosed with CD. We performed strati-
fied analyses by sex and age at CD diagnosis as well as
subanalyses adjusting for education level, psychiatric co-
morbidity and epilepsy. To investigate potential contributing factors to poor
sleep in individuals with CD we performed a post-hoc
analysis adjusting for sleep apnea and RLS since these
two conditions may impair sleep [18-20]. Most studies
have indicated an increased risk of RLS in CD [3,21],
possibly due to prevalent iron deficiency in celiac pa-
tients. There are no data on the specific risk of sleep
apnea in individuals with CD. We identified RLS using
relevant ICD-codes in the National Patient Register
(ICD-8: 787.00; ICD-9: 333X and ICD-10: G25.8) or as
prescription of dopaminergic drugs typically used in the
treatment of RLS (excluding individuals with concomi-
tant Parkinson’s disease) (see Additional file 2 for ATC-
codes). Sleep apnea was identified using the National
Patient Register (ICD-10: G47.3). Ethics This study was conducted in accordance with the na-
tional and institutional standards and was approved by
the Regional Ethical Vetting Board in Stockholm. Since
none of the participants was contacted and individual in-
formation was anonymized prior to the analyses, in-
formed consent was not required by the research ethics
committee. Statistical analyses
d
d SPSS version 20.0 was used for all statistical analyses. We used conditional logistic regression to estimate Odds
Ratios (ORs) and 95% confidence intervals (CIs). Through
conditional logistic regression each stratum (one individ-
ual with CD compared with his/her matched controls)
was analyzed separately before a summary OR was calcu-
lated [22]. This statistical approach eliminated the effect of
sex, age, county and calendar year on our ORs. Practically,
conditional logistic regression works similar as regular
logistic regression except its comparison is carried out
in matched sets. Results We performed stratified analyses by age at CD diagno-
sis (<19 years, 19–39 years, 40–59 years and ≥60 years)
and sex. The therapeutic indications differ between types
of hypnotics. In pre-planned subanalyses we therefore
estimated the association between CD and poor sleep
according to the type of hypnotics used: benzodiazepine-
related drugs, benzodiazepines, other hypnotics and mela-
tonin receptor agonists. We also estimated the dose- and
time-dependent association between use of hypnotics and
subsequent CD: (1) when individuals had received two or
at least three prescriptions of hypnotics and (2) when hyp-
notics had been prescribed in the last year (≤365 days) or
more than one year (>365 days) before CD diagnosis,
respectively. Some 40% of the individuals with CD had been diag-
nosed in childhood and 19% were aged 60 years or more
at time of diagnosis (median age at CD diagnosis was 28
years). Most of the individuals with CD included in our
main analysis were diagnosed in year 2006 and some
60% of the study participants were female (Table 1). Overall, 129 individuals (4.4%) with CD had been
treated for poor sleep, as compared with 487 matched
controls (3.3%), corresponding to an OR of 1.33 (95%
CI = 1.08-1.62) (adjustment for education level: adjusted
OR = 1.34; 95% CI = 1.09-1.64). Because specific hyp-
notics may occasionally also be used in the treatment of
epilepsy and various psychiatric diseases we adjusted for Page 4 of 8 Mårild et al. BMC Gastroenterology (2015) 15:10 Table 1 Descriptive characteristics of individuals with
celiac disease (CD) and matched controls
Celiac disease
Controls
Total
2933
14,571
Sex
Females (%)
1796 (61.2)
8926 (61.3)
Males (%)
1137 (38.8)
5645 (38.7)
Age at study entry,
years (median; rangea)
28; 0-94
28; 0-94
Age groups
≤19 years (%)
1218 (41.5)
6074 (41.7)
20-39 years (%)
566 (19.3)
2809 (19.3)
40-59 years (%)
583 (19.9)
2905 (19.9)
≥60 years (%)
566 (19.3)
2783 (19.1)
Calendar year of CD diagnosis
2005b (%)
819 (27.9)
4062 (27.9)
2006 (%)
1828 (62.3)
9097 (62.4)
2007c (%)
274 (9.3)
1352 (9.3)
2008d (%)
12 (0.4)
60 (0.4)
aTen children were diagnosed of CD during their first year of life and are in
the table referred to as age “0”. bBeginning of study period: July 1st 2005. cMost pathology departments only delivered small-intestinal biopsy data up to
the beginning of 2007. Results Therefore this study includes fewer individuals with CD
diagnosed 2007–2008, as compared with 2005–2006. dEnd of study period: January 29th 2008. Table 1 Descriptive characteristics of individuals with
celiac disease (CD) and matched controls least two prescriptions of hypnotics, as among those
with three or more prescriptions of hypnotics (Table 2). To estimate a possible time-dependent association be-
tween poor sleep and subsequent CD we performed two
subanalyses estimating the association between poor
sleep present in the last year (≤365 days) and poor sleep
present more than one year (≥366 days) before CD diag-
nosis. Some 4.3% of individuals with CD, as compared
with some 3.2% in controls, suffered from poor sleep in
the year before CD diagnosis (OR = 1.35; 95% CI = 1.10-
1.65). The OR remained largely unchanged when we re-
stricted our data to poor sleep present more than one
year before CD diagnosis (OR = 1.23; 95% CI = 0.88-
1.71). In Tables 3 and 4 we describe the association be-
tween CD and poor sleep according to used hypnotics
subgroup and according to time to CD diagnosis. g
g
g
In a number of pre-planned subanalyses we estimated
the risk of poor sleep after CD diagnosis (and corre-
sponding date in controls). During follow-up, 3323 of
26,587 individuals (12.5%) were treated for poor sleep
after CD diagnosis, as compared with 13,067/133,465
(9.8%) among controls, corresponding to a HR of 1.36
(95% CI = 1.30-1.41). The increased risk of poor sleep
after CD diagnosis remained largely unchanged when
adjusted for education level, epilepsy or psychiatric co-
morbidity (adjusted HR = 1.30; 95% CI = 1.25-1.36). Fur-
thermore, individuals with CD had a similar increased
risk poor sleep in the future irrespective of age at CD
diagnosis, sex and according to type of hypnotic used
(Table 5). these comorbidities although with only a minor change
in risk estimates (epilepsy: n = 207; psychiatric disease:
n = 1744; adjusted OR = 1.26; 95% CI = 1.02-1.54). The prevalence of poor sleep treated with prescribed
hypnotics increased with increasing age at CD diagnosis
being most common in those aged 60 years or older;
and there were very few children treated for poor sleep
(Table 2). Results Poor sleep was significantly associated with
CD in individuals aged ≥60 years at time of diagnosis
(OR = 1.40; 95% CI = 1.07-1.83); however the risk esti-
mates did not attain statistical significance in the youn-
ger age groups and with overlapping CIs between age
groups. Overall, poor sleep was more prevalent in fe-
males than in males (Table 2). However, there were only
small differences in risk estimates for poor sleep between
females (OR = 1.38; 95% CI = 1.08-1.76) and males with
CD (OR = 1.21; 95% CI = 0.83-1.75) as compared with
sex-matched controls. Post-hoc analyses To investigate definitive sleep disorders among individ-
uals with CD and poor sleep we identified those with a
previous diagnosis of sleep apnea or with RLS (defined
by relevant ICD-code or prescription of drugs typically
used in RLS). In 19/2933 individuals with CD (0.6%) a
diagnosis of RLS was present, as compared with 89/
14,571 (0.6%) in controls. Sleep apnea was identified in
10/2933 (0.3%) of individuals with CD and in 63/14571
(0.4%) of controls. Adjustment for sleep apnea and RLS
did not influence the risk of poor sleep in CD (adjusted
OR = 1.33; 95% CI = 1.08-1.63). Discussion Table 2 Risk of poor sleep before celiac disease diagnosis Odds ratios estimated through conditional logistic regression modelling. aPoor sleep defined as ≥2 prescriptions of hypnotics before celiac disease diagnosis. Data capture of hypnotics: July 1st 2005-January 29th 2008. bSee Additional file 1: Table S1 for anatomical therapeutic chemical (ATC) codes used to classify hypnotics. cThe category “≥3 prescriptions of hypnotics” is being compared with the category “<3 prescriptions of hypnotics”. d previous data [11] suggests that sleep complaints may be
a manifestation of CD. previous data [11] suggests that sleep complaints may be
a manifestation of CD. The use of biopsy data to identify CD brings several ad-
vantages, such as being able to identify an average CD
population. The use of biopsy data also minimizes the risk
of CD misclassification. In Sweden, small intestinal biopsy
has long been the gold standard for diagnosing CD and
previous data have shown that more than 95% of Swedish
physicians perform a small-intestinal biopsy before diagnos-
ing CD [13]. An earlier validation study found that 95% of
Swedish individuals with villous atrophy have CD [13]. Discussion In this nationwide case–control study we found a mod-
estly increased risk of poor sleep (≥2 prescriptions of
hypnotics) in individuals with CD, both before and after
diagnosis. This association was not influenced by adjust-
ment for education, psychiatric disease, sleep apnea or
RLS. Although, we cannot rule out that surveillance bias
might have influenced our results, i.e. that individuals
with poor sleep also had other complaints resulting in a
CD investigation, the consistency of our results with Benzodiazepine-related drugs (e.g. Zolpidem) were the
most frequently used type of hypnotics, being used by
3.8% of individuals with CD and 2.6% of controls, corre-
sponding to an OR of 1.47 (95% CI = 1.18-1.82). There
were no significant associations between CD and use of
any of the remaining types of hypnotics (Table 2). We
found no indication of a dose–response relationship,
with similar risk estimates among individuals with at Page 5 of 8 Page 5 of 8 Mårild et al. BMC Gastroenterology (2015) 15:10 Table 2 Risk of poor sleep before celiac disease diagnosis
Celiac disease (%)
Controls (%)
Odds ratio
95% CI
Poor sleepa
128/2933 (4.4)
487/14,571 (3.3)
1.33
1.08-1.62
Type of hypnotics usedb
Benzodiazepine-related drugs
111 (3.8)
380 (2.6)
1.47
1.18-1.82
Benzodiazepines
17 (0.6)
65 (0.4)
1.28
0.76-2.16
Other hypnotics
32 (1.1)
143 (1.0)
1.11
0.76-1.62
Melatonin receptor agonistsc
0 (0.0)
2 (0.0)
-
-
≥3 prescriptions of hypnoticsc
Any hypnotics
78 (2.7)
359 (2.5)
1.07
0.84-1.38
Benzodiazepine-related drugs
62 (2.1)
256 (1.8)
1.19
0.90-1.58
Benzodiazepines
7 (0.2)
41 (0.3)
0.85
0.39-1.88
Other hypnotics
12 (0.4)
86 (0.6)
0.70
0.39-1.28
Melatonin receptor agonistsd
0 (0.0)
2 (0.0)
-
-
Age
≤19 years
1/1218 (0.1)
4/6074 (0.1)
1.23
0.15-9.80
20-39 years
8/566 (1.4)
49/2809 (1.7)
0.82
0.39-1.72
40-59 years
39/583 (6.7)
142/2905 (4.9)
1.38
0.96-1.97
≥60 years
80/566 (14.1)
292/2783 (10.5)
1.40
1.07-1.83
Sex
Males
36/1137 (3.2)
150/5645 (2.7)
1.21
0.83-1.75
Females
92/1796 (5.1)
337/8926 (3.8)
1.38
1.08-1.76
Odds ratios estimated through conditional logistic regression modelling. aPoor sleep defined as ≥2 prescriptions of hypnotics before celiac disease diagnosis. Data capture of hypnotics: July 1st 2005-January 29th 2008. bSee Additional file 1: Table S1 for anatomical therapeutic chemical (ATC) codes used to classify hypnotics. cThe category “≥3 prescriptions of hypnotics” is being compared with the category “<3 prescriptions of hypnotics”. dDue to lack of events no odds ratio was estimated. aPoor sleep defined as ≥2 prescriptions of hypnotics before celiac disease diagnosis. Data capture of hypnotics: July 1st 2005-January
bSee Additional file 1: Table S1 for anatomical therapeutic chemical (ATC) codes used to classify hypnotics. Odds ratios estimated through conditional logistic regression modelling. ds ratios estimated through conditional logistic regression modelling. oor sleep defined as ≥2 prescriptions of hypnotics after CD diagnosis. Data capture of hypnotics: July 1st 2005-January 29th 2008.
Addi i
l fil
T bl
S
f
i
l h
i
h
i
l (ATC)
d
d
l
if h
i Strengths and weaknesses Data capture of hypnotics: July 1st 2006-January 29th 2008. bSee Additional file 1: Table S1 for anatomical therapeutic chemical (ATC) codes used to classify hypnotics. cDue to lack of events no odds ratio was estimated. our definition of hypnotics, e.g. sedating antidepressants. Furthermore, non-adherence to prescribed hypnotics as
well as the left truncation of exposure data with no data
on hypnotic use before July 2005, may also cause mis-
classification of our definition of poor sleep. However,
such potential misclassification should not differ by CD
status and therefore not lead to spurious associations,
but might instead somewhat underestimate the true
effect. suffer from some form of sleep complaints [4], such as dif-
ficulty initiating sleep, disrupted sleep and non-restorative
sleep, we lack data on the specific nature of these com-
plaints. Furthermore, previous data have shown that rela-
tively few of those with a self-reported sleeping disorder
receive treatment with hypnotics [5]. Although the patho-
genesis may be similar in mild and more severe sleep com-
plaints, our results may primarily represent severe sleep
complaints requiring hypnotic prescription. Traditionally,
prescription of hypnotics in children with poor sleep has
been limited, and with the short follow-up of children in
this study we may have underestimated the prevalence of
poor sleep in children. Strengths and weaknesses A strength of our study is the use of prospectively recorded
exposure and outcome data, eliminating the risk of recall
bias. Further, the large statistical power allowed for import-
ant subanalyses, such as examination of dose- and time-
dependent associations. We were also able to adjust for sev-
eral potential confounders, e.g. psychiatric comorbidities. This study has some limitations. Although it is likely
that most of the individuals with repeated use of hypnotics Table 3 Risk of poor sleep starting within 1 year before celiac disease diagnosis Table 3 Risk of poor sleep starting within 1 year before celiac disease diagnosis
Celiac disease n = 2,933 (%)
Controls n = 14,571 (%)
Odds ratio
95% CI
Poor sleepa
126 (4.3)
471 (3.2)
1.35
1.10-1.65
Type of hypnotics usedb
Benzodiazepine-related drugs
84 (2.9)
281 (1.9)
1.48
1.16-1.90
Benzodiazepines
5 (0.2)
35 (0.2)
0.71
0.28-1.79
Other hypnotics
7 (0.2)
56 (0.4)
0.63
0.29-1.37
Melatonin receptor agonistsc
0 (0.0)
1 (0.0)
-
-
Odds ratios estimated through conditional logistic regression modelling. aPoor sleep defined as ≥2 prescriptions of hypnotics before celiac disease diagnosis. Data capture of hypnotics: July 1st 2005-January 29th 2008. bSee Additional file 1: Table S1 for anatomical therapeutic chemical (ATC) codes used to classify hypnotics. Table 3 Risk of poor sleep starting within 1 year before celiac disease diagnosis
Celiac disease n = 2,933 (%)
Controls n = 14,571 Odds ratios estimated through conditional logistic regression modelling. Mårild et al. BMC Gastroenterology (2015) 15:10 Page 6 of 8 Table 4 Risk of poor sleep starting more than 1 year before celiac disease diagnosis
Celiac disease n = 1086 (%)
Controls n = 5382 (%)
Odds ratio
95% CI
Poor sleepa
49 (4.5)
200 (3.7)
1.23
0.88-1.71
Type of hypnotics usedb
Benzodiazepine-related drugs
40 (3.7)
151 (2.8)
1.32
0.92-1.90
Benzodiazepines
7 (0.6)
20 (0.4)
1.70
0.73-3.81
Other hypnotics
5 (0.5)
51 (0.9)
0.49
0.20-1.23
Melatonin receptor agonistsc
-
-
-
-
aPoor sleep defined as ≥2 prescriptions of hypnotics before celiac disease diagnosis. Data capture of hypnotics: July 1st 2006-January 29th 2008. bSee Additional file 1: Table S1 for anatomical therapeutic chemical (ATC) codes used to classify hypnotics. cDue to lack of events no odds ratio was estimated. Table 4 Risk of poor sleep starting more than 1 year before celiac disease diagnosis aPoor sleep defined as ≥2 prescriptions of hypnotics before celiac disease diagnosis. oor sleep defined as ≥2 prescriptions of hypnotics after CD diagnosis. Data capture of hypnotics: July 1st 2005 January 29th 2008.
e Additional file 1: Table S1 for anatomical therapeutic chemical (ATC) codes used to classify hypnotics. aPoor sleep defined as ≥2 prescriptions of hypnotics after CD diagnosis. Data capture of hypnotics: July 1st 2005-January 29th 20
bSee Additional file 1: Table S1 for anatomical therapeutic chemical (ATC) codes used to classify hypnotics y
g
y
y
y
for anatomical therapeutic chemical (ATC) codes used to classify hypnotics. Poor sleep defined as ≥2 prescriptions of hypnotics after CD diagnosis. Data capture of hypnotics: July 1st 2005-January 29th 2008.
bSee Additional file 1: Table S1 for anatomical therapeutic chemical (ATC) codes used to classify hypnotics. Hazard ratios estimated through cox regression modelling. Previous studies We know of only one previous study on the risk of poor
sleep in CD. In 2010, Zingone et al. [11] used the Pittsburg
Sleep Quality Index and found that compared with
healthy volunteers, both untreated and treated celiac
patients had significantly poorer sleep quality, including
prolonged sleep latency and shorter duration of sleep. In this study, including 60 patients from a tertiary centre,
sleep quality was inversely associated with both mental To minimize the risk of false positive cases we re-
stricted our definition of poor sleep to repeated (≥2) pre-
scriptions of drugs that are mainly indicated for the
treatment of poor sleep. We acknowledge that some of
the study participants with sleep complaints may have
been prescribed treatments that were not included in Table 5 Risk of poor sleep after celiac disease diagnosis
Celiac disease n = 26,587 (%)
Controls n = 133,465 (%)
Hazard ratio
95% CI
Poor sleepa
3323 (12.5)
13,067 (9.8)
1.36
1.30-1.41
Age at CD diagnosis
≤19 years
298/11,708 (2.5)
1058/57,940 (1.8)
1.41
1.24-1.61
20-39 years
608/5170 (11.8)
2003/25,672 (7.8)
1.54
1.41-1.69
40-59 years
1223/5960 (20.5)
4659/30,252 (15.4)
1.37
1.29-1.47
≥60 years
1194/3749 (31.8)
5347/19,601 (27.3)
1.24
1.16-1.33
Sex
Males
957/9815 (9.8)
3920/49,688 (7.9)
1.32
1.22-1.42
Females
2366/16,772 (14.1)
9147/83,777 (10.9)
1.37
1.31-1.44
Type of hypnotics usedb
Benzodiazepine-related drugs
2782/26,591 (10.5)
10898/133,492 (8.2)
1.35
1.30-1.41
Benzodiazepines
394/26,610 (1.5)
1334/133,547 (1.0)
1.59
1.41-1.78
Other hypnotics
1191/26,604 (4.5)
4847/133,536 (3.6)
1.28
1.20-1.37
Melatonin receptor agonists
148/26,611 (0.6)
445/133,557 (0.3)
1.69
1.40-2.04
Ha ard ratios estimated thro gh co regression modelling Table 5 Risk of poor sleep after celiac disease diagnosis Mårild et al. BMC Gastroenterology (2015) 15:10 Page 7 of 8 excess risk was not influenced by psychiatric comorbidi-
ties. Although, we cannot rule out that surveillance bias
has contributed to our findings, our results are consist-
ent with earlier research suggesting that poor sleep may
be a manifestation of CD. and physical health, but was not associated with the inten-
sity of gastrointestinal symptoms. Our results are consist-
ent with the results of Zingone et al., although the studies
differ by study design. Poor sleep has been recognized as a manifestation of
several gastrointestinal conditions [10,27] including ac-
tive as well as inactive IBD [7]. Previous studies Conversely, poor sleep
may have a detrimental effect on the course of both
functional [28] and inflammatory gastrointestinal condi-
tions, and several studies on IBD have found that sleep
impairment may increase the risk of disease flares
[29,30]. However, even though poor sleep has been
shown to influence ubiquitous pro-inflammatory media-
tors [31] there is no plausible biological explanation for
a causal association between poor sleep and develop-
ment of CD. A non-causal association between CD and
poor sleep is further supported by the lack of a dose-
and time-dependent association in this study. Received: 18 September 2014 Accepted: 20 January 2015 Received: 18 September 2014 Accepted: 20 January 2015 Received: 18 September 2014 Accepted: 20 January 2015 Interpretation of findings The authors declare that they have no competing interests. The authors declare that they have no competing interests. We found an increased risk of poor sleep both before
and after CD diagnosis that was not explained by psychi-
atric comorbidity, sleep apnea or RLS. Our results are
consistent with previous data on CD and other gastro-
intestinal diseases suggesting that poor sleep may be a
manifestation of CD. Individuals with CD suffer from
numerous symptoms that may impair the quality of
sleep [2]. Furthermore, even though most CD manifesta-
tions typically resolve on a gluten-free diet, some symp-
toms may persist, in particularly when dietary adherence
is poor [32]. Regrettably, we lack dietary data in this
study. However, in an earlier validation study off a ran-
dom sample of individuals with villous atrophy, 15/86
(17%) had poor dietary adherence [13]. Authors’ contributions ICMJE criteria for authorship read and met: KM, TIM, VKS, SK, JAM, JFL. Agree
with the manuscript’s results and conclusions: KM, TIM, VKS, SK, JAM, JFL. Designed the experiments/the study: KM, JFL. Collected data: JFL. Analyzed
the data: KM. Wrote the first draft of the paper: KM. Contributed to the
writing of the paper: TIM, VKS, SK, JAM, JFL. Contributed to design of study
and interpretation of the data analyses: TIM, VKS, SK, JAM. Interpretation of
data; approved the final version of the manuscript: KM, TIM, VKS, SK, JAM,
JFL. Responsible for data integrity: KM, JFL. Supervised the project including
data analyses: JFL. Obtained funding: JFL. All authors read and approved the
final manuscript. Acknowledgements This work was supported by: KM: Karolinska Institutet, The Swedish Society of
Medicine; TIM: VKS: The Mayo Foundation; SK: JAM: JFL: The Swedish Society
of Medicine, the Swedish Research Council, the Karolinska Institutet, and the
Swedish Celiac Society. This work was supported by: KM: Karolinska Institutet, The Swedish Society of
Medicine; TIM: VKS: The Mayo Foundation; SK: JAM: JFL: The Swedish Society
of Medicine, the Swedish Research Council, the Karolinska Institutet, and the
Swedish Celiac Society. Finally, we cannot rule out that the association seen in
this study was influenced by surveillance bias with an in-
creased detection of CD in individuals with poor sleep. It is well known that individuals with poor sleep have an
increased health care consumption [33], including in-
creased drug use [9]. In addition, studies have shown a
mean diagnostic delay of more than five years from on-
set of CD symptoms to diagnosis [34], a time associated
with an increased number of outpatients visits [35] and
possibly an increased likelihood of receiving treatment
for poor sleep. Regrettably, we lack data on clinical char-
acteristics of those individuals with CD. However, the
lack of a time-dependent association between poor sleep
and CD diagnosis, argues against surveillance bias as the
sole cause of our findings. Additional files Additional file 2: Anatomical therapeutic chemical codes on drugs
typically used for treatment of restless legs syndrome. Additional file 3: International Classification of Disease (ICD) codes
version 8-10. References
W lk 1. Walker MM, Murray JA, Ronkainen J, Aro P, Storskrubb T, D’Amato M, et al. Detection of celiac disease and lymphocytic enteropathy by parallel
serology and histopathology in a population-based study. Gastroenterology. 2010;139:112–9. 2. Ludvigsson JF, Green PH. Clinical management of coeliac disease. J Intern
Med. 2011;269:560–71. Abbreviations ATC: Anatomical therapeutic chemical; CD: Celiac disease; CI: Confidence
interval; HR: Hazard ratio; IBD: Inflammatory bowel disease; ICD: International
Classification of Disease; OR: Odds ratio; RLS: Restless legs syndrome. Author details
1 1Department of Medical Epidemiology and Biostatistics, Karolinska Institutet,
Stockholm, Sweden. 2Division of Epidemiology, Norwegian Institute of Public
Health, Oslo, Norway. 3Division of Pulmonary and Critical Care Medicine, Center
for Sleep Medicine, Mayo Clinic, Rochester, MN, USA. 4Department of Internal
Medicine, Mayo Clinic, Rochester, MN, USA. 5Departments of Neurology and
Pediatrics, The Center for Sleep Medicine, Mayo Clinic, Rochester, MN, USA. 6Departments of Medicine and Immunology, Mayo Clinic, Rochester, MN, USA. 7Department of Pediatrics, Örebro University Hospital, Örebro, Sweden. 8Astrid
Lindgren Children’s Hospital, Stockholm, Sweden. p
gy,
y
,
,
,
7Department of Pediatrics, Örebro University Hospital, Örebro, Sweden. 8Astrid
Lindgren Children’s Hospital, Stockholm, Sweden. 2.
Ludvigsson JF, Green PH. Clinical management of coeliac disease. J Intern
Med. 2011;269:560–71. 4.
Ohayon MM. Epidemiology of insomnia: what we know and what we still
need to learn. Sleep Med Rev. 2002;6:97–111. Conclusions 3. Moccia M, Pellecchia MT, Erro R, Zingone F, Marelli S, Barone DG, et al. Restless legs syndrome is a common feature of adult celiac disease. Mov
Disord. 2010;25:877–81. 3. Moccia M, Pellecchia MT, Erro R, Zingone F, Marelli S, Barone DG, et al. Restless legs syndrome is a common feature of adult celiac disease. Mov
Disord. 2010;25:877–81. In conclusion, individuals with CD suffer an increased
risk of poor sleep, measured as repeated use of pre-
scribed hypnotics, both before and after diagnosis. The 4. Ohayon MM. Epidemiology of insomnia: what we know and what we still
need to learn. Sleep Med Rev. 2002;6:97–111. Mårild et al. BMC Gastroenterology (2015) 15:10 Page 8 of 8 Page 8 of 8 Page 8 of 8 5. Eder DN, Zou D, Grote L, Hedner J. Self-reported features of sleep, utilization
of medical resources, and socioeconomic position: a Swedish population
survey. Behav Sleep Med. 2011;9:162–72. 31. Opp MR. Cytokines and sleep. Sleep Med Rev. 2005;9:355–64. 32. Leffler DA, Dennis M, Hyett B, Kelly E, Schuppan D, Kelly CP. Etiologies and
predictors of diagnosis in nonresponsive celiac disease. Clin Gastroenterol
Hepatol. 2007;5:445–50. 6. Sateia MJ, Nowell PD. Insomnia. Lancet. 2004;364:1959–73. 33. Hetta J, Broman JE, Mallon L. Evaluation of severe insomnia in the general
population–implications for the management of insomnia: insomnia,
quality of life and healthcare consumption in Sweden. J Psychopharmacol. 1999;13:S35–6. 7. Ranjbaran Z, Keefer L, Farhadi A, Stepanski E, Sedghi S, Kesha 7. Ranjbaran Z, Keefer L, Farhadi A, Stepanski E, Sedghi S, Keshavarzian A. Impact of sleep disturbances in inflammatory bowel disease. J Gastroenterol
Hepatol. 2007;22:1748–53. 8. Seppala M, Hyyppa MT, Impivaara O, Knuts LR, Sourander L. Subjective
quality of sleep and use of hypnotics in an elderly urban population. Aging
(Milano). 1997;9:327–34. eppala M, Hyyppa MT, Impivaara O, Knuts LR, Sourander L. Subjective
f
f 34. Green PHR, Stavropoulos SN, Panagi SG, Goldstein SL, McMahon DJ,
Absan H, et al. Characteristics of adult celiac disease in the USA: results of
a national survey. Am J Gastroenterol. 2001;96:126–31. 9. Leger D, Guilleminault C, Bader G, Levy E, Paillard M. Medical and
socio-professional impact of insomnia. Sleep. 2002;25:625–9. 35. Dickey W, McConnell JB. How many hospital visits does it take before celiac
sprue is diagnosed? J Clin Gastroenterol. 1996;23:21–3. 10. Parish JM. Sleep-related problems in common medical conditions. Chest. 2009;135:563–72. 11. Conclusions Zingone F, Siniscalchi M, Capone P, Tortora R, Andreozzi P, Capone E, et al. The quality of sleep in patients with coeliac disease. Aliment Pharmacol
Ther. 2010;32:1031–6. 12. Wettermark B, Hammar N, Fored CM, Leimanis A, Otterblad Olausson P,
Bergman U, et al. The new Swedish Prescribed Drug Register–opportunities
for pharmacoepidemiological research and experience from the first six
months. Pharmacoepidemiol Drug Saf. 2007;16:726–35. 13. Ludvigsson JF, Brandt L, Montgomery SM, Granath F, Ekbom A. Validation
study of villous atrophy and small intestinal inflammation in Swedish biopsy
registers. BMC Gastroenterol. 2009;9:19. 14. Ludvigsson JF, Montgomery SM, Ekbom A, Brandt L, Granath F. Small-
intestinal histopathology and mortality risk in celiac disease. JAMA. 2009;302:1171–8. 15. Wilén A, Johannesson I. Ett nytt system för Registret över totalbefolkningen:
fler möjligheter och bättre kvalitet [English title: A new Total population
register system : more possiblities and better quality]. Stockholm:
Avdelningen för befolknings- och välfärdsstatistik, Statistiska centralbyrån,
SCB. 2002. 16. WHO Collaborating Centre for Drug Statistics Methodology. ATC: Structure
and principles. 2011 [cited 2012 25th January]; Available from: http://www. whocc.no/atc/structure_and_principles/. 17. Johnell K, Fastbom J. Gender and use of hypnotics or sedatives in old age:
a nationwide register-based study. Int J Clin Pharm. 2011;33:788–93. 18. Kotagal S. Childhood obstructive sleep apnoea. BMJ. 2005;330:978–9. 19. West SD, McBeath HA, Stradling JR. Obstructive sleep apnoea in adults. BMJ. 2009;338:b1165. 20. Allen RP, Picchietti D, Hening WA, Trenkwalder C, Walters AS, Montplaisi J. Restless legs syndrome: diagnostic criteria, special considerations, and
epidemiology. A report from the restless legs syndrome diagnosis and
epidemiology workshop at the National Institutes of Health. Sleep Med. 2003;4:101–19. 21. Weinstock LB, Walters AS, Mullin GE, Duntley SP. Celiac disease is associated
with restless legs syndrome. Dig Dis Sci. 2010;55:1667–73. 22. Rothman KJ, Greenland S, Lash TL. Modern epidemiology. 3rd ed. Philadelphia: Wolters Kluwer Health/Lippincott Williams & Wilkins; 2008. 23. Olen O, Bihagen E, Rasmussen F, Ludvigsson JF. Socioeconomic position
and education in patients with coeliac disease. Dig Liver Dis. 2012;44:471–6. 24. Ludvigsson JF, Zingone F, Tomson T, Ekbom A, Ciacci C. Increased risk of
epilepsy in biopsy-verified celiac disease: a population-based cohort study. Neurology. 2012;78:1401–7. 25. Ludvigsson JF, Reutfors J, Osby U, Ekbom A, Montgomery SM. Coeliac
disease and risk of mood disorders–a general population-based cohort
study. J Affect Disord. 2007;99:117–26. 32.
Leffler DA, Dennis M, Hyett B, Kelly E, Schuppan D, Kelly CP. Etiologies and
predictors of diagnosis in nonresponsive celiac disease. Clin Gastroenterol
Hepatol. 2007;5:445–50. 31.
Opp MR. Cytokines and sleep. Sleep Med Rev. 2005;9:355–64.
32.
Leffler DA, Dennis M, Hyett B, Kelly E, Schuppan D, Kelly CP. Etiologies and
predictors of diagnosis in nonresponsive celiac disease. Clin Gastroenterol
Hepatol. 2007;5:445–50.
33.
Hetta J, Broman JE, Mallon L. Evaluation of severe insomnia in the general
population–implications for the management of insomnia: insomnia,
quality of life and healthcare consumption in Sweden. J Psychopharmacol.
1999;13:S35–6.
34.
Green PHR, Stavropoulos SN, Panagi SG, Goldstein SL, McMahon DJ,
Absan H, et al. Characteristics of adult celiac disease in the USA: results of
a national survey. Am J Gastroenterol. 2001;96:126–31.
35.
Dickey W, McConnell JB. How many hospital visits does it take before celiac
sprue is diagnosed? J Clin Gastroenterol. 1996;23:21–3. 33.
Hetta J, Broman JE, Mallon L. Evaluation of severe insomnia in the general
population–implications for the management of insomnia: insomnia,
quality of life and healthcare consumption in Sweden. J Psychopharmacol.
1999;13:S35–6. Conclusions Submit your next manuscript to BioMed Central
and take full advantage of:
• Convenient online submission
• Thorough peer review
• No space constraints or color figure charges
• Immediate publication on acceptance
• Inclusion in PubMed, CAS, Scopus and Google Scholar
• Research which is freely available for redistribution
Submit your manuscript at
www.biomedcentral.com/submit Submit your next manuscript to BioMed Central
and take full advantage of:
• Convenient online submission
• Thorough peer review
• No space constraints or color figure charges
• Immediate publication on acceptance
• Inclusion in PubMed, CAS, Scopus and Google Scholar
• Research which is freely available for redistribution
Submit your manuscript at
www.biomedcentral.com/submit Submit your next manuscript to BioMed Central
and take full advantage of:
• Convenient online submission
• Thorough peer review
• No space constraints or color figure charges
• Immediate publication on acceptance
• Inclusion in PubMed, CAS, Scopus and Google Scholar
• Research which is freely available for redistribution
Submit your manuscript at
www.biomedcentral.com/submit Submit your next manuscript to BioMed Central
and take full advantage of: Submit your next manuscript to BioMed Central
and take full advantage of: 26. Ludvigsson JF, Andersson E, Ekbom A, Feychting M, Kim JL, Reuterwall C,
et al. External review and validation of the Swedish national inpatient
register. BMC Public Health. 2011;11:450. 27. Orr WC, Chen CL. Sleep and the gastrointestinal tract. Neurol Clin. 2005;23:1007–24. • Convenient online submission 28. Jarrett M, Heitkemper M, Cain KC, Burr RL, Hertig V. Sleep disturbance
influences gastrointestinal symptoms in women with irritable bowel
syndrome. Dig Dis Sci. 2000;45:952–9. 29. Ananthakrishnan AN, Long MD, Martin CF, Sandler RS, Kappelman MD. Sleep disturbance and risk of active disease in patients with Crohn’s
disease and ulcerative colitis. Clin Gastroenterol Hepatol. 2013; 11:965-971 30. Tang Y, Preuss F, Turek FW, Jakate S, Keshavarzian A. Sleep deprivation
worsens inflammation and delays recovery in a mouse model of colitis. Sleep Med. 2009;10:597–603. 30. Tang Y, Preuss F, Turek FW, Jakate S, Keshavarzian A. Sleep deprivation
worsens inflammation and delays recovery in a mouse model of colitis. Sleep Med. 2009;10:597–603.
|
https://openalex.org/W2232281423
|
https://europepmc.org/articles/pmc4702368?pdf=render
|
English
| null |
Comparison of three boosting methods in parent-offspring trios for genotype imputation using simulation study
|
Han'gug dongmul jawon gwahag hoeji/Han-guk dongmul jawon gwahak hoeji/Journal of animal science and technology
| 2,016
|
cc-by
| 6,211
|
© 2016 Mikhchi et al. Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0
International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and
reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to
the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver
(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Comparison of three boosting methods
in parent-offspring trios for genotype
imputation using simulation study Abbas Mikhchi1, Mahmood Honarvar2, Nasser Emam Jomeh Kashan1*, Saeed Zerehdaran3 and Mehdi Aminafshar1 Mikhchi et al. Journal of Animal Science and Technology (2016) 58:1
DOI 10.1186/s40781-015-0081-1 Mikhchi et al. Journal of Animal Science and Technology (2016) 58:1
DOI 10.1186/s40781-015-0081-1 * Correspondence: Nasser_ejk@yahoo.com
1Department of Animal Science, Science and Research Branch, Islamic Azad
University, Tehran, Iran
Full list of author information is available at the end of the article Abstract Background: Genotype imputation is an important process of predicting unknown genotypes, which uses
reference population with dense genotypes to predict missing genotypes for both human and animal genetic
variations at a low cost. Machine learning methods specially boosting methods have been used in genetic studies
to explore the underlying genetic profile of disease and build models capable of predicting missing values of a
marker. Methods: In this study strategies and factors affecting the imputation accuracy of parent-offspring trios compared
from lower-density SNP panels (5 K) to high density (10 K) SNP panel using three different Boosting methods
namely TotalBoost (TB), LogitBoost (LB) and AdaBoost (AB). The methods employed using simulated data to impute
the un-typed SNPs in parent-offspring trios. Four different datasets of G1 (100 trios with 5 k SNPs), G2 (100 trios
with 10 k SNPs), G3 (500 trios with 5 k SNPs), and G4 (500 trio with 10 k SNPs) were simulated. In four datasets all
parents were genotyped completely, and offspring genotyped with a lower density panel. Results: Comparison of the three methods for imputation showed that the LB outperformed AB and TB for imputation
accuracy. The time of computation were different between methods. The AB was the fastest algorithm. The higher SNP
densities resulted the increase of the accuracy of imputation. Larger trios (i.e. 500) was better for performance of LB and TB. Conclusions: The conclusion is that the three methods do well in terms of imputation accuracy also the dense chip is
recommended for imputation of parent-offspring trios. Keyword: Trios, Boosting methods, Imputation accuracy, Computation time Background Machine learn-
ing methods have been used in genetic studies to explore
the underlying genetic profile of disease and build models
capable of predicting missing values of a marker [9, 10]. Boosting is one of Machine learning methods for improv-
ing the predictive performance of classification or regres-
sion procedures which attempts to boost the accuracy of
any given learning algorithm by applying it several times
on slightly modified training data and then combining the
results in a suitable manner [11]. Several methods of
estimation have preceded boosting approach [12]. Com-
mon feature for all methods is that they work out by
extracting samples of a set, calculating the estimate for
each drawn sample group repeatedly and combining the
calculated results into unique one. One of the ways, the
simplest one, to manage estimation is to examine the
statistics of selected available samples from the set and
combine the results of calculation together by averaging
them [11, 12]. The main variation between many Boosting
Algorithms are the method of weighting training data
points and hypotheses. Gradient boosting is typically used
with decision trees of a fixed size as base learners [12]. In
this research the accuracies of three different boosting
methods i.e. (TotalBoost, LogitBoost, and AdaBoost) for
imputation of un-typed-SNPs of parent-offspring trios are
compared. The methods were compared in terms of
imputation accuracy, computation time and factors affect-
ing imputation accuracy. To evaluate the factors affecting
imputation accuracy, sample size and SNP density were
also examined. length. Different marker densities were created for each
simulated data set. The number of SNPs per chromosome
ranged from 1000 to 2000 in various datasets. The Refer-
ence population generated from the HP by mating parent
groups. The parent groups were randomly selected from
the last generation of the HP. Fifty percent of male
offspring selected randomly from each group and were
used as sires for the next generation. Also fifty percent of
female offspring selected randomly as dams to produce
the next generation and the mating scheme continued for
50 generations. The founder population randomly selected
and the haplotypes of offspring generated them. Samples
of 100 parent-offspring trios produced. Each sample was
sequenced at depth of 5 k and 10 k. The sample size of
the second set of simulations consisted 500 trios. Imputation accuracy and running time For each of the methods, the imputation accuracy per un-
typed SNPs were calculated as the correlation between
imputed and observed SNPs, then mean of imputation
accuracy were calculated across the 5 replicates. Compu-
tation time were measured based on running each pro-
gram in second on a windows server with 32 core CPU
Intel, GPU: 192 CUDA Core and a total of 64 GB RAM
by Profiler function in MATLAB. Methods The SNP Density and sample size were considered as
factors that could impact the imputation accuracy. For
each dataset-imputation method combination, imputation
accuracy were averaged across dataset versions NA10,
NA30, NA50, NA70 and NA90 and referred as imputation
accuracy. To assess the effect of the sample size on
imputation accuracy, two groups of 100 and 500 parent-
offspring trios were included the variation in SNP density. For both groups embedded simulated SNPs with two
levels of 5 k and 10 K SNP panels and compared imput-
ation accuracy based on trios sample size. The impact of Background genotypes, which uses reference population with dense
genotypes to predict missing genotypes for both human
and animal genetic variations at a low cost [2, 3]. Geno-
type imputation allows us to accurately evaluate the
evidence for association at genetic markers that are not
directly genotyped [4]. Analysis of un-typed SNPs can
facilitate the localization of disease-causing variants and
permit meta-analysis of association studies with different
genotyping platforms [5]. As un-typed SNPs are not
measured on any study subject, the missing information
cannot be recovered from the study data alone [5]. To
bring down genotyping costs, a reference population can
be genotyped with a high-density panel while other ani-
mals are genotyped with a low-density panel in which
markers are evenly spaced. Then, using information from
the reference population, genotypes for un-typed loci can Background
Innovations in genomic technologies provide new tools
for enhancing productivity and wellbeing of domestic
animals [1]. The technology can genotype some 10
million SNPs in an individual [2]. The availability of
some thousands of SNPs spread across the genome of
different livestock species opens up possibilities to include
genome-wide marker information in prediction of total
breeding values, to perform genomic selection [2]. Also a
major challenge in implementing genomic selection in
most species is the cost of genotyping [2]. Genotype
imputation is an important process of predicting unknown * Correspondence: Nasser_ejk@yahoo.com
1Department of Animal Science, Science and Research Branch, Islamic Azad
University, Tehran, Iran
Full list of author information is available at the end of the article Page 2 of 6 Page 2 of 6 Mikhchi et al. Journal of Animal Science and Technology (2016) 58:1 be inferred for individuals genotyped with the low-
density panel [6]. Phasing and imputation methods
can be divided into family-based methods (which use
linkage information from close relatives) and population-
based methods, which use population linkage disequilib-
rium information [6]. A “trio” data consist of genotypes
from father-mother-child triplets and some phasing algo-
rithms are adapted to be used in this type of data [7]. The
accuracy of imputation depends on several factors, such
as the number of SNPs in the low density panel, the rela-
tionship between the animals genotyped, the effective
population size, and the method used [8]. Background One of the ways, the
simplest one, to manage estimation is to examine the
statistics of selected available samples from the set and
combine the results of calculation together by averaging
them [11, 12]. The main variation between many Boosting
Algorithms are the method of weighting training data
points and hypotheses. Gradient boosting is typically used
with decision trees of a fixed size as base learners [12]. In
this research the accuracies of three different boosting
methods i.e. (TotalBoost, LogitBoost, and AdaBoost) for
imputation of un-typed-SNPs of parent-offspring trios are
compared. The methods were compared in terms of
imputation accuracy, computation time and factors affect-
ing imputation accuracy. To evaluate the factors affecting
imputation accuracy, sample size and SNP density were
also examined. Background Four
different datasets of 100 trios with 5 k SNPs (G1), 100
trios with 10 k SNPs (G2), 500 trios with 5 k SNPs (G3),
and 500 trio with 10 k SNPs (G4) were simulated. Bi-
allelic SNPs were defined on each of homologous chromo-
somes and used “0” and “1” to denote the two alleles at
each SNP site. The allele with high frequency was defined
as ‘0’, and allele with low frequency as ‘1’ and an unknown
value as ‘NaN’. Both parents genotyped for all SNPs, and
offspring were genotyped for some of SNPs (low-density)
(Fig. 1). For each of G1-G4 datasets five versions: NA10,
NA30, NA50, NA70 and NA 90 were created with differ-
ent levels of simulated missing data (10, 30, 50, 70 and
90 % of offspring genotypes). A total of 30 replicates of
each simulated dataset were created. be inferred for individuals genotyped with the low-
density panel [6]. Phasing and imputation methods
can be divided into family-based methods (which use
linkage information from close relatives) and population-
based methods, which use population linkage disequilib-
rium information [6]. A “trio” data consist of genotypes
from father-mother-child triplets and some phasing algo-
rithms are adapted to be used in this type of data [7]. The
accuracy of imputation depends on several factors, such
as the number of SNPs in the low density panel, the rela-
tionship between the animals genotyped, the effective
population size, and the method used [8]. Machine learn-
ing methods have been used in genetic studies to explore
the underlying genetic profile of disease and build models
capable of predicting missing values of a marker [9, 10]. Boosting is one of Machine learning methods for improv-
ing the predictive performance of classification or regres-
sion procedures which attempts to boost the accuracy of
any given learning algorithm by applying it several times
on slightly modified training data and then combining the
results in a suitable manner [11]. Several methods of
estimation have preceded boosting approach [12]. Com-
mon feature for all methods is that they work out by
extracting samples of a set, calculating the estimate for
each drawn sample group repeatedly and combining the
calculated results into unique one. Boosting methods
AdaBoost The AdaBoost algorithm [16] is a well-known method to
build ensembles of classifiers with very good perform-
ance [16]. It has been shown empirically that AdaBoost
with decision trees has excellent performance, being
considered the best off-the-shelf classification algorithm
[16]. This algorithm takes training data and defines weak
classifier functions for each sample of training data. Clas-
sifier function takes the sample as argument and produces
value 0 or 1 in case of a binary classification task and a
constant value - weight factor for each classifier. Gener-
ally, AdaBoost has shown good performance at classifica-
tion. The sensitivity to noisy data and outliers is a weak
feature of AdaBoost. Let X be a set of imputed SNPs, and
y be a vector of observed (‘true’) SNP at an individual. Define M = 100 to be the number of independent classi-
fiers (i.e. the imputation software). Given a training set of
N SNP, there are Z = [(x1, y1), …,(xi, yi), …,(xN, yN)], where
xi ∈X = (xi1, xi2, xi3|i = 1,2, …, N), yi ∈y = (a1, a2), and a1,
a2 are the two alleles at a SNP locus, in question, for SNP
i in the training sample. each of these factors were assessed for each imputation
method. each of these factors were assessed for each imputation
method. The data simulation Four Data sets at different marker densities were sim-
ulated using the statistical software package R [13]. The R package hypred [14] was modified to simulate
of data sets. A Historic Population (HP) was simulated
that half of the animals were female and the other half
male. Mating was performed during 50 generations using
mutation rate of 2.5*10−8 per site by drawing the parents
of an animal randomly from the animals of the previous
generation. The considered genome comprised five chro-
mosomes and each chromosome was set as 1 Morgan Page 3 of 6 Mikhchi et al. Journal of Animal Science and Technology (2016) 58:1 Page 3 of 6 Page 3 of 6 each of these factors were assessed for each imputation
method. Fig. 1 Genotype imputation within a trio the genotype datasets included a matrix P with m individ-
uals and n SNP loci where the P (i, j) indicates the geno-
type of individuals at locus i. The target missing value is
defined as P (i, j) = NaN. The individuals were assumed to
have a known value at locus i, or otherwise it was ex-
cluded from the imputation but to be imputed in exactly
the same way as sample j. On the other hand every other
individuals had a known value at locus i, otherwise it was
excluded from the imputation but to be imputed in
exactly the same way as individuals j. In the imput-
ation methods only parent genotype values at nearby SNP
loci were used in the inference of P (i, j) in offspring. Imputation methods
SNP window Initialize: each SNP was assigned with an equal
weight and All the imputations in this study were done using
MATLAB version (R2014a) [15]. The SNP window is
defined by a fixed number of SNPs to the left and right
(L + R) of the un-typed SNP (except when the un-typed
SNP was near the end of a chromosome). A SNP window
of size L corresponds to L/2 SNPs to the left and L/2 SNPs
to the right of the un-typed SNP. In all imputation
methods, a SNP window of size L centered at marker i to
extend L markers left and right. For SNPs less than L
markers from the beginning or end of a chromosome, the
window extends L SNPs in one direction and to the
boundary of the chromosome in the other. The distance
defined in terms of the index of the SNP or the physical
position on the chromosome, or the genetic distance. A
distance measure fitted to the observed correlation matrix
between markers and selected the best window size of 22
(for 1 k) and 10 (for 5 k and 10 k) for the imputation by
scanning over a large range of windows. For all methods, wi ¼ 1=N; i ∈1; …; N
f
g Result and discussion
Imputation accuracies The Boosting Algorithms in this study were AdaBoost,
LogitBoost and TotalBoost which used the decision trees
Result and discussion
Imputation accuracies
The imputation accuracies in different datasets are
shown in Table 1 for ADA, LB and TOT. The accuracy
of Imputation was high for all Boosting methods. For all
data sets, imputation accuracies always decreased as the
level of missing data increased. In general TOT had the
lowest imputation accuracy compared to other Boosting
methods. The results indicate that LB had the highest
accuracy. A possible reason that TotalBoost was less
accurate than other methods is that the datasets that
used in the experiment may have violated multivariate
Table 1 Mean of imputation accuracy for Boosting methods in
various versions on the four different datasets
Data set
Density
Sample size
Version
AB
LB
TB
5 k
100
NA10
0.9843
0.9954
0.9611
5 k
100
NA30
0.9883
0.9947
0.9638
G1
5 k
100
NA50
0.9822
0.9909
0.9621
5 k
100
NA70
0.9777
0.9829
0.9583
5 k
100
NA90
0.9211
0.9303
0.9246
Mean
0.9707
0.9788
0.9539
10 k
100
NA10
0.9861
0.9981
0.9702
10 k
100
NA30
0.9886
0.9978
0.9697
G2
10 k
100
NA50
0.9912
0.9970
0.9679
10 k
100
NA70
0.9898
0.9939
0.9647
10 k
100
NA90
0.9653
0.9714
0.9523
Mean
0.9842
0.9916
0.9649
5 k
500
NA10
0.9859
0.9967
0.9650
5 k
500
NA30
0.9885
0.9952
0.9650
G3
5 k
500
NA50
0.9877
0.9926
0.9638
5 k
500
NA70
0.9800
0.9848
0.9618
5 k
500
NA90
0.9288
0.9383
0.9362
Mean
0.9741
0.9815
0.9583
10 k
500
NA10
0.9787
0.9983
0.9706
10 k
500
NA30
0.9799
0.9977
0.9692
G4
10 k
500
NA50
0.9830
0.9967
0.9665
10 k
500
NA70
0.9877
0.9959
0.9634
10 k
500
NA90
0.9706
0.9767
0.9552
Mean
0.9799
0.9930
0.9649
NA10: 10 % of genotype is missing per offspring, NA30: 30 % of genotype
is missing per offspring, NA50: 50 % of genotype is missing per offspring,
NA70: 70 % of genotype is missing per offspring, NA90: 90 % of genotype is
missing per offspring, Bold: Mean of different versions in each dataset
AB AdaBoost, LB LogitBoost, TB TotalBoost The imputation accuracies in different datasets are
shown in Table 1 for ADA, LB and TOT. The accuracy
of Imputation was high for all Boosting methods. For all
data sets, imputation accuracies always decreased as the
level of missing data increased. wi ¼ 1=N; i ∈1; …; N
f
g Training: For m = 1, 2… M classifiers Call classifier m, which in turn generates hypothesis
PW (i.e. inferred SNPs in the training set). Calculate the
error of PW: Fit the class probability estimate Pm(x) = Pw(y = 1|x) ∈[0, 1], using weight wi on the
training data. Pm(x) = Pw(y = 1|x) ∈[0, 1], using weight wi on the
training data. Set Hm ¼ 0:5 log
1−Pm x
ð Þ
Pm x
ð Þ
∈R
Update the weight distribution Wi for next classifier as
Set wi ←wi exp(−wiHm(xi)) and renormalize to ∑iwi = 1 Update the weight distribution Wi for next classifier as
Set wi ←wi exp(−wiHm(xi)) and renormalize to ∑iwi = 1 Testing: In the testing set, each Un-typed SNP is
classified via the so-called ‘weighted majority voting’. Briefly, the wrapper program is Page 4 of 6 Mikhchi et al. Journal of Animal Science and Technology (2016) 58:1 as learner [12, 17]. The main tuning parameter, the
optimal number of iterations (or trees), determined and
then the fitensemble function of MATLAB selected and
set the number of decision trees to 100 for all boosting
methods. Output H x
ð Þ ¼ sign
Xm
mHm x
ð Þ
Above, the algorithm maintains a weighted distribution
Wi of training samples xi, for i = 1, …,N, from which a
sequence of training data subsets Zm is chosen for each
consecutive classifier (package) m. Initially, the distribu-
tion of weights is uniform, meaning that all samples
contribute equally to the error rate. Next, the logit Hm of
the rate of correctly classified samples is calculated for
classifier m. A higher Hm is an indicator of better per-
formance. For instance, when Hm = 0.5, Hm takes the
value 0, and increases as Hm →0 [16]. Result and discussion
Imputation accuracies In general TOT had the
lowest imputation accuracy compared to other Boosting
methods. The results indicate that LB had the highest
accuracy. A possible reason that TotalBoost was less
accurate than other methods is that the datasets that
used in the experiment may have violated multivariate LogitBoost LogitBoost is a boosting algorithm that introduces a statis-
tical interpretation to AdaBoost algorithm by using addi-
tive logistic regression model for determining classifier in
each round [12]. Logistic regression is a way of describing
the relationship between one or more factors, in this case
instances from samples of training data, and an outcome,
expressed as a probability. In case of two classes, outcome
can take values 0 or 1. Probability of an outcome being 1
is expressed with logistic function. LogitBoost is a method
to minimize the logistic loss, AdaBoost technique driven
by probabilities optimization. This method requires care
to avoid numerical problems [12]. Table 1 Mean of imputation accuracy for Boosting methods in
i
i
h f
diff
d Table 1 Mean of imputation accuracy for Boosting methods in
various versions on the four different datasets
Data set
Density
Sample size
Version
AB
LB
TB logitBoost algorithm for classification 1. Initialize the weights wi = 1/N, i ∈{1, …, N} 2. For m = 1 to M and while Hm ≠0 Result and discussion
Imputation accuracies onsecutive classifier (package) m. Initially, the distribu
ion of weights is uniform, meaning that all samples
ontribute equally to the error rate. Next, the logit Hm of
he rate of correctly classified samples is calculated for
lassifier m. A higher Hm is an indicator of better per-
ormance. For instance, when Hm = 0.5, Hm takes the
alue 0, and increases as Hm →0 [16]. LogitBoost
LogitBoost is a boosting algorithm that introduces a statis-
ical interpretation to AdaBoost algorithm by using addi-
ive logistic regression model for determining classifier in
ach round [12]. Logistic regression is a way of describing
he relationship between one or more factors, in this case
nstances from samples of training data, and an outcome,
xpressed as a probability. In case of two classes, outcome
an take values 0 or 1. Probability of an outcome being 1
s expressed with logistic function. LogitBoost is a method
o minimize the logistic loss, AdaBoost technique driven
by probabilities optimization. This method requires care
o avoid numerical problems [12]. ogitBoost algorithm for classification
1. Initialize the weights wi = 1/N, i ∈{1, …, N}
2. For m = 1 to M and while Hm ≠0
a) Compute the working response zi = yi −P(xi)/
P(xi)(1 −P(xi)) and weights wi ¼ P xi
ð Þ 1−P xi
ð Þ
ð
Þ
b) Fit Hm(x) by weighted least – squares of zi to yi
with weights wi
c) Set H(x) = H(x) + 0.5 Hm(x) and P(X) =
exp H x
ð Þ
ð
Þ
exp H x
ð Þ
ð
Þþ exp −H x
ð Þ
ð
Þ
3. Output H(x) = sign (∑m
mHm(x))
TotalBoost
General idea of Boosting algorithms, maintaining the dis-
ribution over a given set of examples, has been optimized. A way to accomplish optimization for TotalBoost is to
modify the way measuring the hypothesis goodness, (edge)
s being constrained through iterations. AdaBoost con-
trains the edge with the respect to the last hypothesis to
maximum zero. TotalBoost method is “totally corrective”,
onstraining the edges of all previous hypotheses to max-
mal value that is properly adapted. It is proven that, with
daptive edge maximal value, measurement of confidence
n prediction for a hypothesis weighting increases [12]. Sample size p
The accuracy of imputation increases for all methods under
the condition of low SNP density (5 k), as the number of
trios increase. The results show that under the condition of
high SNP density (10 k), accuracy of imputation increased
for LB and TB as the number of trios increased. The imput-
ation accuracy for AdaBoost (AB) in 10 K SNP panel was
slightly lower. It seems that AdaBoost is suitable for imput-
ation of un-typed SNP in small sample size. However, the
effect of the sample size on imputation accuracy is less than
effect of SNP density on imputation accuracy. The results
show that the sample size of the trios is a substantial impact
on imputation accuracy. We have demonstrated with G3
and G4 datasets that the use of 500 trios produced substan-
tial gain in imputation accuracy and improved imputation
accuracy for LB and TB. The larger sample size will pro-
duce more consistent estimates of measured parameters,
resulting in improved imputation accuracy for various
methods [21]. The performance of any classification de-
pends on sample size, which may be especially so for
present methods, since the number of parameters to be es-
timated is large and low sample size may lead to unstable
results [22]. It was found that larger trios (i.e. 500) could
help to better performance of LB and TB and could be suit-
able for imputation of un-typed SNPs [23]. The LB and TB
showed the large changes with increasing the number of
trios. It is concluded that these methods are suitable for im-
putation of un-typed SNP in large sample. TotalBoost General idea of Boosting algorithms, maintaining the dis-
tribution over a given set of examples, has been optimized. A way to accomplish optimization for TotalBoost is to
modify the way measuring the hypothesis goodness, (edge)
is being constrained through iterations. AdaBoost con-
strains the edge with the respect to the last hypothesis to
maximum zero. TotalBoost method is “totally corrective”,
constraining the edges of all previous hypotheses to max-
imal value that is properly adapted. It is proven that, with
adaptive edge maximal value, measurement of confidence
in prediction for a hypothesis weighting increases [12]. The Boosting Algorithms in this study were AdaBoost,
LogitBoost and TotalBoost which used the decision trees Mikhchi et al. Journal of Animal Science and Technology (2016) 58:1 Page 5 of 6 Page 5 of 6 methods had better performance on the high density
dataset (10 k). We believe this is reasonable since a higher
density provides more neighboring SNPs, and conse-
quently greater linkage disequilibrium, for imputation
purpose [20]. normality. In addition, increasing the total number of trees
can improve boosting ability to impute the un-typed SNP. Nevertheless other reason that affect the decrease of accur-
acy may be due to total number of trees that we used in the
experiment. It was found that LogitBoost had higher accur-
acy than AdaBoost and TotalBoost algorithms because of
LogitBoost was less sensitive to outliers and unlike Ada-
Boost, which uses an exponential function, LogitBoost uses
the binomial log likelihood, which increases linearly rather
than exponentially for strong negative margins. Because of
this, LogitBoost is more robust than AdaBoost when data
are noisy or samples are mislabelled [11]. However, Logit-
Boost can give better performance than AdaBoost and
TotalBoost to impute the un-typed SNP. The imputation
accuracy obtained of this research is not comparable with
the other studies. Because in each study different popula-
tion structure, levels of missing data and levels of LD be-
tween markers are assumed [18]. SNP density The accuracy of imputation increased with the number
of SNPs for all Boosting methods examined. The imput-
ation accuracy was lower for all levels of 5 K SNP panel
compared to 10 k panels. Increasing the SNP density
increased imputation accuracy for two sample size of
trio (100 and 500), especially from 5 k to 10 k. There
was a large increase in the imputation accuracy when
using 10 k SNP panels. As a general trend, mean of
imputation accuracy increased with increasing SNPs
density and increasing sizes of trios (Fig. 2). It seems
that imputation accuracy in all methods more influ-
enced by the SNP density than sample size. Similar to
the current results, Weigel et al. [19] reported mean
imputation accuracy from 80 to 95 % when animals were
genotyped with a medium-density panel (2000–4000
SNPs); less than 80 % when animals were genotyped for
1000 SNPs or less, and greater than 95 % when animals
were genotyped for more than 8000 SNPs. All Boosting Author details
1 1Department of Animal Science, Science and Research Branch, Islamic Azad
University, Tehran, Iran. 2Department of Animal Science, Shahr-e-Qods
Branch, Islamic Azad University, Tehran, Iran. 3Department of Animal Science,
Ferdowsi University of Mashhad, Mashhad, Iran. 1Department of Animal Science, Science and Research Branch, Islamic Azad
University, Tehran, Iran. 2Department of Animal Science, Shahr-e-Qods Branch, Islamic Azad University, Tehran, Iran. 3Department of Animal Science,
Ferdowsi University of Mashhad, Mashhad, Iran. Received: 18 April 2015 Accepted: 28 December 2015 References Genomic selection using low density marker panels with application to a
sire line in pigs. Genet Sel Evol. 2013;45:28. 5. Hu Y, Lin D. Analysis of untyped SNPs: maximum likelihood and imputation
methods. Genet Epidemiol. 2010;34(8):803–15. 6. Sargolzaei M, Jansen GB, Schenkel FS. A new approach for efficient genotype
imputation using information from relatives. BMC Genomics. 2014;15:478. 7. Lu AT, Cantor RM. Identifying rare-variant associations in parent-child trios
using a Gaussian support vector machine. BMC Proc. 2014;8 Suppl 1:S98. 8. Wellmann R, Preuß S, Tholen E, Heinkel J, Wimmers K , Bennewitz J. Genomic selection using low density marker panels with application to a
sire line in pigs. Genet Sel Evol. 2013;45:28. 6. Sargolzaei M, Jansen GB, Schenkel FS. A new approach for efficient genotype
imputation using information from relatives. BMC Genomics. 2014;15:478. 7. Lu AT, Cantor RM. Identifying rare-variant associations in parent-child trios
using a Gaussian support vector machine. BMC Proc. 2014;8 Suppl 1:S98. 8. Wellmann R, Preuß S, Tholen E, Heinkel J, Wimmers K , Bennewitz J. Genomic selection using low density marker panels with application to a
sire line in pigs. Genet Sel Evol. 2013;45:28. 9. Wang Y, Cai Z, Stothard P, Moore S, Goebel R, Wang L, Lin G. Fast accurate
missing SNP genotype local imputation. BMC Res Notes. 2012;5:404. 10. Goddard R, Eccles D, Ennis S, Rafiq S, Tapper W, Fliege J, Collins A. Support
vector machine classifier for estrogen receptor positive and negative early-
onset breast cancer. PLoS One. 2013;8(7), e68606. 11. Dettling M, Bühlmann P. Boosting for tumor classification with gene
expression data. Bioinformatics. 2003;9:1061–9. 12. Sateesh B. Boosting techniques on rarity mining. IJARCSSE. 2012;2:1 Authors’ contributions
All
h
d haplotyping in parent-offspring trios. Genome Res. 2013;23:142–5 All authors participated in its design and analyzed the result. All authors
helped to draft the manuscript, and all authors read and approved the final
manuscript. Computation of time Computation of time
The detailed runtime of the all three methods on four data-
sets at missing rate of 90 % (NA90) presented in Table 2. Fig. 2 The effect of the sample size and SNP de Fig. 2 The effect of the sample size and SNP density on imputation accuracy Fig. 2 The effect of the sample size and SNP density on imputation accuracy Fig. 2 The effect of the sample size and SNP density on imputation accuracy Mikhchi et al. Journal of Animal Science and Technology (2016) 58:1 Page 6 of 6 Page 6 of 6 Page 6 of 6 Table 2 Average imputation runtime on four datasets (seconds)
Data set
Sample size
Density
Version
AB
LB
TB
G1
100
5 K
NA90
2930
3055
6975
G2
100
10 K
NA90
6511
6788
13956
G3
500
5 K
NA90
3460
3665
10221
G4
500
10 K
NA90
7601
7802
23521
NA90: 90 % of genotype is missing per offspring
AB AdaBoost, LB LogitBoost, TB TotalBoost Competing interests Competing interests
The authors declare that they have no competing interests. 22. Sun J, Zhao H. The application of sparse estimation of covariance matrix to
quadratic discriminant analysis. BMC Bioinformatics. 2015;16:48. 23. Chen W, Zhang JG, Li J, Pei YF, Deng HW. Genotype calling and
haplotyping in parent-offspring trios. Genome Res. 2013;23:142–51. Conclusion
h
d 2010;93:2229–38. 20. Van Raden PM, Null DJ, Sargolzaei M, Wiggans GR, Tooker ME, Cole JB, et al. Genomic imputation and evaluation using high-density Holstein genotypes. J Dairy Sci. 2013;96:668–78. 20. Van Raden PM, Null DJ, Sargolzaei M, Wiggans GR, Tooker ME, Cole JB, et al. Genomic imputation and evaluation using high-density Holstein genotypes. J Dairy Sci. 2013;96:668–78. 21. Okser S, Pahikkala T, Airola A, Salakoski T, Ripatti S, Aittokallio T. Regularized
machine learning in the genetic prediction of complex traits. PLoS Genet. 2014;10(11), e1004754. Conclusion
h
d 13. R Development Core Team. R: a language and environment for statistical
computing, Vienna. 2014, Available at: http://www.r-project.org/. In this study we compared the performance of three
Boosting methods based imputation of parent-offspring
trios in terms of imputation accuracy, computation time
and factors affecting imputation accuracy. Simulation of
datasets showed the methods performed well for imput-
ation of un-typed SNPs. The LB had the highest accuracy
of the three imputation methods examined. Accuracy of
imputation increased with the increase of the number of
SNPs and trios. The 10 K SNP panels can be imputed with
high accuracies than 5 k SNP panels. In terms of imput-
ation time, AB outperformed LB and TB. The LB and TB
methods are suitable for imputation of un-typed SNP in
large samples. The results indicated that the methods are
suitable in terms of imputation accuracy and denser chip
is recommended for imputation of parent-offspring trios. 14. Technow AF. hypred: simulation of genomic data in applied genetics. R
package version 0.5. 2015, Available at: http://CRAN.R-project.org/src/
contrib/Archive/hypred/. 14. Technow AF. hypred: simulation of genomic data in applied genetics. R
package version 0.5. 2015, Available at: http://CRAN.R-project.org/src/
contrib/Archive/hypred/. 15. MATLAB; 2014. http://www.mathworks.com. 16. Hastie T, Tibshirani R, Friedman J. The elements of statistical learning, Data
Mining, Inference, and Prediction. Stanford, California.2nd ed. Springer. 2001. 16. Hastie T, Tibshirani R, Friedman J. The elements of statistical learning, Data
Mining, Inference, and Prediction. Stanford, California.2nd ed. Springer. 2001 17. Ogutu JO, Piepho HP, Streeck TS. A comparison of random forests,
boosting and support vector machines for genomic selection. BMC
Proc. 2011;5 Suppl 3:S11. 17. Ogutu JO, Piepho HP, Streeck TS. A comparison of random forests,
boosting and support vector machines for genomic selection. BMC
Proc. 2011;5 Suppl 3:S11. 18. Rutkoski JE, Poland J, Jannink J, Sorrells ME. Imputation of unordered markers
and the impact on genomic selection accuracy. G3 (Bethesda). 2013;3:427–39. 18. Rutkoski JE, Poland J, Jannink J, Sorrells ME. Imputation of unordered markers
and the impact on genomic selection accuracy. G3 (Bethesda). 2013;3:427–39. 19. Weigel KA, Van Tassell CP, O’Connell JR, VanRaden PM, Wiggans GR. Prediction of unobserved single nucleotide polymorphism genotypes of
Jersey cattle using reference panels and population-based imputation
algorithms. J Dairy Sci. 2010;93:2229–38. 19. Weigel KA, Van Tassell CP, O’Connell JR, VanRaden PM, Wiggans GR. Prediction of unobserved single nucleotide polymorphism genotypes of
Jersey cattle using reference panels and population-based imputation
algorithms. J Dairy Sci. 22.
Sun J, Zhao H. The application of sparse estimation of covariance matrix to
quadratic discriminant analysis. BMC Bioinformatics. 2015;16:48. 23.
Chen W, Zhang JG, Li J, Pei YF, Deng HW. Genotype calling and
haplotyping in parent-offspring trios. Genome Res. 2013;23:142–51. References 1. Meuwissen TH, Hayes BJ, Goddard ME. Prediction of total genetic value
using genome-wide dense marker maps. Genetics. 2001;157(4):1819–29. 2. Boichard D, Chung H, Dassonneville R, David X, Eggen A, Fritz S, et al. Design of a bovine low-density SNP array optimized for imputation. PLoS
One. 2012;7(3), e34130. 2. Boichard D, Chung H, Dassonneville R, David X, Eggen A, Fritz S, et al. Design of a bovine low-density SNP array optimized for imputation. PLoS
One. 2012;7(3), e34130. For all data sets, the AB was the fastest algorithm and LB
was next fastest. The TB was always the slowest and needed
more time to impute a dataset. An important factor in
evaluating machine learning algorithms is how quickly
their runtime increases with sample size of dataset. As
number of trios grow, the speed of all eight methods
needed some more time to impute a dataset, especially for
large SNP panel. AdaBoost required less computer time
than the other boosting methods, which may be an advan-
tage among boosting methods when using large data sets
with several thousand markers. The TotalBoost algorithm
seemed to be too time-consuming in large data sets and it
has lowest imputation accuracy than other methods. The
computing time changed with increasing the sample sizes. Increase of sample size from 100 to 500 resulted, the com-
puting time of all methods increased. 3. Chen J, Zhang J-G, Li J, Pei Y-F, Deng H-W. On combining reference data to
improve imputation accuracy. PLoS One. 2013;8(1), e55600. 3. Chen J, Zhang J-G, Li J, Pei Y-F, Deng H-W. On combining reference data to
improve imputation accuracy. PLoS One. 2013;8(1), e55600. 4. Li Y, Willer C, Sanna S, Abecasis G. Genotype imputation. Annu Rev Genomics
Hum Genet. 2010;10:387–406. 4. Li Y, Willer C, Sanna S, Abecasis G. Genotype imputation. Annu Rev Genomics
Hum Genet. 2010;10:387–406. 5. Hu Y, Lin D. Analysis of untyped SNPs: maximum likelihood and imputation
methods. Genet Epidemiol. 2010;34(8):803–15. 5. Hu Y, Lin D. Analysis of untyped SNPs: maximum likelihood and imputation
methods. Genet Epidemiol. 2010;34(8):803–15. 6. Sargolzaei M, Jansen GB, Schenkel FS. A new approach for efficient genotype
imputation using information from relatives. BMC Genomics. 2014;15:478. 7. Lu AT, Cantor RM. Identifying rare-variant associations in parent-child trios
using a Gaussian support vector machine. BMC Proc. 2014;8 Suppl 1:S98. 8. Wellmann R, Preuß S, Tholen E, Heinkel J, Wimmers K , Bennewitz J. Competing interests
The authors declare that they have no competing interests. Acknowledgement g
We would like to express our gratitude to all those who gave us help to
complete this paper, especially for Dr. Y Forghani, Dr. M Kamaei, Dr. Y Bernal
Rubio, constructive suggestions and encouragements help us in all the time
of this research.
|
https://openalex.org/W4210495586
|
https://www.mdpi.com/2073-8994/14/2/277/pdf?version=1644570755
|
English
| null |
A Comparison of Bivariate Zero-Inflated Poisson Inverse Gaussian Regression Models with and without Exposure Variables
|
Symmetry
| 2,022
|
cc-by
| 13,148
|
Keywords: BZIPIGR type II; overdispersion; excess zero; exposure variable; MLE; BHHH; number
of mortalities Citation: Ermawati, E.; Purhadi, P.;
Rahayu, S.P. A Comparison of
Bivariate Zero-Inflated Poisson
Inverse Gaussian Regression Models
with and without Exposure Variables. Symmetry 2022, 14, 277. https://
doi.org/10.3390/sym14020277 Ermawati Ermawati 1,2, Purhadi Purhadi 1,*
and Santi Puteri Rahayu 1 1
Faculty of Science and Data Analytics, Department of Statistics, Institut Teknologi Sepuluh Nopember,
Surabaya 60111, Indonesia; ermawati@uin-alauddin.ac.id (E.E.); santi_pr@statistika.its.ac.id (S.P.R.)
2
Faculty of Science and Technology, Department of Mathematics, Universitas Islam Negeri Alauddin Makassar,
Makassar 92118, Indonesia
*
Correspondence: purhadi@statistika.its.ac.id Abstract: In this paper, we focus on the comparison of the bivariate zero-inflated Poisson inverse
Gaussian regression (BZIPIGR) type II model in two cases: with and without exposure variables. The BZIPIGR type II model is applied to analyze the occurrence of maternal and early neonatal
mortality in South Sulawesi Province, Indonesia using 2019 data, which contain many zero values
and have the issue of overdispersion in the response variable. Furthermore, to analyze the number of
deaths in various areas, the exposure variable is considered. The maximum likelihood estimation
(MLE) is used in parameter estimation, which involves numerical iteration and application of the
Berndt–Hall–Hall–Hausman (BHHH) algorithm. Sum square error (SSE) serves as the criterion of
model selection when exposure variables are included. The existence of exposure variables strongly
affects the model’s accuracy, especially using the BZIPIGR type II model. According to the SSE and
RMSE values, the BZIPIGR type II model with exposure variables performs better than the model
without exposure variables in estimating parameter values. All predictors with exposure variables in
this study had a significant influence on the number of maternal and early neonatal mortalities.
Citation: Ermawati, E.; Purhadi, P.;
Rahayu, S.P. A Comparison of
Bivariate Zero-Inflated Poisson
Inverse Gaussian Regression Models
with and without Exposure Variables. Symmetry 2022, 14, 277. https://
doi.org/10.3390/sym14020277
Academic Editor: Dumitru Baleanu
Received: 12 December 2021
Accepted: 20 January 2022
Published: 29 January 2022 Article
A Comparison of Bivariate Zero-Inflated Poisson Inverse Gaussian
Regression Models with and without Exposure Variables mawati Ermawati 1,2, Purhadi Purhadi 1,*
and Santi Puteri Rahayu 1 symmetry
S
S symmetry
S
S symmetry
S
S 1. Introduction Count data that follow a Poisson distribution can be modeled by Poisson regression. Poisson regression requires the assumption that the mean and variance of the response
variables must be equal (equidispersion), i.e., one parameter gives both the mean and
variance of the distribution. In reality, this assumption may not hold, either with a variance
larger than the mean (overdispersion) or otherwise (underdispersion) [1–5]. Such a violation
can result in errors in decision making in hypothesis testing due to the occurrence of
underestimates [6–8]. A model was developed based on mixed Poisson distribution to
overcome the overdispersion problem, constituting a combination of Poisson distribution
and other distributions, both discrete and continuous, including Poisson Inverse Gaussian
(PIG) distribution [9]. Academic Editor: Dumitru Baleanu Publisher’s Note: MDPI stays neutral
with regard to jurisdictional claims in
published maps and institutional affil-
iations. Poisson regression can be used to model count data, but between events within
a certain period, the data must be independent. For example, maternal mortality at a
certain time does not affect maternal mortality at that time or another time. However, the
number of mothers who die within a certain period in a given area may be influenced
by other factors; for example, more births results in a higher occurrence of maternal
mortality in an area. Thus, when comparing the number of mortalities between one
area and another, it is necessary to consider the existence of other variables that can
affect the number of occurrences, referred to as the exposure variable [10]. Frome and
Morris (1989) [11] conducted research involving exposure variables with Poisson regression
modeling. Violation of the equidispersion assumption, especially overdispersion within Copyright:
© 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article
distributed
under
the
terms
and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/). https://www.mdpi.com/journal/symmetry Symmetry 2022, 14, 277. https://doi.org/10.3390/sym14020277 Symmetry 2022, 14, 277 2 of 18 the Poisson distribution, can be caused by the number of zero values (excess zero) in the
response variable [11–13]. Phang and Loh recommend a zero-inflated Poisson inverse Gaussian regression (ZIP-
IGR) model for data with excess zeros and overdispersion [14]. Zero-inflated count data
contains many zeros in the response variable. Hilbe conducted simulations with data
containing excess zeros and overdispersion. 1. Introduction Several models were included in this simu-
lation, such as zero-inflated negative binomial regression (ZINBR), zero-inflated hurdle
negative binomial regression (ZIHNBR) and zero-inflated generalized regression (ZIGPR). The simulation obtained the lowest AIC and BIC values with the ZIPIGR model, so the
conclusion is that the ZIPIGR model is preferable to the other models [5]. Several other
researchers have also used ZIPIGR to model with count data containing excess zeros and
overdispersion [15,16]. The ZIPIGR model is limited to univariate cases; therefore, if the relevant application
has more than one response variable, excess zeros or overdispersion, this model cannot be
used. In this study, we used two correlated response variables: the number of maternal
and early neonatal mortalities. Maternal mortality markedly affects the survival of the
child, especially in cases of babies younger than one month old (early neonatal), where the
basic needs of the baby are dependent on the mother, such as breast feeding. More than
50% of response variables in this study contained zero values and exhibited overdispersion. Therefore, a bivariate zero-inflated Poisson inverse Gaussian regression (BZIPIGR) model
was developed. p
The BZIPIGR type I model consists of two probability functions that accommodate
the state of the response-variable pairs: all response variables are zero, and all response
variables are not zero. However, not all cases consist of two response conditions. In the case
of the number of maternal and early neonatal mortalities in South Sulawesi Province in 2018,
there were four combinations of zero values in the response variable. Therefore, to identify
the factors that significantly affected the number of maternal and early neonatal mortalities,
a BZIPIGR type II model was developed to accommodate all possible combinations of zero
values in the response variable. This is similar to the formation of the probability function
in the BZIGPR model [17]. Several researchers have used Newton-Raphson iterations to maximize the likelihood
function using the MLE method when the derivative of the likelihood function for the
model parameters was in non-closed form. This algorithm requires a second derivative to
form a Hessian matrix. Therefore, for a more complex probability distribution function, it is
difficult to obtain a second derivative. According to Nugraha, the first-derivative approach
can be used to inform the Hessian matrix using a Berndt–Hall–Hall–Hausman (BHHH)
iteration [18]. The complexity of the probability function in the BZIPIGR type II model
makes it difficult to obtain the second derivative. 1. Introduction Therefore, we used the BHHH iteration
method as an alternative to overcome this problem. 2. Bivariate Zero-Inflated Poisson Inverse Gaussian Regression Type II (BZIPIGR) The BZIPIGR type II model is derived from the ZIPIGR model. The ZIPIGR model is a
mixed of zero-inflated (ZI) and Poisson inverse Gaussian (PIG) distributions [16]. The zero
value in the ZIPIG distribution can be divided into two sources, namely structural zero (zero
state) with probability p0 and sampling zero (mixed Poisson distribution state or PIG state)
with probability (1 −p0). If there are two random variables, Y1 and Y2, with BZIPIG distri-
bution with parameters (ηh, γh, τ) where h = 1, 2 or (Y1, Y2) ∼BZIPIG(η1, η2, τ, γ1, γ2),
and where η1 and η2 are the parameters of the Poisson distribution; τ is the dispersion
parameter, and γ1, γ2 is the parameter of the zero-inflated distribution, the probability
function of Y1 and Y2 can be obtained with: (Y1, Y2)
∼(0, 0) with p0 probability
∼(PIG (η1, τ), 0) with p1 probability
∼(0, PIG (η2, τ)) with p2 probability
∼(BPIG (η1, η2, τ)) with p3 probability (Y1, Y2)
∼(0, 0) with p0 probability
∼(PIG (η1, τ), 0) with p1 probability
∼(0, PIG (η2, τ)) with p2 probability
∼(BPIG (η1, η2, τ)) with p3 probability ∼(BPIG (η1, η2, τ)) with p3 probability Symmetry 2022, 14, 277 3 of 18 3 of 18 where p3 = 1 −p2 −p1 −p0. 2. Bivariate Zero-Inflated Poisson Inverse Gaussian Regression Type II (BZIPIGR) +
(1 −γ1)(1 −γ2)
2
πτ
p
1 + 2τ(η1 + η2)
1
2 η2y2 exp( 1
τ )Ky2−1
2
1
τ
√
1+2τ(η1+η2)
√
1+2τ(η1+η2)
y2 y2! (4) (4) where Ky2−1
2
1
τ
p
1 + 2τη2
is the modified Bessel function of the third kind [20]; F3 =
γ2(1 −γ1)
2
πτ
√1 + α1τ
1
2
η1y1 exp( 1
τ)Ky1−1
2 ( 1
τ
√
1+2τη1)
(√
2η1τ+1)
y1y1! +
(1 −γ1)(1 −γ2)
2
πτ
p
1 + 2τ(η1 + η2)
1
2 η1y1 exp( 1
τ)Ky1−1
2
1
τ
√
1+2τ(η1+η2)
√
1+2τ(η1+η2)
y1y1! (5) (5) where Ky1−1
2
1
τ
p
1 + 2τ(η1 + η2)
is the modified Bessel function of the third kind [20] and where Ky F4 =
(1 −γ1)(1 −γ2)
2
πτ
p
1 + 2τ(η1 + η2)
1
2
η1y1η2y2 exp( 1
τ)
√
1+2τ(η1+η2)
y1+y2y1!y2! Ky1+y2−1
2
1
τ
p
1 + 2τ(η1 + η2)
(6) (6) where Ky1+y2−1
2
1
τ
p
1 + 2τ(η1 + η2)
is the modified Bessel function of the third kind [20]. 2.1. BZIPIGR Type II Model with Exposure Variables 2. Bivariate Zero-Inflated Poisson Inverse Gaussian Regression Type II (BZIPIGR) Hence, the joint probability function of Y1 and Y2 can be
written as
F
i f
0
0 P(Y1 = y1, Y2 = y2) =
F1
;
i f y1 = 0, y2 = 0
F2
;
i f y1 = 0, y2 > 0
F3
;
i f y1 > 0, y2 = 0
F4
;
i f y1 > 0, y2 > 0
(1) (1) where where where F1 = γ1γ2 + γ1(1 −γ2) exp
1
τ
2
πτ
p
1 + 2τη2
1
2 K−1
2
1
τ
p
1 + 2τη2
+
γ2(1 −γ1) exp
1
τ
2
πτ
p
1 + 2τη1
1
2 K−1
2
1
τ
p
1 + 2τη1
+
(1 −γ1)(1 −γ2) exp
1
τ
2
πτ
p
1 + 2τ(η1 + η2)
1
2 K−1
2
1
τ
p
1 + 2τ(η1 + η2)
(2) (2) where K−1
2 (z) = K 1
2 (z) =
π
2(z)
1
2 exp(−z) is the modified Bessel function of the third
kind [19]. Therefore, Equation (1) becomes F1 =
γ1γ2 + γ1(1 −γ2) exp
1
τ
1 −
p
1 + 2τη2
+ γ2(1 −γ1) exp
1
τ
1 −
p
1 + 2τη1
+
(1 −γ1)(1 −γ2) exp
1
τ
1 −
p
1 + 2τ(η1 + η2)
(3) F1 =
γ1γ2 + γ1(1 −γ2) exp
1
τ
1 −
p
1 + 2τη2
+ γ2(1 −γ1) exp
1
τ
1 −
p
1 + 2τη1
+
(1 −γ1)(1 −γ2) exp
1
τ
1 −
p
1 + 2τ(η1 + η2)
(3)
F2 =
γ1(1 −γ2)
2
πτ
p
1 + η2τ
1
2 η2y2 exp( 1
τ )Ky2−1
2 ( 1
τ
√
1+2τη2)
(√
2η2τ+1)
y2 y2! +
(1 −γ1)(1 −γ2)
2
πτ
p
1 + 2τ(η1 + η2)
1
2 η2y2 exp( 1
τ )Ky2−1
2
1
τ
√
1+2τ(η1+η2)
√
1+2τ(η1+η2)
y2 y2! (4) (3) F2 =
γ1(1 −γ2)
2
πτ
p
1 + η2τ
1
2 η2y2 exp( 1
τ )Ky2−1
2 ( 1
τ
√
1+2τη2)
(√
2η2τ+1)
y2 y2! 2.1. BZIPIGR Type II Model with Exposure Variables 2.1. BZIPIGR Type II Model with Exposure Variables The regression model for the BZIPIG distribution is divided into two equations: the
regression model with exposure variable, as in Equation (7) [10]: ηhi = qhi exp
xT
i λh
; h = 1, 2; i = 1, 2, . . . , n
(7) (7) (7) Symmetry 2022, 14, 277 4 of 18 where qhi is the exposure variable for the h-th response variable and the i-th observation; x
is the vector of the predictor variable, and λ is the model parameters. The model for the ZI
distribution can be written as Equation (9): where qhi is the exposure variable for the h-th response variable and the i-th observation; x
is the vector of the predictor variable, and λ is the model parameters. The model for the ZI
distribution can be written as Equation (9): log
γh
1−γh
= exp
xT
i αh
γh =
exp(xT
i αh)
1+exp(xT
i αh) and (1 −γh) =
1
1+exp(xT
i αh)
(8) (8) where α is the vector of the zero-inflated parameters. where α is the vector of the zero-inflated parameters. 2.1. BZIPIGR Type II Model with Exposure Variables p
Based on the model in Equations (7) and (8), the probability function of the BZIPIGR
type II model with exposures variable can be written as follows: Based on the model in Equations (7) and (8), the probability function of the BZIPIGR
type II model with exposures variable can be written as follows: P(Y1i = y1i, Y2i = y2i) =
P1i
;
y1i = 0, y2i = 0
P2i
;
y1i = 0, y2i > 0
P3i
;
y1i > 0, y2i = 0
P4i
;
y1i > 0, y2i > 0
(9) (9) P1i =
1
(1+exp(xTα1))(1+exp(xTα2))
exp
xTα1 + xTα2
+
exp
xTα1 + 1
τ
1 −
p
1 + 2τq exp(xTλ2)
+ exp
xTα2 + 1
τ
1 −
p
1 + 2τq exp(xTλ1)
+
exp
1
τ
1 −
p
1 + 2τq(exp(xTλ1) + exp(xTλ2))
i P1i =
1
(1+exp(xTα1))(1+exp(xTα2))
exp
xTα1 + xTα2
+
exp
xTα1 + 1
τ
1 −
p
1 + 2τq exp(xTλ2)
+ exp
xTα2 + 1
τ
1 −
p
1 + 2τq exp(xTλ1)
+
exp
1
τ
1 −
p
1 + 2τq(exp(xTλ1) + exp(xTλ2))
i P2i =
exp(xTα1)( 2
πτ)
1
2 (q exp(xTλ2))
y2 exp( 1
τ)Ky2−1
2
1
τ
q
1+2τq exp(xTλ2)
(1+exp(xTα1))(1+exp(xTα2))
q
1+2τq exp(xTλ2)
y2−1
2 y2! +
( 2
πτ)
1
2 (q exp(xTλ2))
y2 exp( 1
τ)Ky2−1
2
1
τ
q
1+2τq(exp(xTλ1)+exp(xTλ2))
(1+exp(xTλ1))(1+exp(xTλ2))
q
1+2τq(exp(xTλ1)+exp(xTλ2))
y2−1
2 y2! P3i =
exp(xTα2)( 2
πτ)
1
2 (q exp(xTλ1))
y1 exp( 1
τ)Ky1−1
2
1
τ
q
1+2τq exp(xTλ1)
(1+exp(xTα2))(1+exp(xTα1))
q
1+2τq exp(xTλ1)
y1−1
2 y1! +
( 2
πτ)
1
2 (q exp(xTλ1))
y1 exp( 1
τ)Ky1−1
2
1
τ
q
1+2τq(exp(xTλ1)+exp(xTλ2))
(1+exp(xTα2))(1+exp(xTα1))
q
1+2τq(exp(xTλ1)+exp(xTλ2))
y1−1
2 y1! P4i =
( 2
πτ)
1
2 (q exp(xTλ1))
y1(q exp(xTλ2))
y2 exp( 1
τ)
(1+exp(xTα1))(1+exp(xTα2))
q
1+2τq(exp(xTλ1)+exp(xTλ2))
y1+y2−1
2 y1!y2! Ky1+y2−1
2
1
τ
p
1 + 2τq(exp(xTλ1) + exp(xTλ2))
P3i =
exp(xTα2)( 2
πτ)
1
2 (q exp(xTλ1))
y1 exp( 1
τ)Ky1−1
2
1
τ
q
1+2τq exp(xTλ1)
(1+exp(xTα2))(1+exp(xTα1))
q
1+2τq exp(xTλ1)
y1−1
2 y1! 2.1. BZIPIGR Type II Model with Exposure Variables +
( 2
πτ)
1
2 (q exp(xTλ1))
y1 exp( 1
τ)Ky1−1
2
1
τ
q
1+2τq(exp(xTλ1)+exp(xTλ2))
(1+exp(xTα2))(1+exp(xTα1))
q
1+2τq(exp(xTλ1)+exp(xTλ2))
y1−1
2 y1! P4i =
( 2
πτ)
1
2 (q exp(xTλ1))
y1(q exp(xTλ2))
y2 exp( 1
τ)
(1+exp(xTα1))(1+exp(xTα2))
q
1+2τq(exp(xTλ1)+exp(xTλ2))
y1+y2−1
2 y1!y2! Ky1+y2−1
2
1
τ
p
1 + 2τq(exp(xTλ1) + exp(xTλ2))
2.2. BZIPIGR Type II Model without Exposure Variable 2.2. BZIPIGR Type II Model without Exposure Variable The regression model for the BZIPIGR type II model without exposure variable, as in
Equation (10) is [10] ηhi = exp
xT
i λh
; h = 1, 2; i = 1, 2, . . . , n
(10) (10) where x is the vector of the predictor variable for the i-th observation, and λ is the model
parameters for the h-th response variable. Based on the model in Equations (9) and (10), the probability function of the BZIPIGR
type II model without exposure variables can be written as follows: P(Y1i = y1i, Y2i = y2i) =
Q1i
;
y1i = 0, y2i = 0
Q2i
;
y1i = 0, y2i > 0
Q3i
;
y1i > 0, y2i = 0
Q4i
;
y1i > 0, y2i > 0
(11) (11) Symmetry 2022, 14, 277 5 of 18 Q1i =
1
(1+exp(xTα1))(1+exp(xTα2))
exp
xTα1 + xTα2
+
exp
xTα1 + 1
τ
1 −
p
1 + 2τ exp(xTλ2)
+ exp
xTα2 + 1
τ
1 −
p
1 + 2τ exp(xTλ1)
+
exp
1
τ
1 −
p
1 + 2τ(exp(xTλ1) + exp(xTλ2))
i
Q2i =
exp(xTα1)( 2
πτ)
1
2 (exp(xTλ2))
y2 exp( 1
τ)Ky2−1
2
1
τ
q
1+2τ exp(xTλ2)
(1+exp(xTα1))(1+exp(xTα2))
q
1+2τ exp(xTλ2)
y2−1
2 y2! +
( 2
πτ)
1
2 (exp(xTλ2))
y2 exp( 1
τ)Ky2−1
2
1
τ
q
1+2τ(exp(xTλ1)+exp(xTλ2))
(1+exp(xTλ1))(1+exp(xTλ2))
q
1+2τ(exp(xTλ1)+exp(xTλ2))
y2−1
2 y2!
Q3i =
exp(xTα2)( 2
πτ)
1
2 (exp(xTλ1))
y1 exp( 1
τ)Ky1−1
2
1
τ
q
1+2τ exp(xTλ1)
(1+exp(xTα2))(1+exp(xTα1))
q
1+2τ exp(xTλ1)
y1−1
2 y1! +
( 2
πτ)
1
2 (exp(xTλ1))
y1 exp( 1
τ)Ky1−1
2
1
τ
q
1+2τ(exp(xTλ1)+exp(xTλ2))
(1+exp(xTα2))(1+exp(xTα1))
q
1+2τ(exp(xTλ1)+exp(xTλ2))
y1−1
2 y1! 2.1. BZIPIGR Type II Model with Exposure Variables Q4i =
( 2
πτ)
1
2 (exp(xTλ1))
y1(exp(xTλ2))
y2 exp( 1
τ)
(1+exp(xTα1))(1+exp(xTα2))
q
1+2τ(exp(xTλ1)+exp(xTλ2))
y1+y2−1
2 y1!y2! Ky1+y2−1
2
1
τ
p
1 + 2τ(exp(xTλ1) + exp(xTλ2))
( 2
πτ) 2 (exp(xTλ2))
y2 exp( 1
τ)Ky2−1
2
1
τ
q
1+2τ(exp(xTλ1)+exp(xTλ2))
(1+exp(xTλ1))(1+exp(xTλ2))
q
1+2τ(exp(xTλ1)+exp(xTλ2))
y2−1
2 y2!
Q3i =
exp(xTα2)( 2
πτ)
1
2 (exp(xTλ1))
y1 exp( 1
τ)Ky1−1
2
1
τ
q
1+2τ exp(xTλ1)
(1+exp(xTα2))(1+exp(xTα1))
q
1+2τ exp(xTλ1)
y1−1
2 y1! +
( 2
πτ)
1
2 (exp(xTλ1))
y1 exp( 1
τ)Ky1−1
2
1
τ
q
1+2τ(exp(xTλ1)+exp(xTλ2))
(1+exp(xTα2))(1+exp(xTα1))
q
1+2τ(exp(xTλ1)+exp(xTλ2))
y1−1
2 y1! Q4i =
( 2
πτ)
1
2 (exp(xTλ1))
y1(exp(xTλ2))
y2 exp( 1
τ) Q3i =
exp(xTα2)( 2
πτ)
1
2 (exp(xTλ1))
y1 exp( 1
τ)Ky1−1
2
1
τ
q
1+2τ exp(xTλ1)
(1+exp(xTα2))(1+exp(xTα1))
q
1+2τ exp(xTλ1)
y1−1
2 y1! +
( 2
πτ)
1
2 (exp(xTλ1))
y1 exp( 1
τ)Ky1−1
2
1
τ
q
1+2τ(exp(xTλ1)+exp(xTλ2))
(1+exp(xTα2))(1+exp(xTα1))
q
1+2τ(exp(xTλ1)+exp(xTλ2))
y1−1
2 y1! Q4i =
( 2
πτ)
1
2 (exp(xTλ1))
y1(exp(xTλ2))
y2 exp( 1
τ)
(1+exp(xTα1))(1+exp(xTα2))
q
1+2τ(exp(xTλ1)+exp(xTλ2))
y1+y2−1
2 y1!y2! Ky1+y2−1
2
1
τ
p
1 + 2τ(exp(xTλ1) + exp(xTλ2))
2.3. Parameter Estimation of BZIPIGR Type II Model 2.3. Parameter Estimation of BZIPIGR Type II Model 2.3. Parameter Estimation of BZIPIGR Type II Model 2.3. Parameter Estimation of BZIPIGR Type II Model Parameter estimation using the MLE method begins with taking n random samples,
namely (Y1i, Y2i, X1i, X2i, . . . , Xki), as there are four independent models in the BZIPIGR
type II model, namely (y1 = 0, y2 = 0), (y1 = 0, y2 > 0), (y1 > 0, y2 = 0) and (y1 > 0, y2 > 0),
where the number of observations for each model is n1, n2, n3 and n4, respectively. 2.3. Parameter Estimation of BZIPIGR Type II Model
+
1
(1+exp(xT
i α1))(1+exp(xT
i α2))
2
πτ
q
1 + 2τqi
exp
xT
i λ1
+ exp
xT
i λ2
1
2 qi exp
xT
i λ2
y2i
exp( 1
τ )Ky2−1
2
1
τ
q
1+2τ(exp(xTλ1)+exp(xTλ2))
q
1+2τqi(exp(xT
i λ1)+exp(xT
i λ2))
y2i y2i!
log(L3(λh; αh; τ)) =
n3
∑
i = 1
y1i > 0, y2i = 0
log
"
exp(xT
i α2)( 2
πτ )
1
2
(1+exp(xT
i α2))(1+exp(xT
i α1))
1 + 2τqi exp
xT
i λ1
1
4 −y1i
2
y−1
1i ! exp
1
τ
qi exp
xT
i λ1
y1i Ky1−1
2
1
τ
p
1 + 2τ exp(xTλ1)
+
1
(1+exp(xT
i α2))(1+exp(xT
i α1))
2
πτ
1
2 y−1
1i ! 1 + 2τqi
exp
xT
i λ1
+ exp
xT
i λ2
1
4 −y1i
2 qi exp
xT
i λ1
y1i
exp
1
τ
Ky1−1
2
1
τ
p
1 + 2τ(exp(xTλ1) + exp(xTλ2))
i
log(L4(λh; αh; τ)) =
n4
∑
i = 1
y1i > 0, y2i > 0
h
−log
1 + exp
xT
i α1
−log
1 + exp
xT
i α2
+ 1
2 log
2
πτ
+
1
4 −y1i+y2i
2
log
1 + 2τqi
exp
xT
i λ1
+ exp
xT
i λ2
+
y1i
log qi + xT
i λ1
+ y2i
log qi + xT
i λ2
+ 1
τ −log(y1i!y2i!)+
log
Ky1+y2−1
2
1
τ
p
1 + 2τ(exp(xTλ1) + exp(xTλ2))
i log(L2(λh; αh; τ)) = log(L3(λh; αh; τ)) = If θ =
λT
1
λT
2
αT
1
αT
2
τ
T is a parameter vector of the BZIPIGR type II model,
then, to maximize the log likelihood function in Equation (12), these equations are derived
for each parameter and equalized to zero. The first derivative obtained for each parameter
is in non-closed form (Appendix A), so the maximization process continues with numerical
iteration, which uses the Berndt–Hall–Hall–Hausman (BHHH) algorithm according to the
following steps [21]: Step 1. Determine the initial value of each parameter of the BZIPIGR type II model
θ =
λT
1
λT
2
αT
1
αT
2
τ
T. 2.3. Parameter Estimation of BZIPIGR Type II Model The
likelihood equation can be written as follows: L(λh; αh; τ) =
n1
∏
i = 1
y1i = 0, y2i = 0
P(Yhi = yhi)
n2
∏
i = 1
y1i = 0, y2i > 0
P(Yhi = yhi)
n3
∏
i = 1
y1i > 0, y2i = 0
P(Yhi = yhi)
n4
∏
i = 1
y1i > 0, y2i > 0
P(Yhi = yhi)
L(λ
)
(L (λ
))(L (λ
))(L (λ
))(L (λ
))
(12) (12)
L(λh; αh; τ) = (L1(λh; αh; τ))(L2(λh; αh; τ))(L3(λh; αh; τ))(L4(λh; αh; τ)) Then, the log likelihood function is determined from Equation (11), and the following
equation is obtained: log L(λh; αh; τ) =
log(L1(λh; αh; τ)) + log(L2(λh; αh; τ)) + log(L3(λh; αh; τ))+
log(L4(λh; αh; τ))
(13) log L(λh; αh; τ) =
log(L1(λh; αh; τ)) + log(L2(λh; αh; τ)) + log(L3(λh; αh; τ))+
log(L4(λh; αh; τ))
(13) log L(λh; αh; τ) =
log(L1(λh; αh; τ)) + log(L2(λh; αh; τ)) + log(L3(λh; αh; τ))+
log(L4(λh; αh; τ))
(13) (13) where where
log
h
exp
xT
i α1 + xT
i α2
+ exp
xT
i α1 + 1
τ
1 −
q
1 + 2τqi exp
xT
i λ2
+
log
h
exp
xT
i α1 + xT
i α2
+ exp
xT
i α1 + 1
τ
1 −
q
1 + 2τqi exp
xT
i λ2
+ log(L1(λh; αh; τ)) = ∑
i = 1
y1i = 0, y2i = 0 exp
xT
i α2 + 1
τ
1 −
q
1 + 2τqi exp
xT
i λ1
+
exp
1
τ
1 −
q
1 + 2τqi
exp
xT
i λ1
+ exp
xT
i λ2
i
−log
1 + exp
xT
i α1
−log
1 + exp
xT
i α2
Symmetry 2022, 14, 277 6 of 18 log(L2(λh; αh; τ)) =
n2
∑
i = 1
y1i = 0, y2i > 0
log
exp(xT
i α1)
2
πτ
q
1+2τqi exp(xT
i λ2)
1
2 (qi exp(xT
i λ2))
y2i
(1+exp(xT
i α1))(1+exp(xT
i α2))
exp( 1
τ )Ky2−1
2
1
τ
q
1+2τ exp(xTλ2)
q
1+2τqi exp(xT
i λ2)
y2i y2i! 2.4. Hypothesis Testing of the BZIPIGR Type II Model The test statistic is Zαhj =
ˆαhj
se(ˆαhj). Reject H0 if
Zαhj
> Zδ/2, where δ is the level
of significance. g
c. Partial test of parameter τ g
c. Partial test of parameter τ Partial test of parameter τ
H0 : τ = 0;
H1 : τ ̸= 0 H0 : τ = 0;
H1 : τ ̸= 0 The test statistic is Zτ =
ˆτ
se( ˆτ). Reject H0 if |Zτ| > Zδ/2, where δ is the level
of significance. The test statistic is Zτ =
ˆτ
se( ˆτ). Reject H0 if |Zτ| > Zδ/2, where δ is the level
of significance. 2.3. Parameter Estimation of BZIPIGR Type II Model The initial values of model parameters are obtained
from the regression model parameter values using OLS, whereas the dispersion parameter
is obtained from the variance of the PIG distribution. . Define gradient vector Step 2. Define gradient vector g
θ(m)
=
∂L(θ)
∂λT
1
T
∂L(θ)
∂λT
2
T
∂L(θ)
∂αT
1
T
∂L(θ)
∂αT
2
T
∂L(θ)
∂τ
T Step 3. Determine the Hessian matrix (H*) for the BHHH algorithm H∗
θ(m)
= −
n
∑
i=1
gi
θ(m)
gi
θ(m)T where gi
θ(m)
is the gradient vector at the i-th observation. Step 4. Start the BHHH iteration using the following formula: where gi
θ(m)
is the gradient vector at the i-th observation. Step 4. Start the BHHH iteration using the following formula: where gi
θ(m)
is the gradient vector at the i-th observation. Step 4. Start the BHHH iteration using the following formula: ˆθ(m+1) = ˆθ(m) −H∗−1
ˆθ(m)
g
ˆθ(m) Symmetry 2022, 14, 277 7 of 18 7 of 18 The iteration of the BHHH algorithm starts at m = 0 and stops if ∥ˆθ(m+1) −ˆθ(m)∥≤ε
where ε is a small positive close to 0. 2.4. Hypothesis Testing of the BZIPIGR Type II Model 2.4. Hypothesis Testing of the BZIPIGR Type II Model Simultaneous hypothesis testing for the BZIPIGR model type II parameters can be
performed using the maximum likelihood ratio test (MLRT) method. The hypothesis to be
tested is as follows: H0 : λh0 = λh1 = . . . = λhk = αh0 = αh1 = . . . = αhk = τ = 0; h = 1, 2; j = 0, 1, 2, . . . , k
H1 : there is at least one of λhj ̸= 0, αhj ̸= 0, or τ ̸= 0; h = 1, 2 ; j = 1, 2, . . . , k Before determining the test statistic to be used, the parameter set under the population
(Ω) and H0(ω) are defined as follows: Ω=
n
λh, αh, τ
−∞< λhj < ∞, −∞< αhj < ∞, τ > 0; h = 1, 2; j = 0, 1, 2, . . . , k
o
ω = {λh0, αh0, τ|−∞< λh0 < ∞, −∞< αh0 < ∞, τ > 0} This simultaneous test uses test statistics This simultaneous test uses test statistics G2 = −log
Λ2
= 2
log L
ˆΩ
−log L( ˆω)
. which are compared with χ2
δ,(a−b), where δ is the level of significance; a is the number
of parameters under the population, and b is the number of parameters under H0. H0 is
rejected if G2 > χ2
δ,(a−b). (
)
Partial hypothesis testing to determine the significance of the effect of each predictor
in the model is divided into two tests, namely a partial test of parameters λhj and αhj. The
hypothesis statistical test for each parameter is as follows: a. Partial test of parameter λhj H0 : λhj = 0; h = 1, 2; j = 0, 1, 2, . . . , k
H1 : λhj ̸= 0 λhj =
ˆλhj
se(ˆλhj). Reject H0 if
Zλhj
> Zδ/2, where δ is the level of The test statistic is Zλhj =
ˆλhj
se(ˆλhj). Reject H0 if
Zλhj
> Zδ/2, where δ is the level of
significance. g
b. Partial test of parameter αhj H0 : αhj = 0; h = 1, 2; j = 0, 1, 2, . . . , k
H1 : αhj ̸= 0 The test statistic is Zαhj =
ˆαhj
se(ˆαhj). Reject H0 if
Zαhj
> Zδ/2, where δ is the level
of significance. 3. Application The BZIPIGR Type II model was applied to analyze the number of maternal and
early neonatal mortalities in South Sulawesi Province, Indonesia, in 2019. Modeling
in this study was carried out for two cases: with and without exposure variables. The
response variables used in this study were two correlated variables: the number of maternal Symmetry 2022, 14, 277 8 of 18 8 of 18 mortalities (Y1) and the number of early neonatal mortalities (Y2). There were six predictor
variables: the percentage of visits by pregnant women (X1), the percentage of pregnant
women who received Fe3 tablets (X2), percentage of births assisted by health workers
(X3), the percentage of mother who attended at least three postpartum maternal visits
(X4), the percentage of active integrated service posts (X5) and the percentage of obstetric
complications (X6). The exposure variable used in this study was the number of live births
in each sub-district (q). (q)
Table 1 shows the descriptive statistics of the response variables and predictor variables. The average number of maternal and early neonatal mortalities was 0.587 and 0.533,
respectively. This indicates that the maternal mortality rate was higher than the early
neonatal mortality rate in six districts in South Sulawesi Province, Indonesia. Based on the
coefficient of variance (CoV) value, maternal mortality was more homogeneous than early
neonatal mortality because the CoV value of variable Y1 was lower than that of variable Y2. Table 1. Descriptive statistics of variables. Variable (n = 75)
Mean
SD
Coefficient
of Variance
Min
Max
Number of maternal mortalities (Y1)
0.59
1.02
172.95
0
6
Number of early neonatal mortalities (Y2)
0.53
1.14
214.3
0
8
Percentage of visits by pregnant women (X1)
91.39
8.14
8.91
68.36
100
Percentage of pregnant women who received Fe3
tablets (X2)
82.31
24.83
30.16
15.31
127.76
Percentage of births assisted by health
workers (X3)
91.1
9.15
10.05
49.34
100
Percentage of mother who attended at least three
postpartum maternal visits (X4)
76.54
19.34
25.26
39.45
100
Percentage of active integrated service posts (X5)
73.44
35.71
48.63
0
100
Percentage of obstetric complications (X6)
72.03
29.09
40.39
7.27
128.31 Table 1. Descriptive statistics of variables. able 1. Descriptive statistics of variables. Before applying the BZIPIGIR type II model, it is necessary to test the overdispersion
assumption and calculate the percentage of zero values in the response variable. 3. Application Overdis-
persion occurs when the variance of the response variable is larger than the mean. This
hypothesis can be tested with the following formula [22]: Dh = 2
n
∑
i=1
Yhi log
Yhi
ˆηhi
−(Yhi −ˆηhi) where ˆηhi = exp
xT
i λh
; h = 1, 2; i = 1, 2, . . . , n. Overdispersion occurs if the deviance
value divided by the degrees of freedom is larger than one. Table 2 shows that all response
variables are overdispersed because the value of Dh/df is greater than one, and there are
excess zeros, with more than 61.33% of all observations on Y1 and 68% on Y2 equaling zero. Therefore, the BIZIPIGR type II model should be used to model these data. Table 2. Overdispersion test and percentage of zero values. Variable
% of Zero Value
Dh
df
Dh/df
Number of maternal mortalities (Y1)
61.33
88.637
68
1.303
Number of early neonatal mortalities (Y2)
68
99.099
68
1.457
3.1. Modeling the Number of Maternal and Early Neonatal Mortalities Using the BZIPIGR Type II
Model Involving Exposure Variables Table 2. Overdispersion test and percentage of zero values. 3.1. Modeling the Number of Maternal and Early Neonatal Mortalities Using the BZIPIGR Type II
Model Involving Exposure Variables Based on the results of modeling using the BZIPIGR type II model, the statistical value,
G2 = 1801.85, was greater than χ2
0.95;24 = 36.415, so the conclusion was that at least one of
the predictor variables influenced the number of maternal and early neonatal mortalities Symmetry 2022, 14, 277 9 of 18 in South Sulawesi, Indonesia. The model and partial test statistics for each predictor are
presented in Table 3. in South Sulawesi, Indonesia. The model and partial test statistics for each predictor are
presented in Table 3. Table 3. Partial test statistics of models with exposure variables. Table 3. Partial test statistics of models with exposure variables. 3.2. Modeling the Number of Maternal and Early Neonatal Mortalities Using the BZIPIGR Type II
Model without Exposure Variables Based on the results of modeling using the BZIPIGR type II model without exposure
variables, the statistical value, G2 = 776.754, was larger than χ2
0.95;24 = 36.415, so the
conclusion was that at least one of the predictor variables had an influence on the number
of maternal and early neonatal mortalities in South Sulawesi, Indonesia. The model and
partial test statistics for each predictor are presented seen in Table 4. Based on the results of the partial test in Table 4, several predictor variables have a
p-value > 0.05, namely the X1, X2, X3 and X4 variables for the variable Y2, so the conclusion
was that the effect of these four variables on the number of early neonatal mortalities (Y2)
was not significant. The effect of all variables on the number of maternal mortalities (Y1)
was significant. 3. Application Parameter
Estimate
Standard Error
Z
p-Value
Number of maternal mortalities (Y1)
λ01
−7.308
0.002
−3785.456
p < 0.05
λ11
0.005
0.002
2.860
p < 0.05
λ21
−0.006
0.0003
−20.465
p < 0.05
λ31
0.043
0.002
27.838
p < 0.05
λ41
−0.018
0.0003
−56.475
p < 0.05
λ51
−0.002
0.0003
−7.979
p < 0.05
λ61
0.005
0.0001
25.392
p < 0.05
α01
−5.031
0.0004
−11,937.814
p < 0.05
α11
0.157
0.013
2.159
p < 0.05
α21
−0.004
0.003
−1.539
p > 0.05
α31
−0.101
0.013
−7.535
p < 0.05
α41
0.013
0.003
4.272
p < 0.05
α51
0.006
0.002
2.793
p < 0.05
α61
0.009
0.002
4.985
p < 0.05
Number of early neonatal mortalities (Y2)
λ02
−3.532
0.0438
−80.570
p < 0.05
λ12
−0.036
0.0013
−28.411
p < 0.05
λ22
0.066
0.0010
65.948
p < 0.05
λ32
−0.117
0.0018
−64.065
p < 0.05
λ42
0.064
0.0013
48.034
p < 0.05
λ52
0.007
0.0004
18.887
p < 0.05
λ62
0.008
0.0005
16.890
p < 0.05
α02
−10.340
0.005
−1945.105
p < 0.05
α12
0.143
0.007
21.030
p < 0.05
α22
−0.017
0.003
−4.686
p < 0.05
α32
0.144
0.006
26.107
p < 0.05
α42
−0.116
0.002
−46.762
p < 0.05
α52
−0.009
0.003
−3.670
p < 0.05
α62
−0.016
0.004
−3.977
p < 0.05 Based on the results of the partial test in Table 3, all predictor variables were found to
have a p-value < 0.05, so the conclusion was that all predictor variables had a significant
effect, both for the variable Y1 and Y2, with a significance level of 0.05. 3.2. Modeling the Number of Maternal and Early Neonatal Mortalities Using the BZIPIGR Type II
Model without Exposure Variables 3.3. Comparison of the BZIPIGR Type II Model with and without Exposure Variables The comparison of the BZIPIGR type II model with and without the exposure variables
was performed based on the sum square error (SSE) value between the two models. The
SSE values of these two models are presented in Table 5. Symmetry 2022, 14, 277 10 of 18 10 of 18 Table 4. Partial test statistics without exposure variables. Parameter
Estimate
Standard Error
Z
p-Value
Number of maternal mortalities (Y1)
λ01
−11.274
0.119
−94.127
p < 0.05
λ11
−0.113
0.026
−4.312
p < 0.05
λ21
0.047
0.008
5.948
p < 0.05
λ31
0.229
0.027
8.438
p < 0.05
λ41
−0.037
0.005
−7.323
p < 0.05
λ51
−0.077
0.008
−11.437
p < 0.05
λ61
0.055
0.005
12.159
p < 0.05
α01
−12.243
0.538
−22.760
p < 0.05
α11
0.109
0.009
11.254
p < 0.05
α21
0.096
0.003
35.748
p < 0.05
α31
−0.138
0.009
−14.912
p < 0.05
α41
0.073
0.004
18.467
p < 0.05
α51
−0.084
0.003
−30.094
p < 0.05
α61
0.087
0.003
26.189
p < 0.05
Number of early neonatal mortalities (Y2)
λ02
−4.917
0.296
−16.627
p < 0.05
λ12
0.054
0.036
1.485
p > 0.05
λ22
−0.009
0.023
−0.421
p > 0.05
λ32
−0.046
0.037
−1.231
p > 0.05
λ42
0.020
0.031
0.655
p > 0.05
λ52
−0.013
0.005
−2.713
p < 0.05
λ62
0.051
0.005
10.056
p < 0.05
α02
3.252
0.396
8.218
p < 0.05
α12
−0.159
0.023
−6.904
p < 0.05
α22
−0.059
0.013
−4.626
p < 0.05
α32
0.383
0.035
10.827
p < 0.05
α42
−0.235
0.026
−9.089
p < 0.05
α52
−0.021
0.003
−6.562
p < 0.05
α62
0.048
0.004
12.068
p < 0.05
Table 5. SSE and RMSE values of models with and without exposure variables (real data). BZIPIGR Type II Model
SSE
RMSE
Model with exposure variables
218.8805
1.7811
Model without exposure variables
620.6658
2.9992
The above real data analysis obtained similar results to the simulation data analysis. Data were generated as much as n = 75 and n = 100, with 50% of response variables
containing zero values. Data were analyzed using the BZIPIGR type II model under two
conditions: with and without exposure variables. The simulation SSE and MSE values are
presented in Table 6. Table 6. SSE and RMSE value of models with and without exposure variables (simulation data). N
BZIPIGR Type II Model
SSE
RMSE
N
75
Model with exposure variables
314.7814
2.1359 Table 4. 4. Discussion Based on a comparison of the SSE and RMSE values in Tables 5 and 6, the model
with exposure variables performed better than the model without exposure variables. This
discussion focuses on the BZIPIGR type II model with exposure variables. There are two
BZIPIGR model types II, the log model and the logit model. The log model states that the
probability of Yi from the PIG state is significantly affected by a variable, whereas the logit
model states that the probability of Yi from the zero state is significantly influenced by a
variable. The two models for each response variable are as follows: a. Regression model for the number of maternal mortalities:
log(η1i) = −7.308 + 0.005X1i −0.006X2i + 0.043X3i −0.018X4i −0.002X5i + 0.005X6i
(14)
logit(γ1i) = −5.031 + 0.157X1i −0.004X2i −0.101X3i + 0.013X4i + 0.006X5i + 0.009X6i
(15) a. Regression model for the number of maternal mortalities:
log(η1i) = −7.308 + 0.005X1i −0.006X2i + 0.043X3i −0.018X4i −0.002X5i + 0.005X6i
(14)
logit(γ1i) = −5.031 + 0.157X1i −0.004X2i −0.101X3i + 0.013X4i + 0.006X5i + 0.009X6i
(15) (14) (15) b. Regression model for the number of early neonatal mortalities:
log(η2i) = −3.532 −0.036X1i + 0.066X2i −0.117X3i + 0.064X4i + 0.007X5i + 0.008X6i
(16)
logit(γ2i) = −10.340 −0.143X1i −0.017X2i + 0.144X3i −0.116X4i −0.009X5i −0.016X6i
(17) b. Regression model for the number of early neonatal mortalities:
log(η2i) = −3.532 −0.036X1i + 0.066X2i −0.117X3i + 0.064X4i + 0.007X5i + 0.008X6i
(16)
logit(γ2i) = −10.340 −0.143X1i −0.017X2i + 0.144X3i −0.116X4i −0.009X5i −0.016X6i
(17) Equations (13) and (14) show the model for the number of maternal mortalities. In
this equation, if other variables are held constant, then every 1% increase in variable X1
results in an increase in the number of maternal mortalities by 0.005 times, with the same
explanation for variables X3 and X6. For variable X2, if the other variables are held constant,
then every 1% increase in variable X2 results in a decrease in the number of maternal
mortalities by 0.006 times, with the same explanation for variables X4 and X5. However,
for the logit model in Equation (14), the coefficient values of X1, X3 and X6 are positive. 3.3. Comparison of the BZIPIGR Type II Model with and without Exposure Variables Partial test statistics without exposure variables. Table 5. SSE and RMSE values of models with and without exposure variables (real data). BZIPIGR Type II Model
SSE
RMSE
Model with exposure variables
218.8805
1.7811
Model without exposure variables
620.6658
2.9992 The above real data analysis obtained similar results to the simulation data analysis. The above real data analysis obtained similar results to the simulation data analysis. Data were generated as much as n = 75 and n = 100, with 50% of response variables
containing zero values. Data were analyzed using the BZIPIGR type II model under two
conditions: with and without exposure variables. The simulation SSE and MSE values are
presented in Table 6. Table 6. SSE and RMSE value of models with and without exposure variables (simulation data). Table 6. SSE and RMSE value of models with and without exposure variables (simulation data). N
BZIPIGR Type II Model
SSE
RMSE
N = 75
Model with exposure variables
314.7814
2.1359
Model without exposure variables
369.6821
2.3147
N = 100
Model with exposure variables
584.6494
2.9109
Model without exposure variables
4241.603
7.8404 As shown in Tables 5 and 6, the SSE and RMSE values of the BZIPIGR type II model
with exposure variables are less than those of the model without exposure variables. Therefore, the model with exposure variables performed better than the model without
exposure variables. Symmetry 2022, 14, 277 11 of 18 11 of 18 4. Discussion The means that, if other predictor variables are held constant, with every increase in the
percentage of visits by pregnant women, the probability that maternal mortality will not
occur increases, with the same explanation for X3 (percentage of births assisted by health
workers) and X6 (percentage of obstetric complications). p
g
p
Based on the log and logit models in Equations (15) and (16), the predictor variable
that affects the PIG state also affects the zero state (ZI). Based on the coefficient values of
the predictor variables, the influence of the predictor variable on the response variable can
be interpreted. For example, X2 (percentage of pregnant women who received Fe3 tablets)
in Equation (15) has a positive value, which means that the number of early neonatal
mortalities increases with the percentage of pregnant women who received Fe3 tablets. Variables X4 (percentage of mother who attended at least three postpartum maternal visits),
X5 (percentage of active integrated service posts) and X6 (percentage of obstetric compli-
cations) increase with the number of early neonatal mortalities. That means that if the
other predictor variables are held constant, then every 1% increase in the X2 variable (per-
centage of births assisted by health workers) increases the number of maternal mortalities
by 0.066 times the number of early neonatal mortalities, with the same explanation for
variables X4, X5 and X6. For the logit model in Equation (16), where the coefficient value of
the X1 is negative, every increase in the percentage of visits by pregnant women reduces
the likelihood of early neonatal mortalities not occurring if other predictor variables are
held constant, with the same explanations for X2, X4 and X6. 5. Conclusions The BZIPIGR model is derived from the Poisson regression model and can be used
when the equidispersion assumption is violated and excess zeros are present. A violation
of this assumption occurs when asymmetry exists between the mean and variance of the
response variable, especially when the variance is larger than the mean (overdispersion). The BZIPIGR type I model only accommodates the possibility that all response variables are
zero or that all response variables are non-zero. Therefore, the BZIPIGR type II model was
developed to overcome other possibilities of the zero value of the response variable, namely
that one of the response variables is zero. The BZIPIGR type II model accommodates all
possible combinations of zero values in the response variable. Thus, the BZIPIGR type II Symmetry 2022, 14, 277 12 of 18 12 of 18 model accounts for bivariate cases that experience overdispersion and excess zeros, such as
maternal and early neonatal mortalities in South Sulawesi Province, Indonesia, in 2019. y
According to the SSE and RMSE values of real data in Table 5, the BZIPIGR type
II model with exposure variables performed better than the model without exposure
variables in estimating the parameter values. This conclusion can be generally accepted
based on the results of the analysis using the simulation data in Table 6, where the same
results are obtained, namely lower SSE and RMSE values in the BZIPIGR type II model
with exposure variables. The existence of the exposure variables in the BZIPIGR type II
model strongly affects the model accuracy. The empirical results of modeling the number
of maternal and early neonatal mortalities at the sub-district level in the province of
South Sulawesi, Indonesia with exposure variables using the BZIPIGR type II model
showed that all predictor variables used in this study, i.e., percentage of visits by pregnant
women, percentage of pregnant women who received Fe3 tablets, percentage of births
assisted by health workers, percentage of mothers who attended at least three postpartum
maternal visits, percentage of active integrated service posts and percentage of obstetric
complications, had a significant effect on the response variable. The limited availability of data at the sub-district level means it is possible that there
are other predictor variables that were not included in the study. 5. Conclusions Therefore, for further
research, the BZIPIGR type II model can be applied to the same case by replacing or adding
other variables or other cases that meet the assumptions of this model. The existence of
differences in the characteristics of each region can cause differences in the incidence in
each region; therefore, further research using the BZIPIGR type II model can add spatial
effects to the model. Author Contributions: Conceptualization, E.E., P.P. and S.P.R.; methodology, E.E., P.P. and S.P.R.; soft-
ware, E.E. and P.P.; validation, E.E., P.P. and S.P.R.; formal analysis, E.E., P.P. and S.P.R.; investigation,
E.E., P.P. and S.P.R.; data curation, E.E.; writing—original draft preparation, E.E.; writing—review
and editing, P.P. and S.P.R.; supervision, P.P. and S.P.R.; visualization, E.E. and S.P.R.; project adminis-
tration, P.P. All authors have read and agreed to the published version of the manuscript. Funding: This research received no external funding. Funding: This research received no external funding. Funding: This research received no external funding. Institutional Review Board Statement: Not applicable. Institutional Review Board Statement: Not applicable. Informed Consent Statement: Not applicable. Informed Consent Statement: Not applicable. Data Availability Statement: Dataset not publicly available. Data Availability Statement: Dataset not publicly available. Conflicts of Interest: The authors declare no conflict of interest. Appendix A The likelihood function of the BZIPIGR type II is L(λh, αh, τ; h = 1, 2) =
n
∏
i=1
P(yhi|λh, αh, τ; h = 1, 2) log L(λh, αh, τ; h = 1, 2) =
log(L1(λh, αh, τ; h = 1, 2)) + log(L2(λh, αh, τ; h = 1, 2))+
log(L3(λh, αh, τ; h = 1, 2)) + log(L4(λh, αh, τ; h = 1, 2)) log L(λh, αh, τ; h = 1, 2) =
log(L1(λh, αh, τ; h = 1, 2)) + log(L2(λh, αh, τ; h = 1, 2))+
log(L3(λh, αh, τ; h = 1, 2)) + log(L4(λh, αh, τ; h = 1, 2)) if θ =
λT
1
λT
2
αT
1
αT
2
τ
T is the vector parameters of the BZIPIGR model. Therefore,
the log-likelihood function in Equation (10) is derived from each parameter and equated
with zero, as follows: if θ =
λT
1
λT
2
αT
1
αT
2
τ
T is the vector parameters of the BZIPIGR model. Therefore,
the log-likelihood function in Equation (10) is derived from each parameter and equated
with zero, as follows: ∂log L(•)
∂θ
= ∂log L1(•)
∂θ
+ ∂log L2(•)
∂θ
+ ∂log L3(•)
∂θ
+ ∂log L4(•)
∂θ The first derivative of the first log likelihood (log L(•)) function to parameter λ1 is
−
q1i exp(xT
i λ1)xi
A11+A12+A13+A14
A13
q
1+2τ(q1i exp(xT
i λ1)+q2i exp(xT
i λ2))+ ∂log L1(•)
∂λ1
=
n1
∑
i = 1
y1i = 0, y2i = 0
−
q1i exp(xT
i λ1)xi
A11+A12+A13+A14
A13
q
1+2τ(q1i exp(xT
i λ1)+q2i exp(xT
i λ2))+
A14
q
1+2τq1i exp(xT
i λ1)
! Appendix A pp
The joint probability function of Y1 and Y2 can be written as in the following equation: P(Y1 = y1, Y2 = y2) =
F1
;
i f y1 = 0, y2 = 0
F2
;
i f y1 = 0, y2 > 0
F3
;
i f y1 > 0, y2 = 0
F4
;
i f y1 > 0, y2 > 0 where F1 =
γ1γ2 + γ1(1 −γ2) exp
1
τ
1 −
p
1 + 2τη2
+ γ2(1 −γ1) exp
1
τ
1 −
p
1 + 2τη1
+
(1 −γ1)(1 −γ2) exp
1
τ
1 −
p
1 + 2τ(η1 + η2)
F1 =
γ1γ2 + γ1(1 −γ2) exp
1
τ
1 −
p
1 + 2τη2
+ γ2(1 −γ1) exp
1
τ
1 −
p
1 + 2τη1
+
(1 −γ1)(1 −γ2) exp
1
τ
1 −
p
1 + 2τ(η1 + η2)
F2 =
γ1(1 −γ2)
2
πτ
p
1 + η2τ
1
2
η2y2 exp( 1
τ)Ky2−1
2 ( 1
τ
√
1+2τη2)
(√
2η2τ+1)
y2y2! +
(1 −γ1)(1 −γ2)
2
πτ
p
1 + 2τ(η1 + η2)
1
2 η2y2 exp( 1
τ)Ky2−1
2
1
τ
√
1+2τ(η1+η2)
√
1+2τ(η1+η2)
y2y2! 13 of 18 Symmetry 2022, 14, 277 F3 =
γ2(1 −γ1)
2
πτ
√1 + α1τ
1
2
η1y1 exp( 1
τ)Ky1−1
2 ( 1
τ
√
1+2τη1)
(√
2η1τ+1)
y1y1! +
(1 −γ1)(1 −γ2)
2
πτ
p
1 + 2τ(η1 + η2)
1
2 η1y1 exp( 1
τ)Ky1−1
2
1
τ
√
1+2τ(η1+η2)
√
1+2τ(η1+η2)
y1y1! F4 =
(1 −γ1)(1 −γ2)
2
πτ
p
1 + 2τ(η1 + η2)
1
2
η1y1η2y2 exp( 1
τ)
√
1+2τ(η1+η2)
y1+y2y1!y2! Ky1+y2−1
2
1
τ
p
1 + 2τ(η1 + η2)
The First Derivative of Likelihood functions for parameters of the BZIPIGR type II
model. Appendix A where A11 = exp
xT
i α1 + xT
i α2
, A12 = exp
xT
i α1 + 1
τ
1 −
q
1 + 2τq2i exp
xT
i λ2
A13 = exp
1
τ
1 −
q
1 + 2τ
q1i exp
xT
i λ1
+ q2i exp
xT
i λ2
A14 = exp
xT
i α2 + 1
τ
1 −
q
1 + 2τq1i exp
xT
i λ1
∂log L2(•)
∂λ1
=
n2
∑
i = 1
y1i = 0, y2i > 0
h
B21B22B24xi
A2+B2
1 + 2τ
q1i exp
xT
i λ1
+ q2i exp
xT
i λ2
−3
4 −y2i
2 τq1i exp
xT
i λ1
+
B24
(−z22)Ky2i+ 1
2
(z22)+(y22−1
2)Ky2i−1
2
(z22)
τq1i exp(xT
i λ1)
1+2τ(q1i exp(xT
i λ1)+q2i exp(xT
i λ2))
h
B21B22B24xi
A2+B2
1 + 2τ
q1i exp
xT
i λ1
+ q2i exp
xT
i λ2
−3
4 −y2i
2 τq1i exp
xT
i λ1
+ ∂log L2(•)
∂λ1 ∂log L2(•)
∂λ1 where where A2 = A21A22A23A24A25 Symmetry 2022, 14, 277 14 of 18 14 of 18 A21 =
exp(xT
i α1)
(1+exp(xT
i α1))(1+exp(xT
i α2)); A22 =
2
πτ
1
2 y−1
2i !; A23 =
1 + 2τq2i exp
xT
i λ2
1
4 −y2i
2
A24 = qy2i
2i exp
xT
i λ2y2i + τ−1
; A25 = Ks2(z21); z21 = 1
τ
q
1 + 2τq2i exp
xT
i λ2
B2 =
B21B22B23B24B25
B21 =
1
(1+exp(xT
i α1))(1+exp(xT
i α2)); B23 =
1 + 2τ
q1i exp
xT
i λ1
+ q2i exp
xT
i λ2
1
4 −y2i
2 ;
B22 =
2
πτ
1
2 y−1
2i !; B24 = qy2i
2i exp
xT
i λ2y2i + τ−1
; B25 = Ks2(z22);
z22 = 1
τ
q
1 + 2τ
q1i exp
xT
i λ1
+ q2i exp
xT
i λ2
∂log L3(•)
∂λT
1
=
n3
∑
i = 1
y1i > 0, y2i = 0
1
A3+B3
"
A31A32
"
A34A35( 1
2 −y1i)τq1i exp(xT
i λ1)xi
(1+2τq1i exp(xT
i λ1))
3
4 + y1i
2
+ A33A34xiy1iA35+
A33A34τq1i exp(xT
i λ1)xi
1+2τq1i exp(xT
i λ1)
−z31Ky1i+ 1
2 (z31) +
y1i −1
2
Ky1i−1
2 (z31)
+ A21 =
exp(xT
i α1)
(1+exp(xT
i α1))(1+exp(xT
i α2)); A22 =
2
πτ
1
2 y−1
2i !; A23 =
1 + 2τq2i exp
xT
i λ2
1
4 −y2i
2
A24 = qy2i
2i exp
xT
i λ2y2i + τ−1
; A25 = Ks2(z21); z21 = 1
τ
q
1 + 2τq2i exp
xT
i λ2
B2 =
B21B22B23B24B25
B21 =
1
(1+exp(xT
i α1))(1+exp(xT
i α2)); B23 =
1 + 2τ
q1i exp
xT
i λ1
+ q2i exp
xT
i λ2
1
4 −y2i
2 ;
B22 =
2
πτ
1
2 y−1
2i !; B24 = qy2i
2i exp
xT
i λ2y2i + τ−1
; B25 = Ks2(z22);
z22 = 1
τ
q
1 + 2τ
q1i exp
xT
i λ1
+ q2i exp
xT
i λ2
1
(1+exp(xT
i α1))(1+exp(xT
i α2)); B23 =
1 + 2τ
q1i exp
xT
i λ1
+ q2i exp
xT
i λ2
1
4 −y2i
2 ;
B22 =
2
πτ
1
2 y−1
2i !; B24 = qy2i
2i exp
xT
i λ2y2i + τ−1
; B25 = Ks2(z22);
z22 = 1
τ
q
1 + 2τ
q1i exp
xT
i λ1
+ q2i exp
xT
i λ2
og L3(•)
∂λT
1
=
n3
∑
i = 1
y1i > 0, y2i = 0
1
A3+B3
"
A31A32
"
A34A35( 1
2 −y1i)τq1i exp(xT
i λ1)xi
(1+2τq1i exp(xT
i λ1))
3
4 + y1i
2
+ A33A34xiy1iA35+
A33A34τq1i exp(xT
i λ1)xi
1+2τq1i exp(xT
i λ1)
−z31Ky1i+ 1
2 (z31) +
y1i −1
2
Ky1i−1
2 (z31)
+
B31B32
"
B34B35( 1
2 −y1i)τq1i exp(xT
i λ1)xi
(1+2τ(q1i exp(xT
i λ1)+q2i exp(xT
i λ2)))
3
4 + y1i
2 + B33B34xiy1iB35+ ∂log L3(•)
∂λT
1
= 0, y2i = 0
A33A34τq1i exp(xT
i λ1)xi
1+2τq1i exp(xT
i λ1)
−z31Ky1i+ 1
2 (z31) +
y1i −1
2
Ky1i−1
2 (z31)
+
B31B32
"
B34B35( 1
2 −y1i)τq1i exp(xT
i λ1)xi
(1+2τ(q1i exp(xT
i λ1)+q2i exp(xT
i λ2)))
3
4 + y1i
2 + B33B34xiy1iB35+ B33B34τq1i exp
xT
i λ1
xi
1 + 2τ
q1i exp
xT
i λ1
+ q2i exp
xT
i λ2
−z32 Ky2i+ 1
2 (z32) +
y2i −1
2
Ky2i−1
2 (z32)
##
where where A3 = A31A32A33A34Ks3(z31)
A31 =
exp(xT
i α2)
(1+exp(xT
i α2))(1+exp(xT
i α1)); A32 =
2
πτ
1
2 y−1
1i !; A33 =
1 + 2τq1i exp
xT
i λ1
1
4 −y1i
2
A34 = qy1i
1i exp
xT
i λ1y1i + τ−1
; A35 = Ks3(z31) B3 = B31B32B33B34B35 (1+exp(xi α2))(1+exp(xi α1))
B33 =
1 + 2τ
q1i exp
xT
i λ1
+ q2i exp
xT
i λ2
1
4 −y1i
2 ; B34 = qy1i
1i exp
xT
i λ1y1i + τ−1
;
B35 = Ks3(z32); Ky1i−1
2
1
τ
q
1 + 2τ
q1i exp
xT
i λ1
+ q2i exp
xT
i λ2
= Ks3(z32) (
p( i
2))(
p( i
1))
B33 =
1 + 2τ
q1i exp
xT
i λ1
+ q2i exp
xT
i λ2
1
4 −y1i
2 ; B34 = qy1i
1i exp
xT
i λ1y1i + τ−1
;
B35 = Ks3(z32); Ky1i−1
2
1
τ
q
1 + 2τ
q1i exp
xT
i λ1
+ q2i exp
xT
i λ2
= Ks3(z32) B33 =
1 + 2τ
q1i exp
xT
i λ1
+ q2i exp
xT
i λ2
4
2 ; B34 = qy1i
1i exp
xT
i λ1y1i + τ−1
;
B35 = Ks3(z32); Ky1i−1
2
1
τ
q
1 + 2τ
q1i exp
xT
i λ1
+ q2i exp
xT
i λ2
= Ks3(z32)
∂log L4(•)
∂λT
1
=
n4
∑
i = 1
y1i > 0, y2i > 0
1
4 −y1i+y2i
2
2τq1i exp(xT
i λ1)xi
1+2τ(q1i exp(xT
i λ1)+q2i exp(xT
i λ2)) +y1ixi+
(−z4Ks4+1(z4)+s4Ks4(z4))τq1i exp(xT
i λ1)xi
Ky1i+y2i−1
2
(z4)(1+2τ(q1i exp(xT
i λ1)+q2i exp(xT
i λ2)))
# ∂log L4(•)
∂λT
1
=
n4
∑
i = 1
y1i > 0, y2i > 0
1
4 −y1i+y2i
2
2τq1i exp(xT
i λ1)xi
1+2τ(q1i exp(xT
i λ1)+q2i exp(xT
i λ2)) +y1ixi+
(−z4Ks4+1(z4)+s4Ks4(z4))τq1i exp(xT
i λ1)xi
Ky1i+y2i−1
2
(z4)(1+2τ(q1i exp(xT
i λ1)+q2i exp(xT
i λ2)))
# ∂log L4(•)
∂λT
1
=
n4
∑
i = 1
y1i > 0, y2i > 0
1
4 −y1i+y2i
2
2τq1i exp(xT
i λ1)xi
1+2τ(q1i exp(xT
i λ1)+q2i exp(xT
i λ2)) +y1ixi+
(
K
(
)
K
(
))
( Tλ )
# n4
∑
i = 1
1
4 −y1i+y2i
2
2τq1i exp(xT
i λ1)xi
1+2τ(q1i exp(xT
i λ1)+q2i exp(xT
i λ2)) +y1ixi+ (−z4Ks4+1(z4)+s4Ks4(z4))τq1i exp(xT
i λ1)xi
Ky1i+y2i−1
2
(z4)(1+2τ(q1i exp(xT
i λ1)+q2i exp(xT
i λ2)))
# The first derivative of the log L(•) function to parameter λ2 is obtained as n1
∑
i = 1
y1i = 0, y2i = 0
"
−
q2i exp(xT
i λ2)xi
A11+A12+A13+A14
A12
q
1+2τq2i exp(xT
i λ2)+
A13
q
1+2τ(q1i exp(xT
i λ1)+q2i exp(xT
i λ2))
!# 15 of 18 15 of 18 Symmetry 2022, 14, 277
xi
A2+B2
A21A22
A24A25
1
4 −y2i
2
1 + 2τqi exp
xT
i λ2
−3
4 −y2i
2 2τqi exp
xT
i λ2
+ y1i
, y2i
A23A25y2i
qi exp
xT
i λ2y2i
y2i exp
1
τ
+
A23A24τqi exp(xT
i λ2)xi
1+2τqi exp(xT
i λ2)
−z21 Ky2i+ 1
2 (z21) +
y2i −1
2
Ky2i−1
2 (z21)
i
+ B21B22
"
B24B25( 1
2 −y2i)τqi exp(xT
i λ2)
(1+2τqi(exp(xT
i λ1)+exp(xT
i λ2)))
3
4 + y2i
2 +
B23B25y2i
qi exp
xT
i λ2y2i
y2i exp
1
τ
+
B23B24τqi exp(xT
i λ2)xi
1+2τqi(exp(xT
i λ1)+exp(xT
i λ2))
−z22Ky2i+ 1
2 (z22) +
y2i −1
2
Ky2i−1
2 (z22)
i ∂log L3(•)
∂λT
2
=
n3
∑
i = 1
y1i > 0, y2i = 0
"
1
A3+B3
B31B32B34
( 1
2 −y1i)τq2i exp(xT
i λ2)xi
(1+2τ(q1i exp(xT
i λ1)+q2i exp(xT
i λ2)))
3
4 + y1i
2 B35 +
B33
−z32Ky13+ 1
2 (z32)+(y13−1
2)Ky13−1
2 (z32)
τq2i exp(xT
i λ2)xi
1+2τ(q1i exp(xT
i λ1)+q2i exp(xT
i λ2))
∂log L4(•)
∂λT
2
=
n4
∑
i = 1
y1i > 0, y2i = 0
1
2 −(y1i + y2i)
τq2i exp(xT
i λ2)xi
1+2τ(q1i exp(xT
i λ1)+q2i exp(xT
i λ2)) + y2ixi+
−z4Ky1i+y2i+ 1
2 (z4)+(y1i+y2i−1
2)Ky1i+y2i−1
2 (z4)
τq2i exp(xT
i λ2)xi
Ky1i+y2i−1
2 (z4)[1+2τ(q1i exp(xT
i λ1)+q2i exp(xT
i λ2))]
where
z4 = τ−1q
1 + 2τqi
exp
xT
i λ1
+ exp
xT
i λ2
where z4 = τ−1q
1 + 2τqi
exp
xT
i λ1
+ exp
xT
i λ2
The first derivative of the log-likelihood function to parameter α1 is obtained as ∂log L(•)
∂αT
1
=
xi
n1
∑
i = 1
y1i = 0, y2i = 0
A11+A12
A11+A12+A13+A14 −
exp(xT
i α1)
1+exp(xT
i α1)
+
n2
∑
i = 1
y1i = 0, y2i > 0
(A2−B2 exp(xT
i α1))
(A2+B2)(1+exp(xT
i α1)) −
n3
∑
i = 1
y1i > 0, y2i = 0
exp(xT
i α1)(A3+B3)
(A3+B3)(1+exp(xT
i α1)) +
−
n4
∑
i = 1
y1i > 0, y2i = 0
exp(xT
i α1)
1+exp(xT
i α1)
The first derivative of the log-likelihood function to parameter α2 is ∂log L(•)
∂αT
1
=
xi
n1
∑
i = 1
y1i = 0, y2i = 0
A11+A12
A11+A12+A13+A14 −
exp(xT
i α1)
1+exp(xT
i α1)
+
n2
∑
i = 1
y1i = 0, y2i > 0
(A2−B2 exp(xT
i α1))
(A2+B2)(1+exp(xT
i α1)) −
n3
∑
i = 1
y1i > 0, y2i = 0
exp(xT
i α1)(A3+B3)
(A3+B3)(1+exp(xT
i α1)) +
−
n4
∑
i = 1
y1i > 0, y2i = 0
exp(xT
i α1)
1+exp(xT
i α1)
The first derivative of the log-likelihood function to parameter α2 is 16 of 18 Symmetry 2022, 14, 277 ∂log L1(•)
∂αT
2
=
xi
n1
∑
i = 1
y1i = 0, y2i = 0
A11+A14
A11+A12+A13+A14 −
exp(xT
i α2)
1+exp(xT
i α2)
+
−
n2
∑
i = 1
y1i = 0, y2i > 0
exp(xT
i α2)(A2+B2)xi
(A2+B2)(1+exp(xT
i α2))
−
n3
∑
i = 1
y1i > 0, y2i = 0
(A3+exp(xT
i α1)xiB3)xi
(A3+B3)(1+exp(xT
i α1))
+
−
n4
∑
i = 1
y1i > 0, y2i = 0
exp(xT
i α2)xi
1+exp(xT
i α2)
The first derivative of the first log-likelihood function (log L(•)) to parameter τ is
obtained as ∂log L1(•)
∂τ
=
n1
∑
i = 1
y1i = 0, y2i = 0
1
A11+A12+A13+A14
A12
−1
τ2 +
1+τqi exp(xT
i λ2)
τ2
q
1+2τqi exp(xT
i λ2)
! Appendix A +
A13
−1
τ2 +
1+τqi(exp(xT
i λ1)+exp(xT
i λ2))
τ2
q
1+2τqi(exp(xT
i λ1)+exp(xT
i λ2))
! +
A14
−1
τ2 +
1+τqi exp(xT
i λ1)
τ2
q
1+2τqi exp(xT
i λ1)
!! A13
−1
τ2 +
1+τqi(exp(xT
i λ1)+exp(xT
i λ2))
τ2
q
1+2τqi(exp(xT
i λ1)+exp(xT
i λ2))
! +
A14
−1
τ2 +
1+τqi exp(xT
i λ1)
τ2
q
1+2τqi exp(xT
i λ1)
!! q
!! References 1. Sari, D.N.; Purhadi, P.; Rahayu, S.P.; Irhamah, I. Estimation and Hypothesis Testing for the Parameters of Multivariate Zero
Inflated Generalized Poisson Regression Model. Symmetry 2021, 13, 1876. [CrossRef] 1. Sari, D.N.; Purhadi, P.; Rahayu, S.P.; Irhamah, I. Estimation and Hypothesis Testing for the Parameters of Multivariate Zero
Inflated Generalized Poisson Regression Model. Symmetry 2021, 13, 1876. [CrossRef] g
y
y
2. Zha, L.; Lord, D.; Zou, Y. The Poisson Inverse Gaussian (PIG) Generalized Linear Regression Model for Analyzing Motor Vehicle
Crash Data. J. Transp. Saf. Secur. 2014, 8, 18–35. [CrossRef] g
y
y
2. Zha, L.; Lord, D.; Zou, Y. The Poisson Inverse Gaussian (PIG) Generalized Linear Regression Model for Analyzing Motor Vehicle
Crash Data. J. Transp. Saf. Secur. 2014, 8, 18–35. [CrossRef] 3. Famoye, F.; Sighn, K.P. Zero-Inflated Generalized Poisson Regression Model with an Application to Domestic Violence Data. J. Data Sci. 2006, 4, 117–130. [CrossRef] 3. Famoye, F.; Sighn, K.P. Zero-Inflated Generalized Poisson Regression Model with an Application to Domestic Violence Data. J. Data Sci. 2006, 4, 117–130. [CrossRef] 4. Purhadi; Sari, D.N.; Aini, Q.; Irhamah. Geographically weighted Bivariate Zero Inflated Generalized Po
2021, 7, e07491. [CrossRef] [PubMed] 5. Cameron, A.C.; Trivedi, P.K. Regression Analysis of Count Data; Cambridge University Press: New York, NY, USA, 1998. g
y
f
g
y
6. Hilbe, J.M. Poisson Inverse Gaussian Regression. In Modeling Count Data; Cambridge University Press: New York, NY, USA, 2014. [CrossRef] 6. Hilbe, J.M. Poisson Inverse Gaussian Regression. In Modeling Count Data; Cambridge University Pre
[CrossRef] 7. Ozmen, I.; Famoye, F. Count Regression Models with an Application to Zoological Data Containing Str
2007, 5, 491–502. [CrossRef] [
]
8. McCullagh, P.; Nelder, J.A. Generalized Linear Models, 2nd ed.; Chapman and Hall: London, UK, 1989. 8. McCullagh, P.; Nelder, J.A. Generalized Linear Models, 2nd ed.; Chapman and Hall: London, UK, 1989. 9. Karlis, D.; Xekalaki, E. A Simulation Comparison of Several Procedures for Testing the Poisson Assumption. Statistician 2000, 49,
355–382. [CrossRef] 10. Mardalena, S.; Purhadi; Purnomo, D.T.; Prastyo, D.D. Parameter Estimation and Hypothesis Testing of Multivariate Poisson
Inverse Gaussian Regression. Symmetry 2020, 12, 1738. [CrossRef] 11. Frome, E.L.; Morris, M.D. Evaluating Goodness of Fit of Poisson Regression Model in Cohort Studie 11. Frome, E.L.; Morris, M.D. Evaluating Goodness of Fit of Poisson Regression Model in Cohort Studies. Am. Stat. 1989, 43, 144–147. 12. Leffondré, K.; Jager, K.J.J.; Boucquemont, J.; Stel, V.S.; Heinze, G. Appendix A
−1
2
(τ)−3
2 ; ∂B33
∂τ =
( 1
2 −y1i)(q1i exp(xT
i λ1)+q2i exp(xT
i λ2))
(1+2τ(q1i exp(xT
i λ1)+q2i exp(xT
i λ2)))
3
4 + y1i
2
;
∂B
y
∂B32
∂τ =
2
π
1
2 y−1
1i !
−1
2
(τ)−3
2 ; ∂B33
∂τ =
( 1
2 −y1i)(q1i exp(xT
i λ1)+q2i exp(xT
i λ2))
(1+2τ(q1i exp(xT
i λ1)+q2i exp(xT
i λ2)))
3
4 + y1i
2
;
∂B35
∂τT =
−z32Ky1i+ 1
2
(z32)+(y1i−1
2)Ky1i−1
2
(z32)
τ(1+2τqi(exp(xT
i λ1)+exp(xT
i λ2)))
−
1 + τ
q1i exp
xT
i λ1
+ q2i exp
xT
i λ2
The first derivative of the likelihood for the model without exposure variables is
like the BZIPIGR type II model with exposure variables, but qi is not involved in the
equation above. Appendix A ∂log L2(•)
∂τ
=
n2
∑
i = 1
y1i = 0, y2i > 0
1
A2+B2
A21A22
h
∂A23
∂τ
A24A25 + A23
∂A24
∂τ
A25 + A23A24
∂A25
∂τ
i
+
B21B22
h
∂B23
∂τ
B24B25 + B23
∂B24
∂τ
B25 + B23B24
∂A25
∂τ
i
where
∂A23
∂τ
=
1
2 −y2i
q2i exp(xT
i λ2)
(1+2τq2i exp(xT
i λ2))
3
4 + y2i
2
; ∂A24
∂τ
= −
q2i exp
xT
i λ2
y2i exp
1
τ
τ−2
∂A25
∂τ
= −
1+τq2i exp(xT
i λ2)
τ(1+2τq2i exp(xT
i λ2))
−z21 Ky2i+ 1
2 (z21) +
y2i −1
2
Ky2i−1
2 (z21)
∂B23
∂τ
=
1
2 −y2i
q1i exp
xT
i λ1
+ q2i exp
xT
i λ2
1 + 2τ
q1i exp
xT
i λ1
+ q2i exp
xT
i λ2
3
4 + y2i
2
;
∂B24
∂τ
= −
q2i exp
xT
i λ2
y2i exp
1
τ
τ−2
∂B25
∂τ
=
−z22Ky2i+ 1
2 (z22) +
y2i −1
2
Ky2i−1
2 (z22)
τ
1 + 2τ
q1i exp
xT
i λ1
+ q2i exp
xT
i λ2
−
1 + τq1i exp
xT
i λ1
+ q2i exp
xT
i λ2
∂log L3(•)
∂τ
=
n3
∑
i = 1
y1i > 0, y2i = 0
h
1
A3+B3
A31
h
∂A32
∂τ
A33A34A35 + A32
∂A33
∂τ
A34A35 + A32A33
∂A34
∂τ
A35+
A32A33A34
∂A35
∂τ
i
+ B31B32B34
h
∂B32
∂τ
B33B34B35+ B32
∂B33
∂τ
B34B35+
B32B33
∂B34
∂τ
B35 + B32B33B34
∂B35
∂τ
ii ∂log L2(•)
∂τ
=
n2
∑
i = 1
y1i = 0, y2i > 0
1
A2+B2
A21A22
h
∂A23
∂τ
A24A25 + A23
∂A24
∂τ
A25 + A23A24
∂A25
∂τ
i
+
B21B22
h
∂B23
∂τ
B24B25 + B23
∂B24
∂τ
B25 + B23B24
∂A25
∂τ
i
1
A2+B2
A21A22
h
∂A23
∂τ
A24A25 + A23
∂A24
∂τ
A25 + A23A24
∂A25
∂τ
i
+ ∂log L2(•)
∂τ B21B22
h
∂B23
∂τ
B24B25 + B23
∂B24
∂τ
B25 + B23B24
∂A25
∂τ
i where where ∂A23
∂τ
=
1
2 −y2i
q2i exp(xT
i λ2)
(1+2τq2i exp(xT
i λ2))
3
4 + y2i
2
; ∂A24
∂τ
= −
q2i exp
xT
i λ2
y2i exp
1
τ
τ−2
∂A25
∂τ
= −
1+τq2i exp(xT
i λ2)
τ(1+2τq2i exp(xT
i λ2))
−z21 Ky2i+ 1
2 (z21) +
y2i −1
2
Ky2i−1
2 (z21)
∂B23
∂τ
=
1
2 −y2i
q1i exp
xT
i λ1
+ q2i exp
xT
i λ2
1 + 2τ
q1i exp
xT
i λ1
+ q2i exp
xT
i λ2
3
4 + y2i
2
;
∂B24
∂τ
= −
q2i exp
xT
i λ2
y2i exp
1
τ
τ−2
∂B25
∂τ
=
−z22Ky2i+ 1
2 (z22) +
y2i −1
2
Ky2i−1
2 (z22)
τ
1 + 2τ
q1i exp
xT
i λ1
+ q2i exp
xT
i λ2
−
1 + τq1i exp
xT
i λ1
+ q2i exp
xT
i λ2
∂log L3(•)
∂τ
=
n3
∑
i = 1
y1i > 0, y2i = 0
h
1
A3+B3
A31
h
∂A32
∂τ
A33A34A35 + A32
∂A33
∂τ
A34A35 + A32A33
∂A34
∂τ
A35+
A32A33A34
∂A35
∂τ
i
+ B31B32B34
h
∂B32
∂τ
B33B34B35+ B32
∂B33
∂τ
B34B35+
B32B33
∂B34
∂τ
B35 + B32B33B34
∂B35
∂τ
ii ∂A23
∂τ
=
1
2 −y2i
q2i exp(xT
i λ2)
(1+2τq2i exp(xT
i λ2))
3
4 + y2i
2
; ∂A24
∂τ
= −
q2i exp
xT
i λ2
y2i exp
1
τ
τ−2
∂A25
∂τ
= −
1+τq2i exp(xT
i λ2)
τ(1+2τq2i exp(xT
i λ2))
−z21 Ky2i+ 1
2 (z21) +
y2i −1
2
Ky2i−1
2 (z21)
∂B23
∂τ
=
1
2 −y2i
q1i exp
xT
i λ1
+ q2i exp
xT
i λ2
1 + 2τ
q1i exp
xT
i λ1
+ q2i exp
xT
i λ2
3
4 + y2i
2
;
∂B24
∂τ
= −
q2i exp
xT
i λ2
y2i exp
1
τ
τ−2
∂B25
∂τ
=
−z22Ky2i+ 1
2 (z22) +
y2i −1
2
Ky2i−1
2 (z22)
τ
1 + 2τ
q1i exp
xT
i λ1
+ q2i exp
xT
i λ2
−
1 + τq1i exp
xT
i λ1
+ q2i exp
xT
i λ2
0
h
1
A3+B3
A31
h
∂A32
∂τ
A33A34A35 + A32
∂A33
∂τ
A34A35 + A32A33
∂A34
∂τ
A35+ ∂A23
∂τ
=
1
2 −y2i
q2i exp(xT
i λ2)
(1+2τq2i exp(xT
i λ2))
3
4 + y2i
2
; ∂A24
∂τ
= −
q2i exp
xT
i λ2
y2i exp
1
τ
τ−2
∂A25
∂τ
= −
1+τq2i exp(xT
i λ2)
τ(1+2τq2i exp(xT
i λ2))
−z21 Ky2i+ 1
2 (z21) +
y2i −1
2
Ky2i−1
2 (z21)
∂B23
∂τ
=
1
2 −y2i
q1i exp
xT
i λ1
+ q2i exp
xT
i λ2
T
T
3 + y2i ;
∂B24
∂τ
= −
q2i exp
xT
i λ2
y2i exp
1
τ
τ−2 A32A33A34
∂A35
∂τ
i
+ B31B32B34
h
∂B32
∂τ
B33B34B35+ B32
∂B33
∂τ
B34B35+
B32B33
∂B34
∂τ
B35 + B32B33B34
∂B35
∂τ
ii A32A33A34
∂A35
∂τ
i
+ B31B32B34
h
∂B32
∂τ
B33B34B35+ B32
∂B33
∂τ
B34B35+
B32B33
∂B34
∂τ
B35 + B32B33B34
∂B35
∂τ
ii 17 of 18 Symmetry 2022, 14, 277 where
∂A32
∂τ
=
2
π
1
2 y−1
1i ! Appendix A
−1
2
(τ)−3
2 ; ∂A33
∂τ
=
( 1
2 −y1i)q1i exp(xT
i λ1)
(1+2τq1i exp(xT
i λ1))
3
4 + y1i
2
;
∂A34
∂τ
= −qy1i
1i
exp
xT
i λ1
y1i exp
1
τ
τ−2
∂A32
∂τ
=
2
π
1
2 y−1
1i !
−1
2
(τ)−3
2 ; ∂A33
∂τ
=
( 1
2 −y1i)q1i exp(xT
i λ1)
(1+2τq1i exp(xT
i λ1))
3
4 + y1i
2
;
∂A34
∂τ
= −qy1i
1i
exp
xT
i λ1
y1i exp
1
τ
τ−2
∂A35
∂τ
=
−z31Ky1i+ 1
2 (z31) +
y1i −1
2
Ky1i−1
2 (z31)
−
1+τq1i exp(xT
i λ1)
τ(1+2τq1i exp(xT
i λ1))
∂B32
∂τ =
2
π
1
2 y−1
1i !
−1
2
(τ)−3
2 ; ∂B33
∂τ =
( 1
2 −y1i)(q1i exp(xT
i λ1)+q2i exp(xT
i λ2))
(1+2τ(q1i exp(xT
i λ1)+q2i exp(xT
i λ2)))
3
4 + y1i
2
;
∂B34
∂τ = −qy1i
1i
exp
xT
i λ1
y1i exp
1
τ
τ−2;
∂B35
∂τT =
−z32Ky1i+ 1
2
(z32)+(y1i−1
2)Ky1i−1
2
(z32)
τ(1+2τqi(exp(xT
i λ1)+exp(xT
i λ2)))
−
1 + τ
q1i exp
xT
i λ1
+ q2i exp
xT
i λ2
∂log L4(•)
∂τ
=
n4
∑
i = 1
y1i > 0, y2i > 0
−1
2τ2
+
1
4 −y1i+y2i
2
2(q1i exp(xT
i λ1)+q2i exp(xT
i λ2))
1+2τ(q1i exp(xT
i λ1)+q2i exp(xT
i λ2)) −1
τ2 +
−z4K
y1i+y2i+ 1
2
(z4)+(y1i+y2i−1
2)Ks4(z4)
! (−(1+τ(q1i exp(xT
i λ1)+q2i exp(xT
i λ2))))
Ky1i+y2i−1
2
(z4)τ(1+2τ(q1i exp(xT
i λ1)+q2i exp(xT
i λ2)))
∂A32
∂τ
=
2
π
1
2 y−1
1i !
−1
2
(τ)−3
2 ; ∂A33
∂τ
=
( 1
2 −y1i)q1i exp(xT
i λ1)
(1+2τq1i exp(xT
i λ1))
3
4 + y1i
2
;
∂A34
∂τ
= −qy1i
1i
exp
xT
i λ1
y1i exp
1
τ
τ−2 ∂B32
∂τ =
2
π
1
2 y−1
1i ! 15.
Silva, J.G. Zero-Inflated Mixed Poisson Regression Models. Master’s Thesis, Universidade Federal de Minas Gerais, Belo
Horizonte, Brazil, 17 February 2017. 16.
Gonzalves, J.N.; Souza, W.B. Flexible regression models for counts with high-inflation of zeros. Metron 2020, 78, 71–95. [CrossRef]
17.
Zhang, C.; Tian, G.; Huang, X. Two New Bivariate Zero-Inflated Generalized Poisson Distribution with a Flexible Correlation
Structure Stat Optim Inf Comput 2015 3 105–137 [CrossRef] 22.
Mardalena, S.; Purhadi; Purnomo, D.T.; Prastyo, D.D. Bivariate Poisson Inverse Gaussian Regression Model with Exposure
Variable: Infant and Maternal Death Case Study. J. Phys. Conf. Ser. 2021, 1752, 012016. [CrossRef] g
g
17.
Zhang, C.; Tian, G.; Huang, X. Two New Bivariate Zero-Inflated Generalized Poisson Distribution
Structure. Stat. Optim. Inf. Comput. 2015, 3, 105–137. [CrossRef] y
16.
Gonzalves, J.N.; Souza, W.B. Flexible regression models for counts with high-inflation of zeros. Metron 2020, 78, 71–95. [CrossRef]
17
Zhang C ; Tian G ; Huang X Two New Bivariate Zero-Inflated Generalized Poisson Distribution with a Flexible Correlation 21.
Beren, C.; Thamrin, S.A.; Jaya, A.K. Estimasi Parameter Distribusi Exponentiated Modified Weibull Extension. In Proceedings of
the National Conference Variansi, Makassar, Indonesia, 10 October 2016; ISBN 978-602-99837-2-2. p
f
p
[
]
18.
Nugraha, J.; Guritno, S.; Haryatmi, S. Probit Model on Multivariate Binary Response Using Simulate
Estimator. J. Ilmu Dasar 2010, 11, 70–75. (In Indonesia) ,
,
y
16.
Gonzalves, J.N.; Souza, W.B. Flexible regression models for counts with high-inflation of zeros. Metron 2 [
]
20.
Willmot, G.E. The Poisson-Inverse Gaussian distribution as an alternative to the negative binomial. Scand. Actuar. J. 1987, 1987,
113–127. [CrossRef] 19.
Shoukri, M.M.; Asyali, M.H.; Kelton, D. The Poisson Inverse Gaussian Regression Model of Clustered count Data. J. Data Sci.
2014, 2, 17–32. [CrossRef] References Representation of exposures in regression analysis and
interpretation of regression coefficients: Basic concepts and pitfalls. Nephrol. Dial. Transplant. 2014, 29, 1806–1814. [CrossRef]
[PubMed] [
]
13. Long, D.L.; Preisser, J.S.; Herring, A.H.; Golin, C.E. A Marginalized Zero-Inflated Poisson Regression Model with Overall
Exposure Effects. NIH Public Access 2014, 33, 5151–5165. [CrossRef] [PubMed] 14. Phang, Y.N.; Loh, E.F. Statistical Analysis for Overdispersed Medical Count Data. Int. J. Math. Com 14. Phang, Y.N.; Loh, E.F. Statistical Analysis for Overdispersed Medical Count Data. Int. J. Math. Comput. Sci. 2014, 8, 292–294. 18 of 18 Symmetry 2022, 14, 277 15. Silva, J.G. Zero-Inflated Mixed Poisson Regression Models. Master’s Thesis, Universidade Federal de Minas Gerais, Belo
Horizonte, Brazil, 17 February 2017. y
16. Gonzalves, J.N.; Souza, W.B. Flexible regression models for counts with high-inflation of zeros. Metron 2020, 78, 71–95. [CrossRef]
17. Zhang, C.; Tian, G.; Huang, X. Two New Bivariate Zero-Inflated Generalized Poisson Distribution with a Flexible Correlation
Structure. Stat. Optim. Inf. Comput. 2015, 3, 105–137. [CrossRef] p
f
p
18. Nugraha, J.; Guritno, S.; Haryatmi, S. Probit Model on Multivariate Binary Response Using Simulated Maximum Likelihood
Estimator. J. Ilmu Dasar 2010, 11, 70–75. (In Indonesia) (
)
19. Shoukri, M.M.; Asyali, M.H.; Kelton, D. The Poisson Inverse Gaussian Regression Model of Clustered count Data. J. Data Sci. 2014, 2, 17–32. [CrossRef] 20. Willmot, G.E. The Poisson-Inverse Gaussian distribution as an alternative to the negative binomial. Scand. Actuar. J. 1987, 1987,
113–127. [CrossRef] 20. Willmot, G.E. The Poisson-Inverse Gaussian distribution as an alternative to the negative binomial. Scand. Actuar. J. 1987, 1987,
113–127. [CrossRef] 21. Beren, C.; Thamrin, S.A.; Jaya, A.K. Estimasi Parameter Distribusi Exponentiated Modified Weibull Extension. In Proceedings of
the National Conference Variansi, Makassar, Indonesia, 10 October 2016; ISBN 978-602-99837-2-2. 22. Mardalena, S.; Purhadi; Purnomo, D.T.; Prastyo, D.D. Bivariate Poisson Inverse Gaussian Regression Model with Exposure
Variable: Infant and Maternal Death Case Study. J. Phys. Conf. Ser. 2021, 1752, 012016. [CrossRef]
|
https://openalex.org/W4381432297
|
https://www.frontiersin.org/articles/10.3389/fonc.2023.1171042/pdf
|
English
| null |
Adjuvant composite cold atmospheric plasma therapy increases antitumoral effect of doxorubicin hydrochloride
|
Frontiers in oncology
| 2,023
|
cc-by
| 9,978
|
TYPE Original Research
PUBLISHED 20 June 2023
DOI 10.3389/fonc.2023.1171042 lymphosarcoma, plasma therapy, cold atmospheric plasma (CAP), composite ions,
ion emitter OPEN ACCESS OPEN ACCESS
EDITED BY
Jianqiang Xu,
Dalian University of Technology, China
REVIEWED BY
Rajesh Kumar Gandhirajan,
Sri Ramachandra Institute of Higher
Education and Research, India
Alexios Vlamis,
University of Patras, Greece
*CORRESPONDENCE
Volha Kniazeva
v.kniazeva@realscientist.net
RECEIVED 21 February 2023
ACCEPTED 05 June 2023
PUBLISHED 20 June 2023
CITATION
Kniazeva V, Tzerkovsky D, Baysal Ö,
Kornev A, Roslyakov E and Kostevitch S
(2023) Adjuvant composite cold
atmospheric plasma therapy
increases antitumoral effect of
doxorubicin hydrochloride. Front. Oncol. 13:1171042. doi: 10 3389/fonc 2023 1171042 Volha Kniazeva 1*, Dzmitry Tzerkovsky 2, Ömür Baysal 3,
Alexander Kornev 1, Evgeny Roslyakov 1 and Serhei Kostevitch 1 1Bioresearch Department, R. S. C. Real Scientists Cyprus Ltd., Limassol, Cyprus, 2Laboratory of
Morphology, Molecular and Cellular Biology with a Group of Experimental Medicine, N. N. Alexandrov
National Cancer Center of Belarus, Lesnoy, Belarus, 3Faculty of Science, Department of Molecular
Biology and Genetics, Molecular Microbiology Unit, Mug˘ la Sıtkı Koc¸man University, Kötekli, Türkiye Introduction: Cancer is a global health concern, with a significant impact on
mortality rates. Despite advancements in targeted antitumor drugs, the
development of new therapies remains challenging due to high costs and
tumor resistance. The exploration of novel treatment approaches, such as
combined chemotherapy, holds promise for improving the effectiveness of
existing antitumor agents. Cold atmospheric plasma has demonstrated
antineoplastic effects in preclinical studies, but its potential in combination
with specific ions for lymphosarcoma treatment has not been investigated. Kniazeva V, Tzerkovsky D, Baysal Ö,
Kornev A, Roslyakov E and Kostevitch S
(2023) Adjuvant composite cold
atmospheric plasma therapy
increases antitumoral effect of
doxorubicin hydrochloride. Front. Oncol. 13:1171042. doi: 10.3389/fonc.2023.1171042 doi: 10.3389/fonc.2023.1171042 Frontiers in Oncology KEYWORDS 1 Introduction CAP dosage in clinical practice must
be closely controlled, and this control depends on the treatment
type (41). The CAP devices, dielectric barrier discharge, and plasma jet
demonstrate significant anti-cancer ability in vitro and in vivo
experiments. The method of composite ion emission is also one
of the approaches to generating low-temperature plasmas under
controlled conditions. The method has already found application in
an ion diffuser, PERENIO IONIC SHIELD™, Joule Production,
SIA, Riga, Latvia, created to combat the new type of coronavirus
SARS-CoV-2 (42, 43). Biological tissue is primarily affected by two components of
physical plasma (1): electromagnetic radiation (UV, VIS, IR, high-
frequency electro-magnetic fields, etc.); and (2) ions, electrons, and
reactive chemical species. The technical possibility of generating
physical plasma at low temperatures in an atmospheric environment
opens new chances to use this so-called cold atmospheric plasma
(CAP) for medical therapies (9, 10). The demonstration of the anti-tumor effect in vitro and in vivo
is the foundation of the clinical application of CAP sources. So far,
several CAP sources have been used to directly treat the
subcutaneously xenografted tumors above the skin. It is found
that the growth of a tumor could be effectively halted after the
treatment in most cases, which also results in an extended length of
life and a higher survival rate in mice (7, 44–47). To date, the chemical and molecular mechanisms of the
anticancer action of CAP have not been fully described. However,
there are several studies describing CAP’s anticancer effects through
several mechanisms. According to the current state of knowledge,
plasma effects on biological systems are mainly caused by reactive
oxygen and nitrogen species (ROS and RNS), such as hydrogen
peroxide, superoxide radicals, and nitric oxide. These reactive
species can cause oxidative stress in cancer cells, leading to DNA
damage, mitochondrial dysfunction, and ultimately cell death (11,
12). Induction of apoptosis: CAP can trigger programmed cell
death, known as apoptosis, in cancer cells. It activates signaling
pathways that lead to cell death while leaving healthy cells relatively
unaffected (13–15). Also, the phenomena of ferroptosis—cell death
regulated by iron-dependent lipid peroxidation (16, 17)—and
pyroptosys—a type of lytic programmed cell death (PCD)
characterized by cell swelling with large bubbles bulging from the
plasma cytoplasmic membrane and cell lysis, leading to the release
of pro-inflammatory molecules (18)—were described in in vitro
experiments. COPYRIGHT © 2023 Kniazeva, Tzerkovsky, Baysal, Kornev,
Roslyakov and Kostevitch. This is an open-
access article distributed under the terms of
the Creative Commons Attribution License
(CC BY). The use, distribution or
reproduction in other forums is permitted,
provided the original author(s) and the
copyright owner(s) are credited and that
the original publication in this journal is
cited, in accordance with accepted
academic practice. No use, distribution or
reproduction is permitted which does not
comply with these terms. Methods: An in vivo study was conducted using a Pliss lymphosarcoma rat model
to evaluate the antitumor effects of composite cold plasma and controlled ionic
therapy. Groups of rats were exposed to composite cold plasma for 3, 7, and 14
days, while the control group received no treatment. Additionally, a combination
of chemotherapy with cold plasma therapy was assessed, with doxorubicin
hydrochloride administered at a dosage of 5 mg/kg. PERENIO IONIC SHIELD™
emitted a controlled ionic formula during the treatment period. Results: The in vivo study demonstrated tumor growth inhibition in groups
exposed to composite cold plasma for 3, 7, and 14 days compared to the
control group. Furthermore, combining chemotherapy with cold plasma therapy
resulted in a threefold reduction in tumor volume. The most significant antitumor
effects were observed when doxorubicin hydrochloride at a dosage of 5 mg/kg
was combined with 14 days of PERENIO IONIC SHIELD™ionic therapy. Discussion: The use of composite cold plasma therapy, in conjunction with a
controlled ionic formula emitted by PERENIO IONIC SHIELD™, in the complex
treatment of lymphosarcoma in rats showed promising antitumor effects. The
combination therapy, particularly when combined with doxorubicin
hydrochloride, demonstrated enhanced efficacy. These findings suggest the
potential for utilizing cold atmospheric plasma and controlled ions as an
adjunctive treatment approach in lymphosarcoma therapy. Further research is
warranted to explore the mechanisms underlying these effects and to evaluate
the safety and efficacy in human clinical trials. lymphosarcoma, plasma therapy, cold atmospheric plasma (CAP), composite ions,
ion emitter 01 Frontiers in Oncology frontiersin.org Kniazeva et al. 10.3389/fonc.2023.1171042 1 Introduction Immune system activation: CAP has been shown to
stimulate the immune system by increasing the production of
immune cells and enhancing their activity. This immune response
can help in recognizing and eliminating cancer cells (12, 19). Anti-
angiogenic effects: CAP can inhibit the formation of new blood
vessels (angiogenesis) that are crucial for tumor growth and
metastasis. By disrupting the tumor’s blood supply, CAP can
hinder its progression (20). Unlike traditional anti-cancer
approaches and drugs, CAP is a selective anti-cancer treatment The aim of the study was to evaluate the effectiveness of using
the modified cold plasma emitter PERENIO IONIC SHIELD ™in
the complex therapy of lymphosarcoma. 2 Materials and methods The experimental study was carried out at the laboratory of
photodynamic therapy and hyperthermia with a chemotherapy
group of the N.N. Alexandrov National Cancer Center of Belarus
(Lesnoy, Republic of Belarus). Frontiers in Oncology 1 Introduction method. More research is needed to fully understand its
mechanisms of action, optimize treatment protocols, and evaluate
its safety and efficacy in clinical settings. Oncological diseases pose a serious challenge in the modern
world, ranking second in annual mortality after cardiovascular
disease (1). The treatment of malignant tumors is further
complicated by the increasing number of patients diagnosed with
advanced forms of cancer. Traditional treatment methods for
malignant tumors often yield unsatisfactory results. As a result,
alongside the improvement of existing treatment methods, there is a
constant search for new approaches, one of which involves combining
known chemotherapy drugs with physical and chemical methods (2–
6). In the past decade, cold atmospheric plasma (CAP), an ionized gas
with temperatures similar to room temperature, has shown promising
potential in cancer therapy (7). CAP is generated by adding energy to a
gas, causing the ionization and excitation of gas molecules. Plasma is
formed when the atoms of a substance lose or gain electrons due to
external influences. As a result, the balance of charges is disrupted, and
the once electrically neutral atom transforms into a positively or
negatively charged ion. Plasma displays the characteristics of both a
gas and a liquid, and under certain conditions, it can even exhibit the
properties of a solid despite maintaining a crystalline structure (8). There are different types of devices described for CAP
production, such as plasma jet (21–29), dielectric barrier discharge
(DBD) (30–32), floating-electrode dielectric barrier discharge (FE-
DBD) (33, 34), atmospheric pressure glow discharge torch (APGD-t)
(35, 36), microhollow cathode discharge air plasma jet (37),
microwave plasma torch (38), and nanosecond plasma gun (39, 40). The clinically used and experimentally tested CAP devices are
divided into three main categories: a) direct-discharge, b) indirect-
discharge, and c) hybrid types. CAP dosage in clinical practice must
be closely controlled, and this control depends on the treatment
type (41). There are different types of devices described for CAP
production, such as plasma jet (21–29), dielectric barrier discharge
(DBD) (30–32), floating-electrode dielectric barrier discharge (FE-
DBD) (33, 34), atmospheric pressure glow discharge torch (APGD-t)
(35, 36), microhollow cathode discharge air plasma jet (37),
microwave plasma torch (38), and nanosecond plasma gun (39, 40). The clinically used and experimentally tested CAP devices are
divided into three main categories: a) direct-discharge, b) indirect-
discharge, and c) hybrid types. 2.3 Ethical aspects The experimental studies were conducted in accordance with
the international legislation and the regulatory acts in force in the
Republic of Belarus on conducting experimental studies with
laboratory animals, namely: 1. The European Convention for the Protection of Vertebrate
Animals used for Experimental and other Scientific Purposes
(Strasbourg, France, dated 18.03.1986), as amended in accordance
with the provisions of the Protocol (SED No. 170 of 02.12.2005). 1. The European Convention for the Protection of Vertebrate
Animals used for Experimental and other Scientific Purposes
(Strasbourg, France, dated 18.03.1986), as amended in accordance
with the provisions of the Protocol (SED No. 170 of 02.12.2005). 2.1 Chemicals Doxorubicin hydrochloride (Dox) was obtained from RUE
«Belmedpreparaty» (Minsk, Belarus). Dox was administered once
intraperitoneally at a dose of 5 mg/kg on the 5th day after
tumor transplantation. Hanks’ Balanced Salt Solution (HBSS) was acquired from LT
Biotech Ltd. (Vilnius, Lithuania). The saline (0.9% NaCl solution) was prepared in the
laboratory. Research manuscripts reporting large datasets that
are deposited in a publicly available database should specify
where the data have been deposited and provide the relevant
accession numbers. 02 Frontiers in Oncology frontiersin.org Kniazeva et al. 10.3389/fonc.2023.1171042 2.4 Tumor xenograft models The Pliss lymphosarcoma (PLS) (Vasiliev et al., 2009) strain was
obtained from the Russian Cell Culture Collection, Institute of
Cytology of the Russian Academy of Sciences (St. Petersburg,
Russia). The tumor model in laboratory animals was created by
subcutaneous passivation in vivo. Subcutaneous grafting included
the introduction of 0.5 ml of a 10% suspension of tumor cells in a
20% Hanks’ solution subcutaneously in the left inguinal region. PLS
is one of the rapidly growing tumors with a short latent period of,
on average, 5–7 days. In this connection, rats with tumors were
included in the experiment on the 5th day after inoculation, when
the diameter of the tumor node reached, on average, 3–5 mm. 2.2 Animal models with the observance of humane methods of laboratory animal
treatment in compliance with the AVMA Guidelines for the
Euthanasia of Animals (Hubrecht and Carter, 2019; AVMA
Guidelines for the Euthanasia of Animals, 2020; Directive 2010/
63/EU on the protection of animals used for scientific
purposes, 2010). The pilot study was performed on 50 white mongrel rats of both
sexes obtained from the vivarium of the N.N. Alexandrov National
Cancer Center of Belarus, with a body weight of 250 ± 50 g, aged
2.5–3 months. The duration of quarantine before inclusion in the
experiment was 14 days. Laboratory animals were kept in standard
conditions in terms of food and drinking rations, with a 12-hour
lighting mode, at a temperature of 20–22°C and a humidity of 50%–
60%, in cages with five individuals in each. They were kept under
standard conditions with food and water ad libitum. The indicators
of humidity, temperature, and illumination in the room complied
with the current sanitary rules for the devices, equipment, and
maintenance of vivariums. 2.5 Animal treatment Prior to the treatment, the animals were randomized into eight
experimental groups, with six to seven animals in each
group (Table 1). 2. Directive 2010/63/EU of the European Parliament and of the
European Union on the protection of animals used for scientific
purposes (dated 22.09.2010). 3. TPC 125-2008 «Good Laboratory Practice» (GLP)
(Resolution of the Ministry of Health of the Republic of Belarus
No. 56 of 28.03.2008). TABLE 1 Experimental study design. Frontiers in Oncology 2.6 Chemotherapy Dox (5 mg/kg of animal body weight) in the form of a solution
was administered once at the start of treatment by i.p. injection on
the 5th day after tumor transplantation. The choice of the
doxorubicin dose in our study was based on previous
experiments conducted in a similar experimental model with
rats, which demonstrated relative safety at a single dose of 5
mg/kg (48, 49). All the experiments were carried out as per the guidelines of the
institutional animal ethics committee of the N.N. Alexandrov
National Cancer Center of Belarus and had prior approval from
the same committee (approval 173 dated 07.05.2021). When animals showed terminal signs or after the end of the
observation period for laboratory animals, they were put to death
with generally accepted methods of euthanasia (aether pro narcosi) №. Therapeutic
parameters
Number of animals in groups, n
1
intact control (IK)
7
2
DOX 5 mg/kg
6
3
PEWOW01 3 days
6
4
PEWOW01 7 days
6
5
PEWOW01 14 days
6
6
DOX 5 mg/kg + PEWOW01 3 days
6
7
DOX 5 мг/кг + PEWOW01 7 days
6
8
DOX 5 mg/kg + PEWOW01 14 days
7 №. Therapeutic
parameters
Number of animals in groups, n
1
intact control (IK)
7
2
DOX 5 mg/kg
6
3
PEWOW01 3 days
6
4
PEWOW01 7 days
6
5
PEWOW01 14 days
6
6
DOX 5 mg/kg + PEWOW01 3 days
6
7
DOX 5 мг/кг + PEWOW01 7 days
6
8
DOX 5 mg/kg + PEWOW01 14 days
7 03 Frontiers in Oncology frontiersin.org 10.3389/fonc.2023.1171042 Kniazeva et al. Kniazeva et al. V = 1
6 p d1 d2 d3
(1) 2.7 Tumor characterization parameters Vexperimental group—the average tumor volume in the main
group (cm3). The antitumor activity of the treatment was assessed using the
following parameters: The minimum significant criterion demonstrating the
effectiveness of tumor treatment was considered when ITG >50%. Dynamic changes in tumor volume over the whole course of
treatment. Three mutually perpendicular tumor diameters (d1, d2,
and d3) were measured every 2–3 days, and tumor volume (V, cm3/
mm3) was calculated according to the formula: The frequency of complete regressions (CR) was estimated 60
days after the end of the exposure in the absence of visual and
palpatory signs of tumor growth and was calculated by the formula: Frequency of CR = The number of animals without signs of tumor growth
Total number of animals in the group
100%
(4) FIGURE 1
Comparison of composite cold atmospheric plasma therapy with
different durations (3, 7, and 14 days) and intact control. Frontiers in Oncology
0 (4) FIGURE 1
Comparison of composite cold atmospheric plasma therapy with
different durations (3, 7, and 14 days) and intact control. FIGURE 1
Comparison of composite cold atmospheric plasma therapy with
different durations (3, 7, and 14 days) and intact control. The quantitative criteria for assessing the inhibitory effect on
transplantable tumors in laboratory animals were as follows
(Table 2) (2). 2.6.1 PERENIO IONIC SHIELD™(PEWOW01)
treatment session (1) The Ion Emitter PERENIO IONIC SHIELD™consists of two
major components: where where
d1,2,3—three mutually perpendicular diameters of the
tumor (cm); 1. An electric multiplier with an input of 5 V and up to 2 A
and an output of 14 kV is applied to the anode and cathode
to emit ions of the elements into surrounding air. p/6 = 0.52—constant value; p/6 = 0.52—constant value; V—the tumor volume (cm3). to emit ions of the elements into surrounding air. The coefficient of the absolute tumor growth (K, in relative units
(RU) was calculated bythe formula: 2. Replaceable capsule PEWOW01 with two trays inside, filled
with a stable ionic solution in the porous polymer structure
based on salts of metals Mg2+, K+, Au+, Pt+, and Zn+, and
liquid acrylic polymer is part of the solution. The ionic
solution is stable, does not enter a chemical reaction with
water or air, and does not evaporate under natural
conditions. Based on fundamental knowledge of ionic
channel regulations and biochemical pathways in cancer
cells and literature reviews, a specific formula was created. K = Vt −V0
V0
(2) (2) where: V0—the initial tumor volume (before the start of exposure); V0—the initial tumor volume (before the start of exposure); Vt—the volume of the tumor for a certain period
of observation. At K >0 (V at the corresponding follow-up period exceeded its
initial value), it was regarded as continued tumor growth. The concentration of the ions is up to 15,000 per cubic
centimeter at a distance of 50 cm from the device. At −1 < К < 0 (V at the corresponding observation period was
less than its initial value) as inhibition of tumor growth. Patent application details: Number 500297608, Real Scientists
Limited, Great Britain, Ref. GP16155NLPD. When K = −1, it was regarded as complete regression (CR) of
the tumor. The Ion Emitter in the amount of four units, was set up at a
distance of 50 cm from the cages with laboratory animals. The
devices were operated for 3, 7, and 14 days with a 24-hour mode of
active operation. During the entire period of the experiment, the
operation of the devices was monitored (Figure 1). The coefficient of inhibition of tumor growth (ITG) was
calculated using the formula: ITG = Vcontrol group – Vexperimental group
Vcontrol group
100%
(3) (3) where Vcontrol group—the average tumor volume in the control
group (cm3); 2.8 Statistical analysis Statistical processing of experimental data was carried out
using the software packages Excel, Origin Pro (version 7.0), and
SPSS (version 10.0). Values were expressed as M ± m (mean ±
mean error). To assess the significance of differences, we used the
Mann–Whitney U test. Comparative analysis of survival data
was performed using a nonparametric log-rank test. Differences
were considered statistically significant at a significance level
of p<0.05. 04 Frontiers in Oncology frontiersin.org 10.3389/fonc.2023.1171042 Kniazeva et al. TABLE 2 Criteria for evaluation of tumor growth inhibitory effects (2). durations, including 3 days, 7 days, and 14 days, compared to the
intact control group. Each treatment duration is represented by a
separate bar, and the mean values for tumor growth are indicated by
the height of each bar. The error bars represent the standard
deviation of the measurements. The upper interval of the
standard deviation is also shown, providing an upper bound for
the variation in tumor growth within each treatment duration
group. This interval provides insight into the range of possible
outcomes within each treatment duration. durations, including 3 days, 7 days, and 14 days, compared to the
intact control group. Each treatment duration is represented by a
separate bar, and the mean values for tumor growth are indicated by
the height of each bar. The error bars represent the standard
deviation of the measurements. The upper interval of the
standard deviation is also shown, providing an upper bound for
the variation in tumor growth within each treatment duration
group. This interval provides insight into the range of possible
outcomes within each treatment duration. Criteria efficiency
The values efficiency
ITG<20%
0
ITG<20–50%
±
ITG<51–80%
+
ITG<81–90%
++
ITG<91–100% +<50% CR
+++
ITG<91–100% + >50% CR
++++
ITG < 20%: This range is associated with an efficiency value of 0, meaning there is no observed
effect. ITG < 20-50%: The efficiency level is denoted as ±, indicating a moderate or uncertain effect. ITG < 51-80%: An efficiency level of + suggests a mild or positive effect. TG < 81-90%: The efficiency level ++ indicates a moderate or notable positive effect. ITG < 91-100% (with CR < 50%): This combination of high ITG values and a low complete
response (CR) rate is denoted as +++, representing a substantial positive effect. 2.8 Statistical analysis ITG < 91-100% (with CR > 50%): In this case, the combination of high ITG values and a high
CR rate is denoted as ++++, signifying an extremely high or significant positive effect. The most effective treatment approach was a combined therapy
that included chemotherapy (DOX) and CAP therapy with
PEWOW01 for 14 days, significantly more effective than each of
its components (combined therapy −mean volume = 16.87 ± 4.78
cm3; T/C = 842.50; TGI = 79.53% vs. “DOX” −mean volume =
43.01 ± 8.42 cm3; T/C = 1,432.67; TGI = 47.82% (p = 0.025) and vs. “PEWOW01 14 days” −mean volume = 43.09 ± 3.33 cm3; T/C =
2,153.50; TGI = 47.73% (p = 0.0066) (Figure 2). Figure 2 illustrates a comparison between composite cold
atmospheric plasma therapy with different durations (3, 7, and 14
days) when used in combination with chemotherapy and an intact
control group. The lines present the effects of these treatment
combinations on rats. The error bars represent the standard
deviation of the measurements. The upper interval of the
standard deviation is also shown, providing an upper bound for
the variation in tumor growth within each treatment duration
group. This interval provides insight into the range of tumor
volume within each treatment duration. 3 Results 10.3389/fonc.2023.1171042 FIGURE 3
Comparison of composite cold atmospheric plasma therapy with different durations (3, 7, and 14 days) in combination with chemotherapy versus an
intact control. FIGURE 3
Comparison of composite cold atmospheric plasma therapy with different durations (3, 7, and 14 days) in combination with chemotherapy versus an
intact control FIGURE 3
Comparison of composite cold atmospheric plasma therapy with different durations (3, 7, and 14 days) in combination with chemo
intact control. composite cold atmospheric plasma therapy with different durations (3, 7, and 14 days) in combination with chemotherapy versus an Table 3 shows descriptive statistics for all the studied groups at
the end of treatment (14 days after the start of treatment). Table 3 shows descriptive statistics for all the studied groups at
the end of treatment (14 days after the start of treatment). The blue dots represent the days of observation. Overall, this
diagram provides a visual representation of the spread, central
tendency, and potential outliers within the dataset, allowing for a
quick and easy analysis of the data distribution. In terms of the
combination therapy of PEWOW01 and DOX, the box plot for this
group shows a significantly lower median tumor volume compared
to the chemotherapy and untreated control groups, suggesting that
the combination therapy is more effective in reducing tumor growth
than both chemotherapy and no treatment. Regarding
monotherapy with PEWOW01, the box plot for this group
exhibits a notably lower median tumor volume compared to the
untreated control group, which indicates that monotherapy with
PEWOW01 is more effective in reducing tumor growth than
no treatment. Since the data on tumor volumes in each group for the whole
period of treatment had a normal distribution, the Duncan test was
chosen for group comparison. The levels of statistical significance of
pairwise comparison (Duncan test) of all eight groups at the end of
treatment are presented in Supplementary Material. According to
the test, all seven treatment options resulted in statistically
significant (p ≤0.001) tumor growth inhibition compared to the
control. A statistically significant difference in the antitumor effect
of CAP and its combinations with Dox was observed on the 2nd day
of treatment. CAP treatment as a monotherapy decreased tumor
volume by 1.5–2.4 times compared to the control at the end of
treatment. According to the antitumor effect, the studied CAP
exposures statistically did not differ significantly from each other
or the Dox used alone. 3 Results The antitumor efficacy of various durations of cold plasma
explosion (72 h, 7 days, and 14 days), Dox (5 mg/kg), and their
combinations was studied in the PLS model in rats. The rats were
divided into eight groups: three groups were treated with CAP for 72 h,
7 days, and 14 days; another three groups had a combination of DOX
drug (one injection) and CAP for 72 h, 7 days, and 14 days; one group
received one injection of DOX; and the control group had no therapy. CAP treatment was started 5 days after PLS tumor implantation for the
groups with plasma therapy and followed up to 14 days. Figure 3 depicts a box plot with whiskers showcasing the mean
tumor volume across various groups undergoing different therapy
schemes, including both monotherapy and combination therapy
with PEWOW01. The horizontal axis represents the different
treatment groups, while the vertical axis corresponds to tumor
volume. The boxes in the plot represent the interquartile range, with
the lower and upper quartiles marked by the lower and upper
boundaries of the boxes, respectively. The line inside each box
represents the median tumor volume. The whiskers extending from
the boxes illustrate the range of tumor volumes observed within
each treatment group, excluding any outliers. In this diagram, the
whiskers specifically display the upper interval of the standard
deviation observed in the experiment, indicating the upper limit
of tumor volume variability within each treatment duration group. Mono CAP therapy with PEWOW01 with a duration of 3, 7,
and 14 days showed moderate antitumor efficacy compared to the
intact control group (without intervention; mean volume = 82.43 ±
9.12 cm3; K = 2,746.67). Growth inhibition of transplanted tumors
was observed: for 3 days, the mean volume was 34.12 ± 9.43 cm3; K
= 1136.33; ITG = 58.61% (p = 0.0016); for 7 days, the mean volume
was 53.79 ± 8.92 cm3; K = 2,688.50; ITG = 34.74% (p = 0.0039); for
14 days, the mean volume was 43.09 ± 3.33 cm3; T/C = 2,153.50;
ITG = 47.73% (p = 0.0039) (Figure 1). Figure 1 displays the effects of composite cold atmospheric
plasma therapy on tumor growth over different treatment FIGURE 2
Comparison of composite cold atmospheric plasma therapy with different durations (3, 7, and 14 days) in combination with chemotherapy versus
intact control. 05 05 Frontiers in Oncology frontiersin.org Kniazeva et al. Frontiers in Oncology TABLE 3 Descriptive statistics for the studied groups of animals according to values on the volume of PLS tumors after 14 days from the start of
t
t
t a. Uses Harmonic Mean Sample Size = 6.109.
b. The group sizes are unequal. The harmonic mean of the group sizes is used. Type I error levels are not guaranteed. 4 Discussion In 2012, the World Health Organization estimated the
emergence of more than 14 million new cases of cancer in the
world, while in 2020, more than 18 million new cases were
diagnosed (50). The search for new, effective, low-cost, less-
invasive anticancer treatments with fewer side effects is therefore
in rapid expansion (51). Most CAP devices are used for cancer therapy based on plasma
jet technology. An essential increase in ROS in mouse tumors was
registered in vivo after plasma jet treatment. On the cellular level,
ROS and RNS have various effects on signal transduction; their
excess can result in oxidative damage, cell death, and various
diseases (57–59). In the study we have shown that composite cold atmospheric
plasma has promising potential as a selective anticancer agent,
especially in combination with standard chemotherapy. The growth
of experimental tumors was significantly slowed down with a
duration of explosion for composite CAP (PEWOW01) of up to
14 days. A longer exposure to CAP for 14 days in combination with
DOX chemotherapy caused a slowdown in the growth of the tumor
and increased the frequency of complete regressions at 60 days of
the experiment in comparison to standard monochemotherapy
with DOX. Our approach was based on composite cold
atmospheric plasma produced from a specific ionic gel that
contains complex salts in a chosen combination. The composition
was chosen based on empirical data and literature reviews on non-
organic ions and their ability to interact with cancer cell pathways. Another pioneering approach in the study was to use a non-
invasive, direct, long-term method of plasma application to
reduce side effects and stress in animals. CAP seems to make cancer cells resistant to current treatments. The mechanisms involved seem to depend, inter alia, on p53, NF-
kB, JNK, or caspase pathways (60–63). The effects of electrical fields on cancer cells have been studied
in vitro, in vivo, and in patients. The electrical field causes apoptosis
in cancer cells, inhibits tumor growth, and improves the survival
rate of patients with glioblastoma (64). Moreover, Janigro and
colleagues demonstrated that the treatment of neoplasms by
coupling chemotherapy and electric stimulation improved
therapeutic efficiency, allowing dose reduction of chemotherapy
drugs by inhibiting multidrug resistance pumps (MDR pumps)
(65). Vijayarangan and colleagues concluded that the increased
delivery efficiency of the molecule was related to membrane
permeability resulting from the combined action of the RONS
and the electric field (66). 3 Results When compared with the group treated with
Dox alone, statistically significant differences in tumor volume
values occurred in the control and Dox + PEWOW01 (14 days)
groups. By the end of therapy, the combination of Dox and Overall, among the various treatment schemes, the combination
of cytostatic Dox and CAP with a 14-day plasma exposition
demonstrated the most pronounced antitumor effect. This
suggests that the combined use of Dox and CAP for a longer
duration is highly effective at inhibiting tumor growth. TABLE 3 Descriptive statistics for the studied groups of animals according to values on the volume of PLS tumors after 14 days from the start of
treatment. Test
Treatments
N
Subset for alpha = 0.05
1
2
3
4
Duncana,b
DOX + PEWOW01 (14 days)
7
16.8689
DOX + PEWOW01 (7 days)
6
27.9662
27.9662
PEWOW01 (72 h)
6
34.1240
34.1240
34.1240
DOX + PEWOW01 (3 days)
6
37.5180
37.5180
37.5180
DOX
6
43.0133
43.0133
PEWOW01 (14 days)
6
43.1000
43.1000
PEWOW01 (7 days)
6
53.7873
Intact control
6
82.4250
Sig. .075
.202
.097
1.000
Posthoc = duncan lsd alpha (0.05). Dox, doxorubicine; PLS, Pliss lymphasarcoma; Sig., significance. Means for groups in homogeneous subsets are displayed. a. Uses Harmonic Mean Sample Size = 6.109. b. The group sizes are unequal. The harmonic mean of the group sizes is used. Type I error levels are not guaranteed. 06 frontiersin.org Frontiers in Oncology Kniazeva et al. 10.3389/fonc.2023.1171042 PEWOW01 (14 days) reduced the volume of the PLS tumor by 4.9
times compared with the control (p<0.001) and by 2.5 times
compared with Dox used alone (p = 0.013). pathways and cellular processes. This targeted approach holds
potential for manipulating biochemical reactions within cells,
either increasing or decreasing their activity as needed. The IC for the combined use of Dox + PEWOW01 (14 days) at
the end of treatment was the highest, equal to 79.53%. It differed
significantly (p = 0.15, Student’s t-test) from the IC of Dox equal to
47.82%, which did not surpass the threshold value (Table 4). The ability to modulate biochemical reactions at such a fine
level opens possibilities for therapeutic interventions in various
diseases, including cancer, where specific molecular pathways are
often dysregulated. By influencing these pathways, it may be
possible to restore normal cellular function or disrupt the growth
and survival of cancer cells. 3 Results Among all studied treatment options, the combined use of Dox
+ PEWOW01 (14 days) demonstrated the highest CR value, equal
to 16.7% 60 days following the start of therapy (Table 5). However, it is important to note that while this approach shows
promise, further research is necessary to fully understand the
mechanisms involved and validate its effectiveness and safety in
different disease contexts. Compared with other free radical-
generating therapies, CAP delivers a mild dose of reactive species
and may provide a safer yet potentially slower alternative for cancer
patients unable to tolerate the extensive adverse effects associated
with conventional approaches. It is worth noting that in cases where
an ultimate cure is impossible and prolonging life span and
enhancing life quality are the primary goals, CAP may be a
particularly useful treatment approach. Synergies between CAP
and traditional therapies such as chemotherapy (54) or novel
techniques such as iron oxide-based magnetic nanoparticles (55)
may increase its efficacy in cancer management (56). The study showed promising effects of CAP in lymphoma
therapy, both in mono mode and in combination with
chemotherapy. According to different criteria evaluating the
therapeutic effect, the combined treatment with Dox +
PEWOW01 (14 days) demonstrated the most promising
antitumor activity in the animal model of PLS. The use of
composite cold plasma for 14 days resulted in a significant
enhancement of Dox cytostatic activity. Frontiers in Oncology frontiersin.org 4 Discussion The delay between each pulse of the
electric field plays a key role in permeability. Plasma-induced
chemical species and electric fields make CAP an interesting tool
for optimizing drug delivery (50). The mechanism of CAP interaction with tumors is poorly
understood. Plasma contains a cocktail of different radicals and
ions, reactive oxygen and nitrogen species, and ultraviolet, which
can react with irradiated tissue, as well as an electric field (9, 52, 53). The mechanism of CAP interaction with tumors is poorly
understood. Plasma contains a cocktail of different radicals and
ions, reactive oxygen and nitrogen species, and ultraviolet, which
can react with irradiated tissue, as well as an electric field (9, 52, 53). A new approach to generating controlled plasmas with different
ionic compounds with the Ion Emitter PERENIO IONIC SHIELD™
allows targeting disease pathways at the subcellular and molecular
levels and up or downregulating biochemical reactions. According to the available literature, key features and proposed
benefits of CAP devices in cancer treatment are: selective cancer cell
killing: CAP devices generate a plasma state that consists of various
reactive oxygen and nitrogen species, electric fields, and UV
radiation. This plasma can selectively target and induce apoptosis
(programmed cell death) in cancer cells while sparing healthy
surrounding tissue. Normal cells tend to be more resistant to the
effects of CAP, making it a potentially targeted therapy. Non-
thermal nature: CAP operates at near-room temperature, unlike A new approach to generating controlled plasmas with different
ionic compounds with the Ion Emitter PERENIO IONIC SHIELD™
allows targeting disease pathways at the subcellular and molecular
levels and up or downregulating biochemical reactions. By precisely controlling the composition of the plasma,
including the types and concentrations of ions and other reactive
species, it becomes possible to tailor treatment to specific disease 07 frontiersin.org frontiersin.org 10.3389/fonc.2023.1171042 Kniazeva et al. Levene's
Test for
Equality of
Variances
t-test for Equality of Means
F
Sig. t
df
Sig. (2-tailed)
Mean
Difference
Std. Error Difference
95% Confidence
Interval of the
Difference
Lower
Upper
Day 14
Equal variances assumed
2.024
.185
1.556
10
.151
15.04717
9.67135
-6.50194
36.59627
Equal variances not assumed
1.556
7.909
.159
15.04717
9.67135
-7.29991
37.39424 t-test for Equality of Means The present study highlights the importance of considering the
variations in treatment response among different types of cancer,
including cold plasma therapy. 4 Discussion Our in vitro experiments have
demonstrated the efficacy of the composite cold plasma against
various cancer cell lines, supporting its potential as a promising
treatment strategy (unpublished data). However, our in vivo study
focused solely on Pliss lymphosarcoma, serving as a representative
cancer model, emphasizing the need for further research to evaluate
the response and efficacy of cold plasma treatment in other cancer
types. The intricate nature of cancer involves numerous factors such
as tumor microenvironment, genetic alterations, and signaling
pathways, all of which can contribute to the observed variations
in treatment response. In this regard, our ongoing research aims to
investigate the specific pathways that are upregulated by composite
cold plasma in lymphosarcoma therapy, as compared to other
cancers. Preliminary findings suggest the involvement of immune
modulation, apoptosis, and oxidative stress response pathways. Nonetheless, additional studies are necessary to fully elucidate
these pathways and comprehend their significance in the context
of different cancer types. traditional thermal plasma. This non-thermal characteristic reduces
the risk of damaging healthy tissues and makes it a more feasible
option for biomedical applications (67). Indirect effects: CAP
treatment not only directly affects cancer cells but also triggers a
cascade of secondary effects. These include modulation of the tumor
microenvironment, immune response activation, and angiogenesis
inhibition, which collectively contribute to the anticancer effects. Potential for combination therapy: CAP can be used in combination
with other treatment modalities, such as chemotherapy or
radiotherapy, to enhance their effectiveness. It has been shown to
sensitize cancer cells to radiation and increase the uptake and
efficacy of certain chemotherapeutic agents. Minimal resistance
development: Unlike some conventional cancer treatments, CAP
has shown a low tendency for cancer cells to develop resistance. This may be due to the multiple mechanisms of action involved in
CAP-induced cell death, making it a potentially valuable option for
addressing drug-resistant cancers. Promising preclinical results:
Numerous preclinical studies have demonstrated the efficacy of
CAP in reducing tumor growth, inhibiting metastasis, and
improving overall survival in animal models. Additionally, several
clinical trials are underway to evaluate the safety and effectiveness of
CAP in human patients across different cancer types. Future studies are needed to investigate the anticancer activity
of composite cold atmospheric plasma (CAP), and these studies TABLE 5 Complete tumor regression (CR) frequencies calculated 11 and 60 days after the start of treatment. regression (CR) frequencies calculated 11 and 60 days after the start of treatment. 4 Discussion TABLE 5 Complete tumor regression (CR) frequencies calculated 11 and 60 days after the start of treatment. Group
Number of animals in group
Number
of complete
regressions
(30 days), abs
CR, 30 days, %
Number
of complete regressions
(60 days), abs
CR
(60 days), %
Control
7
0
0
0
0
Dox
6
1
16.67
1
16.67
PEWOW01(72 h)
6
1
16.67
0
0
PEWOW01 (7 days)
6
0
0
0
0
PEWOW01 (14 days)
6
0
0
0
0
Dox + PEWOW01 (72 h)
6
0
0
0
0
Dox + PEWOW01 (7 days)
6
0
0
0
0
Dox + PEWOW01 (14
Days)
7
1
14.29
1
14.29 TABLE 5 Complete tumor regression (CR) frequencies calculated 11 and 60 days after the start of treatment. Group
Number of animals in group
Number
of complete
regressions
(30 days), abs
CR, 30 days, %
Number
of complete regressions
(60 days), abs
CR
(60 days), %
Control
7
0
0
0
0
Dox
6
1
16.67
1
16.67
PEWOW01(72 h)
6
1
16.67
0
0
PEWOW01 (7 days)
6
0
0
0
0
PEWOW01 (14 days)
6
0
0
0
0
Dox + PEWOW01 (72 h)
6
0
0
0
0
Dox + PEWOW01 (7 days)
6
0
0
0
0
Dox + PEWOW01 (14
Days)
7
1
14.29
1
14.29 08 08 Frontiers in Oncology frontiersin.org Kniazeva et al. 10.3389/fonc.2023.1171042 Data availability statement The raw data supporting the conclusions of this article will be
made available by the authors, without undue reservation. Publisher’s note All claims expressed in this article are solely those of the authors
and do not necessarily represent those of their affiliated organizations,
or those of the publisher, the editors and the reviewers. Any product
that may be evaluated in this article, or claim that may be made by its
manufacturer, is not guaranteed or endorsed by the publisher. Conflict of interest Authors VK, AK, ER, and SK was employed by the company R. S. C. Real Scientists Cyprus Ltd. These findings open a novel research direction, which
complements the current focus on the selectivity of composite
CAP in cancer treatment and its high efficiency as adjuvant
therapy. Future studies could explore the specific mechanisms
behind the observed benefits and further investigate the potential
applications of plasma therapy for various cancer types. The remaining authors declare that the research was conducted
in the absence of any commercial or financial relationships that
could be construed as a potential conflict of interest. Funding This research was funded by public funds. Furthermore, CAP can serve as a valuable non-invasive and safe
tool for optimizing drug delivery in cancer treatment. This suggests
that plasma therapy could be utilized to enhance the delivery of
therapeutic drugs, potentially improving their efficacy and reducing
side effects. Author contributions Conceptualization, AK and SK. Methodology, AK, SK, and DT. Software, ER. Validation, DT, SK, and VK. Formal analysis, VK,
DT, and ÖB. Investigation, DT. Resources, DT and ER. Data
curation, VK and DT. Writing—original draft preparation, VK,
DT, and ÖB. Writing—review and editing, VK, DT, ÖB, AK, and
SK. Visualization, VK and DT. Supervision, SK. Project
administration, VK and DT. All authors contributed to the article
and approved the submitted version. 5 Conclusions The study demonstrated that the composite CAP therapy has
the potential to be an effective tool in the treatment of lymphoma
cancer. When used in combination with traditional chemotherapy,
it was found to significantly increase tumor growth inhibition by 4.9
times and the frequency of complete regressions after 60 days
by 14.5%. The use of cold plasma in cancer therapy offers promising
advantages, such as reducing the toxicity associated with standard
therapies and enhancing their effectiveness. By expanding new
horizons in cancer treatment, plasma therapy holds the potential
to revolutionize the field by improving patient outcomes. Ethics statement should incorporate a comprehensive analysis of surrogate markers,
including immunohistochemistry (IHC), to provide further
confirmation of our findings. Regarding future steps in plasma
cancer therapy, we think it might be important to continue
combining plasma treatment with already existing treatment
protocols in oncology. This way, CAP might become an adjuvant
in addition to clinically accredited procedures promoting the
efficiency of the therapy. The animal study was reviewed and approved by The
Institutional Ethics Committee of the N.N. Alexandrov National
Cancer Center of Belarus. 7. Yan D, Sherman JH, Keidar M. Cold atmospheric plasma, a novel promising anti-
cancer treatment modality. Oncotarget (2017) 8(9):15977–95. doi: 10.18632/
oncotarget.13304 Frontiers in Oncology 8. Domonkos M, Tichá P, Trejbal J, Demo P. Applications of cold atmospheric
pressure plasma technology in medicine, agriculture and food industry. Appl Sci (2021)
11:4809. doi: 10.3390/app11114809 4. Olsson H. Curative therapies in cancer: a perspective. Trends Res (2019) 2:1–2.
doi: 10.15761/TR.10001141 3. Hanahan D. Rethinking the war on cancer. Lancet (2014) 383(9916):558–63.
doi: 10.1016/S0140-6736(13)62226-6 2. Cheng Z, Li M, Dey R, Chen Y. Nanomaterials for cancer therapy: current
progress and perspectives. J Hematol Oncol (2021) 14(1):85. doi: 10.1186/s13045-021-
01096-0 7. Yan D, Sherman JH, Keidar M. Cold atmospheric plasma, a novel promising anti-
cancer treatment modality. Oncotarget (2017) 8(9):15977–95. doi: 10.18632/
oncotarget.13304
8. Domonkos M, Tichá P, Trejbal J, Demo P. Applications of cold atmospheric
pressure plasma technology in medicine, agriculture and food industry. Appl Sci (2021)
11:4809. doi: 10.3390/app11114809
9. Braný D, Dvorská D, Halašová E, Škovierová H. Cold atmospheric plasma: a
powerful tool for modern medicine. Int J Mol Sci (2020) 21(8):2932. doi: 10.3390/
ijms21082932
10. Ruchel Khanikar R, Bailung H. Cold atmospheric pressure plasma technology
for biomedical application. Plasma Sci Technol (2022). doi: 10.5772/intechopen.98895
11. Schmidt A, Bekeschus S, Jarick K, Hasse S, von Woedtke T, Wende K. Cold
physical plasma modulates p53 and mitogen-activated protein kinase signaling in
keratinocytes. Oxid Med Cell Longev (2019) 2019:7017363. doi: 10.1155/2019/7017363 9. Braný D, Dvorská D, Halašová E, Škovierová H. Cold atmospheric plasma: a
powerful tool for modern medicine. Int J Mol Sci (2020) 21(8):2932. doi: 10.3390/
ijms21082932 11. Schmidt A, Bekeschus S, Jarick K, Hasse S, von Woedtke T, Wende K. Cold
physical plasma modulates p53 and mitogen-activated protein kinase signaling in
keratinocytes. Oxid Med Cell Longev (2019) 2019:7017363. doi: 10.1155/2019/7017363 10. Ruchel Khanikar R, Bailung H. Cold atmospheric pressure plasma technology
for biomedical application. Plasma Sci Technol (2022). doi: 10.5772/intechopen.98895
11. Schmidt A, Bekeschus S, Jarick K, Hasse S, von Woedtke T, Wende K. Cold
physical plasma modulates p53 and mitogen-activated protein kinase signaling in
keratinocytes. Oxid Med Cell Longev (2019) 2019:7017363. doi: 10.1155/2019/7017363 5. Sung H, Ferlay J, Siegel RL, Laversanne M, Soerjomataram I, Jemal A, et al. A
global cancer statistics 2020: GLOBOCAN estimates of incidence and mortality
worldwide for 36 cancers in 185 countries. CA Cancer J Clin (2021) 71:209–49.
doi: 10.3322/caac.21660 10. Ruchel Khanikar R, Bailung H. Cold atmospheric pressure plasma technology
for biomedical application. Plasma Sci Technol (2022). doi: 10.5772/intechopen.98895 1. The Global Cancer Observatory (GCO). Available at: https://gco.iarc.fr. 1. The Global Cancer Observatory (GCO). Available at: https://gco.iarc.fr.
2. Cheng Z, Li M, Dey R, Chen Y. Nanomaterials for cancer therapy: current
progress and perspectives. J Hematol Oncol (2021) 14(1):85. doi: 10.1186/s13045-021-
01096-0
3. Hanahan D. Rethinking the war on cancer. Lancet (2014) 383(9916):558–63.
doi: 10.1016/S0140-6736(13)62226-6
4. Olsson H. Curative therapies in cancer: a perspective. Trends Res (2019) 2:1–2.
doi: 10.15761/TR.10001141
5. Sung H, Ferlay J, Siegel RL, Laversanne M, Soerjomataram I, Jemal A, et al. A
global cancer statistics 2020: GLOBOCAN estimates of incidence and mortality
worldwide for 36 cancers in 185 countries. CA Cancer J Clin (2021) 71:209–49.
doi: 10.3322/caac.21660
6. Falzone L, Salomone S, Libra M. Evolution of cancer pharmacological treatments at the
turn of the third millennium. Front Pharmacol (2018) 9:1300. doi: 10.3389/fphar.2018.01300 References Jo A, Bae JH, Yoon YJ, Chung TH, Lee EW, Kim YH, et al. Plasma-activated
medium induces ferroptosis by depleting FSP1 in human lung cancer cells. Cell Death
Dis (2022) 13(3):212. doi: 10.1038/s41419-022-04660-9 39. Fridman G, Shereshevsky A, Jost M, Brooks A, Fridman A, Gotsol A, et al. Floating electrode dielectric barrier discharge plasma in air promoting apoptotic
behavior in melanoma skin cancer cell lines. Plasma Chem Plasma Process (2007)
27:163–76. doi: 10.1007/s11090-007-9048-4 17. Furuta T, Shi L, Toyokuni S. Non-thermal plasma as a simple ferroptosis
inducer in cancer cells: a possible role of ferritin. Pathol Int (2018) 68(7):442–3. doi: 10.1111/pin.12665 40. Kolb JF, Price RO, Bowman A, Chiavarini RL, Stacey M, Schoenbach KH, et al. Cold atmospheric pressure air plasma jet for medical applications. Appl Phys Lett
(2008) 92:241501. doi: 10.1063/1.2940325 18. Yang X, Chen G, Yu KN, Yang M, Peng S, Ma J, et al. Cold atmospheric plasma
induces GSDME-dependent pyroptotic signaling pathway via ROS generation in tumor
cells. Cell Death Dis (2020) 11(4):295. doi: 10.1038/s41419-020-2459-3 41. Babaeva NY, Naidis GV, Tereshonok DV, et al. Production of active species in
an argon microwave plasma torch. J Phys D: Appl Phys (2018) 51. doi: 10.1088/1361-
6463/aae455.43 19. Lehmann S, Bien-Möller S, Marx S, Bekeschus S, Schroeder HWS, Mustea A,
et al. Devitalization of glioblastoma cancer cells by non-invasive physical plasma:
modulation of proliferative signalling cascades. Anticancer Res (2023) 43(1):7–18. doi: 10.21873/anticanres.16128 42. Loizou C, Kniazeva V, Apostolou T, Kornev A, Kostevitch S, Roslyakov E, et al. Effect of cold atmospheric plasma on SARS-CoV-2 inactivation: a pilot study in the
hospital environment. COVID (2022) 2:1396–404. doi: 10.3390/covid2100100 20. Haralambiev L, Neuffer O, Nitsch A, Kross NC, Bekeschus S, Hinz P, et al. Inhibition of angiogenesis by treatment with cold atmospheric plasma as a promising
therapeutic approach in oncology. Int J Mol Sci (2020) 21(19):7098. doi: 10.3390/
ijms21197098 43. Kniazeva V, Loizou C, Apostolou T, Kornev A, Kostevitch S, Roslyakov E, et al. Clinical evaluation of prevention of indoor COVID-19 transmission with cold plasma
technology. Phys Med (2022) 104:131–2. 21. Brehmer F, Haenssle HA, Daeschlein G, Ahmed R, Pfeiffer S, Görlitz A, et al. Alleviation of chronic venous leg ulcers with a hand-held dielectric barrier discharge
plasma generator (PlasmaDerm(®) VU-2010): results of a monocentric, two-armed,
open, prospective, randomized and controlled trial (NCT01415622). J Eur Acad
Dermatol Venereol (2015) 29(1):148–55. doi: 10.1111/jdv.12490 44. References Malyavko A, Yan D, Wang Q, Klein AL, Patel KC, Sherman JH, et al. Cold
atmospheric plasma cancer treatment, direct versus indirect approaches. Mater Adv
(2020) 1:1494–505. doi: 10.1039/D0MA00329H 45. Zubor P, Wang Y, Liskova A, Samec M, Koklesova L, Dankova Z, et al. Cold
atmospheric pressure plasma (CAP) as a new tool for the management of vulva cancer
and vulvar premalignant lesions in gynaecological oncology. Int J Mol Sci (2020) 21
(21):7988. doi: 10.3390/ijms21217988 22. Daeschlein G, Scholz S, Lutze S, Arnold A, von Podewils S, Kiefer T, et al. Comparison between cold plasma, electrochemotherapy and combined therapy in a
melanoma mouse model. Exp Dermatol (2013) 22(9):582–6. doi: 10.1111/exd.12201 46. Yan D, Wang Q, Yao X, Malyavko A, Keidar M. Anti-melanoma capability of
contactless cold atmospheric plasma treatment. Int J Mol Sci (2021) 22(21):11728. doi: 10.3390/ijms222111728 23. Xu D, Ning N, Xu Y, Xia W, Liu D, Chen H, et al. Effect of he plasma jet versus
surface plasma on the metabolites of acute myeloid leukemia cells. Front Oncol (2021)
11:552480. doi: 10.3389/fonc.2021.552480 47. Brullé L, Vandamme M, Riès D, Martel E, Robert E, Lerondel S, et al. Effects of a
nonthermal plasma treatment alone or in combination with gemcitabine in a MIA
PaCa2-luc orthotopic pancreatic carcinoma model. PloS One (2012) 7(12):e52653. doi: 10.1371/journal.pone.0052653 24. Thana P, Wijaikhum A, Poramapijitwat P, Kuensaen C, Meerak J,
Ngamjarurojana A, et al. A compact pulse-modulation cold air plasma jet for the
inactivation of chronic wound bacteria: development and characterization. Heliyon
(2019) 5(9):e02455. doi: 10.1016/j.heliyon.2019.e02455 48. Shadyro O, Sosnovskaya A, Edimecheva I, Ihnatovich L, Dubovik B, Krasny S,
et al. In vivo antitumoral effects of linseed oil and its combination with doxorubicin. Front Pharmacol (2022) 13:882197. doi: 10.3389/fphar.2022.882197 25. Daeschlein G, Scholz S, Emmert S, von Podewils S, Haase H, von Woedtke T,
et al. Plasma medicine in dermatology: basic antimicrobial efficacy testing as
prerequisite to clinical plasma therapy. Plasma Med (2012) 2:33–69. doi: 10.1615/
PlasmaMed.2014006217 49. Tarskikh MM. [Experimental investigation of combined action and
hematotoxicity of anti-tumor antibiotic doxorubicin and industrial monomer
acrylonitrile]. Patol Fiziol Eksp Ter (2009) 2):27–30. 26. SeyfiP, Keshavarzi M, Zahedi S, Khademi A, Ghomi H. Low-temperature argon
plasma jet with cascading electrode technique for biological applications. Sci Rep (2022)
12(1):17042. doi: 10.1038/s41598-022-21664-9 50. Dubuc A, Monsarrat P, Virard F, Merbahi N, Sarrette JP, Laurencin-Dalicieux S,
et al. Use of cold-atmospheric plasma in oncology: a concise systematic review. References Ther
Adv Med Oncol (2018) 10:1758835918786475. doi: 10.1177/1758835918786475 27. Daeschlein D, Napp M, von Podewils S, Lutze S, Emmert S, Lange A, et al. In
vitro susceptibility of multidrug resistant skin and wound pathogens against low
temperature atmospheric pressure plasma jet (APPJ) and dielectric barrier discharge
plasma (DBD). Plasma Process Polym (2014) 11:175–83. doi: 10.1002/
ppap.201300070 51. Siegel RL, Miller KD, Jemal A. Cancer statistics, 2020. CA Cancer J Clin (2020)
70(1):7–30. doi: 10.3322/caac.2159 52. Adamovich I, Agarwal S, Ahedo E, Alves LL, Baalrud S, Babaeva N, et al. The
2022 plasma roadmap: low temperature plasma science and technology 2022. J Phys D:
Appl Phys (2022) 55(37):373001. doi: 10.1088/1361-6463/ac5e1c 28. Daeschlein G, Scholz S, Lutze S, Arnold A, et al. Comparison between cold
plasma, electrochemotherapy and combined therapy in a melanoma mouse model. Exp
Dermatol (2013) 22:582–6. doi: 10.1111/exd.12201.30 53. Choi EH, Uhm HS, Kaushik NK. Plasma bioscience and its application to
medicine. AAPPS Bull (2021) 31(1):10. doi: 10.1007/s43673-021-00012-5 29. Emmert S, Brehmer F, Haünßle H, Helmke A, Mertens N, Ahmed R, et al. Atmospheric pressure plasma in dermatology: ulcus treatment and much more. Clin
Plasma Med (2013) 1:24–9. doi: 10.1016/j.cpme.2012.11.002 54. Wang S, Wang Z, Yu G, Zhou Z, Jacobson O, Liu Y, et al. Tumor-specific drug
release and reactive oxygen species generation for cancer Chemo/Chemodynamic
combination therapy. Adv Sci (2019) 6:1801986. doi: 10.1002/advs.201801986 30. Kuchenbecker M, Bibinov N, Kaemling A, Wandke D, Awakowicz P, Viöl W. Characterization of dbd plasma source for biomedical applications. J Phys D: Appl Phys
(2009) 42(4):45212. doi: 10.1088/0022-3727/42/4/045212 55. Li W, Yu H, Ding D, Chen Z, Wang Y, Wang S, et al. Cold atmospheric
plasma and iron oxide-based magnetic nanoparticles for synergetic lung cancer
therapy. Free Radic Biol Med (2019) 130:71–81. doi: 10.1016/
j.freeradbiomed.2018.10.429 31. Panngom K, Baik KY, Nam MK, Han JH, Rhim H, Choi EH. Preferential killing
of human lung cancer cell lines with mitochondrial dysfunction by nonthermal
dielectric barrier discharge plasma. Cell Death Dis (2013) 4(5):e642. doi: 10.1038/
cddis.2013.168 56. Dai X, Bazaka K, Thompson EW, Ostrikov K. Cold atmospheric plasma: a promising
controller of cancer cell states. Cancers (2020) 12:3360. doi: 10.3390/cancers12113360 32. Kaushik NK, Uhm H, Choi EH. Micronucleus formation induced by dielectric
barrier discharge plasma exposure in brain cancer cells. Appl Phys Lett (2012) 100. doi: 10.1063/1.3687172 57. Halliwell B, Whiteman M. References 2. Cheng Z, Li M, Dey R, Chen Y. Nanomaterials for cancer therapy: current
progress and perspectives. J Hematol Oncol (2021) 14(1):85. doi: 10.1186/s13045-021-
01096-0 5. Sung H, Ferlay J, Siegel RL, Laversanne M, Soerjomataram I, Jemal A, et al. A
global cancer statistics 2020: GLOBOCAN estimates of incidence and mortality
worldwide for 36 cancers in 185 countries. CA Cancer J Clin (2021) 71:209–49. doi: 10.3322/caac.21660 6. Falzone L, Salomone S, Libra M. Evolution of cancer pharmacological treatments at the
turn of the third millennium. Front Pharmacol (2018) 9:1300. doi: 10.3389/fphar.2018.01300 09 frontiersin.org Kniazeva et al. 10.3389/fonc.2023.1171042 10.3389/fonc.2023.1171042 Kniazeva et al. 12. Faramarzi F, Zafari P, Alimohammadi M, Moonesi M, Rafiei A, Bekeschus S. Cold physical plasma in cancer therapy: mechanisms, signaling, and immunity. Oxid
Med Cell Longev (2021) 2021:884887. doi: 10.3389/fmed.2022.884887 34. Sakudo A, Misawa T. Antibiotic-resistant and non-resistant bacteria display
similar susceptibility to dielectric barrier discharge plasma. Int J Mol Sci (2020) 21
(17):6326. doi: 10.3390/ijms21176326 13. Wang Y, Mang X, Li X, Cai Z, Tan F. Cold atmospheric plasma induces
apoptosis in human colon and lung cancer cells through modulating mitochondrial
pathway. Front Cell Dev Biol (2022) 10:915785. doi: 10.3389/fcell.2022.915785 35. Kaushik NK, Attri P, Kaushik N, Choi EH. A preliminary study of the effect of
dbd plasma and osmolytes on t98g brain cancer and hek non-malignant cells. Molecules
(2013) 18:4917–28. doi: 10.3390/molecules18054917 14. Golpour M, Alimohammadi M, Sohbatzadeh F, Fattahi S, Bekeschus S, Rafiei A. Cold atmospheric pressure plasma treatment combined with starvation increases
autophagy and apoptosis in melanoma in vitro and in vivo. Exp Dermatol (2022) 31
(7):1016–28. doi: 10.1111/exd.14544 36. Coulombe S, Léveillé V, Yonson S, Leask RL. Miniature atmospheric pressure
glow discharge torch (APGD-t) for local biomedical applications. Pure Appl Chem
(2006) 78:1147–56. doi: 10.1351/pac200678061147 37. Vandamme M, Robert E, Lerondel S, Sarron V, Ries D, Dozias S, et al. Ros
implication in a new antitumor strategy based on non-thermal plasma. Int J Cancer
(2012) 130(9):2185–94. doi: 10.1002/ijc.26252 15. Haralambiev L, Wien L, Gelbrich N, Lange J, Bakir S, Kramer A, et al. Cold
atmospheric plasma inhibits the growth of osteosarcoma cells by inducing apoptosis,
independent of the device used. Oncol Lett (2020) 19(1):283–90. doi: 10.3892/
ol.2019.11115 38. Leduc M, Coulombe S, Leask RL. Atmospheric pressure plasma jet deposition of
patterned polymer films for cell culture applications. IEEE Trans Plasma Sci (2009)
37:927–33. doi: 10.1109/TPS.2008.2010706 16. 62. Köritzer J, Boxhammer V, Schäfer A, Shimizu T, KlämpflTG, Li YF,
et al. Restoration of sensitivity in chemo-resistant glioma cells by cold
atmospheric plasma. PloS One (2013) 8(5):e64498. doi: 10.1371/
journal.pone.0064498 67. Schuster M, Rutkowski R, Hauschild A, Shojaei RK, von Woedtke T, Rana A,
et al. Side effects in cold plasma treatment of advanced oral cancer–clinical data and
biological interpretation. Clin Plasma Med (2018) 10:9–15. doi: 10.1016/
j.cpme.2018.04.001 63. Utsumi F, Kajiyama H, Nakamura K, Tanaka H, Mizuno M, Ishikawa K, et al.
Effect of indirect nonequilibrium atmospheric pressure plasma on anti-proliferative 61. Chang JW, Kang SU, Shin YS, Seo SJ, Kim YS, Yang SS, et al. Combination of
NTP with cetuximab inhibited invasion/migration of cetuximab-resistant OSCC
cells: involvement of NF-kB signaling. Sci Rep (2015) 5:18208. doi: 10.1038/
srep18208 60. Conway GE, Casey A, Milosavljevic V, Liu Y, Howe O, Cullen P, et al. Non-
thermal atmospheric plasma induces ROS-independent cell death in U373MG glioma
cells and augments the cytotoxicity of temozolomide. Br J Cancer (2016) 114:435–43.
doi: 10.1038/bjc.2016.12.63 59. Faramarzi F, Zafari P, Alimohammadi M, Moonesi M, Rafiei A, Bekeschus S.
Cold physical plasma in cancer therapy: mechanisms, signaling, and immunity. Oxid
Med Cell Longev (2021) 2021:9916796. doi: 10.1155/2021/9916796 References Measuring reactive species and oxidative damage in
vivo and in cell culture: how should you do it and what do the results mean? Br J
Pharmacol (2004) 142:231–55. doi: 10.1038/sj.bjp.0705776 33. Kalghatgi S, Kelly CM, Cerchar E, Torabi B, Alekseev O, Fridman A, et al. Effects
of non-thermal plasma on mammalian cells. PloS One (2011) 6(1):e16270. doi: 10.1371/
journal.pone.0016270 58. Szili EJ, Oh J-S, Fukuhara H, Bhatia R, Gaur N, Nguyen CK, et al. Modelling the
helium plasma jet delivery of reactive species into a 3D cancer tumor. Plasma Sour Sci
Technol (2017) 27:014001. doi: 10.1088/1361-6595/aa9b3b 10 Frontiers in Oncology frontiersin.org Kniazeva et al. 10.3389/fonc.2023.1171042 Kniazeva et al. activity against chronic chemo-resistant ovarian cancer cells in vitro and in vivo. PloS
One (2013) 8(12):e81576. doi: 10.1371/journal.pone.0081576 activity against chronic chemo-resistant ovarian cancer cells in vitro and in vivo. PloS
One (2013) 8(12):e81576. doi: 10.1371/journal.pone.0081576 59. Faramarzi F, Zafari P, Alimohammadi M, Moonesi M, Rafiei A, Bekeschus S. Cold physical plasma in cancer therapy: mechanisms, signaling, and immunity. Oxid
Med Cell Longev (2021) 2021:9916796. doi: 10.1155/2021/9916796 64. Kirson ED, Dbalý V, Tovarys F, Vymazal J, Soustiel JF, Itzhaki A, et al. Alternating electric fields arrest cell proliferation in animal tumor models and
human brain tumors. Proc Natl Acad Sci USA (2007) 104:10152–7. doi: 10.1073/
pnas.0702916104 64. Kirson ED, Dbalý V, Tovarys F, Vymazal J, Soustiel JF, Itzhaki A, et al. Alternating electric fields arrest cell proliferation in animal tumor models and
human brain tumors. Proc Natl Acad Sci USA (2007) 104:10152–7. doi: 10.1073/
pnas.0702916104 65. Janigro D, Perju C, Fazio V, Hallene K, Dini G, Agarwal MK, et al. Alternating
current electrical stimulation enhanced chemotherapy: a novel strategy to bypass multidrug
resistance in tumor cells. BMC Cancer (2006) 6:72. doi: 10.1186/1471-2407-6-72 65. Janigro D, Perju C, Fazio V, Hallene K, Dini G, Agarwal MK, et al. Alternating
current electrical stimulation enhanced chemotherapy: a novel strategy to bypass multidrug
resistance in tumor cells. BMC Cancer (2006) 6:72. doi: 10.1186/1471-2407-6-72 66. Vijayarangan V, Delalande A, Dozias S, Povesle JM, Pichon C, Robert E. Cold
atmospheric plasma parameters investigation for efficient drug delivery in HeLa
cells. IEEE Trans Radiat Plasma Med Sci (2017) PP:1–1. doi: 10.1109/TRPMS. 2017.2759322 66. Vijayarangan V, Delalande A, Dozias S, Povesle JM, Pichon C, Robert E. Cold
atmospheric plasma parameters investigation for efficient drug delivery in HeLa
cells. IEEE Trans Radiat Plasma Med Sci (2017) PP:1–1. doi: 10.1109/TRPMS. 66. Vijayarangan V, Delalande A, Dozias S, Povesle JM, Pichon C, Robert E. Cold
atmospheric plasma parameters investigation for efficient drug delivery in HeLa
cells. IEEE Trans Radiat Plasma Med Sci (2017) PP:1–1. doi: 10.1109/TRPMS.
2017.2759322 59. Faramarzi F, Zafari P, Alimohammadi M, Moonesi M, Rafiei A, Bekeschus S.
Cold physical plasma in cancer therapy: mechanisms, signaling, and immunity. Oxid
Med Cell Longev (2021) 2021:9916796. doi: 10.1155/2021/9916796
60. Conway GE, Casey A, Milosavljevic V, Liu Y, Howe O, Cullen P, et al. Non-
thermal atmospheric plasma induces ROS-independent cell death in U373MG glioma
cells and augments the cytotoxicity of temozolomide. Br J Cancer (2016) 114:435–43.
doi: 10.1038/bjc.2016.12.63
61. Chang JW, Kang SU, Shin YS, Seo SJ, Kim YS, Yang SS, et al. Combination of
NTP with cetuximab inhibited invasion/migration of cetuximab-resistant OSCC
cells: involvement of NF-kB signaling. Sci Rep (2015) 5:18208. doi: 10.1038/
srep18208
62. Köritzer J, Boxhammer V, Schäfer A, Shimizu T, KlämpflTG, Li YF,
et al. Restoration of sensitivity in chemo-resistant glioma cells by cold
atmospheric plasma. PloS One (2013) 8(5):e64498. doi: 10.1371/
journal.pone.0064498
63. Utsumi F, Kajiyama H, Nakamura K, Tanaka H, Mizuno M, Ishikawa K, et al.
Effect of indirect nonequilibrium atmospheric pressure plasma on anti-proliferative 65. Janigro D, Perju C, Fazio V, Hallene K, Dini G, Agarwal MK, et al. Alternating
current electrical stimulation enhanced chemotherapy: a novel strategy to bypass multidrug
resistance in tumor cells. BMC Cancer (2006) 6:72. doi: 10.1186/1471-2407-6-72 Kniazeva et al. References 2017.2759322 11 11 Frontiers in Oncology frontiersin.org
|
https://openalex.org/W2125879909
|
https://europepmc.org/articles/pmc3032666?pdf=render
|
English
| null |
Equol an isoflavonoid: potential for improved prostate health, in vitro and in vivoevidence
|
Reproductive biology and endocrinology
| 2,011
|
cc-by
| 8,596
|
© 2011 Lund et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons
Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in
any medium, provided the original work is properly cited. Open Access Open Access Trent D Lund1, Crystal Blake2, Lihong Bu3, Amy N Hamaker2, Edwin D Lephart2* Trent D Lund1, Crystal Blake2, Lihong Bu3, Amy N Hamaker2, Edwin D Lephart2* Trent D Lund1, Crystal Blake2, Lihong Bu3, Amy N Hamaker2, Edwin D Lephart2* Abstract Background: To determine: in vitro binding affinity of equol for 5alpha-dihydrotestosterone (5alpha-DHT), in vitro
effects of equol treatment in human prostate cancer (LNCap) cells, and in vivo effects of equol on rat prostate
weight and circulating levels of sex steroid hormones. Methods: First, in vitro equol binding affinity for 5alpha-DHT was determined using 14C5alpha-DHT combined with
cold 5alpha-DHT (3.0 nM in all samples). These steroids were incubated with increasing concentrations of equol (0-
2,000 nM) and analyzed by Sephadex LH-20 column chromatography. 14C5alpha-DHT peak/profiles were
determined by scintillation counting of column fractions. Using the 14C5alpha-DHT peak (0 nM equol) as a
reference standard, a binding curve was generated by quantifying shifts in the 14C5alpha-DHT peaks as equol
concentrations increased. Second, equol’s in vitro effects on LNCap cells were determined by culturing cells (48
hours) in the presence of increasing concentrations of dimethyl sulfoxide (DMSO) (vehicle-control), 5alpha-DHT,
equol or 5alpha-DHT+equol. Following culture, prostate specific antigen (PSA) levels were quantified via ELISA. Finally, the in vivo effects of equol were tested in sixteen male Long-Evans rats fed a low isoflavone diet. From 190-
215 days, animals received 0.1cc s.c. injections of either DMSO-control vehicle (n = 8) or 1.0 mg/kg (body weight)
of equol (in DMSO) (n = 8). At 215 days, body and prostate weights were recorded, trunk blood was collected and
serum assayed for luteinizing hormone (LH), 5alpha-DHT, testosterone and 17beta-estradiol levels. Results: Maximum and half maximal equol binding to 5alpha-DHT occurred at approximately 100 nM and 4.8 nM
respectively. LNCap cells cultured in the presence of 5alpha-DHT significantly increased PSA levels. However, in the
presence of 5alpha-DHT+equol, equol blocked the significant increases in PSA levels from LNCap cells. In vivo
equol treatment significantly decreased rat prostate weights and serum 5alpha-DHT levels but did not alter LH,
testosterone, and estradiol levels. Conclusions: Equol administration appears to have potential beneficial effects for prostate health and other
5alpha-DHT mediated disorders. Equol administration: reduces PSA levels from LNCap cells under 5alpha-DHT
stimulation, decreases rat prostate size, decreases serum 5alpha-DHT levels and androgen hormone action, while
not altering other circulating sex steroids or LH levels. * Correspondence: edwin_lephart@byu.edu
2The Department of Physiology and Developmental Biology and the
Neuroscience Center, Brigham Young University, Provo, Utah 84602, USA
Full list of author information is available at the end of the article Lund et al. Reproductive Biology and Endocrinology 2011, 9:4
http://www.rbej.com/content/9/1/4 Lund et al. Reproductive Biology and Endocrinology 2011, 9:4
http://www.rbej.com/content/9/1/4 Background phytoestrogens: 1) isoflavones (derived principally from
soybeans), 2) lignans (found in flaxseed in large quanti-
ties), and 3) coumestans (derived from sprouting plants
like alfalfa), human consumption of isoflavones has the
largest impact due to its availability and variety in food
products. Isoflavones have been implicated as potential
treatments for many disorders including cardiovascular
disease, osteoporosis, age-related diseases, and hor-
mone-dependent cancers [3,4]. Polyphenols are a group of chemical substances found
in plants that include berries, grapes, walnuts, peanuts,
pomegranates, and other fruits and vegetables [1]. Many
of these polyphenol preparations are available as dietary
supplements [2]. The largest and best studied polyphe-
nols are the dietary estrogen-like molecules or phytoes-
trogens [1]. Of the three main classifications of However, these isoflavone molecules do not exist at
high levels in their biologically active form in natural
food products, but rather are at high abundance in a * Correspondence: edwin_lephart@byu.edu
2The Department of Physiology and Developmental Biology and the
Neuroscience Center, Brigham Young University, Provo, Utah 84602, USA
Full list of author information is available at the end of the article Page 2 of 9 Page 2 of 9 Lund et al. Reproductive Biology and Endocrinology 2011, 9:4
http://www.rbej.com/content/9/1/4 addressing prostate disorders, such as benign prostatic
hyperplasia (BPH) or prostate cancer. With this in mind,
we sought, in the following experiments, to determine
equol’s effect on 5a-DHT levels in cultured human pros-
tate cancer (LNCap) cells. We also examined equol’s
effect on both prostate weight and circulating hormone
levels in vivo using Long-Evans rats. In brief, the present
results demonstrate that equol: a) (R- and/or S-isomeric
mixtures) has high binding affinity for 5a-DHT making it
a potent selective androgen modulator (SAM), b) blocks
the stimulatory androgen action of 5a-DHT in increasing
prostate specific antigen (PSA) levels in human cancer
(LNCap) cell cultures and, c) significantly decreases
serum 5a-DHT and subsequently prostate weight with-
out altering, testosterone, 17b-estradiol or LH levels. Applications for equol to improve prostate disorders and
other androgen-mediated conditions is also discussed. precursor form [3-5]. For example, daidzin, the precur-
sor of daidzein, is the glycosidic form that contains a
carbohydrate portion of the molecule. Daidzin is meta-
bolized in the gastrointestinal tract by intestinal bacteria,
which hydrolyze the carbohydrate moiety, to the biologi-
cally active isoflavone, daidzein [4,5]. Background Daidzein is then
further metabolized in the intestine to equol at relatively
low or high levels dependent upon several biological,
dietary and presumably environmental factors [4,5]. Although, recent evidence suggests that equol is found
naturally in white cabbage [6]. Equol has recently caught the interest of many research-
ers due to its rich antioxidant activity and implications in
cancer research [4,5]. The chemical structure of equol
contains a stereocenter at carbon number 3 which gives it
two possible enantiomers and it has since been proven
that the production of equol by microflora in mammals or
other animals is selective for the S - enantiomer only [5]. S-equol has unique chemical properties compared to its
R - enantiomer. S-equol has been shown to have a modest
affinity for binding to and mimicking estrogen’s effects on
estrogen beta receptors (ERb) due to its similar structure
to natural estrogens [5,7]. However, S-equol shows little
affinity for estrogen alpha receptors (ERa). Furthermore,
equol (i.e., the R- and/or S-isomer) can act as an anti-
androgen [7]. Equol’s anti-androgen activity is unique as
equol does not bind the androgen receptor (AR) but speci-
fically binds 5a-dihydrotestosterone (5a-DHT) with high
affinity, and thereby prevents DHT from binding the AR
[7], see Figure 1. This finding is reconfirmed and extended
here. Additionally, equol’s mechanism of action, namely,
its ability to specifically bind 5a-DHT and prevent 5a-
DHT’s biological actions in physiological processes, was
studied. Samples In a 20 ml amber glass vial approximately 4,000 to 5,000
dpm of 14C5a- Dihydrotestosterone (DHT; Dupont/
NEN, Boston, MA, USA) was added. Additionally, 3.0
nM (or approximately 865 pg/ml) of cold 5a-DHT
(Sigma/Aldrich Chemical Co., St. Louis, MO, USA) was
added (in all experiments) and where appropriate vary-
ing concentrations of equol (70% R-equol and 30% S-
equol mixture; from 0 to 2,000 nM; obtained from LC
Labs., Woburn, MA,USA or Robert Handa’s Laboratory,
Colorado State University, USA, R-equol or S-equol iso-
mers only) from ethanol stock solutions. This mixture
was vortexed for 30 seconds at room temperature and
then dried down. Subsequently 1 ml of TEGMD (10
mM Tris CL, 1.5 mM EDTA, 10% glycerol, 25 mM
molybdate and 1 mM dithrothreitol) buffer was added,
capped, and mixed by inverting the vial 3-times. This
mixture was incubated at room temperature for 20-24
hours before being placed on a Sephadex column [7]. Columns For example, it is known that prostate cancer cells are
supported in their growth by androgen stimulation and
the androgen-regulated expression of the prostate speci-
fic antigen (PSA) is a biological marker of such stimula-
tion [8]. Logically, any treatment that could decrease PSA
levels in prostate cancer cells, or antagonize specific
androgen hormone action, would have great potential in Sephadex LH-20 (25-100 μm particle diameter) from
Sigma/Aldrich Chemical Co. (St. Louis, MO, USA), was
used to prepare 1.3 cm × 25 cm columns, as described
by Lund et al [7]. Sephadex (0.25 grams) was mixed
with 1.5 ml of TEGMD buffer. The serological pipette
was plugged with a glass bead and 1 ml of Sephadex
was loaded onto the column with TEGMD buffer and
equilibrated. Flow rates under gravity were approxi-
mately 0.5 ml/min and 0.5 ml fractions were collected
in plastic scintillation vials. Figure 1 Biochemical structures of dihydrotestosterone and
equol. Experiment 1. In vitro binding of Equol (Isomer Ratio) to
5a-DHT
S
l Experiment 1. In vitro binding of Equol (Isomer Ratio) to
5a-DHT Quantification Figure 1 Biochemical structures of dihydrotestosterone and
equol. Figure 1 Biochemical structures of dihydrotestosterone and
equol. Five ml of Ultima Gold scintillation fluid (Packard Instr. Co., Meriden, CT, USA) was added to each vial, mixed Page 3 of 9 Page 3 of 9 Lund et al. Reproductive Biology and Endocrinology 2011, 9:4
http://www.rbej.com/content/9/1/4 Lund et al. Reproductive Biology and Endocrinology 2011, 9:4
http://www.rbej.com/content/9/1/4 Experiment 2. In vitro blockade of 5a-DHT stimulated
Prostate Specific Antigen (PSA) levels in LNCap cells by
equol: Human Prostate Cancer Cell Cultures
Human prostate cancer cell line was obtained from
American Type Culture Collection (ATCC; Manassas,
VA, USA; ATCC # CRL-1740, LNCap-FGC) and cul-
tured in a 37° C, humidified incubator with 5% CO2, in
RPMI Medium (Sigma/Aldrich Cat. # R-8758) with 5%
fetal bovine serum (Hyclone/Thermo Sci., Waltham,
MA, USA, Cat. # SH30088.03, Lot number APC20780)
and 5 mM Hepes (Sigma/Aldrich Cat. H-0887), 1× anti-
biotic/antimycotic (Sigma/Aldrich Cat. # A5955). Cells
were expanded in T-150 flasks for three passages until
cryopreservation and storage in liquid nitrogen in RPMI
medium with 10% FBS and 10% DMSO as cryopreserva-
tive. A cryovial was then thawed in a 37° C water bath,
expanded again one or two passages, and then plated at Experiment 2. In vitro blockade of 5a-DHT stimulated
Prostate Specific Antigen (PSA) levels in LNCap cells by
equol: Human Prostate Cancer Cell Cultures thoroughly and counted for 5 min. in a Beckmann LS
6500 scintillation counter. The peak 14C5a-DHT frac-
tion and profile using 0 nM of equol was used as refer-
ence. The shift (advancement) in peak/profile in
subsequent experiments using increasing concentrations
of equol (70% R-isomer and 30% S-isomer in this
experiment) was calculated based upon this reference
(as arbitrary binding units as a percentage of the original
reference, n = 6 for 0 nM of equol). Note: data not
shown, 2,000 nM of equol was tested in this experiment
which yielded 100.0% ± 0.3 (s.e.m.) binding, n = 3. Finally, over thirty steroid hormones were tested (see
Table 1) (in preliminary studies to determine the selec-
tivity and specificity of equol for binding 5a-DHT). The
steroids were purchased from Steraloids, Inc., Newport,
RI, USA or Sigma/Aldrich Chem. Co., St. Louis, MO,
USA. Table 1 Steroid compounds tested in Equol Binding Assays PSA ELISA Tissue culture supernatants were diluted 10-fold in PBS
and stored at -20°C, then thawed at room temperature
prior to assay. A commercial ELISA kit for free PSA
(Bio-Quant, San Diego, CA, USA, Cat. # BQ 067T) was
utilized according to the manufacturer’s instructions,
and data was determined using a Molecular Devices
Vmax 96 well plate reader and SoftMax software. Table 1 Steroid compounds tested in Equol Binding Assays Table 1 Steroid compounds tested in Equol Binding Assays
Chemical Name
Trivial Name
4-ANDROSTEN-17b-OL-3-ONE
TESTOSTERONE
5-ANDROSTEN-3b-OL-17-ONE
DHEA
5a-ANDROSTAN-3a,17b-DIOL
17b-DIHYDROANDROSTERONE
5a-ANDROSTAN-3b,17a-DIOL
NA
5a-ANDROSTAN-3b,17b-DIOL
NA
5a-ANDROSTAN-3,17-DIONE
ANDROSTANEDIONE
5a-ANDROSTAN-17b-OL-3-ONE
5a-DHT
5b-ANDROSTAN-17b-OL
NA
5b-ANDROSTAN-17b-OL-3-ONE
5b-DHT
4-ANDROSTEN-3,17-DIONE
ANDROSTENEDIONE
1,3,5(10)-ESTRATRIEN-3,17a-DIOL
EPIESTRADIOL
1,3,5(10)-ESTRATRIEN-3,17b-DIOL
ESTRADIOL (E2)
1,3,5(10)-ESTRATRIEN-3,16a,17b-TRIOL
ESTRIOL (E3)
1,3,5(10)-ESTRATRIEN-3-OL-17-ONE
ESTRONE (E1)
5a-ESTRAN-3,17-DIONE
5a-DIHYDROANDROSTENEDIONE
5-PREGNEN-3b-OL-20-ONE
PREGNENOLONE
4-PREGENEN-3,20-DIONE
PROGESTERONE (P4)
5a-PREGNAN-3a-OL-20-ONE
ALLOPREGNANOLONE
5a-PREGNAN-11b,21-DIOL-3, 20-DIONE
ALLODIHYDROCOSTERONE
5a-PREGNAN-3a,11b,21-TRIOL-20-ONE
ALLOTETRAHYDROCORTICOSTERONE
5a-PREGNAN-3b, 21b, 21-TRIOL-20-ONE
EPIALLOTETRAHYDROCORTICOSTERONE
5a-PREGNAN-3b,11b,17,21-TETROL-20-ONE
3b, 5a-TETRAHYDROCORTISOL
5a-PREGNAN-11b,17,21-TRIOL-3,20-DIONE
ALLODIHYDROCORTISOL
5a-PREGNAN-17,21-DIOL-3,11,20-TRIONE
ALLODIHYDROCORTISONE
5a-PREGNAN-3b-OL-20-ONE
5a-DIHYDROPREGNANOLONE
5a-PREGNAN-3, 20-DIONE
5a-DIHYDROPROGESTERONE (5a-DHP)
5b-PREGNAN-3a-OL-20-ONE
3a-HYDROXY-5bTETRAHYDROPROGESTONE
5b-PREGNAN-11b,21-DIOL-3,20-DIONE
5b -DIHYDROCORTICOSTERONE
5b-PREGNAN-3a,11b,21-TRIOL-20-ONE
TETRAHYDROCORTICOSTERONE
5b-PREGNAN-3a,11b,17,21-TETROL-20-ONE
TETRAHYDROHYDROCORTISOL
5b-PREGNAN-11b,17,21-TRIOL-3,20-DIONE
5b-DIHYDROCORTISOL
5b-PREGNAN-17,21-DIOL-3,11,20-TRIONE
5b -DIHYDROCORTISONE
5b-PREGNAN-3,20-DIONE
5b-DIHYDROPROGESTERONE (5b-DHP) Table 1 Steroid compounds tested in Equol Binding Assays
Chemical Name
Trivial Name
4-ANDROSTEN-17b-OL-3-ONE
TESTOSTERONE
5-ANDROSTEN-3b-OL-17-ONE
DHEA
5a-ANDROSTAN-3a,17b-DIOL
17b-DIHYDROANDROSTERONE
5a-ANDROSTAN-3b,17a-DIOL
NA
5a-ANDROSTAN-3b,17b-DIOL
NA
5a-ANDROSTAN-3,17-DIONE
ANDROSTANEDIONE
5a-ANDROSTAN-17b-OL-3-ONE
5a-DHT
5b-ANDROSTAN-17b-OL
NA
5b-ANDROSTAN-17b-OL-3-ONE
5b-DHT
4-ANDROSTEN-3,17-DIONE
ANDROSTENEDIONE
1,3,5(10)-ESTRATRIEN-3,17a-DIOL
EPIESTRADIOL
1,3,5(10)-ESTRATRIEN-3,17b-DIOL
ESTRADIOL (E2)
1,3,5(10)-ESTRATRIEN-3,16a,17b-TRIOL
ESTRIOL (E3)
1,3,5(10)-ESTRATRIEN-3-OL-17-ONE
ESTRONE (E1)
5a-ESTRAN-3,17-DIONE
5a-DIHYDROANDROSTENEDIONE
5-PREGNEN-3b-OL-20-ONE
PREGNENOLONE
4-PREGENEN-3,20-DIONE
PROGESTERONE (P4)
5a-PREGNAN-3a-OL-20-ONE
ALLOPREGNANOLONE
5a-PREGNAN-11b,21-DIOL-3, 20-DIONE
ALLODIHYDROCOSTERONE
5a-PREGNAN-3a,11b,21-TRIOL-20-ONE
ALLOTETRAHYDROCORTICOSTERONE
5a-PREGNAN-3b, 21b, 21-TRIOL-20-ONE
EPIALLOTETRAHYDROCORTICOSTERONE
5a-PREGNAN-3b,11b,17,21-TETROL-20-ONE
3b, 5a-TETRAHYDROCORTISOL
5a-PREGNAN-11b,17,21-TRIOL-3,20-DIONE
ALLODIHYDROCORTISOL
5a-PREGNAN-17,21-DIOL-3,11,20-TRIONE
ALLODIHYDROCORTISONE
5a-PREGNAN-3b-OL-20-ONE
5a-DIHYDROPREGNANOLONE
5a-PREGNAN-3, 20-DIONE
5a-DIHYDROPROGESTERONE (5a-DHP)
5b-PREGNAN-3a-OL-20-ONE
3a-HYDROXY-5bTETRAHYDROPROGESTONE
5b-PREGNAN-11b,21-DIOL-3,20-DIONE
5b -DIHYDROCORTICOSTERONE
5b-PREGNAN-3a,11b,21-TRIOL-20-ONE
TETRAHYDROCORTICOSTERONE
5b-PREGNAN-3a,11b,17,21-TETROL-20-ONE
TETRAHYDROHYDROCORTISOL
5b-PREGNAN-11b,17,21-TRIOL-3,20-DIONE
5b-DIHYDROCORTISOL
5b-PREGNAN-17,21-DIOL-3,11,20-TRIONE
5b -DIHYDROCORTISONE
5b-PREGNAN-3,20-DIONE
5b-DIHYDROPROGESTERONE (5b-DHP) Page 4 of 9 Lund et al. Reproductive Biology and Endocrinology 2011, 9:4
http://www.rbej.com/content/9/1/4 200 ppm of isoflavones at the supplier (CRL). At 50
days of age, upon arrival, the male rats were placed on a
diet containing approximately 10 ppm of isoflavones;
referred to hereafter as the low isoflavone diet (Zeigler
Bros., Gardnes, PA, USA; Phytoestrogen Reduced
Rodent Diet II). All animals remained on the low isofla-
vone diet until 216 days of age to exclude the influence
of dietary isoflavones on the measured parameters. At
150 days of age the rats were divided into two groups
(control or equol treatments) that were matched by age
and body weight. Starting at 190 days of age the male
rats received a daily subcutaneous 0.1cc injection at the
nape of the neck of vehicle (n = 8) (dimethyl sulfoxide;
DMSO) or equol (n = 8) at a dose of approximately 1.0
mg/kg for 25 consecutive days. 10,000 cells per 96 well in 0.2 ml medium in 96 well
plates in RPMI 5% FBS medium. After approximately
48 hours, the medium was changed to phenol red-free
DMEM/F12 (Gibco/Invitrogen, Carlsbad, CA, USA, Cat. q
MTT Assay Cytotoxicity was determined by spectrophotometric
detection of reduced 3-(4,5-dimethylthiazol-2-yl)-2,5-
diphenyltetrazolium bromide (MTT, Sigma/Aldrich
Cat # 5655) at 550 nm using a Molecular Devices Vmax
96 well plate reader and SoftMax software (n = 6 per
treatments or controls). Metabolic activity can be used
as a measure of cytotoxicity, in that the intensity of the
reduced form of MTT by live cells is directly propor-
tional to cellular viability, and inversely proportional to
cytotoxicity. Cellular DNA Assay Table 1 Steroid compounds tested in Equol Binding Assays #21041-025) with 2% FBS and 1× antibiotic/antimycotic
and test materials (5a-DHT, equol or 5a-DHT plus
equol; n = 6 per treatment) and DMSO/vehicle controls
were added at the appropriate concentration (either, 0.1
nM, 1.0 nM or 10 nM of 5a-DHT or 100 nM, 10 nM
or 1.0 nM of equol) from 10× stocks for specific experi-
ments. [The change to phenol red-free medium was to
eliminate any potential estrogenic influence]. Cells were
cultured for approximately 48 hours in the presence of
the test materials and controls prior to removal of med-
ium supernatants for prostate specific antigen (PSA)
ELISA quantification. The body weights for each group were recorded
weekly starting at 150 days of age before the treatments
were initiated, with weights obtained immediately before
and after the treatments were administered (there were
no significant differences in body weights between the
control and equol groups at the start of this experi-
ment). At 216 days of age the animals were weighed
[grams (g) ± 0.1 g], then anesthestized with Ketamine/
acepromazine and blood was collected from the heart. Next the ventral prostate organ was dissected and
weighed [milligrams (mg) ± 0.001 mg]. The collected
blood samples were centrifuged and serum was stored
at -20˚ C until assayed. All collection procedures were
performed blind to the treatments. This animal protocol
was approved by the Institutional Animal Care and Use
Committee at Brigham Young University. Reagents and supplies This is an example of equol preventing the stimulatory
effects of 5a-DHT in vivo. Rats were injected with 1 mg
of non-racemic equol (52% S-isomer, 48% R-isomer) for
25 consecutive days, and serum 5a-DHT levels and
prostate weights were measured. Adult (50 day-old)
males (n = 16), purchased from Charles River Labora-
tories (CRL; Wilmington, MA, USA), were caged indivi-
dually and housed in the Brigham Young University
Vivarium and maintained on an 11-dark, 13-hour light
schedule (lights on 0600-1900). Before purchase, the
male animals were fed a diet containing approximately All reagents and supplies not specifically mentioned
(e.g., EDTA, DMSO, glycerol, dithrothreitol, pipettes, 1.3
× 25 cm columns, etc.) were purchased from Sigma/
Aldrich Chem. Co., St. Louis, MO, USA. Cellular DNA Assay Cellular DNA was determined using a CyQuant Cell
Proliferation Kit (Molecular Probes/Invitrogen, Carlsbad,
CA, USA, Cat # C7026) according to the manufacture’s
instructions to quantify cellular proliferation. Serum testosterone, 5a-DHT, and 17b-estradiol were
quantified by radioimmunoassay (RIA) kits purchased
from Diagnostic System Laboratories (Webster, TX, USA). Luteininzing hormone (LH) was quantified by an assay
utilizing standards from the National Institutes of Health
(NIH) USA pituitary hormone program. The samples
were run in duplicate for each RIA, with internal control
samples. In all RIAs, the control values were within nor-
mal ranges. The intra-assay coefficients of variance for the
assays were: testosterone = 6.0%; for 5a-dihydrotestoster-
one = 8%, 17b-estradiol = 5% and LH = 9%. Experiment 1 In preliminary studies, Table 1 shows 33 different ster-
oid compounds that were tested in binding assays (out-
lined above) to determine equol’s affinity for binding to
each. While equol has modest affinity for some 5a-
reduced steroids, equol displayed the highest affinity for
5a-dihydrotestosterone (5a-DHT) and had no affinity
for 5b-dihydrotestosterone (5b-DHT) or some of the
most common natural sex steroids, such as: estradiol,
estrone, estriol, progesterone or testosterone. [Notably:
as tested in these assays, the isoflavones: genistein and
daidzein do not bind 5a-DHT]. As shown in Figure 2, using 14C5a-DHT as the tracer
in the presence of 3.0 nM (or approximately 865 pg/ml)
of cold DHT (in all samples, representing a high-normal
level of circulating 5a-DHT in men) varying concentra-
tions of equol (70% R-equol and 30% S-equol mixture)
from 0 to 2,000 nM yielded the binding curve displayed
in this Figure. [The 2,000 nM dose is not graphically
displayed in this Figure]. Maximum and half maximal
equol binding to 5a-DHT occurred at approximately
100 nM and 4.8 nM, respectively. The number of sam-
ples tested at each concentration of equol examined
ranged between 2 to 6 replicates (see Figure 2). [For
conversion purposes, a concentration of 10.3 nmol/L of
equol is approximately equivalent to 2.5 ng/ml of equol]. Notably, equol at a 70% R-isomer and 30% S-isomer
ratio was tested in this experiment. However, racemic
equol or equol at a 30% R-isomer and 70% S-isomer
ratio when tested in these assays yielded similar results
to that displayed in Figure 2. These data suggest that
equol, regardless of the isomer ratio mixture can bind-
ing selectively 5a-DHT in vitro. As shown in Figure 2, using 14C5a-DHT as the tracer
in the presence of 3.0 nM (or approximately 865 pg/ml)
of cold DHT (in all samples, representing a high-normal
level of circulating 5a-DHT in men) varying concentra-
tions of equol (70% R-equol and 30% S-equol mixture)
from 0 to 2,000 nM yielded the binding curve displayed
in this Figure. [The 2,000 nM dose is not graphically
displayed in this Figure]. Maximum and half maximal
equol binding to 5a-DHT occurred at approximately
100 nM and 4.8 nM, respectively. The number of sam-
ples tested at each concentration of equol examined
ranged between 2 to 6 replicates (see Figure 2). [For
conversion purposes, a concentration of 10.3 nmol/L of
equol is approximately equivalent to 2.5 ng/ml of equol]. Statistical analyses Where appropriate, data were analyzed by analysis of
variance statistics (ANOVA) followed by Newman-Keuls
post hoc tests. Significance was p < 0.05. Curve fitting,
scientific graphing, and analysis were completed using
GraphPad Software (GraphPad Prism 3.0, San Diego,
CA, USA). Where appropriate, data were analyzed by analysis of
variance statistics (ANOVA) followed by Newman-Keuls
post hoc tests. Significance was p < 0.05. Curve fitting,
scientific graphing, and analysis were completed using
GraphPad Software (GraphPad Prism 3.0, San Diego,
CA, USA). Page 5 of 9 Lund et al. Reproductive Biology and Endocrinology 2011, 9:4
http://www.rbej.com/content/9/1/4 Lund et al. Reproductive Biology and Endocrinology 2011, 9:4
http://www.rbej.com/content/9/1/4 Results This is an example of the effects of equol preventing the
stimulatory effects of 5a-DHT in LNCap prostate cancer
cells in vitro from secreting prostate-specific antigen
(PSA), a molecule known to be regulated by 5a-DHT,
as measured by PSA ELISA (Figure 3). Treatment with
0.1, 1 or 10 nM of 5a-DHT; 1, 10 or 100 nM of equol
or a combination of 5a-DHT plus equol (0.1 nM 5a-
DHT and 1 nM equol; 1 nM 5a-DHT and 10 nM equol
or 10 nM 5a-DHT and 100 nM equol). Each obtained
value represents 6 replicates + standard error of the
mean (sem). Cytotoxicity, as assessed by MTT Assay,
did not influence PSA production by LNCap prostate
cancer cells for any of the treatments or controls (data
not shown). Also, while DHT treatment has been
reported to increase cellular proliferation of this cell line
in vitro [9], none of the treatments significantly altered
cellular proliferation as quantified by the cellular DNA
assay (data not shown). Experiment 1 Up black triangle
significantly greater levels in all 5a-DHT treatments compared to all
other treatments (p < 0.0001) Down black triangle, significant
decrease at 100 nM of equol vs. all control values (p < 0.0001). Black
square, slight but significant increase in PSA levels in the equol 10
nM group compared to DMSO-vehicle values (p < 0.008) but not to
the no-treatment group (p < 0.47). Three horizontal lines, significant
decrease at 10 nm of 5a-DHT plus equol at 100 nM compared to
10 nM of 5a-DHT alone (p < 0.0001). Two horizontal lines,
significant decrease at 1 nM of 5a-DHT plus equol at 10 nM
compared to 1 nM of 5a-DHT alone (p < 0.001). One horizontal
line, represent a significant decrease at 0.1 nM of 5a-DHT plus
equol at 1 nM compared to 0.1 nM of 5a-DHT alone (p < 0.042). Figure 3 In vitro evidence: The effects of 5a-DHT or equol
alone and in combination on prostate specific antigen (PSA)
by LNCap Human Prostate Cancer Cells. All data expressed as the Figure 3 In vitro evidence: The effects of 5a-DHT or equol
alone and in combination on prostate specific antigen (PSA)
by LNCap Human Prostate Cancer Cells. All data expressed as the Figure 3 In vitro evidence: The effects of 5a-DHT or equol
alone and in combination on prostate specific antigen (PSA)
by LNCap Human Prostate Cancer Cells. All data expressed as the
mean + the standard error of the mean (sem). Up black triangle
significantly greater levels in all 5a-DHT treatments compared to all
other treatments (p < 0.0001) Down black triangle, significant
decrease at 100 nM of equol vs. all control values (p < 0.0001). Black
square, slight but significant increase in PSA levels in the equol 10
nM group compared to DMSO-vehicle values (p < 0.008) but not to
the no-treatment group (p < 0.47). Three horizontal lines, significant
decrease at 10 nm of 5a-DHT plus equol at 100 nM compared to
10 nM of 5a-DHT alone (p < 0.0001). Two horizontal lines,
significant decrease at 1 nM of 5a-DHT plus equol at 10 nM
compared to 1 nM of 5a-DHT alone (p < 0.001). One horizontal
line, represent a significant decrease at 0.1 nM of 5a-DHT plus
equol at 1 nM compared to 0.1 nM of 5a-DHT alone (p < 0.042). Figure 2 In vitro binding of equol (isomer ratio) to 5a-DHT. Experiment 1 As shown in Figure 3: In the 1, 10 or 100 nM Vehicle
(DMSO) treatments PSA levels did not differ from the
No Treatment baseline group and controls were not sig-
nificantly different from each other. Treatment with 0.1,
1 or 10 nM 5a-DHT stimulated PSA secretion to Figure 3 In vitro evidence: The effects of 5a-DHT or equol
alone and in combination on prostate specific antigen (PSA)
by LNCap Human Prostate Cancer Cells. All data expressed as the
mean + the standard error of the mean (sem). Up black triangle
significantly greater levels in all 5a-DHT treatments compared to all
other treatments (p < 0.0001) Down black triangle, significant
decrease at 100 nM of equol vs. all control values (p < 0.0001). Black
square, slight but significant increase in PSA levels in the equol 10
nM group compared to DMSO-vehicle values (p < 0.008) but not to
the no-treatment group (p < 0.47). Three horizontal lines, significant
decrease at 10 nm of 5a-DHT plus equol at 100 nM compared to
10 nM of 5a-DHT alone (p < 0.0001). Two horizontal lines,
significant decrease at 1 nM of 5a-DHT plus equol at 10 nM
compared to 1 nM of 5a-DHT alone (p < 0.001). One horizontal
line, represent a significant decrease at 0.1 nM of 5a-DHT plus
equol at 1 nM compared to 0.1 nM of 5a-DHT alone (p < 0.042). Notably, equol at a 70% R-isomer and 30% S-isomer
ratio was tested in this experiment. However, racemic
equol or equol at a 30% R-isomer and 70% S-isomer
ratio when tested in these assays yielded similar results
to that displayed in Figure 2. These data suggest that
equol, regardless of the isomer ratio mixture can bind-
ing selectively 5a-DHT in vitro. Figure 2 In vitro binding of equol (isomer ratio) to 5a-DHT. 14C5a- Dihydrotestosterone served as the tracer and 3.0 nM (or
approximately 865 pg/ml) of cold 5a-DHT was added (in all
experiments) and where appropriate varying concentrations of
equol (70% R-equol and 30% S-equol mixture; from 0 to 2,000 (see
methods). All data expressed as the mean of the percent binding ±
SEM (see insert). Figure 3 In vitro evidence: The effects of 5a-DHT or equol
alone and in combination on prostate specific antigen (PSA)
by LNCap Human Prostate Cancer Cells. All data expressed as the
mean + the standard error of the mean (sem). Discussion The binding assays in the present study confirm and
extend previous findings from our laboratories [7]. How-
ever, this protein and cell-free assay represents a system
that is absent of steroid-binding proteins and other biolo-
gicals that may influence the in vivo interaction between
equol and 5a-DHT. Moreover, after testing over 30 ster-
oid hormones with various androgenic, estrogenic or pro-
gestin chemical configurations, equol (racemic or non-
racemic mixtures) specifically binds 5a-DHT with appar-
ent equal affinity since the bind curves obtained were
very similar using different equol isomer ratios. This sug-
gests several important issues. First, all humans have
plasma levels of S-equol circulating in their bloodstream
that may represent a natural modulation of the potent
androgen, 5a-DHT, regardless of whether or not they are
“equol producers” [4,5]. Second, this circulating equol is
derived from consumption of soy- or high content isofla-
vone-containing food products or low abundant isofla-
vone foods such as corn, wheat, cow’s milk, etc., as a
primary consumer. Finally, because grazing or other ani-
mals that produce high levels of S-equol naturally that
are then consumed, such as meat products, by humans
represents a source of equol as a secondary consumer. Evidence to support this last notion is shown in humans
that consume high meat and fat diets display higher
blood isoflavone/equol levels [10]. Experiment 1 However, combinations of 10 nM 5a-DHT plus
100 nM equol, as well as 1 nM 5a-DHT plus 10 nM
equol, and 0.1 nM 5a-DHT plus 1 nM equol treatments
abrogated the increase in PSA secretion, compared to
the 5a-DHT treatments alone, at the respective andro-
gen concentrations (see Figure 3). Taken together with
the in vitro binding demonstrated in Figure 2, these
data from experiment 2, suggest that equol binds the
5a-DHT molecule and biologically inactivates it in
human cancer prostate cells. in the testes, this is not a surprising result. Finally, when
17b-estradiol levels were determined there were no signifi-
cant differences between the treatment groups. All hor-
mone levels were within normal ranges of that expected
for this strain, age and sex of rat. Notably, Long-Evans rats
do not develop spontaneously prostate cancer but enlarge-
ment of prostatic tissue is observed with aging. This in vivo study demonstrates that equol can con-
tact and biologically inactive the 5a-DHT molecule as
shown by the significant decrease in 5a-DHT levels in
blood and significantly reduced prostate weights of
equol-treated male rats. Finally, the in vitro and in vivo
studies reported above demonstrates that equol may be
effective in treating prostate disorders or other andro-
gen-mediated diseases that are regulated by the hor-
mone 5a-DHT. Experiment 1 14C5a- Dihydrotestosterone served as the tracer and 3.0 nM (or
approximately 865 pg/ml) of cold 5a-DHT was added (in all
experiments) and where appropriate varying concentrations of
equol (70% R-equol and 30% S-equol mixture; from 0 to 2,000 (see
methods). All data expressed as the mean of the percent binding ±
SEM (see insert). Page 6 of 9 Page 6 of 9 Page 6 of 9 Lund et al. Reproductive Biology and Endocrinology 2011, 9:4
http://www.rbej.com/content/9/1/4 maximal levels (ranging from 228.3 to 260 ng/ml). PSA
levels from cells treated with 1, 10 or 100 nM equol
were well below DHT-treated values. In fact, the 100
nM equol treatment alone yielded a PSA level at 60 ng/
ml that was significantly decreased below all vehicle-
control values (at approximately 87 ng/ml). The equol
treatment at 10 nM slightly but significantly increased
PSA levels (to 116.7 ng/ml) compared to all the DMSO-
vehicle-control values around 87 ng/ml, but not to the
no-treatment group at 100 ng/ml). [This may be due to
the number of replicates (6) in this experiment where
standard error of the mean (sem) values were relatively
small]. However, combinations of 10 nM 5a-DHT plus
100 nM equol, as well as 1 nM 5a-DHT plus 10 nM
equol, and 0.1 nM 5a-DHT plus 1 nM equol treatments
abrogated the increase in PSA secretion, compared to
the 5a-DHT treatments alone, at the respective andro-
gen concentrations (see Figure 3). Taken together with
the in vitro binding demonstrated in Figure 2, these
data from experiment 2, suggest that equol binds the
5a-DHT molecule and biologically inactivates it in
human cancer prostate cells. maximal levels (ranging from 228.3 to 260 ng/ml). PSA
levels from cells treated with 1, 10 or 100 nM equol
were well below DHT-treated values. In fact, the 100
nM equol treatment alone yielded a PSA level at 60 ng/
ml that was significantly decreased below all vehicle-
control values (at approximately 87 ng/ml). The equol
treatment at 10 nM slightly but significantly increased
PSA levels (to 116.7 ng/ml) compared to all the DMSO-
vehicle-control values around 87 ng/ml, but not to the
no-treatment group at 100 ng/ml). [This may be due to
the number of replicates (6) in this experiment where
standard error of the mean (sem) values were relatively
small]. n = 8 animals per group. NSC = no significant change. The data are expressed
as the mean ± standard error of the mean (sem) for each parameter. ** = p < 0.016 * = p < 0.001. Experiment 3 Also, the apparent half-life
of R- or S-equol in humans is 5 to 6 hours [5], suggest-
ing a viable time interval and mechanism of action. While the present data were derived from in vitro bind-
ing studies that does not represent the complex physio-
logical environment this viewpoint will be covered
below in reference to prostate health. which may be tissue sites for beneficial alterations in
androgen hormone action [26]. For instance, it is known
the androgens decrease, whereas, estrogenic compounds
enhance wound healing [27] that may have effective
applications for topical treatments. The results of the third experiment demonstrated that
equol can significantly decrease serum 5a-DHT along
with a significant decline in prostate weight without
altering, testosterone, 17b-estradiol or LH levels. There
is evidence that Asian cultures have lower prostate can-
cer rates compared to Western cultures [1,3,28]. Furthermore, when data is stratified examining indivi-
duals that produce equol naturally after soy consump-
tion suggest that lower cancer rates are due to the
beneficial influence of this isoflavonoid molecule [29]. Further evidence presented in this study for equol bind-
ing 5a-DHT is presented in the experiment where equol
blocked the stimulatory androgen action of 5a-DHT in
increasing PSA levels in human cancer cell cultures. Although, this may be the major mechanism of action, it
is known that equol (particularly S-equol) has an affinity
for ERb [5]. Previous laboratories, including ours, have
shown that isoflavones binding to ERb in the prostate can
down regulate the androgen receptor (AR) and thus
decrease androgen hormone action [15]. Since PSA is an
androgen-regulated gene this idea may have merit [16,17]. In addition to equol directly binding 5a-DHT, it may alter
AR expression and hence significantly decreased PSA
levels in the present study. Support for this view comes
from the significant decrease in PSA levels with equol
treatment alone (@ 100 nM) in experiment 2 that dis-
played values even below baseline (vehicle) PSA concen-
trations. However, it could be argued that the fetal bovine
serum provided a background androgenic stimulation and
this may also account for this observation. In any event,
the results of this experiment demonstrate the positive
influence equol has on PSA levels and implicates potential
applications of equol for prostate disorders. Experiment 3 To determine whether equol can bind 5a-DHT in vivo,
adult male Long-Evans rats were treated with equol for
25 consecutive days (@ 1 mg/kg via s.c. injections). Equol-injected animals displayed an approximately 50%
decrease in serum 5a-DHT compared to (DMSO) vehi-
cle-injected control animals (Table 2). This finding cor-
responds with the significant decrease in androgen
hormone action which is known to regulate prostatic
cell proliferation, and hence, prostate weight. Prostate
weights were significantly decreased by approximately
20% in the equol-injected males compared to control
rats (either alone or when prostate weights were stan-
dardized with body weights). When LH and testosterone were quantified between the
treatment groups there were no significant differences in
these hormone levels (Table 2). Since LH is the gonado-
trophin regulating testosterone synthesis from Leydig cells Table 2 Serum 5a-DHT levels, prostate weights and
reproductive parameters in Equol-treated male rats Table 2 Serum 5a-DHT levels, prostate weights and
reproductive parameters in Equol-treated male rats
Parameter measured
Vehicle
Equol
Change
Prostate Weight (PW), mg
535 ± 23
429 ± 30**
20%↓
PW/100 g Body Weight
76 ± 4
61 ± 5**
20%↓
Luteinizing Hormone (LH), ng/ml
1.6 ± 0.2
1.3 ± 0.1
NSC
Serum Testosterone, ng/ml
2.1 ± 0.4
2.3 ± 0.5
NSC
Serum 5a-DHT, pg/ml
100 ± 18
52 ± 5*
50%↓
Serum 17b-Estradiol, pg/ml
3.4 ± 0.6
4.8 ± 0.7
NSC
n = 8 animals per group NSC = no significant change The data are expressed This denotes an interesting ‘natural’ modulation of
5a-DHT in humans, since all individuals consume food
products that contain the precursor daidzein molecule
or equol itself (i.e., soy, corn, wheat, cow’s milk or ani-
mal meat products, etc.) [4,6,11-13]. In fact, whether or
not swine were fed a soy-containing or protein-supple-
mented diet the serum equol levels between the dietary
treatment groups were essential the same [14]. Since the
apparent half-maximal binding of equol to 5a-DHT
occurred at 4.8 nM (or approximately 1.2 ng/ml) the Page 7 of 9 Page 7 of 9 Lund et al. Reproductive Biology and Endocrinology 2011, 9:4
http://www.rbej.com/content/9/1/4 Lund et al. Reproductive Biology and Endocrinology 2011, 9:4
http://www.rbej.com/content/9/1/4 very low level of circulating equol may alter androgen
hormone action in a manner previous unknown in nor-
mal physiological conditions. Experiment 3 Finally, phase
II clinical trial data examining dietary intervention with
isoflavone supplementation in men with recurrent prostate
cancer display a decline in PSA levels that support our
present in vitro data [18]. Moreover, while equol has been shown to have posi-
tive benefits in human prostate health, there is evidence
that R-, rather than the S - enantiomer is responsible
for the in vivo chemoprotective properties of equol [30]. Additionally, from pilot data, serum 5a-DHT levels
significantly decreased by 10 to 15% in men (50 to 60
years of age) and 20 to 26% in postmenopausal women
(60 to 65 years of age) when an equol oral dose of 3 mg
per day was administered (without side effects), suggest-
ing that equol effectively binds and/or alters 5a-DHT
levels. This pilot data is supported by a recent study
showing isoflavone supplementation that increased the
production of serum equol in equol-producers resulted
in a decline in serum 5a-DHT levels in men by approxi-
mately 18% versus before supplementation values [31]. Conclusions In summary, our in vitro studies reveal equol binds spe-
cifically 5a-DHT and prevents increases in PSA from
LNCap cells, while the in vivo studies demonstrate that
equol decreases rat prostate size, decreases serum 5a-
DHT levels and androgen hormone action, while not
altering other circulating sex steroids or LH levels. Taken together these data suggest a potential use of
equol as a therapeutic agent for men’s prostate health. For example, benign prostate hyperplasia (BPH or pros-
tate enlargement) that is manifested by nocturia (need
to urinate frequently during the sleep-cycle), frequency
of urination during the day or night, decreased voided
urine volume, sensory urgency to urinate, inability to
start and stop a urine stream, etc., where equol may
mediate androgen hormone action of 5a-DHT and/or
androgen steroid receptor inactivation is influenced by
binding ERb in a positive manner. As mentioned earlier, equol has the ability to bind ERb. This may explain how isoflavones accumulate in prostate
tissue and prostatic fluid after oral supplementation
[19-21] and it has been shown that binding ERb in the
prostate decreases inflammation and carcinogenesis [22]. Furthermore, this not only applies to prostate health but
to other tissue-specific sites in the body where ERb is
present. For example, there is an abundance of ERb in
the brain, specifically in the frontal cortex and raphe
nuclei as well as other brain regions that may be involved
in mechanisms of anxiety and depression [23-25]. Preli-
minary data from our laboratory suggest that equol treat-
ment in aged-rats after natural ovarian failure is effective
as an anti-depressive agent at low concentrations that
significantly increase serotonin levels. Additionally, the
skin represents the largest organ of the body where ERb
is abundant in the epidermal and dermal regions, espe-
cially the keratinocytes and fibroblasts, respectively References 1. Manach C, Williamson G, Morand C, Scalbert A, Rémésy C: Bioavailability 1. Manach C, Williamson G, Morand C, Scalbert A, Rémésy C: Bioavailability
and bioefficacy of polyphenols in humans. I. Review of 97 bioavailability
studies. Am J Clin Nutr 2005, 81(Suppl):230S-242S. 1. Manach C, Williamson G, Morand C, Scalbert A, Rémésy C: Bioavailability
and bioefficacy of polyphenols in humans. I. Review of 97 bioavailability
studies. Am J Clin Nutr 2005, 81(Suppl):230S-242S. 24. Nomura M, Akama KT, Alves SE, Korach KS, Gustafsson JA, Pfaff DW,
Ogawa S: Differential distribution of estrogen receptor (ER)-alpha and
ER-beta in the midbrain raphe nuclei and periaqueductal gray in male
mouse: Predominant role of ER-beta in midbrain serotonergic systems. Neuroscience 2005, 130:445-456. and bioefficacy of polyphenols in humans. I. Review of 97 bioavailability
studies. Am J Clin Nutr 2005, 81(Suppl):230S-242S. 2. Jamison JR: Clinical Guide To Nutrition and Dietary Supplements in Disease
Amsterdam, Netherland, Churchhill Livingstone/Elsevier; 2003. 2. Jamison JR: Clinical Guide To Nutrition and Dietary Supplements in Disease
Amsterdam, Netherland, Churchhill Livingstone/Elsevier; 2003. 3. Adlercreutz H: Phyto-oestrogens and cancer. Lancet Oncol 2002, 3:364-373. 3. Adlercreutz H: Phyto-oestrogens and cancer. Lancet Oncol 2002, 3:364-373. 25. Halbreich U, Kahn LS: Selective oestrogen receptor modulators-current
and future brain and behavior applications. Expert Opin Pharmacother
2001, 1:1385-1398. 4. Setchell KD, Brown NM, Lydeking-Olsen E: The clinical importance of the
metabolite equol - a clue to the effectiveness of soy and its isoflavones. J Nutr 2002, 132:3577-3584. 26. Thornton MJ, Taylor AH, Mulligan K, Al-Azzawi F, Lyon CC, O’Driscoll J,
Messenger AG: The distribution of estrogen receptor beta is distinct to
that of estrogen receptor alpha and the androgen receptor in human
skin and the pilosebaceous unit. J Investig Dermatol Symp Proc 2003,
8:100-103. 5. Setchell KD, Clerici C, Lephart ED, Cole SJ, Heenan C, Castellani D, Wolfe BE,
Nechemias-Zimmer L, Brown NM, Lund TD, Handa RJ, Heubi JE: S-equol, a
potent ligand for estrogen receptor beta, is the exclusive enantiomeric
form of the soy isoflavone metabolite produced by human intestinal
bacterial flora. Am J Clin Nutr 2005, 81:1072-1079. 5. Setchell KD, Clerici C, Lephart ED, Cole SJ, Heenan C, Castellani D, Wolfe BE,
Nechemias-Zimmer L, Brown NM, Lund TD, Handa RJ, Heubi JE: S-equol, a
potent ligand for estrogen receptor beta, is the exclusive enantiomeric
form of the soy isoflavone metabolite produced by human intestinal
bacterial flora. Am J Clin Nutr 2005, 81:1072-1079. 27. Authors’ contributions TDL designed and performed initial binding studies, analyzed data sets and
authored portions of this paper. CB analyzed data sets and authored
portions of this paper. LB conducted and performed the blood steroid/
hormone assays in all the studies and performed portions of the in vivo
animal studies, analyzed data sets and authored portions of this paper. ANH
analyzed data sets and authored portions of this paper. EDL designed and
performed the in vitro binding assays, LNCap studies and in vivo animal
studies, analyzed data sets, authored the paper and obtained funding. All
authors read and approved the final manuscript. 15. Lund TD, Munson DJ, Adlercreutz H, Handa RJ, Lephart ED: Androgen
receptor expression in the rat prostate is down-regulated by dietary
phytoestrogens. Reprod Biol Endocrinol 2004, 2:5. TDL designed and performed initial binding studies, analyzed data sets and
authored portions of this paper. CB analyzed data sets and authored
portions of this paper. LB conducted and performed the blood steroid/
hormone assays in all the studies and performed portions of the in vivo
animal studies, analyzed data sets and authored portions of this paper. ANH
analyzed data sets and authored portions of this paper. EDL designed and
performed the in vitro binding assays, LNCap studies and in vivo animal
studies, analyzed data sets, authored the paper and obtained funding. All
authors read and approved the final manuscript. 16. Pike J, Holmes D, Kamalati T, Davies D, Tolhurst R, Mazhar D, Fishpool S, al-
Jehani R, Waxman J, Zelent A, Lemoine NR, Ali S, Buluwela L: Silencing of
androgen-regulated genes using a fusion of AR with the PLZF
transcriptional repressor. Oncogene 2004, 23:7561-7570. 17. Legg RL, Tolman JR, Lovinger CT, Lephart ED, Setchell KD, Christensen MJ:
Diets high in selenium and isoflavones decrease androgen-regulated
gene expression in healthy rat dorsolateral prostate. Reprod Biol
Endocrinol 2008, 6:57. Authors’ information 18. Pendleton JM, Tan WW, Anai S, Chang M, Hou W, Shiverick KT, Rosser CJ:
Phase II trial of isoflavone in prostate-specific antigen recurrent prostate
cancer after previous local therapy. BMC Cancer 2008, 8:132. Trent D. Lund, PhD, is president of Stoelting Co., Wood Dale, IL 60191, USA. Crystal Blake, PhD, is a research associate in Physiology and Developmental
Biology at Brigham Young University (BYU), Provo, Utah 84602, USA. Lihong
Bu, PhD, is MRDDRC Imaging Core manager and instructor, Department of
Neurobiology, Children’s Hospital Boston and Harvard Medical School,
Boston, MA 02115, USA. Amy N Hamaker, is an undergraduate student
majoring in Neuroscience at Brigham Young University, Provo, Utah 84602,
USA. Edwin D. Lephart, is professor of Physiology and Developmental
Biology and The Neuroscience Center at BYU, Provo, Utah, 84602, USA. Trent D. Lund, PhD, is president of Stoelting Co., Wood Dale, IL 60191, USA. Crystal Blake, PhD, is a research associate in Physiology and Developmental
Biology at Brigham Young University (BYU), Provo, Utah 84602, USA. Lihong
Bu, PhD, is MRDDRC Imaging Core manager and instructor, Department of
Neurobiology, Children’s Hospital Boston and Harvard Medical School,
Boston, MA 02115, USA. Amy N Hamaker, is an undergraduate student 19. Hedlund TE, Maroni PD, Ferucci PG, Dayton R, Barnes S, Jones K, Moore R,
Ogden LG, Wähälä K, Sackett HM, Gray KJ: Long-term dietary habits affect
soy isoflavone metabolism and accumulation in prostatic fluid in
caucasian men. J Nutr 2005, 135:1400-1406. majoring in Neuroscience at Brigham Young University, Provo, Utah 84602,
USA. Edwin D. Lephart, is professor of Physiology and Developmental
Biology and The Neuroscience Center at BYU, Provo, Utah, 84602, USA. 20. Rannikko A, Petas A, Rannikko S, Adlercreutz H: Plasma and prostate
phytoestrogen concentrations in prostate cancer patients after oral
phytoestogen supplementation. Prostate 2006, 66:82-87. List of abbreviations used List of abbreviations used
(5α-DHT): 5α- dihydrotestosterone; (AR): androgen receptor; (BPH): benign
prostate hyperplasia; (PSA): prostate specific antigen; (LNCap): lymph node
metastatic lesion of human prostatic adenocarcinoma; (DMSO): dimethyl Page 8 of 9 Page 8 of 9 Lund et al. Reproductive Biology and Endocrinology 2011, 9:4
http://www.rbej.com/content/9/1/4 sulfoxide; (SAM): selective androgen modulator; (ERβ): estrogen receptor
beta; (LH): luteinizing hormone; (ELISA): enzyme-linked immunosorbent
assay. prostate-specific antigen in androgen-sensitive prostate cancer cells. Endocrinology 1995, 136:796-803. gy
10. Ozasa K, Nakao M, Watanabe Y, Hayashi K, Miki T, Mikami K, Mori M,
Sakauchi F, Washio M, Ito Y, Suzuki K, Kubo T, Wakai K, Tamakoshi A: JACC
Study Group: Association of serum phytoestrogen concentration and
dietary habits in a sample set of the JACC Study. J Epidemiol 2005,
15(Suppl 2):S196-202. Competing interests
h
h
d
l
h 21. Gardner CD, Oelrich B, Liu JP, Feldman D, Franke AA, Brooks JD: Prostatic
soy isoflavone concentrations exceed serum levels after dietary
supplementation. Prostate 2009, 69:719-726. The authors declare that they have no competing interests. The authors declare that they have no competing interests. Received: 3 December 2010 Accepted: 13 January 2011 22. Ellem SJ, Risbriger GP: The dual, opposing role of estrogens in the
prostate. Ann NY Acad Sci 2009, 1155:174-186, Steroid Enzymes & Cancer. Published: 13 January 2011 Published: 13 January 2011 23. Lephart ED, Porter JP, Hedges DW, Lund TD, Setchell KD: Phytoestrogens:
implications in neurovascular research. Curr Neurovasc Res 2004,
1:455-464. implications in neurovascular research. Curr Neurovasc Res 2004,
1:455-464. Author details
1S
l i
C
W 12. Lundh TJO, Pettersson HI, Martinsson KA: Comparative levels of free and
conjugated plant estrogens in blood plasma of sheep and cattle fed
ostrogenic silage. J Agric Food Chem 1990, 38:1530-1534. 1Stoelting Co., Wood Dale, IL 60191, USA. 2The Department of Physiology
and Developmental Biology and the Neuroscience Center, Brigham Young
University, Provo, Utah 84602, USA. 3MRDDRC Imaging Core, Department of
Neurobiology, Children’s Hospital Boston and Harvard Medical School,
Boston, MA 02115, USA. 1Stoelting Co., Wood Dale, IL 60191, USA. 2The Department of Physiology
and Developmental Biology and the Neuroscience Center, Brigham Young
University, Provo, Utah 84602, USA. 3MRDDRC Imaging Core, Department of
Neurobiology, Children’s Hospital Boston and Harvard Medical School,
Boston, MA 02115, USA. 13. Common RH, Ainsworth L: Identification of equol in the urine of the
domestic fowl. Biochim Biophys Acta 1961, 53:403-404. 14. Kuhn G, Hennig U, Kalbe C, Rehfeldt C, Ren MQ, Moors S, Degen GH: Growth
performance, carcass characteristics and bioavailability of isoflavones in
pigs fed soy bean based diets. Arch Anim Nutr 2004, 58:265-276. Acknowledgements This work was supported, in part, by a grant from the United States
Department of Agriculture (USDA 2004-01811 to EDL). We thank Tim Aucoin
for his participation in the binding experiments and Frank Zeigler for
assistance with the LNCap cell culture experiment. 11. Hoikkala A, Mustonen E, Saastamoinen I, Jokela T, Taponen J, Saloniemi H,
Wähälä K: High levels of equol in organic skimmed Finnish cow milk. Mol
Nutr Food Res 2007, 51:782-786. doi:10.1186/1477-7827-9-4
Cite this article as: Lund et al.: iEquol an isoflavonoid: potential for
improved prostate health, in vitro and in vivo evidence. Reproductive
Biology and Endocrinology 2011 9:4. References Ashcroft GS, Mills SJ, Lei K, Gibbons L, Jeong MJ, Taniguchi M, Burow M,
Horan MA, Wahl SM, Nakayama T: Estrogen modulates cutaneous wound
healing by downregulating macrophage migration inhibitory factor. J Clin Invest 2003, 111:1309-1318. 6. Hounsome N, Hounsome B, Tomos D, Edward-Jones G: Changes in
Antioxidant Compounds in White Cabbage During Winter Storage. Postharvest Biol Tech 2009, 52:173-179. 6. Hounsome N, Hounsome B, Tomos D, Edward-Jones G: Changes in
Antioxidant Compounds in White Cabbage During Winter Storage. Postharvest Biol Tech 2009, 52:173-179. 7. Lund TD, Munson DJ, Haldy ME, Setchell KDR, Lephart ED, Handa RJ: Equol
is a novel anti-androgen that inhibits prostate growth and hormone
feedback. Biol Reprod 2004, 70:1188-1195. 7. Lund TD, Munson DJ, Haldy ME, Setchell KDR, Lephart ED, Handa RJ: Equol
is a novel anti-androgen that inhibits prostate growth and hormone
feedback. Biol Reprod 2004, 70:1188-1195. 28. Vij U, Kumar A: Phyto-oestrogens and prostatic growth. Natl Med J India
2004, 17:22-26. 29. Akaza H, Miyanaga N, Takashima N, Naito S, Hirao Y, Tsukamoto T, Fujioka T,
Mori M, Kim WJ, Song JM, Pantuck AJ: Comparisons of percent equol
producers between prostate cancer patients and controls: case-
controlled studies of isoflavones in Japanese, Korean and American
residents. Jpn J Clin Oncol 2004, 34:86-89. 8. Yang Q, Fung KM, Day WV, Kropp BP, Lin HK: Androgen receptor signaling
is required for androgen-sensitive human prostate cancer cell
proliferation and survival. Cancer Cell International 2005, 5:8. 9. Lee C, Sutkowski DM, Sensibar JA, Zelner D, Kim I, Amsel I, Shaw N,
Prins GS, Kozlowski JM: Regulation of proliferation and production of Page 9 of 9 Page 9 of 9 Lund et al. Reproductive Biology and Endocrinology 2011, 9:4
http://www.rbej.com/content/9/1/4 Lund et al. Reproductive Biology and Endocrinology 2011, 9:4
http://www.rbej.com/content/9/1/4 30. Magee PJ, Raschke M, Steiner C, Duffin JG, Pool-Zobel BL, Jokela T,
Wahala K, Rowland IR: Equol: a comparison of the effects of the racemic
compound with that of the purified S-enantiomer on the growth,
invasion, and DNA integrity of breast and prostate cells in vitro. Nutr
Cancer 2006, 54:232-242. 31. Tanaka M, Fujimoto K, Chihara Y, Torimoto K, Yoneda T, Tanaka N,
Hirayama A, Miyanaga N, Akaza H, Hirao Y: Isoflavone supplements
stimulated the production of serum equol and decreased the serum
dihydrotestosterone levels in healthy male volunteers. Prostate Cancer
Prostatic Dis 2009, 12:247-252. doi:10.1186/1477-7827-9-4
Cite this article as: Lund et al.: iEquol an isoflavonoid: potential for
improved prostate health, in vitro and in vivo evidence. Reproductive
Biology and Endocrinology 2011 9:4. 30. Magee PJ, Raschke M, Steiner C, Duffin JG, Pool-Zobel BL, Jokela T,
Wahala K, Rowland IR: Equol: a comparison of the effects of the racemic
compound with that of the purified S-enantiomer on the growth,
invasion, and DNA integrity of breast and prostate cells in vitro. Nutr
Cancer 2006, 54:232-242. 30. Magee PJ, Raschke M, Steiner C, Duffin JG, Pool-Zobel BL, Jokela T,
Wahala K, Rowland IR: Equol: a comparison of the effects of the racemic
compound with that of the purified S-enantiomer on the growth,
invasion, and DNA integrity of breast and prostate cells in vitro. Nutr
Cancer 2006, 54:232-242. 31. Tanaka M, Fujimoto K, Chihara Y, Torimoto K, Yoneda T, Tanaka N,
Hirayama A, Miyanaga N, Akaza H, Hirao Y: Isoflavone supplements
stimulated the production of serum equol and decreased the serum
dihydrotestosterone levels in healthy male volunteers. Prostate Cancer
Prostatic Dis 2009, 12:247-252. 31. Tanaka M, Fujimoto K, Chihara Y, Torimoto K, Yoneda T, Tanaka N,
Hirayama A, Miyanaga N, Akaza H, Hirao Y: Isoflavone supplements
stimulated the production of serum equol and decreased the serum
dihydrotestosterone levels in healthy male volunteers. Prostate Cancer
Prostatic Dis 2009, 12:247-252. doi:10.1186/1477-7827-9-4
Cite this article as: Lund et al.: iEquol an isoflavonoid: potential for
improved prostate health, in vitro and in vivo evidence. Reproductive
Biology and Endocrinology 2011 9:4. Submit your next manuscript to BioMed Central
and take full advantage of:
• Convenient online submission
• Thorough peer review
• No space constraints or color figure charges
• Immediate publication on acceptance
• Inclusion in PubMed, CAS, Scopus and Google Scholar
• Research which is freely available for redistribution
Submit your manuscript at
www.biomedcentral.com/submit
|
https://openalex.org/W3118376525
|
https://europepmc.org/articles/pmc7789152?pdf=render
|
English
| null |
Contribution of sarcomere gene mutations to left atrial function in patients with hypertrophic cardiomyopathy
|
Cardiovascular ultrasound
| 2,021
|
cc-by
| 7,964
|
Contribution of sarcomere gene mutations
to left atrial function in patients with
hypertrophic cardiomyopathy Hyemoon Chung1†, Yoonjung Kim2†, Chul Hwan Park3, In-Soo Kim4, Jong-Youn Kim4, Pil-Ki Min4,
Young Won Yoon4, Tae Hoon Kim4, Byoung Kwon Lee4, Bum-Kee Hong4, Se-Joong Rim4, Hyuck Moon Kwon4,
Kyung-A Lee2* and Eui-Young Choi4* * Correspondence: KAL1119@yuhs.ac; choi0928@yuhs.ac * Correspondence: KAL1119@yuhs.ac; choi0928@yuhs.ac p
y
y
†Hyemoon Chung and Yoonjung Kim contributed equally to this work. 2Department of Laboratory Medicine, Gangnam Severance Hospital, Yonsei
University College of Medicine, 211 Eonju-Ro, Gangnam-Gu, Seoul 06273,
Republic of Korea p
4Division of Cardiology, Gangnam Severance Hospital, Yonsei University
College of Medicine, 211 Eonju-Ro, Gangnam-Gu, Seoul 06273, Republic of
Korea Full list of author information is available at the end of the article Abstract Background: Left atrial (LA) enlargement and dysfunction are related to clinical course in patients with hypertrophic
cardiomyopathy (HCM). We aimed to investigate genetic contribution to LA structural and functional remodeling. Methods: Two hundred twelve patients were consecutively enrolled, and echocardiography and extensive genetic
analysis were performed. Cardiac magnetic resonance (CMR) was performed in 135 patients. Echocardiography was
also performed in controls (n = 30). Results: Patients with HCM had lower late-diastolic mitral annular velocity (a’) and higher LA volume index (LAVI)
than controls. Patients with pathogenic or likely pathogenic sarcomere gene mutations (PSM, n = 67, 32%) had
higher LAVI and lower CMR-derived LA total emptying fraction (37.0 ± 18.5 vs. 44.2 ± 12.4%, p = 0.025). In patients
without AF (n = 187), the PSM had lower a’ (6.9 ± 2.0 vs. 7.8 ± 1.9 cm/s, p = 0.004) than others. The PSM had
higher prevalence and amount of late gadolinium enhancement (LGE) in the left ventricle (LV). In multivariate
analysis, PSM was significantly related to lower a’ independent of E/e’, LV mass index, and LAVI. However, the
relation significantly attenuated after adjustment for the extent of LGE in the LV, suggesting common myopathy
in the LV and LA. In addition, PSM was significantly related to lower LA total emptying fraction independent of
age, E/e’, s’, LV ejection fraction, LV myocardial global longitudinal strain and %LGE mass. Conclusions: PSM was related to LA dysfunction independent of LV filling pressure and LAVI, suggesting its
contribution to atrial myopathy in HCM. Keywords: Hypertrophic cardiomyopathy, Genetic mutation, Left atrial function Chung et al. Cardiovascular Ultrasound (2021) 19:4
https://doi.org/10.1186/s12947-020-00233-y Chung et al. Cardiovascular Ultrasound (2021) 19:4
https://doi.org/10.1186/s12947-020-00233-y Chung et al. Cardiovascular Ultrasound
https://doi.org/10.1186/s12947-020-00233-y Open Access DNA preparation, library construction and sequencing of
HCM gene panel and mtDNA
The details are described in Supplementary Method 1 to
3. Identification of potential pathogenic mtDNA variants
Mitochondrial genome databases, including MITOMAP
[8]
and
Human
Mitochondrial
Genome
Database
(mtDB) [9] and Phylotree [10] were referred to validate
the detected variants. Novel and rare non-haplogroup-
associated variants were further evaluated for their po-
tential pathogenicity based on the variant’s location,
changes in the amino acid sequence, and evolutionary
conservation [11]. We have assessed potential pathogen-
icity using multiple software programs including Poly-
phen2, Fathmmw, Mutation Assessor, and PROVEAN. When the majority of computational evidence supported
a deleterious effect, we have assigned novel and rare
non-haplogroup-associated
variants
as
damaging
mtDNA variants [12]. Data analysis of mitochondrial
genome are described in Supplementary Method 4. © The Author(s). 2021 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License,
which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give
appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if
changes were made. The images or other third party material in this article are included in the article's Creative Commons
licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons
licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain
permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the
data made available in this article, unless otherwise stated in a credit line to the data. Chung et al. Cardiovascular Ultrasound (2021) 19:4 Page 2 of 11 Chung et al. Cardiovascular Ultrasound (2021) 19:4 Background HCM-specific panel, consisting of 82 nuclear DNAs
(nDNAs: 33 sarcomere-associated genes, 5 phenocopy
genes, and 44 nuclear genes linked to mitochondrial car-
diomyopathy) and 37 mitochondrial DNAs (mtDNAs),
was analyzed (Supplementary Table S1). g
Left atrial (LA) enlargement and dysfunction are related
to the development of atrial fibrillation (AF) and exercise
intolerance in patients with hypertrophic cardiomyop-
athy (HCM) [1–3]. LA dysfunction is usually related to
chronic structural remodeling owing to long-standing
pressure and volume overload. In addition, LA structural
remodeling such as enlargement or fibrosis is a marker
of development of AF in HCM [4]. Recent studies also
showed that LA function measured by strain imaging is
an independent predictor of poor outcome in patients
with HCM [1, 2]. However, it has been reported that in
HCM, LA dysfunction can occur as a myopathic process
itself, irrespective of loading conditions [4, 5]. In this re-
gard, genetic contribution to LA structural and func-
tional remodeling has not been well investigated. Tissue
Doppler imaging-based late mitral annular velocity (a’)
and LA phasic function have been validated as good LA
functional parameters reflecting active contraction in
several studies and are widely used [6]. Herein, we
sought to evaluate the relationship between related gene
mutations and LA function in patients with HCM. In
addition, we aimed to study the direct effect on LA func-
tion as a process of LA myopathy, independent of LA
loading conditions. DNA preparation, library construction and sequencing of
HCM gene panel and mtDNA
The details are described in Supplementary Method 1 to
3. Materials and methods
Study population y p p
A total of 432 patients treated at a single center were en-
rolled in an HCM Registry from 2006 to 2014. Among
them, 220 patients were excluded owing to insufficient
data, follow-up loss, or declining study enrollment. Fi-
nally, 212 patients underwent genetic testing. The pa-
tients had maximal left ventricular (LV) hypertrophy
greater than 13 mm and a ratio of maximal thickness to
posterior wall thickness greater than 1.3 without an
underlying cause of hypertrophy, such as uncontrolled
hypertension or aortic stenosis. Patterns of LV hyper-
trophy were classified as ApHCM and non-ApHCM
(asymmetrical hypertrophy, diffuse hypertrophy, and
focal segmental hypertrophy). All patients underwent
screening for Fabry disease and were confirmed negative
for the galactosidase alpha variant. For comparison, con-
ventional echocardiography was performed in controls. The study protocol conformed to the ethical guidelines
of the 1975 Declaration of Helsinki, and it was approved
by institutional review board of Gangnam Severance
Hospital (3–2015-0019). Written informed consent was
obtained from all participants. Data analysis of the HCM gene panel
The Burrows–Wheeler aligner algorithm with default
parameters was used to align reads to the human refer-
ence genome sequence GRCh37 [13]. SAMTools was
used to convert the sequence alignment map file to the
BAM format [14]. Duplicates were sorted and removed
using the Picard tool (http://broadinstitute.github.io/
picard/). The Genome Analysis Toolkit was used for
indel realigning and base-quality score re-calibration
[15]. Variants were further filtered with altered allele fre-
quency > 30%,
50×
coverage. For
33
HCM
genes,
annotated variants using ANNOVAR [16] and Variant
Effect Predictor (http://asia.ensembl.org/info/docs/tools/
vep/index.html) were classified as pathogenic and likely
pathogenic based on refined American College of Med-
ical Genetics and Genomics (ACMG) standards and
guidelines for inherited cardiac conditions [12, 17–19]. For 44 mitochondria-related nuclear DNA genes (reces-
sive conditions), annotated variants were classified as
pathogenic and likely pathogenic based on ACMG
guidelines [17]. And we adapted gnomAD AF cutoff
0.01% as the moderate level of evidence supporting
pathogenicity (ACMG/AMP criterion PM2) based on
maximum credible population AF (http://cardiodb.org/
allelefrequencyapp/) [17, 18]. Genetic testing and analysis
HCM gene panel (nDNA) and mtDNA design
HCM genes consisted of 8 validated sarcomere genes
and 25 putative HCM genes [7]. A comprehensive Genetic testing and analysis
HCM gene panel (nDNA) and mtDNA design HCM gene panel (nDNA) and mtDNA design
HCM genes consisted of 8 validated sarcomere genes
and 25 putative HCM genes [7]. Echocardiographic analysis equation,
as
follows:
4 × (peak
velocity
across
the
LVOT)2 [21]. It was measured at rest and during Val-
salva maneuver. LVOT obstruction was defined as a sys-
tolic pressure gradient ≥30 mmHg. g
p
y
Comprehensive echo-Doppler evaluation was performed
according to the current American Society of Echocardi-
ography guidelines [20]. LV ejection fraction was mea-
sured by biplane Simpson’s method. LA volume was
measured at the end-systole by the ellipsoidal method,
and LA volume index was calculated as LA volume/body
surface area (BSA). Peak early (E) and late (A) diastolic
mitral inflow velocities were measured in apical four-
chamber view. The filter was set to exclude high fre-
quency signal, and the Nyquist limit was adjusted to a
range of 15 to 20 cm/s. Gain and sample volume were
minimized to allow for a clear tissue signal with minimal
background noise. Systolic (s’), early (e’) and late (a’) dia-
stolic velocities of the mitral annulus were measured
from the apical 4-chamber view with a 2- to 5-mm sam-
ple volume placed at the septal corner of the mitral an-
nulus. (Fig. 1) The ratio of E/e′ was calculated. LV wall
thickness was measured in all cross-sectional planes. Continuous wave Doppler was used to measure the peak
velocity across the LV outflow tract (LVOT), and the
pressure gradient was calculated using the Bernoulli Materials and methods
Study population A comprehensive Page 3 of 11 Chung et al. Cardiovascular Ultrasound (2021) 19:4 Chung et al. Cardiovascular Ultrasound (2021) 19:4 Cardiac magnetic resonance imaging (CMR) and LV
chamber performance assessment CMR was performed using a 1.5-T scanner (Magnetom
Avanto®; Siemens Medical Solutions, Erlangen, Germany)
with a phased array body coil. (Supplementary Method 5). Genetic characteristics Genetic characteristics
Based on the American College of Medical Genetics
and Genomics guidelines, [19] 67 of 212 (32%) pa-
tients had 71 pathogenic or likely pathogenic variants
in 33 sarcomere-associated genes (33 MYBPC3, 19
MYH7, 14 TNNI3, 2 MYH6, 1 JPH2, 1 TNNC1, and 1
MYL3). Four patients harbored more than one vari-
ants in HCM genes. Homozygous or compound het-
erozygous variants in MYBPC3 were identified in one
patient and co-variants were identified in three pa-
tients (two had MYBPC3 and MYH7, and one had
MYBPC3 and JPH2). In total, 26 (12%) patients had
probably damaging mtDNA variants (one patient was
missed due to non-analysis of mtDNA), 15 (7%) had
mitochondria-related nDNA variants, and 1 had a
pathogenic variant in GAA. Seven patients (3%) had
both
pathogenic
sarcomere
gene
variants
and
mitochondria-related variants. Detailed genetic vari-
ants
of
detected
sarcomere
genes,
mitochondria-
related nDNAs and damaging mtDNAs are shown in
Supplementary Table S2, S3 and S4 [12]. The patients
with pathogenic or likely pathogenic sarcomere gene
mutations (PSM) had higher prevalence of LGE (90%
vs. 63% vs. 50%, p = 0.001) and number of LGE-
involved segments (4.9 ± 2.8 vs. 2.9 ± 3.3 vs. 2.6 ± 3.8,
p = 0.005) than patients without any mutations and
those with mitochondria-related mutations. Extent of LGE assessment From the LGE images, the LV was divided into 16 seg-
ments [22]. Presence of LGE involvement in each seg-
ment and the total number of LGE-involving segments
were measured. In addition, the percentage of LGE in
LV mass were measured using dedicated quantitative
analysis
software
(Qmass®MR
8.1,
Medis,
Leiden,
Netherland) using LGE images with PSIR sequence. To
improve the reproducibility, experienced radiologist and
cardiologist with more than 10 years of experience ana-
lyzed the LGE sizes. In each short-axis slice image, the Fig. 1 Tissue Doppler index (right panel) and phasic left atrial function measurement with cardiac magnetic resonance imaging (right panel) in a
patient with pathogenic sarcomere gene mutation (TNNI3 mutation, a, b) and without sarcomere mutations (c, d) Fig. 1 Tissue Doppler index (right panel) and phasic left atrial function measurement with cardiac magnetic resonance imaging (right panel) in a
patient with pathogenic sarcomere gene mutation (TNNI3 mutation, a, b) and without sarcomere mutations (c, d) ig. 1 Tissue Doppler index (right panel) and phasic left atrial function measurement with cardiac magnetic resonance imag
atient with pathogenic sarcomere gene mutation (TNNI3 mutation, a, b) and without sarcomere mutations (c, d) Fig. 1 Tissue Doppler index (right panel) and phasic left atrial function measurement with cardiac magnetic resonance i
patient with pathogenic sarcomere gene mutation (TNNI3 mutation, a, b) and without sarcomere mutations (c, d) Page 4 of 11 Chung et al. Cardiovascular Ultrasound (2021) 19:4 Chung et al. Cardiovascular Ultrasound (2021) 19:4 performed with variables with p < 0.05 in univariate
analysis to check the independence of the variables. All statistical analyses were performed using SPSS
version 25.0 (IBM Corp., Armonk, NY, USA). A two-
sided
p-value
< 0.05
was
considered
statistically
significant. boundaries of contrast-enhanced areas were automatic-
ally traced. On LGE-MR images, myocardium with ab-
normal enhancement was defined as an area of hyper-
enhancement more than five standard deviations from
the remote myocardium. Remote myocardium was de-
fined as non-enhanced myocardium, opposite of the
hyper-enhanced myocardium [23]. The maximum signal
was determined by computer-assisted window threshold-
ing of the enhanced area. Obvious artifacts, such as
those caused by motion, were excluded using a tool from
the software package. Total LGE amount was calculated
by summation of all slice volumes of enhancement. Myocardial strain analysis using feature tracking CMR Myocardial strain analysis using feature tracking CMR
Myocardial strain analysis using feature tracking CMR
was performed in 135 patients with semi-automated
software (Qstrain®MR 2.0, Medis, Leiden, Netherland). LV endocardial borders were manually drawn at a ref-
erence frame. LV endocardial and epicardial borders
were manually traced in 2-, 3-, 4-chamber long-axis
views at end-systolic and end-diastolic phase. LV glo-
bal longitudinal strain (GLS) was obtained from aver-
aging longitudinal strains of apical 2-, 3-, and 4-
chamber view. LA endocardial border was manually
traced in 4-chamber long-axis view using LV end-
diastole as reference phase. LA global longitudinal
strain is defined as the average peak strain value. LA
maximal, pre-contraction (pre-A in cases without AF)
and minimal volume were measured [24]. LA total
emptying fraction was calculated as (LA maximal vol-
ume – LA minimal volume) / LA maximal volume. LA reservoir fraction, as (LA maximal volume – LA
minimal volume / LA minimal volume), LA conduit
fraction as (LA maximal volume – LA pre-A volume)
/ LA maximal volume, and LA active emptying frac-
tion as (LA pre-A volume – LA minimal volume) /
LA pre-A volume [6]. (Fig. 1). Baseline characteristics The mean age of the patients was 59 ± 14 years, and 63
(30%) of them were female. Of the total patients, 49
(23%) had obstructive HCM; 100 patients (47%) had
ApHCM, and 64 (64%, 64/100) of these had pure-type
ApHCM. Patients with HCM had higher LA volume
index and lower a’ than those of age- and sex- matched
controls. CMR was performed in 135 patients (LGE in
133 patients). The mean LV mass index was 85.7 ± 22.9
g/m2; 92 (69%) patients showed evidence of LGE. The
average number of LGE segments and %LGE mass were
3.5 ± 3.4 and 8.0 ± 9.7%, respectively (Table 1). Statistical analysis Continuous variables with normal distributions are re-
ported as the mean ± standard deviation or 95% confi-
dence interval. Student’s t-tests were used to compare
the means of continuous variables that were approxi-
mately normally distributed between the two groups. Normality was determined using the Shapiro–Wilk
test. Categorical variables are reported as counts (or
percentages)
and
were
compared
using
chi-square
tests. For comparison of more than two groups, ana-
lysis of variance was performed with post-hoc analysis
(LSD) for subgroup comparison. For the correlates of
LA function, Pearson’s correlation coefficients were
determined and Pearson’s correlation analyses were
performed. For the multivariable analysis, linear re-
gression analysis or logistic regression analysis was Effects of sarcomere mutations on LA function
The PSM (n = 67, 32%) had higher LA volume index
(41.0 ± 22.0 vs. 34.9 ± 15.0 mL/m2, p = 0.045) and ten-
dency of higher prevalence of AF (16.4% vs. 9.7%, p =
0.156). Even in patients without AF (n = 187), PSM had Chung et al. *p < 0.05; †p < 0.01, compared with total patients with hypertrophic cardiomyopathy; ‡p, comparisons between patients with and without sarcomere gene
mutations; HCM hypertrophic cardiomyopathy; ApHCM apical HCM; AF atrial fibrillation; NSVT non-sustained ventricular tachycardia; SCD sudden cardiac death; LV
left ventricle; LVEDV LV end-diastolic volume; LVESV LV end-systolic volume; LVEF LV ejection fraction; LA left atrial; e’ early diastolic mitral annular velocity; a’ late
diastolic mitral annular velocity; s’, systolic mitral annular tissue velocity; E early diastolic transmitral inflow velocity; A late diastolic transmitral inflow velocity; LGE
late gadolinium enhancement, GLS global longitudinal strain Statistical analysis Cardiovascular Ultrasound (2021) 19:4 Page 6 of 11 (7.13 ± 1.96 vs. 7.68 ± 1.96 cm/s, p = 0.099) compared
with those without thick filament mutations (n = 141). (7.13 ± 1.96 vs. 7.68 ± 1.96 cm/s, p = 0.099) compared
with those without thick filament mutations (n = 141). significantly lower A (61.9 ± 19.6 vs. 72.4 ± 20.8 cm/s,
p = 0.001) and a’ (6.9 ± 2.0 vs. 7.8 ± 1.9 cm/s, p = 0.004). PSM had lower LA total emptying fraction by CMR than
others (44.2 ± 12.4 vs. 37.0 ± 18.5%, p = 0.025). However,
patients with only mitochondria-related mutations had
higher a’, compared with PSM (8.6 ± 3.3 vs. 6.9 ± 2.0 cm/
s, p < 0.001) (Fig. 2). Within PSM, those with TNNI3
mutation and MYH6 had lower a’, A and higher LA vol-
ume index than others (Fig. 3). Patients with thick fila-
ment mutations (n = 46) had a lower A (64.0 ± 19.0 vs. 71.0 ± 21.2 cm/s, p = 0.048) and a tendency to lower a’ Statistical analysis Cardiovascular Ultrasound (2021) 19:4 Page 5 of 11 Table 1 Clinical, echocardiographic, electrocardiographic, and cardiac magnetic resonance imaging findings in controls and patients
with hypertrophic cardiomyopathy according to sarcomere gene mutations
Control
(n = 30)
HCM (n = 212)
Presence of sarcomere protein gene variant
group (n = 67)
Absence of sarcomere protein gene variant group
(n = 145)
‡p
Age, years
58.9 ±
2.6
54.8 ± 14.3
61.3 ± 12.8
0.001
Women, n (%)
10 (33)
25 (37)
38 (26)
0.106
Hypertension, n (%)
11 (37)*
28 (42)
91 (63)
0.004
Diabetes, n (%)
5 (17)
12 (18)
27 (19)
0.884
Persistent AF at echo, n (%)
11 (16)
14 (10)
0.156
NSVT at 24-h Holter, n (%) (n = 123)
10 (22)
7 (9)
0.049
5-year SCD risk, % (n = 123)
2.64 ± 1.51
2.01 ± 1.66
0.040
Echocardiography
ApHCM, n (%)
21 (31)
79 (55)
0.001
Dynamic obstruction, n (%)
13 (19)
36 (25)
0.370
LV ejection fraction, %
65.7 ±
3.8
63.8 ± 7.3
64.9 ± 5.8
0.239
LA volume index, mL/m2
19.8 ±
3.4†
41.0 ± 22.0
34.9 ± 15.0
0.045
e’, cm/s
6.9 ± 1.5†
5.1 ± 1.8
5.1 ± 1.7
0.960
a’, cm/s
10.3 ±
1.7†
6.9 ± 2.0
7.8 ± 1.9
0.004
s′, cm/s
8.8 ± 1.6†
6.6 ± 1.6
6.8 ± 1.8
0.408
E, cm/s
64.5 ±
14.1
71.0 ± 23.2
68.7 ± 18.0
0.463
A, cm/s
75.7 ±
15.9
61.9 ± 19.6
72.4 ± 20.8
0.001
E/e′
9.8 ± 2.8†
15.4 ± 6.7
14.6 ± 5.5
0.371
Maximal thickness, mm
19.7 ± 3.7
18.6 ± 3.4
0.035
CMR (n = 135)
LVEDV, mL
133.6 ± 28.6
135.7 ± 28.2
0.698
LVESV, mL
51.7 ± 23.1
46.9 ± 18.5
0.208
LV mass index, g/m2
88.0 ± 21.1
84.8 ± 23.8
0.456
LV mass/volume ratio
1.19 ± 0.32
1.14 ± 0.32
0.469
Presence of LGE in LV (n = 133)
37 (90)
55 (60)
<
0.001
Number of LGE segments in
LV(n = 133)
4.9 ± 2.8
2.9 ± 3.5
0.002
% LGE amount of LV(n = 133)
10.6 ± 10.1
6.4 ± 9.3
0.040
LV myocardial GLS, %
−14.6 ± 5.7
−16.2 ± 4.0
0.100
LV endocardial GLS, %
−17.5 ± 5.5
−20.6 ± 4.2
0.058
LA minimal volume index, mL/m2
49.0 ± 35.9
37.3 ± 20.4
0.053
LA GLS, %
17.7 ± 11.5
19.8 ± 9.1
0.288
LA total emptying fraction, %
37.0 ± 18.5
44.2 ± 12.4
0.025
LA reserve fraction, %
73.1 ± 52.1
89.3 ± 50.8
0.090
LA conduit fraction, % (n = 119)
20.0 ± 11.7
21.8 ± 8.3
0.433
LA active emptying fraction, %
(n = 119)
27.0 ± 14.8
31.3 ± 11.3
0.093
*p < 0.05; †p < 0.01, compared with total patients with hypertrophic cardiomyopathy; ‡p, comparisons between patients with and without sarcomere gene
mutations; HCM hypertrophic cardiomyopathy; ApHCM apical HCM; AF atrial fibrillation; NSVT non-sustained ventricular tachycardia; SCD sudden cardiac death; LV
left ventricle; LVEDV LV end-diastolic volume; LVESV LV end-systolic volume; LVEF LV ejection fraction; LA left atrial; e’ early diastolic mitral annular velocity; a’ late
diastolic mitral annular velocity; s’, systolic mitral annular tissue velocity; E early diastolic transmitral inflow velocity; A late diastolic transmitral inflow velocity; LGE
late gadolinium enhancement, GLS global longitudinal strain Table 1 Clinical, echocardiographic, electrocardiographic, and cardiac magnetic resonance imaging findings in controls and patients
with hypertrophic cardiomyopathy according to sarcomere gene mutations *p < 0.05; †p < 0.01, compared with total patients with hypertrophic cardiomyopathy; ‡p, comparisons between patients with and without sarcomere gene
mutations; HCM hypertrophic cardiomyopathy; ApHCM apical HCM; AF atrial fibrillation; NSVT non-sustained ventricular tachycardia; SCD sudden cardiac death; LV
left ventricle; LVEDV LV end-diastolic volume; LVESV LV end-systolic volume; LVEF LV ejection fraction; LA left atrial; e’ early diastolic mitral annular velocity; a’ late
diastolic mitral annular velocity; s’, systolic mitral annular tissue velocity; E early diastolic transmitral inflow velocity; A late diastolic transmitral inflow velocity; LGE
late gadolinium enhancement, GLS global longitudinal strain Chung et al. Correlates and independent relationship of sarcomere
mutations to LA function Patients with ApHCM had higher a’ than that of patients
without ApHCM. In univariate analysis, a’ was signifi-
cantly correlated to LA volume index, E/e’, e’, and s’. In
addition, LV mass index, LA minimal volume, LA total
emptying fraction, LA active emptying fraction, 5-year Fig. 2 Comparison of prevalence of atrial fibrillation (a), left atrial volume index (b), and a’ velocity (c) and left atrial total emptying fraction (d)
between sarcomere gene mutation-positive group and mutation-negative group. Comparison of left atrial volume index (e), and a’ (f) velocity
between sarcomere- and mitochondria-related gene mutation groups*. HCM, hypertrophic cardiomyopathy; AF, atrial fibrillation, LA, left atrial, TDI,
tissue Doppler imaging; a’, peak late-diastolic septal mitral annular; CMR, cardiac magnetic resonance imaging; bar indicates standard error; *one
patient was missed due to non-analysis of mtDNA Fig. 2 Comparison of prevalence of atrial fibrillation (a), left atrial volume index (b), and a’ velocity (c) and left atrial total emptying fraction (d)
between sarcomere gene mutation-positive group and mutation-negative group. Comparison of left atrial volume index (e), and a’ (f) velocity
between sarcomere- and mitochondria-related gene mutation groups*. HCM, hypertrophic cardiomyopathy; AF, atrial fibrillation, LA, left atrial, TDI,
tissue Doppler imaging; a’, peak late-diastolic septal mitral annular; CMR, cardiac magnetic resonance imaging; bar indicates standard error; *one
patient was missed due to non-analysis of mtDNA Chung et al. Cardiovascular Ultrasound (2021) 19:4 Page 7 of 11 Fig. 3 Comparisons of a’ (a), A (b) velocity and left atrial volume index (c) within patients with individual sarcomere gene mutations Fig. 3 Comparisons of a’ (a), A (b) velocity and left atrial volume index (c) within patients with individual sarcomere gene mutations mparisons of a’ (a), A (b) velocity and left atrial volume index (c) within patients with individual sarcomere gene mutations conditions. Although we did not measure LA fibrosis by
LGE quantification, this finding suggests that sarcomere
gene mutation may contribute to both LV and LA my-
opathy, and may then induce LA dysfunction. This
sarcomere gene mutation- related LA dysfunction is not
just through elevated LV filling pressure, because even
after adjustment for E/e’, LA volume index, and LV mass
index, sarcomere gene mutation-related LA dysfunction
remained significant. Correlates and independent relationship of sarcomere
mutations to LA function In addition, the findings that PSM
were significantly younger than those in the non-
sarcomere-related group and the LA function of patients
with HCM was significantly lower than controls despite
same age and sex distribution, suggest that LA dysfunc-
tion in HCM is a unique myopathic process and that
sarcomere gene mutation significantly contribute to LA
dysfunction. Previous basic studies showed that sarco-
mere gene mutation directly induces myocyte hyper-
trophy, myocyte disarray, and fibrosis in the LV in HCM
mouse models [25]. Some studies showed that TGF β1-
medicated LA fibrosis also developed in sarcomere gene
mutation-derived mouse HCM models only with dys-
functional LV [25], which supports our speculation of a
common myopathic process in the LV and LA. This
finding also suggests that current anti-hypertrophy and
anti-fibrotic agents [26] would affect LA structural and
functional remodeling, thereby reducing AF. In our
study, mitochondria-related mutations were not signifi-
cantly related to LA dysfunction and LV fibrosis; there-
fore, phenotypic contribution and myocardial fibrosis is
mainly contributed by pathogenic sarcomere gene muta-
tions. Despite
the
small
number
of
subjects
with
validated pathogenic sarcomere gene mutations in our
study, the degree of genetic contribution to LA function
seemed to vary between mutations. According to our
study, patients with TNNI3 mutation had worse LA
function than patients with MYBPC3 mutation did,
whereas mitochondria-related gene mutations showed
more benign phenotypes in terms of LA function. Therefore,
gene-targeted
therapy
needs
to
be
individualized. Prevention of sudden cardiac death due to ventricular risk of sudden cardiac death, and the presence and num-
ber of LGE segments were significantly correlated to a’
but not with LV global longitudinal strain. In multivari-
ate analysis, PSM was significantly related to a’ inde-
pendent of the presence of ApHCM, E/e’, s’, and LA
volume index. When using the definition of a’ < 6 cm/s
(less than 3rd quartile value) as LA dysfunction, PSM
was also significantly correlated to LA dysfunction inde-
pendent of ApHCM, E/e’, s’ and LA volume index by
echocardiography (Table 2). However, the relation sig-
nificantly attenuated (p > 0.05) after adjustment for the
presence of LGE or number of LGE segment in LV,
which suggests common myopathy in LV and LA. In
subgroup who underwent CMR, LA total emptying frac-
tion was significantly correlated to age, E/e’, s’, a’, %LGE
mass, LV ejection fraction by CMR, LV myocardial GLS
and presence of sarcomere mutations. Correlates and independent relationship of sarcomere
mutations to LA function In multivariate
analysis, presence of sarcomere mutations was signifi-
cantly correlated to LA total emptying fraction inde-
pendent of age, E/e’, s’, LV ejection fraction by CMR, LV
myocardial GLS and %LGE mass (Table 3). Discussion This study shows the following major findings. First, LA
function was significantly reduced in patients with
HCM, compared with controls despite same age and sex
distribution. Second, among patients with HCM, the
PSM had significantly reduced LA function (measured
by a’ in echo and LA total emptying fraction by CMR),
compared with patients without sarcomere gene muta-
tions. Third, worse LA function in the PSM was inde-
pendent of the LA volume index, LV mass index, and E/
e’. Fourth, LA function, measured by a’, was significantly
correlated to the extent of LGE in LV, and the PSM had
significantly higher extent of LGE. Although, the rela-
tionship between sarcomere gene mutation and a’ was
significantly attenuated after adjustment for the extent
of LGE in LV, the LA total emptying fraction was signifi-
cantly correlated to sarcomere gene mutation independ-
ent of %LGE mass, which suggests that LA dysfunction
is a unique finding of HCM as an LA myopathy related
to sarcomere gene mutations independent of LA loading Prevention of sudden cardiac death due to ventricular
tachycardia is a major goal of management of HCM Chung et al. Discussion Second, owing to
lack of long-term follow-up data, the relationship between
LA dysfunction and future development of new-onset AF
or heart failure was not proved. However, several previous
studies showed that LA structural or functional remodel-
ing was significantly related to poor outcomes, especially
to the development of new-onset AF and heart failure [1,
2]. The findings of this study could help elucidate this re-
lationship. In addition, our results attempt to explain the
reason behind the worse prognosis of PSM among those
with HCM. Third, due to limited number of CMR, LA ac-
tive emptying fraction did not reach the statistical differ-
ences between sarcomere groups. However, LA total
emptying fraction was significantly lower in sarcomere
mutation groups even in the limited patients, which [27]. However, several recent studies showed that the
general prognosis of HCM is better than that in previous
reports, and active primary prevention has reduced the
rate of sudden deaths related to HCM [28]. With the ad-
vancing age of patients with HCM, heart failure due to
advanced diastolic dysfunction or development of AF be-
come a major concern [29]. In fact, several studies
showed that concomitant or newly developed AF is a
major prognostic factor for heart failure admission, new-
onset stroke, and mortality in HCM [30, 31]. Therefore,
prevention of AF should be one of the goals of manage-
ment of HCM. In this regard, several new attempts for
upstream genetic modulation using CRISPR/Cas9, [32] a
sarcomere-targeted
anti-hypertrophic
or
anti-fibrotic
agent, [26] could have potential preventive effects on the
development of AF and heart failure. [27]. However, several recent studies showed that the
general prognosis of HCM is better than that in previous
reports, and active primary prevention has reduced the
rate of sudden deaths related to HCM [28]. With the ad-
vancing age of patients with HCM, heart failure due to
advanced diastolic dysfunction or development of AF be-
come a major concern [29]. In fact, several studies
showed that concomitant or newly developed AF is a
major prognostic factor for heart failure admission, new-
onset stroke, and mortality in HCM [30, 31]. Therefore,
prevention of AF should be one of the goals of manage-
ment of HCM. In this regard, several new attempts for
upstream genetic modulation using CRISPR/Cas9, [32] a
sarcomere-targeted
anti-hypertrophic
or
anti-fibrotic
agent, [26] could have potential preventive effects on the
development of AF and heart failure. Discussion Cardiovascular Ultrasound (2021) 19:4 Page 8 of 11 Table 2 Univariate and multivariate analysis for a’ velocity Table 2 Univariate and multivariate analysis for a’ velocity
a’
LA dysfunction (a’ < 6 cm/s)
Univariate analysis
Multivariate analysis
Univariate analysis
Multivariate analysis
(r)
(p)
(β)
(p)
OR (95% CI)
p
OR (95% CI)
p
Age, per year
0.102
0.166
0.98 (0.95–1.01)
0.098
Dynamic obstruction
−0.069
0.348
3.68 (1.60–8.45)
0.002
ApHCM
0.182
0.013
0.057
0.341
0.38 (0.16–0.90)
0.029
1.34 (0.39–4.57)
0.638
LVEF-by echo, per %
0.030
0.683
0.93 (0.88–0.99)
0.015
E, per cm/s
−0.260
< 0.001
26.39 (3.40–204.94)
0.002
A, per cm/s
0.291
< 0.001
0.06 (0.006–0.49)
0.009
e’, per 100 cm/s
0.232
0.001
0.74 (0.57–0.97)
0.030
E/e’
0.411
< 0.001
−0.089
0.190
1.17 (1.08–1.25)
< 0.001
1.07 (0.97–1.17)
0.182
s’, per 100 cm/s
0.592
< 0.001
0.484
< 0.001
0.47 (0.34–0.65)
< 0.001
0.52 (0.34–0.78)
0.001
LA volume index, per mL/m2
−0.418
< 0.001
−0.200
0.002
1.05 (1.03–1.07)
< 0.001
1.04 (1.01–1.07)
0.004
Presence of sarcomere mutations
0.211
0.004
−0.138
0.017
4.76 (2.05–11.02)
< 0.001
5.33 (1.71–16.61)
0.004
LV mass index by CMR (n = 135)
−0.214
0.021
1.02 (1.00–1.05)
0.046
Presence of LGE (n = 133)
−0.248
0.007
1.89 (0.49–7.21)
0.353
No of LGE segment (n = 133)
−0.334
< 0.001
1.26 (1.07–1.49)
0.006
%LGE mass (n = 133)
−0.132
0.156
1.03 (0.97–1.09)
0.312
LV myocardial GLS (n = 135)
−0.045
0.627
1.05 (0.93–1.18)
0.435
LA GLS (n = 135), per %
0.076
0.413
0.92 (0.85–0.99)
0.034
LA minimal volume (n = 135), per mL
−0.443
< 0.001
1.03 (1.02–1.05)
< 0.001
LA maximal volume (n = 135), per mL
−0.427
< 0.001
1.02 (1.01–1.04)
0.001
LA total emptying fraction (n = 135), per %
0.251
0.006
0.93 (0.89–0.97)
0.001
LA active emptying fraction (n = 119), per %
0.248
0.006
0.93 (0.88–0.97)
0.003
5-yr-SCD risk (n = 123)
−0.314
0.001
1.45 (1.06–1.99)
0.020
Presence of NSVT (n = 123)
−0.214
0.036
5.27 (1.58–17.66)
0.007
Abbreviations are same as in Table 1 LA dysfunction (a’ < 6 cm/s) involves extensive variability, and 3D-LGE is not routinely
measured in most institutions. Hence, use of the most
popular parameters with less measurement variability
would be justified. Further studies are warranted to de-
velop a more accurate and reproducible method for asses-
sing the LA myopathic process in HCM. Authors’ contributions Hyemoon Chung:Writing - Original Draft, Formal analysis, Investigation;
Yoonjung Kim: Investigation, Formal analysis Writing - Original Draft; Chul-
Hwan Park: Methodology, Validation; In-Soo Kim, Jong-Youn Kim, Pil-Ki Min,
Young Won Yoon: Writing - Review & Editing, Supervision; Tae Hoon Kim:
Methodology; Byoung Kwon Lee, Bum-Kee Hong, Se-Joong Rim, Hyuck
Moon Kwon:Supervision, Investigation; Kyung-A Lee: Methodology, Investiga-
tion, Formal analysis, Writing - Review & Editing; Eui-Young Choi:
Conceptualization, Formal analysis, Writing - Review & Editing, Funding ac-
quisition. The author(s) read and approved the final manuscript. Supplementary Information
h
l
l y
The online version contains supplementary material available at https://doi. org/10.1186/s12947-020-00233-y. Additional file 1: Method S1. DNA preparation. Method S2. Library
construction and sequencing of the HCM gene panel. Method S3. Library construction and mtDNA sequencing. Method S4. Data analysis
of the mitochondrial genome. Method S5. CMR. Table S1. Summary of
82 genes associated with hypertrophic cardiomyopathy. Supplementary
Table S2. Likely pathogenic or Pathogenic variants in the 33 sarcomere
associated genes classified according to the refined American College of
Medical Genetics and Genomics (ACMG) standards and guidelines for
inherited cardiac conditions. Supplementary Table S3. Likely
pathogenic or pathogenic variants in the 6 non-sarcomere genes and the
44 mitochondria-related nuclear genes. Supplementary Table S4. Non-
haplogroup-associated variants with a GenBank frequency < 0.1%. Additional file 1: Method S1. DNA preparation. Method S2. Library
construction and sequencing of the HCM gene panel. Method S3. Library construction and mtDNA sequencing. Method S4. Data analysis
of the mitochondrial genome. Method S5. CMR. Table S1. Summary of
82 genes associated with hypertrophic cardiomyopathy. Supplementary
Table S2. Likely pathogenic or Pathogenic variants in the 33 sarcomere
associated genes classified according to the refined American College of
Medical Genetics and Genomics (ACMG) standards and guidelines for
inherited cardiac conditions. Supplementary Table S3. Likely
pathogenic or pathogenic variants in the 6 non-sarcomere genes and the
44 mitochondria-related nuclear genes. Supplementary Table S4. Non-
haplogroup-associated variants with a GenBank frequency < 0.1%. Acknowledgements
None. Acknowledgements
None. Funding This work was supported by the Basic Science Research Program through
the National Research Foundation of Korea (NRF) funded by the Ministry of
Education (2014R1A1A2055872). Abbreviations LA: Left atrial; AF: Atrial fibrillation; HCM: Hypertrophic cardiomyopathy;
a’: Tissue Doppler imaging-based late mitral annular velocity; LV: Left
ventricular; ApHCM: Apical hypertrophic cardiomyopathy; nDNA: Nuclear
DNA; mtDNA: Mitochondrial DNA; CMR: Cardiac magnetic resonance
imaging; LGE: Late gadolinium enhancement; PSM: Patients with pathogenic
or likely pathogenic sarcomere gene mutations; E: Early diastolic transmitral
inflow velocity; s’: Tissue Doppler imaging-based systolic mitral annular tissue
velocity; GLS: Global longitudinal strain Availability of data and materials The datasets of the current study are available from the corresponding
author on reasonable request. The datasets of the current study are available from the corresponding
author on reasonable request. Conclusion Presence of pathogenic sarcomere mutation was signifi-
cantly related to LA dysfunction independent of LV fill-
ing pressure and LAVI, suggesting its contribution to
atrial myopathy in HCM. Therefore, our study results
may provide an explanation why HCM patients with
sarcomere gene mutations had poorer prognosis, regard-
ing higher prevalence of AF and hospitalization due to
heart failure [34]. Discussion This study has several limitations. First, we did not
evaluate the myopathic process of the LA itself, such as
direct tissue characterization or CMR-derived LA fibrosis
measurement [33]. However, measurement of LA-LGE Chung et al. Cardiovascular Ultrasound (2021) 19:4 Page 9 of 11 Table 3 Univariate and multivariate analysis for left atrial total emptying fraction
LA total emptying fraction (%) by CMR
Univariate analysis
Multivariate analysis
(r)
(p)
(β)
(p)
Age, per year
−0.334
< 0.001
−0.265
< 0.001
Dynamic obstruction
0.163
0.069
ApHCM
0.148
0.089
E/e’
−0.178
0.038
−0.052
0.497
s’, per cm/s
0.370
< 0.001
0.164
0.043
a’, per cm/sa
0.251
0.006
Presence of sarcomere mutations
−0.226
0.008
−0.171
0.015
LV mass index by CMR, g/m2
−0.040
0.647
Presence of LGE (n = 133)
−0.086
0.218
No of LGE segment (n = 133)
−0.172
0.048
%LGE mass (n = 133)
−0.182
0.036
0.042
0.566
LVEF by CMR, per %
0.274
0.001
0.103
0.186
LV myocardial GLS, per %
−0.537
< 0.001
−0.429
< 0.001
LA GLS, per %a
0.824
< 0.001
LA minimal volume, per mLa
−0.677
< 0.001
LA maximal volume, per mLa
−0.408
< 0.001
LA active emptying fraction (n = 119), per %a
0.879
< 0.001
aDue to collinearity to LA total emptying fraction, they were not included in the multivariate analysis. See abbreviations in Table 1 Table 3 Univariate and multivariate analysis for left atrial total emptying fraction supports our observation measured by a’ velocity. Larger
number of echocardiographic or CMR studies with phasic
LA functional analyses warrant further investigation. References Sivalokanathan S, Zghaib T, Greenland GV, Vasquez N, Kudchadkar SM,
Kontari E, Lu DY, Dolores-Cerna K, van der Geest RJ, Kamel IR, et al. Hypertrophic cardiomyopathy patients with paroxysmal atrial fibrillation
have a high burden of left atrial fibrosis by cardiac magnetic resonance
imaging. JACC Clinical electrophysiology. 2019;5(3):364–75. 23. Park CH, Chung H, Kim Y, Kim JY, Min PK, Lee KA, Yoon YW, Kim TH, Lee BK,
Hong BK, et al. Electrocardiography based prediction of hypertrophy pattern
and fibrosis amount in hypertrophic cardiomyopathy: comparative study
with cardiac magnetic resonance imaging. Int J Cardiovasc Imag. 2018;
34(10):1619–28. 5. Ko T. Left atrium as an active component of the pathophysiology in HCM. Int Heart J. 2018;59(5):906–8. 6. Blume GG, Mcleod CJ, Barnes ME, Seward JB, Pellikka PA, Bastiansen PM,
Tsang TSM. Left atrial function: physiology, assessment, and clinical
implications. Eur Heart J - Cardiovascular Imaging. 2011;12(6):421–30. 24. Li L, Chen X, Yin G, Yan W, Cui C, Cheng H, Lu M, Zhao S. Early detection of
left atrial dysfunction assessed by CMR feature tracking in hypertensive
patients. Eur Radiol. 2020;30(2):702–11. 7. Walsh R, Buchan R, Wilk A, John S, Felkin LE, Thomson KL, Chiaw TH, Loong
CC, Pua CJ, Raphael C, et al. Defining the genetic architecture of
hypertrophic cardiomyopathy: re-evaluating the role of non-sarcomeric
genes. Eur Heart J. 2017. 25. Nakajima H, Nakajima Hisako O, Salcher O, Dittiè Andrea S, Dembowsky K,
Jing S, Field Loren J. Atrial but not ventricular fibrosis in mice expressing a
mutant transforming growth factor-β1 transgene in the heart. Circ Res. 2000;86(5):571–9. 8. Brandon MC, Lott MT, Nguyen KC, Spolim S, Navathe SB, Baldi P, Wallace
DC. MITOMAP: a human mitochondrial genome database--2004 update. Nucleic Acids Res. 2005;33(Database issue):D611–3. 26. Green EM, Wakimoto H, Anderson RL, Evanchik MJ, Gorham JM, Harrison BC,
Henze M, Kawas R, Oslob JD, Rodriguez HM, et al. A small-molecule
inhibitor of sarcomere contractility suppresses hypertrophic cardiomyopathy
in mice. Science (New York, NY). 2016;351(6273):617–21. 9. Ingman M. Gyllensten U: mtDB: human mitochondrial genome database, a
resource for population genetics and medical sciences. Nucleic Acids Res. 2006;34(Database issue):D749–51. 27. Marian AJ, Braunwald E. Hypertrophic cardiomyopathy: genetics,
pathogenesis, clinical manifestations, diagnosis, and therapy. Circ Res. 2017;
121(7):749–70. 10. Chaitanya L, Ralf A, van Oven M, Kupiec T, Chang J, Lagace R, Kayser M. Simultaneous whole mitochondrial genome sequencing with short
overlapping amplicons suitable for degraded DNA using the ion torrent
personal genome machine. Received: 19 October 2020 Accepted: 14 December 2020 Received: 19 October 2020 Accepted: 14 December 2020 19. Whiffin N, Walsh R, Govind R, Edwards M, Ahmad M, Zhang X, Tayal U,
Buchan R, Midwinter W, Wilk AE, et al. CardioClassifier: disease- and gene-
specific computational decision support for clinical genome interpretation. Genetics in medicine : official journal of the American College of Medical
Genetics. 2018;20(10):1246–54. Consent for publication Consent for publication 15. McKenna A, Hanna M, Banks E, Sivachenko A, Cibulskis K, Kernytsky A,
Garimella K, Altshuler D, Gabriel S, Daly M, et al. The genome analysis
toolkit: a MapReduce framework for analyzing next-generation DNA
sequencing data. Genome Res. 2010;20(9):1297–303. All the authors read final version of manuscript and approved publication. sequencing data. Genome Res. 2010;20(9):1297–303. Ethics approval and consent to participate The study protocol conformed to the ethical guidelines of the 1975
Declaration of Helsinki, and it was approved by institutional review board of
Gangnam Severance Hospital (3–2015-0019). Written informed consent was
obtained from all participants. The study protocol conformed to the ethical guidelines of the 1975
Declaration of Helsinki, and it was approved by institutional review board of
Gangnam Severance Hospital (3–2015-0019). Written informed consent was
obtained from all participants. Page 10 of 11 Page 10 of 11 Chung et al. Cardiovascular Ultrasound (2021) 19:4 References 1. Hinojar R, Zamorano JL, Fernandez-Mendez M, Esteban A, Plaza-Martin M,
Gonzalez-Gomez A, Carbonell A, Rincon LM, Nacher JJJ, Fernandez-Golfin C. Prognostic value of left atrial function by cardiovascular magnetic
resonance feature tracking in hypertrophic cardiomyopathy. Int J Cardiovasc
Imaging. 2019. 20. Lang RM, Badano LP, Mor-Avi V, Afilalo J, Armstrong A, Ernande L,
Flachskampf FA, Foster E, Goldstein SA, Kuznetsova T, et al. Recommendations for Cardiac Chamber Quantification by
Echocardiography in Adults: An Update from the American Society of
Echocardiography and the European Association of Cardiovascular Imaging. J Am Soc Echocardiography. 2015;28(1):1–39 e14. 2. Vasquez N, Ostrander BT, Lu DY, Ventoulis I, Haileselassie B, Goyal S,
Greenland GV, Vakrou S, Olgin JE, Abraham TP, et al. Low left atrial strain is
associated with adverse outcomes in hypertrophic cardiomyopathy patients. J Am Soc Echocardiography. 2019. 2. Vasquez N, Ostrander BT, Lu DY, Ventoulis I, Haileselassie B, Goyal S,
Greenland GV, Vakrou S, Olgin JE, Abraham TP, et al. Low left atrial strain is
associated with adverse outcomes in hypertrophic cardiomyopathy patients. J Am Soc Echocardiography. 2019. 21. Sasson Z, Yock PG, Hatle LK, Alderman EL, Popp RL. Doppler
echocardiographic determination of the pressure gradient in hypertrophic
cardiomyopathy. J Am Coll Cardiol. 1988;11(4):752–6. 3. Debonnaire P, Joyce E, Hiemstra Y, Mertens Bart J, Atsma Douwe E, Schalij
Martin J, Bax Jeroen J, Delgado V, Marsan Nina A. Left atrial size and
function in hypertrophic cardiomyopathy patients and risk of new-onset
atrial fibrillation. Circ Arrhythm Electrophysiol. 2017;10(2):e004052. 3. Debonnaire P, Joyce E, Hiemstra Y, Mertens Bart J, Atsma Douwe E, Schalij
Martin J, Bax Jeroen J, Delgado V, Marsan Nina A. Left atrial size and
function in hypertrophic cardiomyopathy patients and risk of new-onset
atrial fibrillation. Circ Arrhythm Electrophysiol. 2017;10(2):e004052. 22. Cerqueira MD, Weissman NJ, Dilsizian V, Jacobs AK, Kaul S, Laskey WK,
Pennell DJ, Rumberger JA, Ryan T, Verani MS. Standardized myocardial
segmentation and nomenclature for tomographic imaging of the heart. A
Statement for Healthcare Professionals From the Cardiac Imaging
Committee of the Council on Clinical Cardiology of the American Heart
Association. 2002;105(4):539–42. 4. Sivalokanathan S, Zghaib T, Greenland GV, Vasquez N, Kudchadkar SM,
Kontari E, Lu DY, Dolores-Cerna K, van der Geest RJ, Kamel IR, et al. Hypertrophic cardiomyopathy patients with paroxysmal atrial fibrillation
have a high burden of left atrial fibrosis by cardiac magnetic resonance
imaging. JACC Clinical electrophysiology. 2019;5(3):364–75. 4. Author details
1 17. Richards S, Aziz N, Bale S, Bick D, Das S, Gastier-Foster J, Grody WW, Hegde
M, Lyon E, Spector E, et al. Standards and guidelines for the interpretation
of sequence variants: a joint consensus recommendation of the American
College of Medical Genetics and Genomics and the Association for
Molecular Pathology. 2015;17(5):405–24. 1Department of Internal Medicine, Division of Cardiology, Kyung Hee
University School of Medicine, Seoul, South Korea. 2Department of
Laboratory Medicine, Gangnam Severance Hospital, Yonsei University College
of Medicine, 211 Eonju-Ro, Gangnam-Gu, Seoul 06273, Republic of Korea. 3Department of Radiology, Gangnam Severance Hospital, Yonsei University
College of Medicine, Seoul, South Korea. 4Division of Cardiology, Gangnam
Severance Hospital, Yonsei University College of Medicine, 211 Eonju-Ro,
Gangnam-Gu, Seoul 06273, Republic of Korea. 18. Whiffin N, Minikel E, Walsh R, O'Donnell-Luria AH, Karczewski K, Ing AY,
Barton PJR, Funke B, Cook SA, MacArthur D, et al. Using high-resolution
variant frequencies to empower clinical genome interpretation. Genetics in
medicine : official journal of the American College of Medical Genetics. 2017;19(10):1151–8. Competing interests 16. Wang K, Li M, Hakonarson H. ANNOVAR: functional annotation of genetic
variants from high-throughput sequencing data. Nucleic Acids Res. 2010;
38(16):e164. The authors declare that there is no competing interest. IMPACT OF ATRIAL FIBRILLATION IN HYPERTROPHIC CARDIOMYOPATHY.
2016;67(13 Supplement):1515. 32. Mosqueira D, Mannhardt I, Bhagwan JR, Lis-Slimak K, Katili P, Scott E, Hassan
M, Prondzynski M, Harmer SC, Tinker A, et al. CRISPR/Cas9 editing in human
pluripotent stem cell-cardiomyocytes highlights arrhythmias,
hypocontractility, and energy depletion as potential therapeutic targets for
hypertrophic cardiomyopathy. Eur Heart J. 2018;39(43):3879–92. 33. Oakes Robert S, Badger Troy J, Kholmovski Eugene G, Akoum N, Burgon
Nathan S, Fish Eric N, Blauer Joshua JE, Rao Swati N, DiBella Edward VR,
Segerson Nathan M, et al. Detection and quantification of left atrial
structural remodeling with delayed-enhancement magnetic resonance
imaging in patients with atrial fibrillation. Circulation. 2009;119(13):1758–67. 34. Ho CY, Day SM, Ashley EA, Michels M, Pereira AC, Jacoby D, Cirino AL, Fox
JC, Lakdawala NK, Ware JS, et al. Genotype and lifetime burden of disease in
hypertrophic cardiomyopathy: insights from the Sarcomeric human
cardiomyopathy registry (SHaRe). Circulation. 2018;138(14):1387–98. References Hum Mutat. 2015;36(12):1236–47. 28. Weissler-Snir A, Adler A, Williams L, Gruner C, Rakowski H. Prevention of
sudden death in hypertrophic cardiomyopathy: bridging the gaps in
knowledge. Eur Heart J. 2016;38(22):1728–37. 11. Zaragoza MV, Brandon MC, Diegoli M, Arbustini E, Wallace DC. Mitochondrial cardiomyopathies: how to identify candidate pathogenic
mutations by mitochondrial DNA sequencing, MITOMASTER and phylogeny. European journal of human genetics : EJHG. 2011;19(2):200–7. 29. members ATF, Elliott PM, Anastasakis A, Borger MA, Borggrefe M, Cecchi F,
Charron P, Hagege AA, Lafont A, Limongelli G, et al. 2014 ESC guidelines on
diagnosis and management of hypertrophic cardiomyopathy: the task force
for the diagnosis and Management of Hypertrophic Cardiomyopathy of the
European Society of Cardiology (ESC). Eur Heart J. 2014;35(39):2733–79. 12. Chung H, Kim Y, Cho SM, Lee HJ, Park CH, Kim JY, Lee SH, Min PK, Yoon
YW, Lee BK, et al. Differential contributions of sarcomere and mitochondria-
related multigene variants to the endophenotype of hypertrophic
cardiomyopathy. Mitochondrion. 2020;53:48–56. 30. Kubo T, Kitaoka H, Okawa M, Hirota T, Hayato K, Yamasaki N, Matsumura Y, Yabe T,
Takata J, Doi YL. Clinical impact of atrial fibrillation in patients with hypertrophic
cardiomyopathy. Results from Kochi RYOMA study. Circulation journal : official
journal of the Japanese Circulation Society. 2009;73(9):1599–605. 13. Li H, Durbin R. Fast and accurate short read alignment with Burrows-
Wheeler transform. Bioinformatics (Oxford, England). 2009;25(14):1754–60. 31. Guerrero JCC, Jimenez-Baena E, Martinez-Martinez A, Valle-Racero J,
Francisco L-P, Lopez-Haldon J, Urbano-Moral J. REVISITING PROGNOSTIC 14. Yang J, Ding X, Sun X, Tsang SY, Xue H. SAMSVM: a tool for misalignment
filtration of SAM-format sequences with support vector machine. J
Bioinforma Comput Biol. 2015;13(6):1550025. 14. Yang J, Ding X, Sun X, Tsang SY, Xue H. SAMSVM: a tool for misalignment
filtration of SAM-format sequences with support vector machine. J
Bioinforma Comput Biol. 2015;13(6):1550025. Page 11 of 11 Chung et al. Cardiovascular Ultrasound (2021) 19:4 Chung et al. Cardiovascular Ultrasound (2021) 19:4 IMPACT OF ATRIAL FIBRILLATION IN HYPERTROPHIC CARDIOMYOPATHY. 2016;67(13 Supplement):1515. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations.
|
https://openalex.org/W4388126005
|
https://www.frontiersin.org/articles/10.3389/frsfm.2023.1257688/pdf?isPublishedV2=False
|
English
| null |
Delivery of probiotics and enzymes in self-assemblies of lipids and biopolymers based on colloidal principles
|
Frontiers in soft matter
| 2,023
|
cc-by
| 18,923
|
TYPE Review
PUBLISHED 01 November 2023
DOI 10.3389/frsfm.2023.1257688 TYPE Review
PUBLISHED 01 November 2023
DOI 10.3389/frsfm.2023.1257688 OPEN ACCESS OPEN ACCESS
EDITED BY
Kai Yang,
Soochow University, China
REVIEWED BY
Naiyan Lu,
Jiangnan University, China
Laurent Sagalowicz,
Nestle, Switzerland
*CORRESPONDENCE
Jens Risbo,
jri@food.ku.dk
Tommy Nylander,
Tommy.Nylander@fkem1.lu.se
Motomu Tanaka,
tanaka@uni-heidelberg.de
RECEIVED 12 July 2023
ACCEPTED 11 October 2023
PUBLISHED 01 November 2023 Jens Risbo1*, Tommy Nylander2,3,4,5* and Motomu Tanaka6,7* 1Department of Food Science, University of Copenhagen, Frederiksberg, Denmark, 2Division of Physical
Chemistry, Department of Chemistry, Lund University, Lund, Sweden, 3NanoLund, Lund University, Lund,
Sweden, 4LINXS Institute of Advanced Neutron and X-Ray Science, Lund, Sweden, 5School of Chemical
Engineering and Translational Nanobioscience Research Center, Sungkyunkwan University, Suwon,
Republic of Korea, 6Physical Chemistry of Biosystems, Institute of Physical Chemistry, Heidelberg
University, Heidelberg, Germany, 7Center for Integrative Medicine and Physics, Institute for Advanced
Study, Kyoto University, Kyoto, Japan RECEIVED 12 July 2023
ACCEPTED 11 October 2023
PUBLISHED 01 November 2023 Food is a complex soft matter, because various components, such as proteins,
lipids,
and
carbohydrates,
are
self-assembled
via
non-covalent,
colloidal
interactions and form hierarchical structures at multiple length scales. Soft
matter scientists have shown an increasing interest in understanding the
general principles governing the food structure formation. During the last
several decades, an increasing number of studies have shown that the
maintenance of healthy gastrointestinal tract and its microbiome is essential
for human health and wellbeing. The realization of the importance of the
gastrointestinal microbiome has led to the development of probiotics, which
are defined as living bacteria that confer a health benefit on the host. Probiotic
bacteria and enzymes can be delivered to the intestinal system by formulating
appropriate carriers and including these into food ingested by humans. Despite
this simple statement, it involves many challenges in the field of soft matter
science. This review aims to highlight how the key concepts in soft matter science
can be used to design, characterize, and evaluate self-assembled formulations of
probiotics and enzymes based on lipids and biopolymers. The topics covered in
this review includes the emulsification of oil-water mixtures, the self-assembly of
lipids and polymers at interfaces, the electrostatics and viscoelasticity of
interfaces, and the wetting/adhesion of colloidal particles. CITATION
Risbo J, Nylander T and Tanaka M (2023),
Delivery of probiotics and enzymes in
self-assemblies of lipids and biopolymers
based on colloidal principles. Front. Soft Matter 3:1257688. doi: 10.3389/frsfm.2023.1257688 COPYRIGHT
© 2023 Risbo, Nylander and Tanaka. This
is an open-access article distributed
under the terms of the Creative
Commons Attribution License (CC BY). KEYWORDS
self-assembly, lipids, polymers, colloids, probiotics, enzymes, foods self-assembly, lipids, polymers, colloids, probiotics, enzymes, foods frontiersin.org OPEN ACCESS The use, distribution or reproduction in
other forums is permitted, provided the
original author(s) and the copyright
owner(s) are credited and that the original
publication in this journal is cited, in
accordance with accepted academic
practice. No use, distribution or
reproduction is permitted which does not
comply with these terms. Frontiers in Soft Matter 1 Introduction: food science and soft matter Bile salts
have been shown to interact not only with lipids but also with a
range of biopolymers (Maldonado-Valderrama et al., 2011). Numerous strategies have been proposed to overcome these
challenges based on the understanding of colloidal aspects
(McClements, 2015). in understanding the general principles governing the food structure
formation. As Assenza and Mezzenga pointed out, the application of
theoretical and experimental tools from soft condensed matter
physics to describe foods and macronutrients is challenging
because the length scales involved is ranging from angstroms to
tens of micrometers (Assenza and Mezzenga, 2019). The use of
techniques, such as neutron and x-ray scattering, has provided new
insights (Gilbert, 2019). An important aspect that controls the nutritional value of food is
that the structure is broken down with the help of enzymes in the
gastro-intestinal tract. The structural changes are directly linked to
physiological events, as reported for digestion of lipids by Patton and
Carrey in the 1970s (Patton and Carey, 1979). Modern analytical
tools, such as synchrotron x-ray scattering, have provided new
insights on the changes in the lipid assembly structures that
occur due to lipolysis (Barauskas and Nylander, 2008; Salentinig
et al., 2011; Wadsäter et al., 2018), as well as for more complex
substrates, such as milk, under simulated in vivo gastrointestinal
tract conditions (Salentinig et al., 2013; Salentinig et al., 2015). The
structure formed during digestion depends on the lipid composition
of the food, as shown by Pham et al. for different infant formulas
(Pham et al., 2020). Such studies offer the possibility to design
human milk-like systems that control nutrient transport and the
uptake of bioactive substances. Because lipolytic processes occur
under physiological conditions involving a water soluble enzyme
acting on water insoluble lipid substrates, digestion always occurs at
the interfaces and is strongly influenced by pH and other solution
conditions (Humphreys et al., 2022). X-ray fluorescence microscopy,
X-ray and neutron scattering were applied to study the digestion of
protein gels formed from canola seed storage proteins in vitro, which
enabled to follow the process on the individual protein and the
structure connectivity (Pasquier et al., 2019). This review article is structured as follows. In Section 2, we
provide an overview of how one can characterize and modify the
surface and colloidal properties of probiotic bacteria and how
probiotic bacteria can be formulated in colloidal food materials. 1 Introduction: food science and soft matter In Section 3, some representative methodological strategies are
introduced to show how to formulate probiotics and enzymes in
lipid self-assembliess, silica-based particles, and polymer-based
microcapsules. Section 4 highlights recent results on the physical
characterization and functional delivery of probiotics and enzymes
using the above-mentioned formulations. Section 5 covers how to
assess the functions of probiotics and enzymes after delivery using
the simulator of the human gastrointestinal tract and the in vitro
model of intestinal epithelium integrated into microfludic devices. After this, review is summarized with some future perspectives and
challenges for soft matter sciences. Another key factor is the effect of diet on physiological processes
and health. Food supplements like anthocyanins from blueberries
have been demonstrated to hamper weakening of neuronal
functions but are challenging to use due to their limited
bioavailability (Tran and Tran, 2021). It should be noted
Porphyromonas gingivalis, which is a pathogen associated chronic
periodontitis also has been identified in the brain of Alzheimer’s
disease patients (Dominy et al., 2019). The use of brain-penetrant
inhibitors can reduce P. gingivalis brain infection and hamper
neurodegeneration in e.g. AD patients (Dominy et al., 2019). Again, formulation is a challenge and might possible through
food supplement. The brain contains a rich variety of lipids,
assembled to form intriguing structures for which the links to the
brain functions are not yet fully understood (Siegel et al., 1999). These structures are dependent on the cholesterol content, the
composition of salts, and temperature (Alfredsson et al., 2021). Lipidomics
studies
have
demonstrated
that
a
certain
lipid
composition pattern in the blood can be linked to the clinical
diagnosis of Alzheimer Disease (Proitsi et al., 2017). It is
tempting to assume that this in turn is related to differences in
the lipid self-assembly structure. Some of the work has been
concerned with the Alzheimer disease and the possibility to use
food
components,
such
as
extracts
from
mulberry,
as
neuroprotective agents in this respect as it is has been considered
to
inhibit
neuroinflammation
(Liu
and
Du,
2020). Our
understanding of the uptake of nutrients and drugs in the 1 Introduction: food science and soft matter Food is a complex material, consisting of macromolecules, such as proteins, lipids, and
carbohydrates, together with water, minerals, and many other minor but nutritionally
significant compounds, such as vitamins and polyphenols. These components are self-
assembled via colloidal forces and form hierarchical structures at multiple length scales. The
understanding of such forces is fundamental to not only understand the structures naturally
present in plant and animal tissues, but also how they can be disrupted and reassembled into
new structures with higher nutritional value by employing appropriate processing. It is
noteworthy that food structures are not constant but are prone to changes during harvesting,
storage, distribution, and digestion. Soft matter scientists have shown an increasing interest Frontiers in Soft Matter 01 frontiersin.org frontiersin.org Risbo et al. 10.3389/frsfm.2023.1257688 intestine is complicated by the fact that many food components,
such as fat, starch, and fibers, are insoluble. The content of the lumen
should be considered as a dispersed colloidal system rather than a
simple solution. The maintenance of healthy gastrointestinal tract
and its microbiome is essential for human health and wellbeing. The
total number of cells in the microbiome in the gastrointestinal tract
is larger than that of the human/mammalian cells and it has
therefore been suggested that the human body should be
considered a “superorganism” (Woolfson, 2016). The realization
of the importance of the gastrointestinal microbiome has led to the
development of so-called probiotics. Probiotics are defined as living
bacteria that, when administered in adequate amounts, confer a
health benefit on the host (FAO/WHO, 2001). A mechanistic
understanding of the uptake of bio-actives (comprising nutrients
and other functional components such as drugs) in the intestine also
requires knowledge of the interaction within in the mucus layer,
which is partly lacking. In particular, the link between the
interactions on the molecular, colloidal, and macroscopic levels
are often missing partly due to the lack of suitable model
systems. Some studies on mucus expressing cell cultures have
provided some new insights (Gagnon et al., 2013; Kleiveland,
2015; Ferraretto et al., 2018; Eshrati et al., 2019). The digestive
system is a challenging environment where the pH varies from pH =
1 in the stomach to pH = 8 in the intestine. Additionally, lipases and
bile salts challenge any self-assembled structure used to orally deliver
any functional ingredients (Assenza and Mezzenga, 2019). Frontiers in Soft Matter frontiersin.org 2 Surface and colloidal properties of
probiotic bacteria Probiotic bacteria belonging to either the Gram-negative or
Gram-positive groups are micron-sized, living colloidal particles
(Behnsen et al., 2013). This review primarily focuses on Gram-
positive bacteria, which are the largest group claiming to have
probiotic properties. The bacterial surface properties, formation
of biofilms on surfaces, aggregation of bacteria (in microbiology
termed auto-aggregation) are important, but also constitute the 02 frontiersin.org Risbo et al. 10.3389/frsfm.2023.1257688 FIGURE 1
Cell wall structures of Gram-positive and Gram-Negative bacteria. (A) Gram positive-bacteria highlighting the plasma membrane, membrane
proteins, thick peptidoglycan layer, S-layer proteins, teichoic acid, and polysaccharides (B) Gram-negative cell wall highlighting the plasma membrane,
membrane proteins, thin peptidoglycan layer, outer membrane with trans-membrane pores, and LPSs facing the external environment. CAP, covalently
attached protein; IMP, integral membrane protein; LP, lipoprotein; LPS, lipopolysaccharide; LTA, lipoteichoic acid; OMP, outer membrane protein;
WTA, wall teichoic acid. Adapted from (Silhavy et al., 2010), with permission from Cold Spring Harbor Publications. acteria (A) Gram positive bacteria highlighting the plasma membrane membrane Cell wall structures of Gram-positive and Gram-Negative bacteria. (A) Gram positive-bacteria highlighting the plasma membrane, membrane
proteins, thick peptidoglycan layer, S-layer proteins, teichoic acid, and polysaccharides (B) Gram-negative cell wall highlighting the plasma membrane,
membrane proteins, thin peptidoglycan layer, outer membrane with trans-membrane pores, and LPSs facing the external environment. CAP, covalently
attached protein; IMP, integral membrane protein; LP, lipoprotein; LPS, lipopolysaccharide; LTA, lipoteichoic acid; OMP, outer membrane protein;
WTA, wall teichoic acid. Adapted from (Silhavy et al., 2010), with permission from Cold Spring Harbor Publications. Revealing the structure and properties of bacterial surfaces are
non-trivial both experimentally and theoretically. Small- and wide-
angle X-ray scattering measurements have been widely used to
determine
the
supramolecular
self-assembly
of
purified
lipopolysaccharides
(LPSs)
in
suspension
experimentally
(Brandenburg et al., 1992; Garidel et al., 2005). Specular neutron
and X-ray reflectivity at the solid/water and air/water interface have
been used to further investigate the molecular-level structures in the
z-direction
perpendicular
to
the
membrane
xy-plane. The
monolayer of LPSs at the air/water interface can be used to study
the structural integrity of the outermost LPSs (Oliveira et al., 2010;
Thoma et al., 2020), whereas asymmetric bilayers formed at the
solid/aqueous interface can be used as a more realistic model of the
outer membrane, demonstrating the ion-specific collapse of the
saccharide head groups (Schneck et al., 2009b; Clifton et al.,
2013). 2 Surface and colloidal properties of
probiotic bacteria The
neutron-based
techniques
provide
structural
information with Å resolution by optimizing the contrast in the
scattering length densities, by selective deuteration and using
appropriate mixtures of H2O and D2O. However, these methods
are less sensitive to specific ions/elements. Grazing-incidence X-ray
fluorescence (GIXF) was proposed as a promising strategy to analyze
the electrostatics of LPS surfaces covered by charged saccharide head
groups because the X-ray fluorescence signals measured near the
critical angle of total reflection can be used to calculate the density
profiles of ions in the head group of LPS (Schneck et al., 2010;
Abuillan et al., 2013). It is notable that the charge density near the
LPS surface is neither uniform nor static, and charged LPS moieties
can be crosslinked using multivalent ions. These considerations
render the classical Derjaguin-Landau-Verwey-Overbeek theory
inadequate. Moreover, atomistic molecular dynamics simulations
are not suited to handle the typical length scale (>103 nm3) and
characteristic time window of diffusion-limited reactions (~10–3 s). Pink et al. employed a coarse-grained, minimal model of LPS foundation for various approaches for the formulation and the
colloidal behavior of bacteria in multiphase systems, such as
foams and emulsions. Frontiers in Soft Matter frontiersin.org 2.1 Characterization of Gram-negative
bacteria cell walls The cell wall of Gram-positive bacteria (Figure 1A) is
characterized by its main entities, namely, an inner cytoplasmic
membrane and an outer peptidoglycan layer facing the external
environment. In
contrast,
Gram-negative
bacteria
have
a
significantly thinner peptidoglycan layer that is surrounded by a
second outer phospholipid membrane, which is decorated with
poly-saccharides facing the extracellular environment (Figure 1B)
(Rohde, 2019). The peptidoglycan layer is a water-containing cross-
linked system of di-saccharides formed by N-acetylglucosamine
(NAG) and N-acetylmuramic acid (NAM), which are cross-
linked by flexible pentapeptide amino acid chains forming a
mesh-like system (Vollmer et al., 2008), and has embedded
proteins and teichoic acid. The main component of teichoic acid
is a co-polymer (poly-di-ester) of glycerol and phosphate groups,
which is negatively charged at neutral pH. Therefore, the surface of
Gram-positive bacteria is considered to be hydrophilic and negatively
charged due to the nature of the peptidoglycan layer and the teichoic
acid content. However, as discussed in the later section, the natural
variation of the cell wall composition also manifests itself by varying its
surface charge (from almost neutral to negatively charged) and
hydrophobicity. The joint electrostatic effect of the charged groups
on the bacterial surface are routinely characterized using zeta-potential
measurements, either performed at neutral (physiological) pH or as
function of pH to reveal the effect of the protonation and deprotonation
of bacterial surface groups (Jiang et al., 2023a). 03 frontiersin.org Risbo et al. 10.3389/frsfm.2023.1257688 mechanical parameters of outer membranes, such as the bending
rigidity and vertical compression modulus (Schneck et al., 2009a). Measuring the response function under the application of a shear
stress, which is called interfacial stress rheology, is another technique
that is employed to measure the complex viscoelastic modulus of
model membranes (Brooks et al., 1999; Schneider et al., 2002). Compared to the commonly used rotating disk rheometry, a small
probe-membrane contact enables sensitivity for determining the loss
and storage moduli of “soft” membranes, such as the membranes of
LPSs of Gram-negative bacteria and lipids mimicking lipoteic acid of
Gram-positive bacteria (Tanaka et al., 2004; Herrmann et al., 2015). Experiments using model systems, as compared to experiments
using living (or fixed) bacteria, enable the understanding of the
contributions of the outermost surfaces, such as the viscoelasticity of
the peptidoglycans and inner membranes, as well as the turgor
pressure pushing the cell wall from the inside. 2.1 Characterization of Gram-negative
bacteria cell walls The Lewis base character (A−) is defined as the
difference between the MATS values for chloroform and hexadecane
and the Lewis acid character (A+) is derived from the differences of
the solvent pair, acetylacetate, and decane. Using contact angle
measurements in combination with reference solvents, e.g. water,
formamide, and 1-bromo-naphtalene, and the van Oss−Good
model, the Lewis acid and base components of the bacterial
interfacial tensions can be separated and the bacterial-water
interfacial tensions can be estimated. In this approach, the
bacterial cohesive free energy density, ΔGBWB −2γBW, is the
underlying driving free energy change for bacterial aggregation. 2.1 Characterization of Gram-negative
bacteria cell walls membranes
and
calculated
the
ion
density
profiles
and
conformational changes within the framework of a linearized
Poisson-Boltzmann theorem derived by Netz (Netz, 1999). They
demonstrated that the divalent cations bound to core saccharides
form electrostatic potential barrier (Pink et al., 2003; Oliveira et al.,
2009), which explains the data obtained by GIXF and reflectivity
experiments. Microbial adhesion to hexadecane (MATH) or microbial
adhesion
to
solvents
(MATS)
are
microbiological
standard
techniques that provide indirect information about the bacterial
surface properties (Rosenberg et al., 1980; Geertsema-Doornbusch
et al., 1993). Upon vigorous shaking of a two-phase system, a
fraction of the bacteria is trapped in the organic phase and a
MATH
value
is
estimated
from
the
optical
density
(OD)
measurements. A larger MATH value means that the surface is
predominantly hydrophobic. It has been observed that a coarse and
creaming emulsion that retains a fraction of bacteria through a
Pickering mechanism is formed (Rosenberg, 1984; Rosenberg,
2006). Unlike real distribution coefficients, no simple equilibrium
thermodynamic interpretation can me made using MATH values. Yet, despite the major progress made over the past decades as
described above, there is a clear demand for experimental tools that
bridge the well-defined model systems and more complex, real
bacterial surfaces. On one hand, well-defined two-dimensional
model systems, such as LPS membranes, enable the precise
measurement of detailed structural features, ion density profiles,
and interfacial viscoelasticity. On the other hand, most of the model
systems cannot address the effect of the physico-chemical properties
of food on biological functions. Moreover, from the theoretical
viewpoint, the development of multiscale modeling that connects
atomistic simulations and coarse-scale models is required to cover
several orders of magnitude in space and time. y
g
The contact angle at the three-phase contact is related to
interfacial tensions of wetting liquid, between the liquid and the
bacterial film, and the bacterial film and air through the Young’s
equation, that is, in principle the thermodynamic quantities that can
be extracted from contact angle measurements. It should be noted
that the contact angle of the “dry” bacterial layer in contact to air
does not represent the hydrated surfaces of bacteria under water;
these measurements represent the contact angles with the “collapsed
and dry constituents” of the bacterial cell wall (Van Oss et al., 1988;
Jiang et al., 2023a). 2.2 Modification of bacterial surfaces via
chemical and physical approaches Different approaches have recently been used to change the
native surface properties of bacteria, with a particular focus on
increasing hydrophobicity and thereby the adhesion behavior of
Gram-positive bacteria to promote the ability of bacteria to be active
in building structure and texture of food by for example be active
Pickering stabilizers at surfaces and interphases. Three approaches
can be implemented, namely i) adsorption of macro molecules, ii)
chemical grafting of molecules to the bacterial wall components, and
iii) selective removal of chemical constituents from the bacterial cell
wall using enzymes. The first category contains studies by means of
the deposition of cationic and anionic polymers (Priya et al., 2011;
Yucel Falco et al., 2017b). This procedure changes the sign of the
surface potential of the bacterial surface, depending on the effective
charge of the outer polyelectrolyte layer. As the polymers used in
these studies were mostly hydrophilic, the bacterial surface remained
hydrophilic. Atomic force microscopy (AFM) can be used to characterize the
bacterial cell wall and isolated cell wall components such as the
peptidoglycan layers. For a review, see (Dufrêne et al., 2021). AFM
can be used to detect the surface morphology of the bacterial cell
wall, revealing that Gram-positive bacteria exhibit a regular
architecture of a peptidoglycan layer and adhesins (Dupres et al.,
2005). Moreover, the force-distance curves can be obtained with the
µm-scale spatial resolution. A small deformation can be used to
detect the elasticity of the cell wall, whereas a larger deformation
provides information on the cell stiffness, reflecting the turgor
pressure of the cell. Frontiers in Soft Matter frontiersin.org 2.2.1 Modifications of group i In addition to AFM, defined model systems can be used to study
the mechanical properties of bacterial surfaces more quantitatively. Schneck et al. fabricated planar stacks of purified LPS membranes
with various strains and preformed off-specular neutron scattering
at defined disjoining pressures. Although the model consisted of
symmetric bilayer stacks of LPSs, this approach enabled quantitative
insights into how saccharide chain modifications modulate the Falco et al. adsorbed milk proteins (sodium caseinate or β-
casein) on Lactobacillus acidophilus (La-5) at a low pH, where the
proteins and bacteria are oppositely charged (Yucel Falco et al.,
2017a). The modified bacteria showed higher surface activities at the
air/water interface, and the change in the bacterial surface potential
was assessed by the zeta potential measurements, and Shekarforoush
et al. used the adsorption of hydrophobic corn protein zein to 04 Risbo et al. 10.3389/frsfm.2023.1257688 FIGURE 2
Properties of modified bacterial surfaces. (A) Zeta potential as a function of pH for UM, HA-, and OCT-modified LGG and LBC measured at various
pHs (B) ΔGBWB values of UM and modified LGG and LBC. (A, B): Adapted from (Jiang et al., 2023a), licensed CC-BY-4.0. (C) MATH values and contact
angles for the unmodified and modified to different degrees of LGG treated with lauroylchloride at different concentrations. Adapted from (Jiang et al.,
2021), with permission from Elsevier. FIGURE 2
Properties of modified bacterial surfaces. (A) Zeta potential as a function of pH for UM, HA-, and OCT-modified LGG and LBC measured at various
pHs (B) ΔGBWB values of UM and modified LGG and LBC. (A, B): Adapted from (Jiang et al., 2023a), licensed CC-BY-4.0. (C) MATH values and contact
angles for the unmodified and modified to different degrees of LGG treated with lauroylchloride at different concentrations. Adapted from (Jiang et al.,
2021), with permission from Elsevier. enhance the hydrophobicity (Shekarforoush et al., 2022). The
MATH values were moderately enhanced without lowering the
colony forming unit (CFU) of the strains. The MATH values for
Lactobacillus rhamnosus (LGG) and Lactobacillus delbruekii subs. lactis ATCC 4797 (LBD) increased from approximately 7%–20%
and from 12% to 45%, respectively. prior water removal, such as, lyophilization. Using such a procedure
can increase the MATH values of LGG from approximately 15%–
75%, depending on the concentration of lauryl chloride (Jiang et al.,
2021). 2.2.1 Modifications of group i At a 20% dose, the water contact angle continuously increases
to approximately 100°, whereas the MATH values remained
unchanged at the dose of ≥3%. These data suggest that the
MATH
may
not
be
sensitive
enough
to
discriminate
the
difference between hydrophobic bacteria. The most hydrophobic
bacteria could not be fully dispersed in water and appeared as
“bacterial chunks” in the microscopy images. CFU measurements
revealed that the chemical treatment was not detrimental beyond the
effect of the initial lyophilization step, which caused a marked
decrease in CFU. Frontiers in Soft Matter frontiersin.org 2.2.2 Modifications of group ii used β-casein- and sodium caseinate-modified La-5 to produce dry
foams (foams showing fast drainage of the serum phase) that
exhibited stability for more than 180 min (Yucel Falco et al.,
2017a). Jiang et al. modified the same bacteria using OSA to
produce dry foams that were stable for more than 180 min with
superior stability, as compared to those using UM bacteria (Jiang et al.,
2019). Because the air bubbles in food foam are typically significantly
larger than the typical oil droplets of emulsions and due to the micron
size of bacteria, bacterial stabilization may be more suited for
foams than for emulsions. The new field of bacteria as Pickering
stabilizers for emulsions and foams lacks systematic studies of the
interrelationships between bacterial surface properties of natural or
modified bacteria and the properties of emulsions. Enzymatic degradation of the poly-saccharide and protein
components of Gram-positive cell walls is a standard procedure
for making protoplasts, which are cell wall-free bacteria only having
a bacterial membrane and are spherical despite their original shape
(Martin, 1963). Protoplasts are used for extracting or introducing
DNA to bacteria as the cell wall constitutes a major barrier against
DNA extraction and the protoplast can regrow their cell wall again
after mild manipulation (such as DNA insertion). Shakarfourosh
et al. used an even milder enzymatic treatment employing
polysaccharide
degrading
enzymes
(such
as
lysozyme)
and
proteolytic
enzymes
(such
as
protease
K)
to
degrade
the
components of the LGG and LBD bacterial cell walls without
forming protoplasts (Shekarforoush et al., 2022). Despite the fact
that polysaccharides are predominantly hydrophilic and proteins are
more
hydrophobic,
both
enzyme
treatments
increased
the
hydrophobicity, as quantified using MATH measurements. This
surprising effect was attributed to the fact that the MATH method is
based on a Pickering effect, and thus changes the repulsive steric
effect of the peptidoglycan layer (Shekarforoush et al., 2022). A more complex structure was fabricated using bacteria as colloidal
building blocks (Jiang et al., 2022; Jiang et al., 2023b). Jiang et al. noted
that fat globules and lactic acid have the same approximate size and
share functionality with respect to aggregation and the ability to adhere
at the water/air interphase, and proposed that the fat globules of a
whippeable emulsion could be substituted with lactic acid bacteria
(Jiang et al., 2023b). 2.2.2 Modifications of group ii A whipping cream was formulated using bacteria
(35% (V/V), either the hydrophobic LBC or more hydrophilic LBD, as
well as sodium caseinate, and a thickening agent, both in low
concentrations. The naturally hydrophobic LBC imparted the foam
structure with bacteria densely covering the air phase with a network of
bacterial aggregates, producing foam with a high modulus and good
drainage stability (Figures 3E–I). In contrast, the LBD-based foam was
softer with poorer drainage resistance and exhibited a structure where
the internal air/water interphase was depleted of bacteria. The bacteria
culturability for both formulations was intact upon whipping, although
the aggregation of LBC lowered the apparent CFU quantification. 2.2.2 Modifications of group ii The outer surface of the oil droplets (diameter ≈60 µm)
was stabilized by a densely packed layer of adsorbed bacteria,
whereas the inner water droplets (diameter ≈5 µm) were jointly
stabilized by the bacteria and unknown low molecular weight
surfactants (Figure 3D). Foams can also be made using natural
and modified lactic acid bacteria as Pickering stabilizers. Falco et al. used β-casein- and sodium caseinate-modified La-5 to produce dry
foams (foams showing fast drainage of the serum phase) that
exhibited stability for more than 180 min (Yucel Falco et al.,
2017a). Jiang et al. modified the same bacteria using OSA to
produce dry foams that were stable for more than 180 min with
superior stability, as compared to those using UM bacteria (Jiang et al.,
2019). Because the air bubbles in food foam are typically significantly
larger than the typical oil droplets of emulsions and due to the micron
size of bacteria, bacterial stabilization may be more suited for
foams than for emulsions. The new field of bacteria as Pickering
stabilizers for emulsions and foams lacks systematic studies of the
interrelationships between bacterial surface properties of natural or
modified bacteria and the properties of emulsions. increasing pH, and the difference between UM, OCT, and HA
became more distinct under higher pH conditions. As shown in
Figure 2B, UM, LGG, and LBC exhibited a large difference in
ΔGBWB, but the difference significantly decreased after chemical
modification with OCT and HA. The clear decrease in the free
energy suggests that the surfaces of both LGG and LBC become
equally hydrophobic. Jiang et al. demonstrated that the MATH value
and water contact angle of LGG treated with laurylchloride at
different concentrations exhibited a clearly positive correlation
(Figure 2C) (Jiang et al., 2021). MATH and contact angle
measurements
with
different
solvents
showed
that
the
predominantly Lewis-basic character of both strains diminished
when the hydroxyl and amino groups were modified (Jiang et al.,
2023a). (Figures 3A–C using a simple one-step procedure (Jiang et al.,
2021). The outer surface of the oil droplets (diameter ≈60 µm)
was stabilized by a densely packed layer of adsorbed bacteria,
whereas the inner water droplets (diameter ≈5 µm) were jointly
stabilized by the bacteria and unknown low molecular weight
surfactants (Figure 3D). Foams can also be made using natural
and modified lactic acid bacteria as Pickering stabilizers. Falco et al. Frontiers in Soft Matter 2.2.2 Modifications of group ii Fatty acid anhydrides and fatty acid chlorides can be used as
hydrophobizing agents by forming either fatty acid amides or esters
with the amino or hydroxyl groups of the macro molecules under
investigation. The component octanyl succinic anhydride (OSA) is
used in the food industry to convert hydrolyzed starch to
components with emulsifying properties. OSA is only mildly
reactive with water and the reaction can occur under aqueous
conditions. Jiang et al. noted that Gram-positive cell walls
contain both amino and hydroxy groups capable of reacting with
OSA, and used this principle to enhance the hydrophobicity of LA-5
(Jiang et al., 2019). The dosing was determined according to the
percentage of OSA per dry bio-mass of the bacteria. An OSA dose
between 0% and 10% increased the MATH hydrophobicity from
approximately 40%–60%. CFU measurements at a dose of 6%
resulted in a substantial increase in hydrophobicity without
compromising
culturability,
whereas
higher
dosing
did
not
increase the MATH values significantly but was accompanied by
massive loss of culturability. A recent study involved bacterial modification of LGG and
Lactobacillus
crispatus
(LBC)
in
both
the
exponential
and
stationary phase (Jiang et al., 2023a). The surface of LGG and
LBC were chemically modified using either octanal or hexanoic
anhydride. The octanal selectively reacts with amino groups,
whereas hexanoic anhydride is reactive with both the amino and
hydroxy groups. X-ray photoelectron spectroscopy showed that
LGG has a higher O/C ratio and a lower N/C ratio, suggesting
that the LGG surface was more enriched with polysaccharides than
that of LBC. In contrast, the LBC surface contained higher amount
of proteins than the LGG surface. Figure 2A shows the zeta
potentials of unmodified (UM), octanal-treated (OCT), and
hexanoic anhydride-treated (HA) LGG and LBC measured under
different pH conditions. LGG exhibited no remarkable difference
between UM, OCT, and HA under all pH conditions. In contrast,
LBC exhibited a clear decrease in the zeta potential with an Fatty acid chlorides, such as lauryl chloride, are very reactive
with water, and bacterial modifications must be performed in water-
free but non-harmful environments, such as MCT oil, and involve 05 frontiersin.org Risbo et al. 10.3389/frsfm.2023.1257688 10.3389/frsfm.2023.1257688 10.3389/frsfm.2023.1257688 (Figures 3A–C using a simple one-step procedure (Jiang et al.,
2021). frontiersin.org FIGURE 3 FIGURE 3
Colloidal food materials containing bacteria. Emulsions and double emulsions stabilized using 10% laurylchloride-treated LGG bacteria. (A) Primary
oil droplet with inner water droplets. The oil and lipid phases are labeled with BODIPY (green). (B) Maximum intensity projections of a z-stack series of
30 images (DAPI, blue). (C) Inner droplet partially stabilized by bacteria (D) Average droplet diameter plotted as a function of time, measured in the
presence of different laurylchloride concentrations. The stability of emulsions depends on the bacterial surface properties. (A-D): Adapted from
(Jiang et al., 2021), with permission from Elsevier. (E) Rheological properties of whipping foam based on hydrophobic LBC and hydrophilic LBD (F)
Drainage of the two types of bacterial whipping foam. The reference line indicates drainage of ordinary dairy-based whipping foam. (G) Overrun (degree
of air incorporation) of bacterial whipping foam, as compared to standard dairy whipping foam. (H) Microstructure of the whipping foam showing the
adsorption of hydrophobic LBC to air bubbles (I) Structure of hydrophilic LBD-based foam with almost no surface adsorption of bacteria. (E-I): Adapted
from (Jiang et al., 2023b), licensed CC-BY-4.0. (J) Colloidosome based on the self-assembled hydrophobic LBD (green). The outer most surface is coated
with a highly anionic PCP layer (red). Insert: scanning electron microscopy image of the surface, showing both cylindrical LBD and spherical PCP. Adapted
from (Jiang et al., 2022), licensed CC-BY-4.0. interplay between the hydrocarbon chain and polar head group
interactions controls the assembly in aqueous solvents. This can be
understood by the packing parameter of the amphiphilic molecule
(Israelachvili et al., 1976; Mitchell and Ninham, 1981). The packing
parameter (v/al) is defined as the ratio between the volume of the
hydrophobic chain (v) and the product of the head group area (a)
and the chain length (l). It is important to note that the packing
parameter is not a fixed molecular parameter but depends on the
environment. Therefore, different lipid liquid crystalline (LLC)
structures illustrated in Figure 4 can be formed by both changing
the
lipid
composition
and
environmental
factors,
such
as
temperature, pH, and ionic strength (Larsson, 1989; Larsson
et al., 2006; Barriga et al., 2019). The so-called inverse LLC,
where the lipid aqueous interface is curved towards the aqueous
solvent, forming aqueous cavities, is of particular interest. This is
because the size of the cavity often matches or can be formulated to
match the size of functional protein or an enzyme. 2.3 Colloidal food materials containing
bacteria “Colloidosomes” are hollow, water-filled structures analogue to
liposomes, but formed by colloidal particles rather than lipid
molecules (Thompson et al., 2015). Simple stable colloidosomes
were generated by spontaneous water ingression into micro-bubbles
Pickering stabilized by OSA-modified LBC bacteria (Jiang et al.,
2022). The fact that the structure does not disintegrate after removal
of the air/water interphase indicate that a stabilizing principle
beyond Pickering phenomena is in action, that is, inter-bacterial
attraction of the modified hydrophobic bacteria. The basic anionic
bacterial colloidosome was modified using the layer-by-layer
method involving cationic and anionic polymers or hybrid
combinations of polymers and a second layer of the strongly
anionic bacterial strain Pediococcus pentosaceus (PCP), thus
forming a bacterial double layer of two strains (Figure 3J). The
permeability
of
the anionic
fluorescein
permeant
could
be
rationalized by electrostatic repulsions (capsule zeta potential) for
anionic colloidosomes and porosity for cationic capsules (Jiang et al.,
2022). Probiotic lactic acid bacteria can be incorporated into food
structures and have the dual role of being a functional part of
the food structure as well as delivering the probiotic function of the
bacteria. Several studies have focused on Pickering emulsions
stabilized by bacteria adhering to the oil/water interface using
natural
UM
bacteria
(Firoozmand
and
Rousseau,
2016;
Muhammed et al., 2023) and chemically modified bacteria (Jiang
et al., 2019; Jiang et al., 2021). The emulsions exhibited large
emulsions droplets of 10—100 μm, which was one to two orders
of magnitude larger than the size of bacteria (~1 µm). For food
emulsions, the size limitation mainy comes from the size of bacteria
that are used as the stabilizers. Mayonnaise-like emulsions were
produced using food grade, Streptococcus thermophilus with a high
internal oil phase load exceeding 40% (Firoozmand and Rousseau,
2016). MCT oil in water emulsions were produced using a selection
of 31 naturally hydrophobic lactobacillus bacteria (Muhammed
et al., 2023), as well as enzymatically treated LGG and LBD
bacteria (Jiang et al., 2021). The highly hydrophobic LGG
bacteria modified by laurylchloride exhibited the surprising
ability to promote the formation of W/O/W double emulsions Finally is should be mentioned that the health effect of ingestion
of
modified
bacteria
is
currently
understudied. Physical
modification based on food components such as proteins will not 06 frontiersin.org Risbo et al. 10.3389/frsfm.2023.1257688 FIGURE 3
Colloidal food materials containing bacteria. Emulsions and double emulsions stabilized using 10% laurylchloride-treated LGG bacteria. (A) Primary
oil droplet with inner water droplets. FIGURE 3 Varying the lipid
composition results in an LLC with different structures and
curvatures of the aqueous lipid interface, including inverse require approval and is unlikely to induce any toxic effects. Chemical
modification by amide and ester bonded fatty acids will most like
require approval from authorities, however it can be argued that
upon hydrolysis natural fatty acids will be released and a toxic
response is unlikely. An interesting future research topic is that the
relationship between bacterial surface properties, their binding to
the intestinal interphase and the ability of bacteria to colonize the
intestine. Frontiers in Soft Matter frontiersin.org 2.3 Colloidal food materials containing
bacteria The oil and lipid phases are labeled with BODIPY (green). (B) Maximum intensity projections of a z-stack series of
30 images (DAPI, blue). (C) Inner droplet partially stabilized by bacteria (D) Average droplet diameter plotted as a function of time, measured in the
presence of different laurylchloride concentrations. The stability of emulsions depends on the bacterial surface properties. (A-D): Adapted from
(Jiang et al., 2021), with permission from Elsevier. (E) Rheological properties of whipping foam based on hydrophobic LBC and hydrophilic LBD (F)
Drainage of the two types of bacterial whipping foam. The reference line indicates drainage of ordinary dairy-based whipping foam. (G) Overrun (degree
of air incorporation) of bacterial whipping foam, as compared to standard dairy whipping foam. (H) Microstructure of the whipping foam showing the
adsorption of hydrophobic LBC to air bubbles (I) Structure of hydrophilic LBD-based foam with almost no surface adsorption of bacteria. (E-I): Adapted
from (Jiang et al., 2023b), licensed CC-BY-4.0. (J) Colloidosome based on the self-assembled hydrophobic LBD (green). The outer most surface is coated
with a highly anionic PCP layer (red). Insert: scanning electron microscopy image of the surface, showing both cylindrical LBD and spherical PCP. Adapted
from (Jiang et al., 2022), licensed CC-BY-4.0. 3.1 Formulations of probiotic bacteria and
enzymes in lipid self-assemblies Polar lipids are key components of cell membranes that on their
own form liquid crystals that possess the long-range orientational
order of crystals and the high molecular mobility of liquids. The 07 frontiersin.org Risbo et al. 10.3389/frsfm.2023.1257688 FIGURE 4
Various lipid liquid crystalline phases (LLC). From left to right: micellar cubic (I1), hexagonal (H1), bicontinuous cubic (V1), lamellar (Lα), inversed
bicontinuous cubic (V2), inversed hexagonal (H2), and inversed micellar cubic (I2). The micellar (L1), inversed micellar (L2), and typical location of the
“sponge” or L3 phase are also indicated. Adapted from (Barauskas and Nylander, 2008), with permission from Elsevier. FIGURE 4
Various lipid liquid crystalline phases (LLC). From left to right: micellar cubic (I1), hexagonal (H1), bicontinuous cubic (V1), lamellar (Lα), inversed
bicontinuous cubic (V2), inversed hexagonal (H2), and inversed micellar cubic (I2). The micellar (L1), inversed micellar (L2), and typical location of the
“sponge” or L3 phase are also indicated. Adapted from (Barauskas and Nylander, 2008), with permission from Elsevier. more macroscopically fluid and swollen than the V2 phases. A
dispersing agent is usually added to form LLC nanoparticles in a
large excess of water (>90%), but the inner structure is closely related
to the corresponding fully hydrated bulk phase (Valldeperas et al.,
2016). bicontinuous cubic (Q2), sponge phase (L3), or inverse hexagonal
(H2) structure (Larsson, 1989; Seddon and Templer, 1995; Mulet
et al., 2013; Gilbert et al., 2019). The phase behavior of polar lipids
with decreasing water content and increasing temperature often
follows the sequence (left to right in Figure 4): hexagonal phase (HI)
→lamellar phase (Lα) for water soluble lipids and lamellar phase
(Lα) →reversed hexagonal phase (HII) →inverse micellar phase for
lipids (L2) for lipids with low water solubility. Different types of
phases with cubic symmetry (micellar cubic I1 and I2, bicontinuous
cubic V1 and V2) often occur in between these phases. When it
comes to bicontinuous cubic phases the two systems of water
channels allow the substrates to be transported in and out of the
structure. The self-assembled character of the structure also offers
advantages when it comes to formulation. The flexibility and
softness of the structure makes it easier to include enzyme that
are as large or even larger than the water channel dimensions. Entrapping
proteins,
including
enzymes,
with
retained
bioactivity is challenging due to their large size, complexity, and
limited stability outside their natural biological environment. 3.1 Formulations of probiotic bacteria and
enzymes in lipid self-assemblies A lipid
self-assembly
with
large
water
pores
is
required
for
the
encapsulation of such large biomolecules. Inverse bicontinuous
cubic (V2) and sponge (L3) phases have been successfully applied
for this purpose (Razumas et al., 1994; Angelova et al., 2011; Sun
et al., 2015; Valldeperas et al., 2019c; Gilbert et al., 2019). A recent
summary shows that cubosomes can be loaded with a wide range of
compounds from small drug molecules (<1 kDa) to large proteins
with a molecular weight up to 174 kDa (Barriga et al., 2019). However, the activity of entrapped enzymes depends on the size
of the water channels in relation to size of the enzyme molecules
(Razumas et al., 1994; Sun et al., 2015). It has been shown that the
long-term stability from days to a month in terms of retained activity
of biosensors based on electrodes coated with monoolein and
entrapped enzyme increase in the urease (molecular weight, Mw
≈590 kDa) < creatinine deiminase (Mw ≈200 kDa) ≈glucose
oxidase (Mw ≈160 kDa) < lactate oxidase (Mw ≈80 kDa)
(Razumas et al., 1994). The enzyme stability can be improved
when the water channels are larger than the protein size (Sun
et al., 2015). It should be noted that the bicontinuous cubic
phase used here are liquid crystalline and can adopt to size of
the water channel to the enzyme size. This is even easier for the more
flexible sponge (L3) phase. Here the inclusion of enzymes was found
to increase the bending rigidity of the lipid bilayer of the sponge
phase as determined by neutron spin echo measurements (Gilbert
et al., 2022). The size of the water channels in the LLC structure The LLC phases can be dispersed into colloidally stable
nanoparticles,
such
as
cubosomes,
spongosomes/sponge-like
nanoparticles, and hexosomes, which are more convenient for
many
scientific
and
technological
applications
that
require
continuous processing, such as dispensing or injection (Larsson,
2000; Mulet et al., 2013; Barriga et al., 2019; Valldeperas et al.,
2019b). Inverse bicontinuous phases (V2) are nanostructures where
the lipid bilayer curves to form water channels in a three-
dimensional network (Larsson, 1989). The curvature of such a
bilayer can be described as a minimal surface, which is as
concave as it is convex, that is, it has a zero mean curvature. Frontiers in Soft Matter frontiersin.org 3.2 Formulation of probiotic bacteria and
enzymes in silica-based particles Amorphous forms of silica and silicates are recognized by the US
Food and Drug Administration (FDA) and the European Food
Safety Authority (EFSA) as the safe ingredients for oral delivery up
to the dose of 1.5 g per day (Diab et al., 2017). In biological systems,
silica can be found in a variety of nano- and micro-structures, such
as silica diatoms (Sumper and Brunner, 2006) and spicules of
sponges (Uriz et al., 2003), which inspired the synthesis of
various hybrid materials (Nassif and Livage, 2011). Mesoporous
silica nanoparticles are extensively used for the encapsulation of
drugs with small molecular weights (Manzano and Vallet-Regí,
2020) and for the imaging of biological specimens (Tasciotti
et al., 2008). The sol-gel process is a promising and effective
strategy to synthesize mesoporous silica for food applications
because
of
the
compatibility
and
non-cytotoxicity
to
bio-
macromolecules, bioactive enzymes, and probiotic bacteria (Diab
et al., 2017). p
p y
y
g
p
As PA−microcapsules are not sufficiently stable to protect
probiotic bacteria under gastric conditions, the deposition of
additional protective layers is desired (Matricardi et al., 2008). For example, the layer-by-layer deposition of cationic
and
anionic polyelectrolytes (Lvov et al., 1993; Decher, 1997) has
been used for the stable encapsulation of enzymes (Caruso et al.,
2000), antigens (De Koker et al., 2009), and living cells (Diaspro
et al., 2002). 1H NMR data showed that polyelectrolyte multilayers
act as a diffusion barrier by modulating the mobility of water
molecules, especially when are stacked (Schwarz and Schönhoff,
2002). Notably, only one additional layer of chitosan on PA−
capsules increases the survival rate of Bifidobacterium and
Lactobacillus
in
gastric
fluids
by
two
to
three
orders
of
magnitude (Figure 5C) (Krasaekoopt et al., 2004; Cook et al., Mesoscopic pores can be used as cavities that facilitate the
stability of the encapsulated bioactive enzymes (or proteins in
general) under broader pH and temperature conditions (Hudson
et al., 2008; Carlsson et al., 2014). As the pore size, typically in the
order of several nm to 10 nm, is large enough to allow large
molecules to enter and diffuse inside the pores. Therefore,
enzymes or their aggregates bind not only to the outer surface of
particles, but many of them can also bind to the significantly larger,
inner pore surfaces either by physisorption or by chemisorption. 3.1 Formulations of probiotic bacteria and
enzymes in lipid self-assemblies The
bicontinuous lipid sponge phase can be regarded as a “melted”
bicontinuous cubic phase, and as such, it does not present the long-
range order of cubic phases (Anderson et al., 1989). The scattering
pattern only features two broad diffraction peaks, indicating the
mobility of the system (Porcar et al., 2003). It has a more flexible
bilayer structure than that of the V2 phases, and consequently, it is 08 frontiersin.org Risbo et al. 10.3389/frsfm.2023.1257688 could also be tuned by changing the environmental conditions,
which also opens up the possibility to control the encapsulation and
release of enzymes using external stimuli (Fong et al., 2016). Oral
delivery
of
lipid-based
carriers,
such
as
liquid
crystalline
nanoparticles, vesicles, or nano-emulsions, also need to survive
the harsh environment through the gastrointestinal tract, where
lipolysis is a particular challenge as pointed out by Dully et al. (2020). The explored a strategy using lipase inhibitors, pointed out
that there are other strategies like use of non-digestible lipids or use
of capsules. If the lipid particle survives the harsh conditions in the
stomach all the way down the small intestine the cargo need to
traverse the mucus layer prior to absorption of the drug. In some
cases, such as in inflammatory bowel disease, local delivery to the
intestine is desirable. In this case, absorption is not needed, but the
drug or nanoparticle still must penetrate the mucus (Date et al.,
2018). In other cases, mucus adhesion is needed, where the
nanoparticle
adheres
to
the
mucus
for
a
prolonged
time,
releasing the drug and resulting in a prolonged effect (Gamboa
and Leong, 2013). Therefore, the colloidal aspect is clearly crucial for
the optimization of the formulation. This is a very wide field of
research beyond the scope of this review. A comprehensive review of
the topic and associated challenges has been presented by Zhai et al. (2019). substrates and improves the enzymatic activity, but the enzymes
can also exit confinement and diffuse away from the formulation. Therefore, the chemical coupling of enzymes to the inner pore
surface or the use of crosslinked or aggregated enzymes is desirable
to keep the enzymes inside the mesopores (Wang et al., 2010). 3.3 Formulation of probiotic bacteria in
polymer-based microcapsules A variety of formulations have been developed and examined for
the delivery of probiotics, such as spray-drying, freeze-drying,
extrusion, and emulsification (Gouin, 2004; Champagne and
Fustier, 2007; Gasperini et al., 2014). Among these methods,
emulsification with the aid of surfactants is a gentle technique
that causes minimal damage to microorganisms (Krasaekoopt
et al., 2003). When an aqueous polymer solution is mixed with
oil under stirring, the presence of surfactants (such as polyglycerol
polyricinoleate (PGPR)) further contributes to stabilizing the water
droplets by reducing the surface tension at the oil/water interface. If
the polymers in aqueous droplets can be crosslinked, the
microparticles of crosslinked hydrogels can easily be extracted
from the oil-in-water emulsions using centrifugation. g
g
The use of crosslinked hydrogels based on biopolymers
approved by the US FDA is a straightforward strategy to deliver
such formulations to the gastrointestinal tract (Figure 5A). For this
purpose, probiotic bacteria can be encapsulated in particles of
hydrogels based on anionic polysaccharides, such as alginate,
pectin, and xanthan gum, to form hydrogels crosslinked by
divalent
cations,
such
as
Ca2+
(Figure
5B)
(Kwiecień
and
Kwiecień, 2018). Polyalginate (PA−), which is crosslinked by
divalent cations such as Mg2+ and Ca2+ by the formation of “egg-
box” junctions hosting divalent cations between the stacks of
guluronic acid blocks (Katchalsky et al., 1961), is one of the most
common encapsulation matrices for probiotic bacteria and enzymes
(Orive et al., 2002). PA−is an FDA-approved food ingredient and is a
linear polysaccharide extracted from the cell wall of marine brown
algae, consisting of disaccharide units based on L-guluronic and
D-mannuronic acids (Lee and Mooney, 2012). It is expected that
formulations based on anionic polysaccharides, such as PA−, are
protonated and hence are less soluble at low pH. The formulations
are meant to protect probiotic bacteria under gastric conditions,
whereas the bacteria are released in the intestines because the de-
protonated polyelectrolytes are more soluble at higher pHs. Frontiers in Soft Matter frontiersin.org 3.2 Formulation of probiotic bacteria and
enzymes in silica-based particles It
should be noted that the immobilization of the bioactive enzymes
into small pores or into cavities with only one opening can block the
diffusion of substrates and products, which could reduce the
catalytic activity of the confined enzymes (Lee et al., 2009). In
contrast, the use of “macropores” facilitates the diffusion of 09 frontiersin.org Risbo et al. 10.3389/frsfm.2023.1257688 FIGURE 5
Formulation of probiotic bacteria in polymer-based microcapsules. (A) Schematic illustration of the delivery of probiotic bacteria using polymer-
based formulations (B) Examples of polymer-based formulations. (C) Encapsulation of Bifidobacterium breve in PA−particles coated with one chitosan
layer. Scale bar: 100 µm. Note that the size is significantly larger than the sensory limit of the human tongue (80 µm). Adapted from (Cook et al., 2011), with
permission from the American Chemical Society. FIGURE 5
Formulation of probiotic bacteria in polymer-based microcapsules. (A) Schematic illustration of the delivery of probiotic bacteria using polymer-
based formulations (B) Examples of polymer-based formulations. (C) Encapsulation of Bifidobacterium breve in PA−particles coated with one chitosan
layer. Scale bar: 100 µm. Note that the size is significantly larger than the sensory limit of the human tongue (80 µm). Adapted from (Cook et al., 2011), with
permission from the American Chemical Society. 2011). However, it should also be noted that the size of the particles
used in these studies was one order of magnitude larger than the
sensory limit of the human tongue (Krasaekoopt et al., 2004), or
were not even controlled (Cook et al., 2011). Consequently, several
key factors remain unclear, such as (a) the minimum requirements
for stable formulations, (b) the status of the probiotic bacteria release
from the polymer capsules after reaching the target colon, and (c)
whether the functionality of the probiotics is sustained after delivery. Systematic investigations addressing these points are still largely
missing. food formulations intended for lactose-intolerant consumers. The
ability to deliver iron is becoming increasingly important as the shift
to more plant-based foods can lead to iron deficiency. The delivery
of iron-carrying heme proteins has become important for food iron
supply (Gilbert et al., 2023). Although the curvature of the lipid
aqueous interfaces in the non-lamellar phases determines the size of
the aqueous cavities, and hence the space given to the enzyme, the
interaction between the enzyme and the lipid layer is an important
factor that controls the efficiency of the encapsulation. Frontiers in Soft Matter 3.2 Formulation of probiotic bacteria and
enzymes in silica-based particles Mixtures of food-grade acylglycerides and acyldiglycerides can
form highly swollen sponge phases (L3) with aqueous pores up to
13 nm in diameter, and can form well defined nanoparticles in
excess water with the help of the dispersing agent polysorbate 80
(P80) (Figure 6) (Valldeperas et al., 2016). Aspartic protease (Mw =
34 kDa) assayed by milk clotting capability (78) and β-galactosidase
(Mw = 476 kDa, tetramer) assayed by monitoring the Ortho-
Nitrophenyl-β-galactoside - > O-nitrophenol (yellow) transition
was found to be still active after inclusion into the sponge phase
lipid matrix even after 3 months at 25°C (Gilbert et al., 2019). This is
not possible if the enzyme is delivered from a solution where the
stability at 25°C is a matter of days. It should here be noted that the
lipid matrix used only consisted by approved food grade lipids. Interestingly, the two proteins affect the phase behavior of the lipid
system differently (Figure 7). The smaller enzyme has a larger effect
on the phase structure. The enzyme “likes” the lipid bilayer to be
wrapped around itself, indicating that the protein is inside the water
channels and that the change in curvature effect is induced by
protein-lipid interaction. frontiersin.org 4.1 Formulations of enzymes in lipid self-
assemblies As examples of the delivery of bioactive molecules using lipid-
based formulations, the encapsulation of two key types of enzymes
of different sizes, namely aspartic protease (34 kDa or a diameter of
about 5.7 nm) and β-galactosidase (460 kDa or a diameter of about
14 nm), are discussed (Valldeperas et al., 2019c; Gilbert et al., 2019). Here we note that the typical water channel diameter for fully
swollen, bicontinuous cubic phase is about 5 nm while the
corresponding value for a sponge phase is about 13 nm, bearing
mind that both phases are flexible structures. Both enzymes are
important for food processing, but the storage and delivery of the
bioactivity is challenging. Aspartic protease from C. parasitica is a
milk-clotting enzyme that is highly specific for κ-casein and
promotes
curd
formation
as
well
as
cheese
ripening. β-
galactosidase from K. lactis (465 kDa, size = 14 nm) hydrolyses
lactose to a mixture of glucose and galactose; therefore, it is used in The particles adsorb at the interface to form a lipid bilayer, as
shown using quartz crystal microbalance with dissipation (QCM-D)
and neutron reflectometry (Valldeperas et al., 2019a). Raman
spectroscopy results for the sponge phases show large similarities
in the lipid chain confirmation and head group interactions as in the 10 frontiersin.org Risbo et al. 10.3389/frsfm.2023.1257688 FIGURE 6
Phase diagram of the diglycerol monoolein (DGMO), monoolein, and polysorbate 80 system and the molecular structure of the components. Adapted from (Valldeperas et al., 2016), with permission from the American Chemical Society. FIGURE 6 FIGURE 6 FIGURE 6
Phase diagram of the diglycerol monoolein (DGMO), monoolein, and polysorbate 80 system and the molecular structure of the components. Adapted from (Valldeperas et al., 2016), with permission from the American Chemical Society. FIGURE 7
Effect of enzymes on LLC structures. (A) Small-angle x-ray scattering data of lipid LLC structures measured at different aspartic protease
concentrations, indicating the transition from a sponge phase to a cubic phase at higher enzyme concentrations (B) Addition of β-galactosidase leads to a
shift of main peak to a larger q with an increasing enzyme concentration, suggesting a decrease in the degree of hydration. 4.1 Formulations of enzymes in lipid self-
assemblies (C) Aspartic protease is still
active after inclusion into the lipid matrix, even after 3 months at 25°C as judged by its ability to clot milk (D) β-galactosidase maintained its function,
that is, hydrolyses of galactose analogue in the lipid system, even after 3 months, that is o-nitrophenyl-β-galactoside (ONPG) is hydrolyzed, in the
presence of β-galactosidase to o-nitrophenol (D). (A,C): Adapted from (Valldeperas et al., 2019c), with permission from Elsevier. (B,D): Adapted from
(Gilbert et al., 2019), with permission from the Royal Society of Chemistry. FIGURE 7
Effect of enzymes on LLC structures. (A) Small-angle x-ray scattering data of lipid LLC structures measured at different aspartic protease
concentrations, indicating the transition from a sponge phase to a cubic phase at higher enzyme concentrations (B) Addition of β-galactosidase leads to a
shift of main peak to a larger q with an increasing enzyme concentration, suggesting a decrease in the degree of hydration. (C) Aspartic protease is still
active after inclusion into the lipid matrix, even after 3 months at 25°C as judged by its ability to clot milk (D) β-galactosidase maintained its function,
that is, hydrolyses of galactose analogue in the lipid system, even after 3 months, that is o-nitrophenyl-β-galactoside (ONPG) is hydrolyzed, in the
presence of β-galactosidase to o-nitrophenol (D). (A,C): Adapted from (Valldeperas et al., 2019c), with permission from Elsevier. (B,D): Adapted from
(Gilbert et al., 2019), with permission from the Royal Society of Chemistry. 11 11 Frontiers in Soft Matter frontiersin.org Risbo et al. 10.3389/frsfm.2023.1257688 FIGURE 8
Molecular dynamics simulations of aspartic protease encapsulated in LLC suggests the penetration of enzymes into the lipid bilayers. This was
confirmed using neutron spin echo measurements, which that showed an increase in the membrane bending rigidity upon the encapsulation of enzymes
in the lipid sponge phase (data not shown). Reproduced from (Gilbert et al., 2022), licensed CC-BY-3.0. FIGURE 8 FIGURE 8
Molecular dynamics simulations of aspartic protease encapsulated in LLC suggests the penetration of enzymes into the lipid bilayers. This was
confirmed using neutron spin echo measurements, which that showed an increase in the membrane bending rigidity upon the encapsulation of enzymes
in the lipid sponge phase (data not shown). Reproduced from (Gilbert et al., 2022), licensed CC-BY-3.0. lamellar and reverse bicontinuous cubic phase in the same lipid
system (Talaikis et al., 2019). 4.1 Formulations of enzymes in lipid self-
assemblies Size exclusion chromatography
demonstrated efficient encapsulation of both enzymes, yet they
retained their enzymatic activity over months, surpassing the
storage stability of pure enzymes in solution (Valldeperas et al.,
2019c; Gilbert et al., 2019). This may be owing to the penetration of
the enzymes into the formed lipid bilayer, as shown using Raman
spectroscopy (Valldeperas et al., 2019c) and neutron reflectometry
(Valldeperas et al., 2019a). Neutron spin echo and molecular
dynamics simulations have confirmed this (Figure 8) (Gilbert
et al., 2022). aligned mesopores and scalable macropores (Blin et al., 2014). Pavel et al. applied this formulation (mesopore diameter: 9 nm,
macropore diameter: 0.2 µm) to encapsulate β-galactosidase (Pavel
et al., 2017). The adsorption of enzymes to meso- or macropores
show a dependence on concentration and thus the aggregation state
of enzymes. Notably, the oligomers of β-galactosidase adsorbed on
the surface of macropores exhibit a high catalytic activity. The encapsulation of probiotic bacteria requires another
strategy because the pore size is too small to accommodate µm-
sized cells. For this purpose, a protective silicate layer can be
mineralized
on
polymer-based
microparticles
using
electro-
spraying or emulsification. Callone et al. (Callone et al., 2008)
reported
the
encapsulation
of
Saccharomyces
cerevisiae
and
Oenococcus oeni in PA−particles by injecting cell/PA−droplets
into
an
M2+-loaded
bath,
resulting
in
the
formation
of
400–500 µm-sized particles. Subsequently, the surface was coated
with amorphous silica synthesized using three methods, namely (i)
dipping the particles in silica sol, (ii) consolidating the particles with
tetraethoxysilane in an apolar solvent, and (iii) coating the particle
surface with methyltriethoxysilane in the gas phase. The particles
prepared using (ii) exhibited cytotoxicity, but those prepared using
(i) and (iii) exhibited an increase in the function of the cells, such as
the fermentation of glucose by Saccharomyces and maleic acid by
Oenococcus. Haffner et al. (Haffner et al., 2016) implemented a
similar strategy and fabricated Lactobacillus/PA−particles by
electrospraying
into
a
bath
containing
Ca2+
ions,
yielding
200 µm-sized hydrogel particles encapsulating Lactobacillus. The
protective amorphous silica layer was deposited by the electrostatic
adsorption of aminotrimethoxysilane, followed by the condensation
of tetramethyl orthosilicate. The bacterial viability was assessed by
counting the colony forming units (CFUs) before and after
incubation in solutions mimicking gastric and intestinal fluids
(Minekus et al., 2014). The Lactobacillus encapsulated in silica- Frontiers in Soft Matter frontiersin.org 4.2 Formulations of probiotic bacteria and
enzymes in silica-based particles The formulation of bioactive enzymes in porous silica particles is
a highly promising approach (Lee et al., 2009). To address the trade-
off between “mesopores” vs. “macropores” (Section 3.2), the use of
materials displaying both types of pores has been proposed. Yang
et al. reported the “fish-in-net” encapsulation of enzymes by the
condensation of pre-formed mesoporous silica particles around the
enzyme aggregates in the presence of glycerol (Yang et al., 2006). Bernal et al. synthesized meso-macroporous silica materials
(mesopore diameter: 10–40 nm, macropore diameter: 0.1–20 µm)
by the polycondensation of sodium silicate using the micelles of
cationic surfactants (cetyltriammoniumbromide (CTAB)) in the
presence
of
ethylacetate
as
the
template
and
encapsulated
bioactive enzymes, such as β-galactosidase and lipase (Bernal
et al., 2011; Bernal et al., 2014). Blin et al. employed double
emulsions based on water/oil/water (W/O/W) or water/solid lipid
nanoparticles/water (W/SLPs/W) as templates and synthesized
meso-macroporous
silica
particles,
resulting
in
hexagonally 12 frontiersin.org 10.3389/frsfm.2023.1257688 Risbo et al. layer deposition of positively charged chitosan (Ch+) and
negatively charged PA– (Figure 9A) significantly increases the
viability of LGG (Caruso et al., 2000; Cook et al., 2011). Remarkably,
the deposition of only three polyelectrolyte layers increases the CFU
values by a factor of more than 106 in stomach fluid and 105 in small
intestine fluid (Figure 9B). It is notable that a further increase in the layer
number did not improve the viability, suggesting that the deposition of
three polyelectrolyte layers (chitosan + PA + chitosan), corresponding to
a total thickness of only 20 nm, is the minimum requirement to achieve
the highest protective function (Eshrati et al., 2018). PA−particles exhibited only very minor leakages and maintained the
viability of the bacteria, whereas all the naked bacteria with no
protection died in both fluids. The metabolic function of bacteria
can be further evaluated by using an in vitro gastrointestinal
simulator (see Section 5.1 for more details) (Haffner et al., 2017). The coating of organic particles with “porous” silica layers further
allows for the diffusion of molecules, which helps the bacteria
maintain their functions and stabilizes the encapsulated enzymes
(Blin et al., 2014; Haffner et al., 2016). Unfortunately, chitosan is not an ingredient approved by the
FDA or EFSA, and it is claimed that chitosan is antimicrobial (No
et al., 2002; Tsai and Hwang, 2004). 4.3 Formulations of probiotic bacteria in
composite polymer microcapsules As summarized in Section 3.3, probiotic bacteria can be
encapsulated in formulations based on biopolymer-based hydrogel
droplets using emulsification (Krasaekoopt et al., 2003). Maintaining
the size of the formulation particles to smaller than the sensory limit of
the human tongue (80 μm, according to (Tyle, 1993)) is an important
requirement for food applications, otherwise consumers would feel
the formulations as “additional, foreign ingredients”. For this purpose,
it is necessary to optimize the emulsification process. For example, the
combination of shear-thinning agents and surfactants enables the
formation of small and stable droplets even at a significantly lower
emulsification speed. However, this has been either difficult to control
or even overlooked in previous studies (Krasaekoopt et al., 2004; Cook
et al., 2011). Moreover, hydrogel-based microcapsules prepared by
emulsification, such as those based on PA−hydrogels crosslinked by
Ca2+, cannot maintain their structural integrity in gastrointestinal
environments or sustain the functions of the encapsulated probiotic
bacteria (Eshrati et al., 2018). The deposition of additional polymer
layers is a possible strategy to increase the stability of PA−
microparticles and hence the protection of the probiotic bacteria
(Matricardi et al., 2008). For example, the layer-by-layer deposition of
cationic and anionic polyelectrolytes (Lvov et al., 1993; Decher, 1997)
has been used for the stable encapsulation of enzymes (Caruso et al.,
2000), antigens (De Koker et al., 2009), and living cells (Diaspro et al.,
2002). The deposition of one chitosan layer increases the survival rate
of Bifidobacterium and Lactobacillus encapsulated in PA particles by a
factor of 100–1,000 under gastric conditions, suggesting that even one
additional polymer layer could increase the stability of the
formulations (Krasaekoopt et al., 2004; Cook et al., 2011). Moreover, layer-by-layer coating with polyelectrolytes can be
applied on the surfaces of the negatively charged cell wall of bacteria,
enabling the use of a single bacteria as the smallest delivery vehicle
(Priya et al., 2011; Yucel Falco et al., 2017b). Falco et al. coated
individual bacteria with sulfated β-glucan and Ch+, because β-glucan
can resist the human enzymatic system but not the bacterial
glycanases in the small intestine (Yucel Falco et al., 2017b). The
layer-by-layer deposition was confirmed on planar model surfaces
using
QCM-D
and
spectroscopic
ellipsometry,
whereas
the
deposition on the bacteria was confirmed using zeta-potential
measurements. 4.3 Formulations of probiotic bacteria in
composite polymer microcapsules Although the formulations could resist the acidic
gastric conditions, they were partially degraded during exposure to a
combination of simulated intestinal fluid containing glucanase,
indicating the release mechanism of bacteria in the intestine. Eshrati et al. demonstrated the encapsulation of LGG into small,
highly monodisperse PA−particles by the systematic optimization of
the preparatory parameters. To avoid mechanical damage to
probiotic bacteria during emulsification, it is necessary to stir the
oil/water mixture at a low speed (such as 300 rpm) in the presence of
surfactants (such as PGPR) and shear-thinning agents (such as
xanthan
gum). Although
LGG
encapsulated
in
crosslinked
PA–particles exhibits an increase in viability, which was assessed
by the number of CFUs, by four orders of magnitude, as compared
to “naked” LGG, the fraction of viable LGG after 2 h in gastric fluid
(3 g/L pepsin and 85 mM NaCl, pH 2.0) is still 0.1%. The fraction of
viable LGG in crosslinked PA–particles is 0.01% after incubating
LGG/PA−particles in intestinal fluid (1 g/L pancreatin, 1 g/L bile
salt, and 85 mM NaCl, pH 6.5) for 3 h. These results clearly
indicate
that
PA–particle
cores
alone
do
not
have
a
sufficiently high protective function. In contrast, the layer-by- 4.2 Formulations of probiotic bacteria and
enzymes in silica-based particles The use of food-derived milk
proteins is a promising strategy for the stabilization of PA–particles
by the layer-by-layer deposition of charged macromolecules because
milk proteins possess both positive and negative net charges near
neutral pH, such as lactoferrin (+), caseinate (−), and whey protein
isolates (−). Falco et al. used several milk proteins for the layer-by-
layer coating of 150 µm-diameter PA−particles encapsulating LGG
(Yucel Falco et al., 2019). The layer-by-layer deposition was
monitored using the change in the zeta potential and QCM-D. The coating with only milk proteins did not lead to sufficiently high
stability. However, the deposition of Ch+ as the first layer improved
the stability. The treatment of PA–particles coated with Ch+ and
additional milk protein layers in gastric fluid for 30 min resulted in a
decrease in the CFU number by three orders of magnitude. Unfortunately, a longer exposure to gastric fluids resulted in the
death of all bacteria, indicating insufficient stability. The use of
materials approved by the FDA and EFSA is a necessary step to
apply such (bio)polymer-based formulations as food ingredients. Frontiers in Soft Matter frontiersin.org 4.4 Formulations of probiotic bacteria in
cellulose-based particles Cellulose derivatives, such as carboxymethyl cellulose (CMC)
and hydroxyethyl cellulose (HEC), have been used for the
development of probiotic delivery systems based on crosslinked
edible films (Singh et al., 2019), water in water (W/W) emulsions
with gelatin (Singh et al., 2018a), and coacervates with Ch+ with
genepin (Singh et al., 2018b) or without (Singh et al., 2017). Food
grade cellulose derivatives are water soluble, typically due to the
presence of carboxylic acid groups, and serve as thickeners in the
food industry. Therefore, these components require additional 13 frontiersin.org Risbo et al. 10.3389/frsfm.2023.1257688 FIGURE 9
Encapsulation of Probiotic Bacteria in Composite Polymer Formulations. (A) Scanning electron micrographs of LGG encapsulated in a PA−
microcapsule and PA–-coated with three additional polyelectrolyte layers. The LGGs were well-protected by the 20 nm-thick polyelectrolyte layers. The
particle size was <40 μm, which is not sensed by the human tongue (B) Photographs of Petri dishes and normalized CFUs after treatment with fluids
simulating gastric (left) and small intestine environments. The PA–core with three additional layers increased the survival rates by 107 and 104 in the
stomach and small intestine environments, respectively. (C) Metabolic activity of the “released” LGG assessed by the lactate synthesis determined using a
gastrointestinal simulator. The composite polymer formulations increased the lactate production by 105. Adapted from (Eshrati et al., 2018), with
permission from the American Chemical Society. FIGURE 9
Encapsulation of Probiotic Bacteria in Composite Polymer Formulations. (A) Scanning electron micrographs of LGG encapsulated in a PA−
microcapsule and PA–-coated with three additional polyelectrolyte layers. The LGGs were well-protected by the 20 nm-thick polyelectrolyte layers. The
particle size was <40 μm, which is not sensed by the human tongue (B) Photographs of Petri dishes and normalized CFUs after treatment with fluids
simulating gastric (left) and small intestine environments. The PA–core with three additional layers increased the survival rates by 107 and 104 in the
stomach and small intestine environments, respectively. (C) Metabolic activity of the “released” LGG assessed by the lactate synthesis determined using a
gastrointestinal simulator. The composite polymer formulations increased the lactate production by 105. Adapted from (Eshrati et al., 2018), with
permission from the American Chemical Society. coacervation route by the dropwise addition of a CMC solution into a
chitosan solution (Singh et al., 2018b) in combination with a mild
crosslinker, namely genipin. 4.4 Formulations of probiotic bacteria in
cellulose-based particles The addition of sodium caseinate
prevented particle aggregation due to electrostatic repulsion of the
otherwise nearly neutral particles. The formulation significantly
improved the survival rate of the bacteria when exposed to
simulated gastric and intestinal conditions, as compared with the
unformulated LGG bacteria. Since probiotics can be administered
through food, the long-term culturability was tested over 30 days,
showing a significantly improved stability, as compared with the
unformulated samples. However still up to five log units of loss was
detected even for the improved CMC-chitosan formulations which
still limits the practical applications. gelling agents or crosslinkers to form stable particles or films that do
not readily dissolve or disintegrate. CMC- and HEC-based edible
films were cast and crosslinked under mild conditions by drying at
50°C after the addition of citric acid (Singh et al., 2019). The lactic
acid bacteria in LGG could be absorbed into the films and the
pH dependency of film swelling could be tuned to provide a low
degree of swelling at gastric pH and high degree of swelling at
intestinal pH by adjusting the CMC/HEC ratio and citric acid
content. Overall, the films could be loaded with a high quantity
of bacteria and provided good protection. Water
in
water
emulsion
is
a
phenomenon
driven
by
thermodynamic incompatibility between pairs of water-soluble
polymers. The
interfacial
tensions
are
extremely
small,
the
interphase has a large width, and both inhibit the stabilization
with small (molecular) emulsifiers; therefore, Pickering stabilization
using nano- or micron-scaled particles must rather be employed. Singh et al. (2018a) used CMS and gelatin at temperatures above 35°C
to avoid the gelation of gelatin and to introduce LGG bacteria in
dispersed CMC droplets. The bacteria exhibited a specific affinity for
the CMC phase and was intensively present at the interphase of the
particles, which was attributed to Pickering stabilization phenomena
and the high desorption energy of the large size of the bacteria. The
formulation strategy resulted in the enhanced survival of entrapped
LGG when exposed to simulated gastric and intestinal conditions. Sub-millimeter CMC particles were produced following a polymer Frontiers in Soft Matter frontiersin.org 5.1 In vitro gastrointestinal simulator As the physical access to the digestive product after each
compartment in the gastrointestinal tract in vivo is not realistic,
several types of in vitro systems have been developed to simulate
the entire gastrointestinal tract (Van de Wiele et al., 2015). In vitro
gastrointestinal simulators are powerful tools to not only assess the 14 frontiersin.org Risbo et al. 10.3389/frsfm.2023.1257688 protective capacity of formulations but also to quantify the metabolic
function of LGG during its passage through the entire gastrointestinal
tract in the presence of nutrients (starch and other carbohydrates) and
other bacteria. This assay is distinct from the previous CFU counting
assay because only the probiotic bacteria that are released from the
formulation can be detected using flow cytometry. Namely, the
bacteria that are encapsulated and protected by the formulation
are not detected. Once the LGG/PA−particles were coated with
three polyelectrolyte layers, no released bacteria could be detected
throughout the entire gastric process (1 h at pH 4.8, 0.5 h at pH 3.0,
and 0.5 h at pH 2.0). Notably, the residence in the small intestine
chamber showed an increase in the released LGG by a factor of five, as
compared with the initial level, suggesting the release of LGG that
proliferated in the formulations by the deprotonation of chitosan
(pKb = 6.5) (Guarino et al., 2015). Most significantly, the amount of
lactate (metabolic product) produced increased in the ileum, small
intestine, and colon chambers by five orders of magnitude (105 ×), as
compared with the free LGG (Eshrati et al., 2018), demonstrating that
the 40 µm composite polymer formulations significantly improved
the delivery of functional probiotic bacteria without being detected by
the tongue of the consumer (Figure 9C). mucin-functionalized
surface
exhibits
a
markedly
stronger
interaction with silica particles coated with phosphatidylcholine
than those coated with positively charged lipids, suggesting that
the mucin-phosphatidylcholine interaction is not driven by
electrostatic interactions, as suggested by a previous isothermal
calorimetry study (Korytowski et al., 2017). As a more realistic model for the interaction of the released
probiotic bacteria with the intestinal tract surfaces, Eshrati et al. fabricated a polarized monolayer of enterocyte-like cells and goblet
cell-like cells differentiated from CaCo2 and HT29 MTX cells
(Figure 10A). After the establishment of a polarized epithelial
monolayer with tight junctions, the sample was transferred into a
parallel plate flow chamber. 5.2 Adhesion of released probiotics on
in vitro model of intestinal mucosa After their release from the formulations, probiotic bacteria adhere
on the surface of the colon epithelium and create colonies (Sonnenburg
et al., 2004). The surface of the gastrointestinal tract is covered with a
mucosal layer consisting of mucin, which protects against the invasion
of pathological bacteria (Pelaseyed et al., 2014). An increasing
number of studies have demonstrated that the protective function of
mucus is supported by the presence of phospholipids, especially,
phosphatidylcholine and lysophosphatidylcholine (Ehehalt et al., 2004). AFM is a commonly used method to reveal the mechanism of the
initial step, which is the adhesion of LGG on the mucosal surface of
intestinal epithelial cells. Earlier studies have shown that bacteria can be
immobilized on a cantilever surface by polycations, such as poly (lysine)
and poly (ethyleneimine), and the force-distance curves can be
measured on various substrates (Cao et al., 2006; Boks et al., 2008;
Rowan-Nash et al., 2019). Dague et al. coated plastic dish surfaces with
porcine gastric mucin using physisorption as a model of the
gastrointestinal tract surfaces, and measured the force-distance
curves with a Lactococcus lactis-coated cantilever (Dague et al.,
2010). The force-distance curves measured using bacteria-coated
cantilevers are distinct from those measured using intact Si3N4
cantilevers. As the surface density of mucin is hardly controllable by
physisorption, Amadei et al. reported the immobilization of biotin-
functionalized mucin onto planar, supported lipid membranes
incorporating
biotinylated
lipids
via
neutravidin
crosslinkers
(Amadei et al., 2018). A high-affinity binding of biotin and
neutravidin (KD ~ 10–15 M) enables the grafting of biopolymers by
controlling the self-assembly of biotinylated lipids (Kaindl et al.,
2012; Rieger et al., 2015). Instead of measuring interfacial
forces using AFM, the microinterferometry technique has been
used to calculate the mean force potential from the height
fluctuation of cell-sized beads (Higaki et al., 2017). Although
mucin is highly glycosylated, carrying negative net charges, the Finally, adhered bacteria start forming new microbiomes by forming
colonies. To characterize the interactions between bacteria, the timescale
of bacterial auto-aggregation is commonly used in microbiology. The
assay involves monitoring the time evolution of the loss of OD due to
sedimentation of the aggregated bacteria (Thomas et al., 2018). Freely
suspended single bacteria remain in suspension for a significantly longer
time because the sedimentation velocity (v) of a particle with a radius (r)
increases with v ∝r−2. 5.1 In vitro gastrointestinal simulator The adhesion of LGG on the surface was
monitored under the shear stress that covers the bacteria in vivo,
0.1–0.5 Pa (Eshrati et al., 2019). It should be noted that the cells need
to be cultured in Transwell® culture until they establish the apico-
basal polarity because the commonly used cell culture on plastic or
glass substrates is not able to establish the polarized epithelial cell
monolayer connected with tight junctions to express the layer of
mucin proteins. Therefore, the differentiated cell monolayer with a
sufficiently high transepithelial electric resistance (≥400 Ω cm2) was
transferred into a microfluidic chamber (Fröhlich et al., 2021). In
addition to the establishment of tight junctions (Figure 10B), the
expression of mucin 2, which is a major constituent of mucus layers
that build up a protective barrier together with phospholipids on the
epithelial surfaces against pathogens (Pelaseyed et al., 2014;
Korytowski et al., 2017), was confirmed by immunofluorescence
imaging (Figure 10B). Subsequently, the Young’s modulus of the
mucin layer was measured using AFM nano-indentation, which
indicated that the goblet cell-like cells derived from HT29 MTX cells
produced thicker and softer mucin layers, EHT29 = 0.2 kPa and
ECaCo2 = 10 kPa. Interestingly, the surface density of adhered LGG
(Figure 10C) exhibited a clear maximum at approximately 0.3 Pa,
suggesting that the adhesion of LGG to the mucosal surface in the
colon is shear-dependent. Moreover, the persistence of LGG
adhesion was further examined by a stepwise increase in the
shear stress exerted on the adherent LGG, demonstrating that
some LGG cells can withstand high shear stress even beyond the
physiological levels. The presence of such “strong binders” might
contribute to the maintenance of the chronic homeostasis of
microbiomes in the gastrointestinal tract. 5.2 Adhesion of released probiotics on
in vitro model of intestinal mucosa The adhesion force between bacteria can also be
measured both on the cellular level and on the single molecular level
using optical tweezers and single molecule force spectroscopy,
respectively (Verbelen et al., 2007; S Bezryadina et al., 2016). Frontiers in Soft Matter frontiersin.org 6 Final remarks and perspectives Probiotic bacteria and enzymes can be delivered to the intestinal
system by formulating appropriate carriers and including these into 15 frontiersin.org Risbo et al. 10.3389/frsfm.2023.1257688 FIGURE 10
Adhesion of Probiotic Bacteria (LGG) to the in vitro Intestinal Epithelium Model. (A) Schematic representation of the adhesion of “released” probiotic
bacteria (LGG) to the surface of the cultured epithelium model expressing heavily glycosylated mucin proteins. LGG binds to the glycan chains of mucin
using SpaC expressed in pili. (B) Immunofluorescence top and side view images of the cultured colon epithelium model. Red: ZO-1 (tight junction), blue:
DAPI (cell nucleus), and green: mucin (C) Left: monitoring of LGG adhesion to the in vitro model of the intestinal epithelium. The polarized epithelial
monolayer was transferred from the Transwell
® culture to the flow chamber after confirming the establishment of tight junctions. Right: a representative
image of LGG adhered to the surface of a model epithelium. Adapted from (Eshrati et al., 2019), with permission from the American Chemical Society. FIGURE 10
Adhesion of Probiotic Bacteria (LGG) to the in vitro Intestinal Epithelium Model. (A) Schematic representation of the adhesion of “released” probiotic
bacteria (LGG) to the surface of the cultured epithelium model expressing heavily glycosylated mucin proteins. LGG binds to the glycan chains of mucin
using SpaC expressed in pili. (B) Immunofluorescence top and side view images of the cultured colon epithelium model. Red: ZO-1 (tight junction), blue:
DAPI (cell nucleus), and green: mucin (C) Left: monitoring of LGG adhesion to the in vitro model of the intestinal epithelium. The polarized epithelial
monolayer was transferred from the Transwell
® culture to the flow chamber after confirming the establishment of tight junctions. Right: a representative
image of LGG adhered to the surface of a model epithelium. Adapted from (Eshrati et al., 2019), with permission from the American Chemical Society. The protection of the functions of probiotic bacteria and
enzymes throughout delivery can be evaluated using in vitro
systems that simulate the microenvironments of the human
gastrointestinal tract. For example, the survival of bacteria can be
monitored by counting the CFUs after their passage through
chambers mimicking the stomach, small intestine, ileum, and
colon
environments,
whereas
the
function
of
bacteria
and
enzymes can be assessed by measuring the amount of bacteria or
enzyme products. Frontiers in Soft Matter frontiersin.org References Behnsen, J., Deriu, E., Sassone-Corsi, M., and Raffatellu, M. (2013). Probiotics:
properties, examples, and specific applications. Cold Spring Harb. Perspect. Med. 3
(3), a010074. doi:10.1101/cshperspect.a010074 Abuillan, W., Schneck, E., Körner, A., Brandenburg, K., Gutsmann, T., Gill, T., et al. (2013). Physical interactions of fish protamine and antisepsis peptide drugs with
bacterial membranes revealed by combination of specular x-ray reflectivity and
grazing-incidence x-ray fluorescence. Phys. Rev. E 88 (1), 012705. doi:10.1103/
PhysRevE.88.012705 Bernal,
C.,
Illanes,
A.,
and
Wilson,
L. (2014). Heterofunctional
hydrophilic–hydrophobic
porous
silica
as
support
for
multipoint
covalent
immobilization of lipases: application to lactulose palmitate synthesis. Langmuir 30
(12), 3557–3566. doi:10.1021/la4047512 Alfredsson, V., Lo Nostro, P., Ninham, B., and Nylander, T. (2021). Morphologies and
structure of brain lipid membrane dispersions. Front. Cell Dev. Biol. 9, 675140. doi:10. 3389/fcell.2021.675140 Bernal, C., Sierra, L., and Mesa, M. (2011). Application of hierarchical porous silica
with a stable large porosity for β-galactosidase immobilization. ChemCatChem 3 (12),
1948–1954. doi:10.1002/cctc.201100174 Amadei, F., Fröhlich, B., Stremmel, W., and Tanaka, M. (2018). Nonclassical
interactions of phosphatidylcholine with mucin protect intestinal surfaces: a
microinterferometry
study. Langmuir
34
(46),
14046–14057. doi:10.1021/acs. langmuir.8b03035 Bezryadina, A., C Preece, D., Chen, J. C., and Chen, Z. (2016). Optical disassembly of
cellular clusters by tunable ‘tug-of-war’ tweezers. Light Sci. Appl. 5 (10), e16158. doi:10. 1038/lsa.2016.158 Anderson, D., Wennerstroem, H., and Olsson, U. (1989). Isotropic bicontinuous
solutions in surfactant-Solvent systems: the L3 phase. J. Phys. Chem. 93 (10), 4243–4253. doi:10.1021/j100347a067 Blin, J.-L., Jacoby, J., Kim, S., Stébé, M.-J., Canilho, N., and Pasc, A. (2014). A meso-
macro compartmentalized bioreactor obtained through silicalization of “green” double
emulsions: W/O/W and W/SLNs/W. . Chem. Commun. 50 (80), 11871–11874. doi:10. 1039/C4CC06007E Angelova, A., Angelov, B., Mutafchieva, R., Lesieur, S., and Couvreur, P. (2011). Self-
assembled multicompartment liquid crystalline lipid carriers for protein, peptide, and
nucleic acid drug delivery. Accounts Chem. Res. 44 (2), 147–156. doi:10.1021/ar100120v Boks, N. P., Busscher, H. J., van der Mei, H. C., and Norde, W. (2008). Bond-
strengthening in staphylococcal adhesion to hydrophilic and hydrophobic surfaces
using atomic force microscopy. Langmuir 24 (22), 12990–12994. doi:10.1021/la801824c Assenza, S., and Mezzenga, R. (2019). Soft condensed matter physics of foods and
macronutrients. Nat. Rev. Phys. 1 (9), 551–566. doi:10.1038/s42254-019-0077-8 Brandenburg, K., Koch, M. H. J., and Seydel, U. (1992). Phase diagram of deep rough
mutant lipopolysaccharide from Salmonella Minnesota R595. J. Struct. Biol. 108 (2),
93–106. doi:10.1016/1047-8477(92)90010-8 Barauskas, J., and Nylander, T. (2008). Publisher’s note All claims expressed in this article are solely those of the authors and
do
not
necessarily
represent
those
of
their
affiliated
organizations, or those of the publisher, the editors and the
reviewers. Any product that may be evaluated in this article,
or claim that may be made by its manufacturer, is not guaranteed
or endorsed by the publisher. Conflict of interest The authors declare financial support was received for the
research, authorship, and/or publication of this article. Much of
the work presented in this review were performed with the
support by EU FP7 under and REA grant agreement no
606713 BIBAFOODS. JR is thankful to the Independent
Research Fund Denmark for support (grant 8022-00139B). TN
acknowledges the financial support from the Swedish Research
Council
under
contract
2017-06716
and
2020-05421
and
NanoLund. MT thanks JSPS (JP19H05719) and Nakatani
Foundation for their support. The authors declare that the research was conducted in the
absence of any commercial or financial relationships that could be
construed as a potential conflict of interest. The authors MT and TN declared that they were editorial
board members of Frontiers at the time of submission. This had
no impact on the peer review process and the final decision. Acknowledgments The authors didicate this manuscript to the memory of Maria
Miguel, who passed away too early. The authors thank the following 6 Final remarks and perspectives 10.3389/frsfm.2023.1257688 people for fruitful collaboration and stimulating discussions: Björn
Lindman, Maria Miguel, Andreea Pasc, Marité Cardenas, Stefan
Kaufmann,
Tom
Van
Wiele,
Bruno
Medronho,
Stefan
Ulvenlund, Gemma Montalvo, Sam Possemiers, Jordi Esquena,
Hans van den Brink. Dennis S. Nielsen, Nils Arneborg, and
Kathryn A. Whitehead. The authors are especially thankful to
the graduate students and junior and senior researchers who
contributed to the work presented here, including M. Valldeperas
Badell, J. Gilbert, I. Ermilova, J. Swenson, M. Nagao, J. Barauskas,
C. Yucel Falco, I. Pavel, F. Haffner, M. Eshrati, F. Amadei, P. Singh, M. Veschgini, W. Abuillan, X. Jiang, E. Shekarforoush,
M.K. Muhammed. We thank Editage (www.editage.jp) for
English language editing. bacteria and enzymes could be a more direct approach. The
development of efficient carriers for the delivery of gut bacteria
opens a new challenge because the natural gut bacteria are strictly
anaerobic, unlike Lactobacillus in yoghurt; therefore, they need to be
protected against oxygen before ingestion. 6 Final remarks and perspectives Upon reaching their destination, the delivered
probiotic bacteria must adhere to the surface of the colon epithelium
coated with mucin and form a new microbiome. The initial phase,
that is, the adhesion of bacteria onto the mucin surface, can be
monitored either by coating the surface of dishes or microfluidic
channels with isolated mucin proteins or by transferring the
polarized monolayers of enterocyte-like cells and goblet cell-like
cells expressing mucin. This enables the determination of the
binding/unbinding force and critical shear stress for adhesion/
detachment. food ingested by humans. Despite this simple statement, it involves
many challenges in the field of soft matter science. This review aimed
to highlight how the key concepts in soft matter science can be used
to design, characterize, and evaluate self-assembled formulations
based on lipids and biopolymers. The topics covered in this review
included the emulsification of oil-water mixtures, the self-assembly
of lipids and polymers at interfaces, the electrostatics and
viscoelasticity of interfaces, and the wetting/adhesion of colloidal
particles. One of the specific challenges in food applications is the series of
diverse chemical conditions and timescales that the carriers
experience throughout their journeys from storage to the colon
through the entire gastrointestinal tract. This calls for the design of
responsive, food-grade materials that ensure the protection of the
probiotic bacteria and enzymes under highly acidic gastric conditions
(pH 2, pepsin) and the intestinal conditions exposed to pancreatic
juice (pH 6.5, pancreatin and bile acid). Many approaches introduced
here have demonstrated that exploiting the self-assembly of lipids and
biopolymers, such as the sponge phase of lipids and layer-by-layer
coating with nm-thick polyelectrolyte layers, opens the possibility to
tune the encabsulation capability and stability of the carriers. This
offers the possibility not only to protect the bacteria and enzymes
under the harsh gastrointestinal environments, but also to release the
active component at the right place. The increasing knowledge on the key roles of compositions of
healthy and less healthy microbiomes in sustaining health and
wellbeing
has
drawn
increasing
attention
in
improving
microbiomes by the delivery of probiotic bacteria and enzymes. Changing the composition of the human diet (such as a more plant-
based diet) is one direction; however, the delivery of missing gut 16 frontiersin.org Risbo et al. Author contributions JR: Conceptualization, Writing–original draft, Writing–review
and
editing. TN:
Conceptualization,
Writing–original
draft,
Writing–review
and
editing. MT:
Conceptualization,
Writing–original draft, Writing–review and editing. References Nanoscale 14 (18), 6990–7002. doi:10.1039/
D2NR00882C Decher, G. (1997). Fuzzy nanoassemblies: toward layered polymeric multicomposites. Science 277, 1232–1237. doi:10.1126/science.277.5330.1232 Diab, R., Canilho, N., Pavel, I. A., Haffner, F. B., Girardon, M., and Pasc, A. (2017). Silica-based systems for oral delivery of drugs, macromolecules and cells. Adv. Colloid
Interface Sci. 249, 346–362. doi:10.1016/j.cis.2017.04.005 Gilbert, J., Valldeperas, M., Dhayal, S. K., Barauskas, J., Dicko, C., and Nylander, T. (2019). Immobilisation of β-galactosidase within a lipid sponge phase: structure,
stability and kinetics characterisation. Nanoscale 11 (44), 21291–21301. doi:10.1039/
C9NR06675F Diaspro, A., Silvano, D., Krol, S., Cavalleri, O., and Gliozzi, A. (2002). Single living cell
encapsulation in nano-organized polyelectrolyte shells. Langmuir 18, 5047–5050. doi:10.1021/la025646e G. J. Siegel, B. W. Agranoff, R. Wayne Albers, S. K. Fisher, and M. D. Uhler (Editors)
(1999). Basic neurochemistry (Philadelphia: Lippincott-Raven). Dominy, S. S., Lynch, C., Ermini, F., Benedyk, M., Marczyk, A., Konradi, A., et al. (2019). Porphyromonas gingivalis in Alzheimer’s disease brains: evidence for disease
causation and treatment with small-molecule inhibitors. Sci. Adv. 5, eaau3333. doi:10. 1126/sciadv.aau3333 Gouin, S. (2004). Microencapsulation: industrial appraisal of existing technologies
and trends. Trends Food Sci. Technol. 15, 330–347. doi:10.1016/j.tifs.2003.10.005 Guarino, V., Caputo, T., Altobelli, R., and Ambrosio, L. (2015). Degradation
properties and metabolic activity of alginate and chitosan polyelectrolytes for drug
delivery and tissue engineering applications. AIMS Mater. Sci. 2 (4), 497–502. doi:10. 3934/matersci.2015.4.497 Dufrêne, Y. F., Viljoen, A., Mignolet, J., and Mathelié-Guinlet, M. (2021). AFM in
cellular and molecular microbiology. Cell. Microbiol. 23 (7), e13324. doi:10.1111/cmi. 13324 Haffner, F. B., Girardon, M., Fontanay, S., Canilho, N., Duval, R. E., Mierzwa, M., et al. (2016). Core–shell alginate@silica microparticles encapsulating probiotics. J. Mater. Chem. B 4 (48), 7929–7935. doi:10.1039/C6TB02802K Dully, M., Brasnett, C., Djeghader, A., Seddon, A., Neilan, J., Murray, D., et al. (2020). Modulating the release of pharmaceuticals from lipid cubic phases using a lipase
inhibitor. J. colloid interface Sci. 573, 176–192. doi:10.1016/j.jcis.2020.04.015 Haffner, F. B., van de Wiele, T., and Pasc, A. (2017). Original behavior of L. rhamnosus GG encapsulated in freeze-dried alginate–silica microparticles revealed
under simulated gastrointestinal conditions. J. Mater. Chem. B 5 (38), 7839–7847. doi:10.1039/C7TB02190A Dupres, V., Menozzi, F. D., Locht, C., Clare, B. H., Abbott, N. L., Cuenot, S., et al. (2005). Nanoscale mapping and functional analysis of individual adhesins on living
bacteria. Nat. Methods 2 (7), 515–520. doi:10.1038/nmeth769 Ehehalt, R., Wagenblast, J., Erben, G., Lehmann, W. D., Hinz, U., Merle, U., et al. (2004). References M., and Leong, K. W. (2013). In vitro and in vivo models for the study of
oral delivery of nanoparticles. Adv. drug Deliv. Rev. 65 (6), 800–810. doi:10.1016/j.addr. 2013.01.003 Clifton, L. A., Skoda, M. W. A., Daulton, E. L., Hughes, A. V., Le Brun, A. P., Lakey,
J. H., et al. (2013). Asymmetric phospholipid: lipopolysaccharide bilayers; a Gram-
negative bacterial outer membrane mimic. J. R. Soc. Interface 10 (89), 20130810. doi:10. 1098/rsif.2013.0810 Garidel, P., Rappolt, M., Schromm, A. B., Howe, J., Lohner, K., Andrä, J., et al. (2005). Divalent cations affect chain mobility and aggregate structure of lipopolysaccharide from
Salmonella Minnesota reflected in a decrease of its biological activity. Biochimica Biophysica
Acta (BBA) - Biomembr. 1715 (2), 122–131. doi:10.1016/j.bbamem.2005.07.013 Cook, M. T., Tzortzis, G., Charalampopoulos, D., and Khutoryanskiy, V. V. (2011). Production and evaluation of dry alginate-chitosan microcapsules as an enteric delivery
vehicle for probiotic bacteria. Biomacromolecules 12, 2834–2840. doi:10.1021/
bm200576h Gasperini, L., Mano, J. F., and Reis, R. L. (2014). Natural polymers for the
microencapsulation of cells. J. R. Soc. Interface 11, 20140817. doi:10.1098/rsif.2014.0817 Dague, E., Le, D. T. L., Zanna, S., Marcus, P., Loubière, P., and Mercier-Bonin, M. (2010). Probing in vitro interactions between Lactococcus lactis and mucins using AFM. Langmuir 26 (13), 11010–11017. doi:10.1021/la101862n Geertsema-Doornbusch, G. I., van der Mei, H. C., and Busscher, H. J. (1993). Microbial cell surface hydrophobicity the involvement of electrostatic interactions in
microbial adhesion to hydrocarbons (MATH). J. Microbiol. Methods 18 (1), 61–68. doi:10.1016/0167-7012(93)90072-P Date, A. A., Halpert, G., Babu, T., Ortiz, J., Kanvinde, P., Dimitrion, P., et al. (2018). Mucus-penetrating budesonide nanosuspension
enema
for
local treatment
of
inflammatory bowel disease. Biomaterials 185, 97–105. doi:10.1016/j.biomaterials. 2018.09.005 Gilbert, E. P. (2019). Small-angle X-Ray and neutron scattering in food colloids. Curr. Opin. Colloid & Interface Sci. 42, 55–72. doi:10.1016/j.cocis.2019.03.005 Gilbert, J., Christensen, S., Nylander, T., and Bülow, L. (2023). Encapsulation of sugar
beet phytoglobin BvPgb 1.2 and myoglobin in a lipid sponge phase system. Front. Soft
Matter 3. doi:10.3389/frsfm.2023.1201561 De Koker, S., De Geest, B. G., Singh, S. K., De Rycke, R., Naessens, T., Van Kooyk, Y.,
et al. (2009). Polyelectrolyte microcapsules as antigen delivery vehicles to dendritic cells:
uptake, processing, and cross-presentation of encapsulated antigens. Angew. Chem. 121,
8637–8641. doi:10.1002/ange.200903769 Gilbert, J., Ermilova, I., Nagao, M., Swenson, J., and Nylander, T. (2022). Effect of
encapsulated protein on the dynamics of lipid sponge phase: a neutron spin echo and
molecular dynamics simulation study. References “Lyotropic liquid crystals as delivery vehicles
for food ingredients,” in Delivery and controlled release of bioactives in foods and
nutraceuticals. Editor N. Garti (Cambridge: Woodhead Publishing), 107–131. Brooks, C. F., Fuller, G. G., Frank, C. W., and Robertson, C. R. (1999). An interfacial
stress rheometer to study rheological transitions in monolayers at the Air−Water
interface. Langmuir 15 (7), 2450–2459. doi:10.1021/la980465r Barriga, H. M. G., Holme, M. N., and Stevens, M. M. (2019). Cubosomes: the next
generation of smart lipid nanoparticles? Angew. Chem. Int. Ed. 58 (10), 2958–2978. doi:10.1002/anie.201804067 17 Frontiers in Soft Matter frontiersin.org Risbo et al. Risbo et al. 10.3389/frsfm.2023.1257688 as an in vitro model of human intestinal epithelium. Biosci. Rep. 38 (2). doi:10.1042/
bsr20171497 Callone, E., Campostrini, R., Carturan, G., Cavazza, A., and Guzzon, R. (2008). Immobilization of yeast and bacteria cells in alginate microbeads coated with silica
membranes: procedures, physico-chemical features and bioactivity. J. Mater. Chem. 18
(40), 4839–4848. doi:10.1039/B807301E Firoozmand, H., and Rousseau, D. (2016). Microbial cells as colloidal particles:
pickering oil-in-water emulsions stabilized by bacteria and yeast. Food Res. Int. 81,
66–73. doi:10.1016/j.foodres.2015.10.018 Cao, T., Tang, H., Liang, X., Wang, A., Auner, G. W., Salley, S. O., et al. (2006). Nanoscale investigation on adhesion of E. coli to surface modified silicone using
atomic force microscopy. Biotechnol. Bioeng. 94 (1), 167–176. doi:10.1002/bit. 20841 Fong, W.-K., Negrini, R., Vallooran, J. J., Mezzenga, R., and Boyd, B. J. (2016). Responsive self-assembled nanostructured lipid systems for drug delivery and
diagnostics. J. Colloid Interface Sci. 484, 320–339. doi:10.1016/j.jcis.2016.08.077 Carlsson, N., Gustafsson, H., Thörn, C., Olsson, L., Holmberg, K., and Åkerman, B. (2014). Enzymes immobilized in mesoporous silica: a physical–chemical perspective. Adv. Colloid Interface Sci. 205, 339–360. doi:10.1016/j.cis.2013.08.010 Fröhlich, B., Dasanna, A. K., Lansche, C., Czajor, J., Sanchez, C. P., Cyrklaff, M., et al. (2021). Functionalized supported membranes for quantifying adhesion of P. falciparum-infected erythrocytes. Biophysical J. 120 (16), 3315–3328. doi:10.1016/j. bpj.2021.07.003 Caruso, F., Trau, D., Möhwald, H., and Renneberg, R. (2000). Enzyme encapsulation
in layer-by-layer engineered polymer multilayer capsules. Langmuir 16, 1485–1488. doi:10.1021/la991161n Gagnon, M., Zihler Berner, A., Chervet, N., Chassard, C., and Lacroix, C. (2013). Comparison of the Caco-2, HT-29 and the mucus-secreting HT29-MTX intestinal cell
models to investigate Salmonella adhesion and invasion. J. Microbiol. Methods 94 (3),
274–279. doi:10.1016/j.mimet.2013.06.027 Champagne, C. P., and Fustier, P. (2007). Microencapsulation for the improved
delivery of bioactive compounds into foods. Curr. Opin. Biotechnol. 18, 184–190. doi:10. 1016/j.copbio.2007.03.001 Gamboa, J. Frontiers in Soft Matter References 2 Mol. Chem. Phys. 77 (4), 601–629. doi:10.1039/
f29817700601 Jiang, X., Shekarforoush, E., Muhammed, M. K., Bovet, N., Ceccato, M., Whitehead,
K. A., et al. (2023a). Probing surface properties of lactic acid bacteria - comparative
modification by anhydride and aldehyde grafting. Surfaces Interfaces 38, 102848. doi:10. 1016/j.surfin.2023.102848 Muhammed, M. K., Jiang, X., Shekarforoush, E., Whitehead, K., Vogensen, F. K.,
Risbo, J., et al. (2023). Ex-<em>Lactobacillus</em> strains with intrinsic propensity to
stabilize pickering oil-in-water emulsions. bioRxiv, 540633. doi:10.1101/2023.05.13. 540633 Jiang, X., Shekarforoush, E., Muhammed, M. K., Whitehead, K. A., Arneborg, N., and
Risbo, J. (2023b). Lactic acid bacteria as structural building blocks in non-fat whipping
cream analogues. Food Hydrocoll. 135, 108137. doi:10.1016/j.foodhyd.2022.108137 Mulet, X., Boyd, B. J., and Drummond, C. J. (2013). Advances in drug delivery and
medical imaging using colloidal lyotropic liquid crystalline dispersions. J. Colloid
Interface Sci. 393, 1–20. doi:10.1016/j.jcis.2012.10.014 Jiang, X., Shekarforoush, E., Muhammed, M. K., Whitehead, K., Simonsen, A. C.,
Arneborg, N., et al. (2021). Efficient chemical hydrophobization of lactic acid
bacteria – one-step formation of double emulsion. Food Res. Int. 147, 110460. doi:10.1016/j.foodres.2021.110460 Nassif, N., and Livage, J. (2011). From diatoms to silica-based biohybrids. Chem. Soc. Rev. 40 (2), 849–859. doi:10.1039/C0CS00122H Netz, R. R. (1999). Debye-Hückel theory for interfacial geometries. Phys. Rev. E 60 (3),
3174–3182. doi:10.1103/PhysRevE.60.3174 Jiang, X., Yucel Falco, C., Dalby, K. N., Siegumfeldt, H., Arneborg, N., and Risbo, J. (2019). Surface engineered bacteria as Pickering stabilizers for foams and emulsions. Food Hydrocoll. 89, 224–233. doi:10.1016/j.foodhyd.2018.10.044 No, H. K., Park, N. Y., Lee, S. H., and Meyers, S. P. (2002). Antibacterial activity of
chitosans and chitosan oligomers with different molecular weights. Int. J. food
Microbiol. 74 (1-2), 65–72. doi:10.1016/s0168-1605(01)00717-6 Kaindl, T., Rieger, H., Kaschel, L.-M., Engel, U., Schmaus, A., Sleeman, J., et al. (2012). Spatio-temporal patterns of pancreatic cancer cells expressing CD44 isoforms on
supported membranes displaying hyaluronic acid oligomers arrays. PLoS One 7,
e42991. doi:10.1371/journal.pone.0042991 Oliveira, R. G., Schneck, E., Quinn, B. E., Konovalov, O. V., Brandenburg, K.,
Gutsmann, T., et al. (2010). Crucial roles of charged saccharide moieties in survival
of gram negative bacteria against protamine revealed by combination of grazing
incidence x-ray structural characterizations and Monte Carlo simulations. Phys. Rev. E 81 (4), 041901. doi:10.1103/PhysRevE.81.041901 Katchalsky, A., Cooper, R. E., Upadhyay, J., and Wassermann, A. (1961). 1028. Counter-ion fixation in alginates. J. Chem. Soc. (Resumed) (0), 5198–5204. doi:10.1039/
JR9610005198 Oliveira, R. G., Schneck, E., Quinn, B. E., Konovalov, O. References doi:10.1021/j100358a010 Larsson, K., Quinn, P., Sato, K., and Tyberg, F. (2006). Lipids: structure, physical
properties and functionality. Bridgewater: The Oily Press. Pham, A. C., Peng, K.-Y., Salim, M., Ramirez, G., Hawley, A., Clulow, A. J., et al. (2020). Correlating digestion-driven self-assembly in milk and infant formulas with
changes in lipid composition. ACS Appl. Bio Mater. 3 (5), 3087–3098. doi:10.1021/
acsabm.0c00131 Lee, C.-H., Lin, T.-S., and Mou, C.-Y. (2009). Mesoporous materials for encapsulating
enzymes. Nano today 4 (2), 165–179. doi:10.1016/j.nantod.2009.02.001 Pink, D. A., Truelstrup Hansen, L., Gill, T. A., Quinn, B. E., Jericho, M. H., and
Beveridge, T. J. (2003). Divalent calcium ions inhibit the penetration of protamine
through the polysaccharide brush of the outer membrane of gram-negative bacteria. Langmuir 19 (21), 8852–8858. doi:10.1021/la030193e Lee, K. Y., and Mooney, D. J. (2012). Alginate: properties and biomedical applications. Prog. Polym. Sci. 37, 106–126. doi:10.1016/j.progpolymsci.2011.06.003 Liu, D., and Du, D. (2020). Mulberry fruit extract alleviates cognitive impairment by
promoting the clearance of amyloid-β and inhibiting neuroinflammation in Alzheimer’s
disease mice. Neurochem. Res. 45 (9), 2009–2019. doi:10.1007/s11064-020-03062-7 Porcar, L., Hamilton, W. A., Butler, P. D., and Warr, G. (2003). Scaling of structural
and rheological response of L3 sponge phases in the “sweetened” cetylpyridinium/
hexanol/dextrose/brine
system. Langmuir
19
(26),
10779–10794. doi:10.1021/
la034661w Lvov, Y., Decher, G., and Moehwald, H. (1993). Assembly, structural
characterization, and thermal behavior of layer-by-layer deposited ultrathin
films of poly (vinyl sulfate) and poly (allylamine). Langmuir 9, 481–486. doi:10. 1021/la00026a020 Priya, A. J., Vijayalakshmi, S. P., and Raichur, A. M. (2011). Enhanced survival of
probiotic
lactobacillus
acidophilus
by
encapsulation
with
nanostructured
polyelectrolyte layers through layer-by-layer approach. J. Agric. Food Chem. 59 (21),
11838–11845. doi:10.1021/jf203378s Maldonado-Valderrama, J., Wilde, P., Macierzanka, A., and Mackie, A. (2011). The
role of bile salts in digestion. Adv. Colloid Interface Sci. 165 (1), 36–46. doi:10.1016/j.cis. 2010.12.002 Proitsi, P., Kim, M., Whiley, L., Simmons, A., Sattlecker, M., Velayudhan, L., et al. (2017). Association of blood lipids with Alzheimer’s disease: a comprehensive
lipidomics analysis. Alzheimer’s Dementia 13 (2), 140–151. doi:10.1016/j.jalz.2016. 08.003 Manzano, M., and Vallet-Regí, M. (2020). Mesoporous silica nanoparticles for drug
delivery. Adv. Funct. Mater. 30 (2), 1902634. doi:10.1002/adfm.201902634 Martin, H. H. (1963). Bacterial protoplasts—a review. J. Theor. Biol. 5 (1), 1–34. doi:10.1016/0022-5193(63)90034-1 Razumas, V., Kanapieniené, J., Nylander, T., Engström, S., and Larsson, K. (1994). Electrochemical biosensors for glucose, lactate, urea, and creatinine based on enzymes
entrapped in a cubic liquid crystalline phase. Anal. Chim. acta 289 (2), 155–162. References Phosphatidylcholine and lysophosphatidylcholine in intestinal mucus of
ulcerative colitis patients. A quantitative approach by nanoelectrospray-tandem
mass
spectrometry. Scand. J. gastroenterology
39
(8),
737–742. doi:10.1080/
00365520410006233 Herrmann, M., Schneck, E., Gutsmann, T., Brandenburg, K., and Tanaka, M. (2015). Bacterial
lipopolysaccharides
form
physically
cross-linked,
two-
dimensional gels in the presence of divalent cations. Soft Matter 11 (30),
6037–6044. doi:10.1039/C5SM01002K Higaki, Y., Frohlich, B., Yamamoto, A., Murakami, R., Kaneko, M., Takahara, A., et al. (2017). Ion-specific modulation of interfacial interaction potentials between solid
substrates and cell-sized particles mediated via zwitterionic, super-hydrophilic poly
(sulfobetaine) brushes. J. Phys. Chem. B 121 (6), 1396–1404. doi:10.1021/acs.jpcb.6b11540 Eshrati, M., Amadei, F., Staffer, S., Stremmel, W., and Tanaka, M. (2019). Shear-
enhanced dynamic adhesion of lactobacillus rhamnosus GG on intestinal epithelia:
correlative effect of protein expression and interface mechanics. Langmuir 35 (2),
529–537. doi:10.1021/acs.langmuir.8b02931 Hudson, S., Cooney, J., and Magner, E. (2008). Proteins in mesoporous silicates. Angew. Chem. Int. Ed. 47 (45), 8582–8594. doi:10.1002/anie.200705238 Eshrati, M., Amadei, F., Van de Wiele, T., Veschgini, M., Kaufmann, S., and Tanaka,
M. (2018). Biopolymer-based minimal formulations boost viability and metabolic
functionality of probiotics lactobacillus rhamnosus GG through gastrointestinal
passage. Langmuir 34 (37), 11167–11175. doi:10.1021/acs.langmuir.8b01915 Humphreys, B. A., Campos-Terán, J., Arnold, T., Baunsgaard, L., Vind, J., Dicko, C.,
et al. (2022). The influence of pH on the lipase digestion of nanosized triolein, diolein
and monoolein films. Front. Soft Matter 2. doi:10.3389/frsfm.2022.929104 FAO/WHO (2001). Evaluation of Health and Nutritional Properties of Probiotics in
Food Including Powder Milk with Live Acid Bacteria. report: report of a Joint FAO/
WHO Expert Consultation, Córdoba, Argentina. Available at: http://www.who.int/
foodsafety/publications/fs_management/en/probiotics.pdf?ua=1. Israelachvili, J. N., Mitchell, D. J., and Ninham, B. W. (1976). Theory of self-assembly
of hydrocarbon amphiphiles into micelles and bilayers. J. Chem. Soc. Faraday Trans. 2 Mol. Chem. Phys. 72, 1525–1568. doi:10.1039/f29767201525 Ferraretto, A., Bottani, M., De Luca, P., Cornaghi, L., Arnaboldi, F., Maggioni, M.,
et al. (2018). Morphofunctional properties of a differentiated Caco2/HT-29 co-culture Ferraretto, A., Bottani, M., De Luca, P., Cornaghi, L., Arnaboldi, F., Maggioni, M.,
et al. (2018). Morphofunctional properties of a differentiated Caco2/HT-29 co-culture Jiang, X., Martens, H. J., Shekarforoush, E., Muhammed, M. K., Whitehead, K. A.,
Arneborg, N., et al. (2022). Multi-species colloidosomes by surface-modified lactic acid 18 frontiersin.org Risbo et al. 10.3389/frsfm.2023.1257688 bacteria with enhanced aggregation properties. J. Colloid Interface Sci. 622, 503–514. doi:10.1016/j.jcis.2022.04.136 Mitchell, D. J., and Ninham, B. W. (1981). Micelles, vesicles and microemulsions. J. Chem. Soc. Faraday Trans. References V., Brandenburg, K., Seydel,
U., et al. (2009). Physical mechanisms of bacterial survival revealed by combined
grazing-incidence X-ray scattering and Monte Carlo simulation. Comptes Rendus Chim. 12 (1), 209–217. doi:10.1016/j.crci.2008.06.020 Kleiveland, C. R. (2015). “Co-Cultivation of caco-2 and HT-29mtx,” in The Impact of
Food Bioactives on Health: in vitro and ex vivo models. Editors K. Verhoeckx, P. Cotter,
I. López-Expósito, C. Kleiveland, T. Lea, A. Mackie, et al. (Cham: Springer International
Publishing), 135–140. Orive, G., Ponce, S., Hernández, R. M., Gascón, A. R., Igartua, M., and Pedraz, J. L. (2002). Biocompatibility of microcapsules for cell immobilization elaborated with
different type of alginates. Biomaterials 23, 3825–3831. doi:10.1016/s0142-9612(02)
00118-7 Korytowski, A., Abuillan, W., Amadei, F., Makky, A., Gumiero, A., Sinning, I., et al. (2017). Accumulation of phosphatidylcholine on gut mucosal surface is not dominated
by electrostatic interactions. Biochimica Biophysica Acta (BBA) - Biomembr. 1859 (5),
959–965. doi:10.1016/j.bbamem.2017.02.008 Pasquier, J., Brûlet, A., Boire, A., Jamme, F., Perez, J., Bizien, T., et al. (2019). Monitoring food structure during digestion using small-angle scattering and imaging
techniques. Colloids Surfaces A Physicochem. Eng. Aspects 570, 96–106. doi:10.1016/j. colsurfa.2019.02.059 Krasaekoopt, W., Bhandari, B., and Deeth, H. (2003). Evaluation of encapsulation
techniques of probiotics for yoghurt. Int. Dairy J. 13, 3–13. doi:10.1016/s0958-6946(02)
00155-3 Patton, J. S., and Carey, M. C. (1979). Watching fat digestion. Science 204 (4389),
145–148. doi:10.1126/science.432636 Krasaekoopt, W., Bhandari, B., and Deeth, H. (2004). The influence of coating
materials on some properties of alginate beads and survivability of microencapsulated
probiotic bacteria. Int. Dairy J. 14, 737–743. doi:10.1016/j.idairyj.2004.01.004 Pavel, I.-A., Prazeres, S. F., Montalvo, G., Garcia Ruiz, C., Nicolas, V., Celzard, A.,
et al. (2017). Effect of meso vs macro size of hierarchical porous silica on the adsorption
and activity of immobilized β-galactosidase. Langmuir 33 (13), 3333–3340. doi:10.1021/
acs.langmuir.7b00134 Kwiecień, I., and Kwiecień, M. (2018). Application of polysaccharide-based hydrogels
as probiotic delivery systems. Gels 4 (2), 47. doi:10.3390/gels4020047 Larsson, K. (2000). Aqueous dispersions of cubic lipid–water phases. Curr. Opin. Colloid & Interface Sci. 5 (1), 64–69. doi:10.1016/S1359-0294(00)00040-6 Pelaseyed, T., Bergström, J. H., Gustafsson, J. K., Ermund, A., Birchenough, G. M.,
Schütte, A., et al. (2014). The mucus and mucins of the goblet cells and enterocytes
provide the first defense line of the gastrointestinal tract and interact with the immune
system. Immunol. Rev. 260 (1), 8–20. doi:10.1111/imr.12182 Larsson, K. (1989). Cubic lipid-water phases: structures and biomembrane aspects. J. Phys. Chem. 93 (21), 7304–7314. References doi:10. 1016/0003-2670(94)80098-7 Matricardi, P., Meo, C. D., Coviello, T., and Alhaique, F. (2008). Recent advances and
perspectives on coated alginate microspheres for modified drug delivery. Expert Opin. Drug Deliv. 5, 417–425. doi:10.1517/17425247.5.4.417 McClements, D. J. (2015). Encapsulation, protection, and release of hydrophilic active
components: potential and limitations of colloidal delivery systems. Adv. Colloid
Interface Sci. 219, 27–53. doi:10.1016/j.cis.2015.02.002 Rieger, H., Yoshikawa, H. Y., Quadt, K., Nielsen, M. A., Sanchez, C. P., Salanti, A.,
et al. (2015). Cytoadhesion of Plasmodium falciparum–infected erythrocytes to
chondroitin-4-sulfate is cooperative and shear enhanced. Blood, J. Am. Soc. Hematol. 125 (2), 383–391. doi:10.1182/blood-2014-03-561019 Minekus, M., Alminger, M., Alvito, P., Ballance, S., Bohn, T., Bourlieu, C., et al. (2014). A standardised static in vitro digestion method suitable for food – an
international consensus. Food & Funct. 5 (6), 1113–1124. doi:10.1039/C3FO60702J Rohde, M. (2019). The gram-positive bacterial cell wall. Microbiol. Spectr. 7(3), doi:10. 1128/microbiolspec.gpp1123-0044-2018 19 Frontiers in Soft Matter frontiersin.org Risbo et al. Risbo et al. 10.3389/frsfm.2023.1257688 Rosenberg, M. (1984). Bacterial adherence to hydrocarbons: a useful technique for
studying cell surface hydrophobicity. FEMS Microbiol. Lett. 22 (3), 289–295. doi:10. 1111/j.1574-6968.1984.tb00743.x Tasciotti, E., Liu, X., Bhavane, R., Plant, K., Leonard, A. D., Price, B. K., et al. (2008). Mesoporous silicon particles as a multistage delivery system for imaging and therapeutic
applications. Nat. Nanotechnol. 3 (3), 151–157. doi:10.1038/nnano.2008.34 Rosenberg, M., Gutnick, D., and Rosenberg, E. (1980). Adherence of bacteria to
hydrocarbons: a simple method for measuring cell-surface hydrophobicity. FEMS
Microbiol. Lett. 9 (1), 29–33. doi:10.1111/j.1574-6968.1980.tb05599.x Thoma, J., Abuillan, W., Furikado, I., Habe, T., Yamamoto, A., Gierlich, S., et al. (2020). Specific localisation of ions in bacterial membranes unravels physical mechanism of effective
bacteria killing by sanitiser. Sci. Rep. 10 (1), 12302. doi:10.1038/s41598-020-69064-1 Rosenberg, M. (2006). Microbial adhesion to hydrocarbons: twenty-five years of doing
MATH. FEMS Microbiol. Lett. 262 (2), 129–134. doi:10.1111/j.1574-6968.2006.00291.x Thomas, T., Hawzeen, S. K., and Jack, C. L. (2018). Bacterial autoaggregation. AIMS
Microbiol. 4 (1), 140–164. doi:10.3934/microbiol.2018.1.140 Thompson,K.L.,Williams, M.,and Armes,S.P.(2015).Colloidosomes:synthesis, properties
and applications. J. Colloid Interface Sci. 447, 217–228. doi:10.1016/j.jcis.2014.11.058 Rowan-Nash, A. D., Korry, B. J., Mylonakis, E., and Belenky, P. (2019). Cross-domain
and viral interactions in the microbiome. Microbiol. Mol. Biol. Rev. 83 (1),
e00044–e00018. doi:10.1128/MMBR.00044-18 Tran, P. H. L., and Tran, T. T. D. (2021). Blueberry supplementation in neuronal
health and protective technologies for efficient delivery of blueberry anthocyanins. Biomolecules 11 (1), 102. doi:10.3390/biom11010102 Salentinig, S., Phan, S., Hawley, A., and Boyd, B. J. (2015). References Rheology of glycocalix model
at air/water interface. Phys. Chem. Chem. Phys. 4 (10), 1949–1952. doi:10.1039/B110631G Valldeperas, M., Wiśniewska, M., Ram-On, M., Kesselman, E., Danino, D., Nylander,
T., et al. (2016). Sponge phases and nanoparticle dispersions in aqueous mixtures of
mono- and diglycerides. Langmuir 32 (34), 8650–8659. doi:10.1021/acs.langmuir. 6b01356 Schwarz, B., and Schönhoff, M. (2002). A 1H NMR relaxation study of hydration water in
polyelectrolyte mono and multilayers adsorbed to colloidal particles. Colloids Surfaces A
Physicochem. Eng. Aspects 198, 293–304. doi:10.1016/s0927-7757(01)00945-1 Van de Wiele, T., Van den Abbeele, P., Ossieur, W., Possemiers, S., and Marzorati, M. (2015). The simulator of the human intestinal microbial ecosystem (SHIME®). Impact
Food Bioact. Health vitro ex vivo models 2015, 305–317. Seddon, J., and Templer, R. (1995). “Polymorphism of lipid-water systems,” in
Handbook of biological physics (Elsevier), 97–160. Van Oss, C. J., Chaudhury, M. K., and Good, R. J. (1988). Interfacial Lifshitz-van der
Waals and polar interactions in macroscopic systems. Chem. Rev. 88 (6), 927–941. doi:10.1021/cr00088a006 Shekarforoush, E., Jiang, X., Kedir Muhammed, M., Whitehead, K. A., Arneborg, N.,
and Risbo, J. (2022). Enzymatic modification and adsorption of hydrophobic zein
proteins on lactic acid bacteria stabilize Pickering emulsions. Food Res. Int. 161, 111783. doi:10.1016/j.foodres.2022.111783 Verbelen, C., Raze, D., Dewitte, F., Locht, C., and Dufrene, Y. F. (2007). Single-
molecule force spectroscopy of mycobacterial adhesin-adhesin interactions. J. Bacteriol. 189 (24), 8801–8806. doi:10.1128/jb.01299-07 Silhavy, T. J., Kahne, D., and Walker, S. (2010). The bacterial cell envelope. Cold
Spring Harb. Perspect. Biol. 2 (5), a000414. doi:10.1101/cshperspect.a000414 Vollmer, W., Blanot, D., and De Pedro, M. A. (2008). Peptidoglycan structure and
architecture. FEMS Microbiol. Rev. 32 (2), 149–167. doi:10.1111/j.1574-6976.2007.00094.x Singh, P., Magalhães, S., Alves, L., Antunes, F., Miguel, M., Lindman, B., et al. (2019). Cellulose-based edible films for probiotic entrapment. Food Hydrocoll. 88, 68–74. doi:10.1016/j.foodhyd.2018.08.057 Wadsäter, M., Barauskas, J., Tiberg, F., and Nylander, T. (2018). The lipolytic
degradation
of
highly
structured
cubic
micellar
nanoparticles
of
soy
phosphatidylcholine and glycerol dioleate by phospholipase A2 and triacylglycerol
lipase. Chem. Phys. Lipids 211, 86–92. doi:10.1016/j.chemphyslip.2017.11.011 Singh, P., Medronho, B., Alves, L., da Silva, G. J., Miguel, M. G., and Lindman, B. (2017). Development
of
carboxymethyl
cellulose-chitosan
hybrid
micro-
and
macroparticles for encapsulation of probiotic bacteria. Carbohydr. Polym. 175,
87–95. doi:10.1016/j.carbpol.2017.06.119 Wang, M., Qi, W., Yu, Q., Su, R., and He, Z. (2010). Cross-linking enzyme aggregates
in the macropores of silica gel: a practical and efficient method for enzyme stabilization. Biochem. Eng. J. 52 (2-3), 168–174. References doi:10.1016/j.bej.2010.08.003 Singh, P., Medronho, B., Miguel, M. G., and Esquena, J. (2018a). On the encapsulation
and viability of probiotic bacteria in edible carboxymethyl cellulose-gelatin water-in-
water emulsions. Food Hydrocoll. 75, 41–50. doi:10.1016/j.foodhyd.2017.09.014 Woolfson, A. (2016). Microbiology: mob rule. Nature 536 (7615), 146–147. doi:10. 1038/536146a Singh, P., Medronho, B., Santos, T. d., Nunes-Correia, I., Granja, P., Miguel, M. G., et al. (2018b). On the viability, cytotoxicity and stability of probiotic bacteria entrapped in cellulose-
based particles. Food Hydrocoll. 82, 457–465. doi:10.1016/j.foodhyd.2018.04.027 Yang, X.-Y., Li, Z.-Q., Liu, B., Klein-Hofmann, A., Tian, G., Feng, Y.-F., et al. (2006). Fish-
in-Net encapsulation of enzymes in macroporous cages for stable, reusable, and active
heterogeneous biocatalysts. Adv. Mater. 18 (4), 410–414. doi:10.1002/adma.200502003 Sonnenburg, J. L., Angenent, L. T., and Gordon, J. I. (2004). Getting a grip on things:
how do communities of bacterial symbionts become established in our intestine? Nat. Immunol. 5 (6), 569–573. doi:10.1038/ni1079 Yucel Falco, C., Amadei, F., Dhayal, S. K., Cárdenas, M., Tanaka, M., and Risbo, J. (2019). Hybrid coating of alginate microbeads based on protein-biopolymer multilayers
for encapsulation of probiotics. Biotechnol. Prog. 35 (3), e2806. doi:10.1002/btpr.2806 Sumper, M., and Brunner, E. (2006). Learning from diatoms: nature’s tools for the production
of nanostructured silica. Adv. Funct. Mater. 16 (1), 17–26. doi:10.1002/adfm.200500616 Yucel Falco, C., Geng, X., Cárdenas, M., and Risbo, J. (2017a). Edible foam based on
Pickering effect of probiotic bacteria and milk proteins. Food Hydrocoll. 70, 211–218. doi:10.1016/j.foodhyd.2017.04.003 Sun, W., Vallooran, J. J., and Mezzenga, R. (2015). Enzyme kinetics in liquid crystalline
mesophases: size matters, but also topology. Langmuir 31 (15), 4558–4565. doi:10.1021/acs. langmuir.5b00579 Yucel Falco, C., Sotres, J., Rascón, A., Risbo, J., and Cárdenas, M. (2017b). Design of a
potentially prebiotic and responsive encapsulation material for probiotic bacteria based
on chitosan and sulfated β-glucan. J. Colloid Interface Sci. 487, 97–106. doi:10.1016/j. jcis.2016.10.019 Talaikis, M., Valldeperas, M., Matulaitienė, I., Borzova, J. L., Barauskas, J., Niaura, G.,
et al. (2019). On the molecular interactions in lipid bilayer–water assemblies of different
curvatures. J. Phys. Chem. B 123 (12), 2662–2672. doi:10.1021/acs.jpcb.8b11387 Tanaka, M., Schiefer, S., Gege, C., Schmidt, R. R., and Fuller, G. G. (2004). Influence of
subphase conditions on interfacial viscoelastic properties of synthetic lipids with
gentiobiose head groups. J. Phys. Chem. B 108 (10), 3211–3214. doi:10.1021/jp0367934 Zhai, J., Fong, C., Tran, N., and Drummond, C. J. (2019). Non-lamellar lyotropic
liquid crystalline lipid nanoparticles for the next generation of nanomedicine. References Self-assembly structure
formation during the digestion of human breast milk. Angew. Chem. Int. Ed. 54 (5),
1600–1603. doi:10.1002/anie.201408320 Tsai, G.-J., and Hwang, S.-P. (2004). In vitro and in vivo antibacterial activity of
shrimp chitosan against some intestinal bacteria. Fish. Sci. 70 (4), 675–681. doi:10.1111/
j.1444-2906.2004.00856.x Salentinig, S., Phan, S., Khan, J., Hawley, A., and Boyd, B. J. (2013). Formation of
highly organized nanostructures during the digestion of milk. ACS Nano 7 (12),
10904–10911. doi:10.1021/nn405123j Tyle, P. (1993). Effect of size, shape and hardness of particles in suspension on oral
texture and palatability. Acta Psychol. 84, 111–118. doi:10.1016/0001-6918(93)90077-5 Salentinig, S., Sagalowicz, L., Leser, M. E., Tedeschi, C., and Glatter, O. (2011). Transitions in the internal structure of lipid droplets during fat digestion. Soft Matter 7
(2), 650–661. doi:10.1039/C0SM00491J Uriz, M. J., Turon, X., Becerro, M. A., and Agell, G. (2003). Siliceous spicules and
skeleton frameworks in sponges: origin, diversity, ultrastructural patterns, and
biological functions. Microsc. Res. Tech. 62 (4), 279–299. doi:10.1002/jemt.10395 Schneck, E., Oliveira, R. G., Rehfeldt, F., Demé, B., Brandenburg, K., Seydel, U., et al. (2009a). Mechanical properties of interacting lipopolysaccharide membranes from
bacteria mutants studied by specular and off-specular neutron scattering. Phys. Rev. E 80 (4), 041929. doi:10.1103/PhysRevE.80.041929 Valldeperas, M., Dabkowska, A. P., Pálsson, G. K., Rogers, S., Mahmoudi, N.,
Carnerup, A., et al. (2019a). Interfacial properties of lipid sponge-like nanoparticles
and the role of stabilizer on particle structure and surface interactions. Soft Matter 15
(10), 2178–2189. doi:10.1039/C8SM02634C Schneck, E., Papp-Szabo, E., Quinn, B. E., Konovalov, O. V., Beveridge, T. J., Pink, D. A., et al. (2009b). Calcium ions induce collapse of charged O-side chains of
lipopolysaccharides from Pseudomonas aeruginosa. J. R. Soc. Interface 6 (5),
S671–S678. doi:10.1098/rsif.2009.0190.focus Valldeperas, M., Salis, A., Barauskas, J., Tiberg, F., Arnebrant, T., Razumas, V., et al. (2019b). Enzyme encapsulation in nanostructured self-assembled structures: toward
biofunctional supramolecular assemblies. Curr. Opin. Colloid & Interface Sci. 44,
130–142. doi:10.1016/j.cocis.2019.09.007 Schneck, E., Schubert, T., Konovalov, O. V., Quinn, B. E., Gutsmann, T.,
Brandenburg, K., et al. (2010). Quantitative determination of ion distributions in
bacterial lipopolysaccharide membranes by grazing-incidence X-ray fluorescence. Proc. Natl. Acad. Sci. 107 (20), 9147–9151. doi:10.1073/pnas.0913737107 Valldeperas, M., Talaikis, M., Dhayal, S. K., Velička, M., Barauskas, J., Niaura, G., et al. (2019c). Encapsulation of aspartic protease in nonlamellar lipid liquid crystalline
phases. Biophysical J. 117 (5), 829–843. doi:10.1016/j.bpj.2019.07.031 Schneider, M. F., Lim, K., Fuller, G. G., and Tanaka, M. (2002). References ACS
Nano 13 (6), 6178–6206. doi:10.1021/acsnano.8b07961 Frontiers in Soft Matter 20 frontiersin.org
|
https://openalex.org/W2991466654
|
https://europepmc.org/articles/pmc6925370?pdf=render
|
English
| null |
Tuberculosis of the Left Wrist Joint and Spine
|
Curēus
| 2,019
|
cc-by
| 3,516
|
Tuberculosis of the Left Wrist Joint and Parthiban Sivasamy , Mohd Yazid Bajuri , Abdul Wahab Ghani
1
2
3 1. Orthopaedics, KPJ Healthcare University College, Nilai, MYS 2. Orthopaedics and Traumatology,
Universiti Kebangsaan Malaysia, Kuala Lumpur, MYS 3. Orthopaedics, Ampang Hospital, Kuala Lumpur,
MYS Corresponding author: Mohd Yazid Bajuri, ezeds007@yahoo.com.my Categories: Orthopedics, Infectious Disease Categories: Orthopedics, Infectious Disease
Keywords: tuberculosis, wrist arthritis, wrist tuberculosis, tubercular arthritis, septic arthrit g
p
,
Keywords: tuberculosis, wrist arthritis, wrist tuberculosis, tubercular arthritis, septic arthritis © Copyright 2019
Sivasamy et al. This is an open
access article distributed under the
terms of the Creative Commons
Attribution License CC-BY 3.0., which
permits unrestricted use, distribution,
and reproduction in any medium,
provided the original author and
source are credited. DOI: 10.7759/cureus.6203 Tuberculosis of the Left Wrist Joint and Introduction Skeletal involvement occurs in 1% to 3% of overall patients with tuberculosis (TB) [1]. It is
estimated that 50% of skeletal TB is of the spine and extraspinal articulating joints [2]. The
wrist joint is among the rare sites. TB of the joints has a prolonged onset and is rarely
diagnosed before developing into severe arthritis [3]. Poncet’s disease or TB rheumatism is differentiated from TB arthritis, as the former disease is a
non-destructive variant of joint inflammation, which is observed during the acute phase of TB. TB arthritis commonly affects a single joint and is where the organism can be obtained [4]. TB
arthritis begins with synovitis, leads to periarticular demineralization, subsequently to
marginal erosions, and eventually to damage the joint [5]. Quicker progression from synovial
inflammation damage of the joint is seen in weight-bearing patients. In the presence of a
superinfection, i.e. Staphylococcus aureus, the acceleration of joint destruction occurs along
with a systemic inflammatory response [2]. Received 10/15/2019
Review began 10/17/2019
Review ended 11/11/2019
Published 11/19/2019 The delay in diagnosis is contributed by a non-directional misleading presentation of patients
such as malaise and constitutional symptoms [2]. These delays in mycobacterial infection
diagnosis establishment and delayed initiation of treatment could result in the destruction of
more portions of the bone, adjacent bones, or joints [5]. Therefore, tubercular arthritis is to be
understood clearly by its manifestation, diagnosis establishment, and treatment. Abstract The incidence of wrist tuberculosis is rare. Clinical features and radiographs are not conclusive
in the beginning, happen to delay the achievement of the diagnosis, and thus result in poor
treatment. We present a case report of wrist tuberculosis that had delayed diagnosis. Hence,
the initiation of antituberculous treatment was delayed, as the initial investigations were not
conclusive of tuberculous infection. The patient was treated surgically multiple times before
tuberculosis was diagnosed. Antituberculous chemotherapy was started for the patient for one
year until she became afebrile and infective markers returned to normal. However, the patient
developed left wrist stiffness due to arthrofibrosis and bony destruction of the wrist joint. Received 10/15/2019
Review began 10/17/2019
Review ended 11/11/2019
Published 11/19/2019
© Copyright 2019
Sivasamy et al. This is an open
access article distributed under the
terms of the Creative Commons
Attribution License CC-BY 3.0., which
permits unrestricted use, distribution,
and reproduction in any medium,
provided the original author and
source are credited. Case Presentation Case Presentation A 70-year-old woman, admitted on May 26, 2017, to the orthopedic department, presented with A 70-year-old woman, admitted on May 26, 2017, to the orthopedic department, presented with How to cite this article
Sivasamy P, Bajuri M, Ghani A (November 19, 2019) Tuberculosis of the Left Wrist Joint and Spine. Cureus
11(11): e6203. DOI 10.7759/cureus.6203 How to cite this article
Sivasamy P, Bajuri M, Ghani A (November 19, 2019) Tuberculosis of the Left Wrist Joint and Spine. Cureus
11(11): e6203. DOI 10.7759/cureus.6203 2019 Sivasamy et al. Cureus 11(11): e6203. DOI 10.7759/cureus.6203 2019 Sivasamy et al. Cureus 11(11): e6203. DOI 10.7759/cureus.6203 symptoms of left wrist pain and swelling, associated with fever in the past two days. The
swelling was significant enough to expand over the palmar and dorsal aspect of the hand. It was
soft on palpation, with minimal fluctuation, mild edema over the dorsum, tender, and warm. She had underlying diabetes mellitus and ischemic heart disease. There was no past or family
history of TB. She had no complaints of respiratory symptoms or hemoptysis. Blood investigations revealed a total white cell count (TWC) of 38,000/UL, erythrocyte
sedimentation rate (ESR) of 45 mm/h, and C-reactive protein (CRP) of 455 mg/L. The blood
culture and sensitivity revealed Streptococcus pyogenes. A left wrist X-ray was performed with
no abnormal findings (Figures 1A-1B). The chest radiograph was normal. FIGURE 1: X-ray film of the left wrist on admission; increased
soft tissue shadow with no obvious abnormality
A: antero-posterior view; B: lateral view
She was diagnosed with a left wrist abscess and a left wrist incision and drainage with flexor
retinaculum release was performed (Figure 2). FIGURE 1: X-ray film of the left wrist on admission; increased FIGURE 1: X-ray film of the left wrist on admission; increased
soft tissue shadow with no obvious abnormality A: antero-posterior view; B: lateral view She was diagnosed with a left wrist abscess and a left wrist incision and drainage with flexor
retinaculum release was performed (Figure 2). She was diagnosed with a left wrist abscess and a left wrist incision and drainage with flexor
retinaculum release was performed (Figure 2). 2019 Sivasamy et al. Cureus 11(11): e6203. DOI 10.7759/cureus.6203 2019 Sivasamy et al. Cureus 11(11): e6203. DOI 10.7759/cureus.6203 2 of 9 FIGURE 2: Swollen left hand picture taken after the first
drainage and debridement FIGURE 2: Swollen left hand picture taken after the first
drainage and debridement A dense yellow fluid was seen during the operation. Intraoperative specimens were sent to
microbiology. The pus culture and sensitivity were negative. Based on blood culture and
sensitivity, a Streptococcus pyogenes antibiotic therapy, intravenous (IV) meropenem 1 g tds
was administered over a course of four weeks, and discharged with another two weeks of oral
unasyn 375 mg bd. The white blood cell (WBC), C-reactive protein (CRP), and erythrocyte
sedimentation rate (ESR) were all normal readings by the end of six weeks of IV meropenem. However, the wrist was warm and she developed recurrent swelling in the course of six weeks
after discharge. She also complained of back pain. A thoracolumbar X-ray revealed erosion of
the T8 and T9 vertebral bodies (Figures 3A-3B). 3 of 9 2019 Sivasamy et al. Cureus 11(11): e6203. DOI 10.7759/cureus.6203 FIGURE 3: X-ray film of the spine at the two months follow-up. There is reduction of T8 and T9 vertebral body height, erosion
of the superior endplate of T9 noted as shown in Figure 2B. The T8/T9 intervertebral disc space is narrowed. Appearance
suggestive of discitis with paravertebral soft tissue collection
at T8/T9 level and erosion of the T8 and T9 vertebra. A : antero-posterior view; B : lateral view FIGURE 3: X-ray film of the spine at the two months follow-up. There is reduction of T8 and T9 vertebral body height, erosion
of the superior endplate of T9 noted as shown in Figure 2B. The T8/T9 intervertebral disc space is narrowed. Appearance
suggestive of discitis with paravertebral soft tissue collection
at T8/T9 level and erosion of the T8 and T9 vertebra. A : antero-posterior view; B : lateral view The repeated left wrist X-ray revealed erosive arthritis of the wrist and carpal joints (Figures
4A-4B). The repeated left wrist X-ray revealed erosive arthritis of the wrist and carpal joints (Figures
4A-4B). 4 of 9 FIGURE 4: X-ray film of the left wrist after one year of
antituberculous therapy. Based on Figure 4B, there is erosion
and a lytic lesion around the bony skeleton at the wrist. Mild
reduction of the bony density is seen at the carpals and
metacarpals. A : lateral view; B : antero-posterior view
Magnetic resonance imaging (MRI) thoracolumbar done was suggestive of spondylodiscitis of
T8 and T9, with suspicion of tuberculosis infection (Figures 5A-5H). 2019 Sivasamy et al. Cureus 11(11): e6203. DOI 10.7759/cureus.6203 FIGURE 4: X-ray film of the left wrist after one year of
antituberculous therapy. Based on Figure 4B, there is erosion
and a lytic lesion around the bony skeleton at the wrist. Mild
reduction of the bony density is seen at the carpals and
metacarpals. A : lateral view; B : antero-posterior view Magnetic resonance imaging (MRI) thoracolumbar done was suggestive of spondylodiscitis of
T8 and T9, with suspicion of tuberculosis infection (Figures 5A-5H). 5 of 9 2019 Sivasamy et al. Cureus 11(11): e6203. DOI 10.7759/cureus.6203 A-D : coronal view; E-H : axial view MRI : magnetic resonance imaging The involvement of the spine in diagnostic imaging led to the suspicion of tuberculous
infection. Left wrist debridement and arthrotomy were performed and samples sent for biopsy. A spine T8 and T9 transpedicular biopsy was performed. The blood markers were TWC 10,000/UL, ESR 70 mm/h, and CRP 46 mg/L. The Mantoux test,
serum polymerase chain reaction (PCR) tuberculosis (TB), and Quantiferon TB Gold test were
performed and the results are negative. The left wrist tissue acid-fast Bacilli (AFB) was revealed
as positive during the second-look debridement on October 5, 2017. She underwent antituberculous therapy for a duration of one year after being diagnosed with
left wrist and spine TB. Her left wrist swelling and back pain started to reduce as the
antituberculous treatment was initiated and then resolved completely by the end of treatment. However, she developed left wrist stiffness due to arthrofibrosis and bony destruction of the left
wrist joint. FIGURE 5: MRI films of the spine at the two months follow-up,
suggestive of spondylodiscitis of T8 and T9 with suspicion of
tuberculosis infection
A-D : sagittal view; E-H : axial view FIGURE 5: MRI films of the spine at the two months follow-up,
suggestive of spondylodiscitis of T8 and T9 with suspicion of
tuberculosis infection
A-D : sagittal view; E-H : axial view
MRI of the left wrist was suggestive of erosive arthritis with a gas-forming abscess along the
ulnar side. It showed fluid collection in the intraarticular and extraarticular spaces of the wrist
(Figures 6A-6H). FIGURE 6: MRI film of the left wrist at the two months follow-
up, suggestive of erosive arthritis with a gas-forming abscess
along the ulnar side
): e6203. DOI 10.7759/cureus.6203
6 of FIGURE 5: MRI films of the spine at the two months follow-up,
suggestive of spondylodiscitis of T8 and T9 with suspicion of
tuberculosis infection A-D : sagittal view; E-H : axial view MRI of the left wrist was suggestive of erosive arthritis with a gas-forming abscess along the
ulnar side. It showed fluid collection in the intraarticular and extraarticular spaces of the wrist
(Figures 6A-6H). MRI of the left wrist was suggestive of erosive arthritis with a gas-forming abscess along the
ulnar side. It showed fluid collection in the intraarticular and extraarticular spaces of the wrist
(Figures 6A-6H). (Figures 6A 6H). FIGURE 6: MRI film of the left wrist at the two months follow-
up, suggestive of erosive arthritis with a gas-forming abscess
along the ulnar side FIGURE 6: MRI film of the left wrist at the two months follow-
up, suggestive of erosive arthritis with a gas-forming abscess
along the ulnar side FIGURE 6: MRI film of the left wrist at the two months follow-
up, suggestive of erosive arthritis with a gas-forming abscess
along the ulnar side 2019 Sivasamy et al. Cureus 11(11): e6203. DOI 10.7759/cureus.6203 6 of 9 2019 Sivasamy et al. Cureus 11(11): e6203. DOI 10.7759/cureus.6203 swelling, with no involvement of the tendons along with constitutional symptoms. To confirm the diagnosis of TB, as per the gold standard, the histological pattern of the tissue
specimen or any body fluid is needed to detect the presence of acid-fast Bacillus (AFB) [2]. A
classic triad of radiological findings, known as the Phemister’s triad, which includes the
presence of juxta-articular osteoporosis, peripheral bony erosions, and gradual joint space
narrowing suggests the presence of tubercular arthritis. However, such a typical triad may not
be seen all the time. Hence, there are high chances of early stages of tubercular arthritis being
overlooked by clinicians due to its non-specific clinical signs [12]. In cases where a bone scan
was to be performed, it may show in the increased uptake, but this is not specific. To confirm the diagnosis of TB, as per the gold standard, the histological pattern of the tissue
specimen or any body fluid is needed to detect the presence of acid-fast Bacillus (AFB) [2]. A
classic triad of radiological findings, known as the Phemister’s triad, which includes the
presence of juxta-articular osteoporosis, peripheral bony erosions, and gradual joint space
narrowing suggests the presence of tubercular arthritis. However, such a typical triad may not
be seen all the time. Hence, there are high chances of early stages of tubercular arthritis being
overlooked by clinicians due to its non-specific clinical signs [12]. In cases where a bone scan
was to be performed, it may show in the increased uptake, but this is not specific. The other radiological examination is MRI, which may showcase the synovial inflammation,
fluid collection, bony erosions, and bony lesion. Thus, MRI is useful in revealing the severity of
the disease and could be repeated to follow the progress. However, the MRI is nonspecific. The
Mantoux test or tuberculin skin test (TST) is used as a supportive examination. Interferon-
gamma release assays (IGRA), such as Quantiferon-TB Gold, is specific to detect TB infections
but unable to differentiate between actively ongoing disease and latent infection of TB [2]. PCR tests are highly sensitive by detecting amplified TB DNA but unable to distinguish live
bacilli from dead bacilli. PCR turns up to 50% to 60% positive in culture-negative groups of
cases [2]. The PCR test is specific and faster for obtaining an analysis of synovial fluid, bone,
and soft tissue of joints [4]. Discussion TB has always remained a significant infection in developing countries. There are a number of
clinical and radiographic features of osseous TB, which mimics a wide range of pathologies,
hence, the involvement of extrapulmonary features can be particularly difficult to diagnose [6]. Apart from this, there are also numerous possible differential diagnoses linked, such as subacute
or chronic suppurative arthritis, rheumatoid arthritis, benign bone tumors, osteochondrosis,
and Kaposi sarcoma, which further makes it difficult to come to a conclusive diagnosis [7]. The
lesion also can be misdiagnosed as chronic osteomyelitis as presented by this case in which
bone destruction is present. The treatment for cases that involve bone destruction is the
eradication of infection by debridement and the usage of antibiotic cement spacer if there is a
defect [8]. In our case, it presented as a wrist abscess. Sometimes, if there is a concomitant
infection that came before the diagnosis was made, the use of an antibiotic-impregnated
medium is very effective, as it delivers high concentrations of antibiotics suffice to treat the
local infection site [9]. Articular TB is a chronic and worsening disease that often affects the load-bearing joints. Small
joint presentations are not common and diagnosis is usually delayed due to low suspicion. For
extrapulmonary involvement, osteoarticular TB is in the fourth position after vertebral,
urogenital, and ganglionic localization. Vertebral localization remains the most frequent one
[10]. Primary foci reactivation and secondary spreading via the bloodstream are the two
gateways to extrapulmonary TB [11]. The discovery of spine involvement two months later in
our patient questions which could be the primary foci. The most common type of presentation of hand TB is tenosynovitis [11]. Tubercular
tenosynovitis usually is of gradual onset in nature and progresses slowly. Generally, it is
presented with swelling accompanied by mild pain and limitation of movement in the affected
area. The swelling will progress through tendons mostly without any other systemic evidence of
TB. The most common findings are compound palmar or dumbbell ganglion of the ulnar bursa,
“sausage digit,” and carpal tunnel syndrome [11]. Wrist joint TB often presents as swelling without pain that tracks along tendons and is rarely
associated with systemic symptoms [1]. However, our case presented with painful progressive 2019 Sivasamy et al. Cureus 11(11): e6203. DOI 10.7759/cureus.6203 7 of 9 A synovial biopsy is the current gold standard as a diagnostic test for tubercular arthritis. Approximately, 80% of cases are detected via synovial biopsy, which gives accurate
characteristics of lymphocytes, caseating granulomas, and giant cells with caseation. The treatment is easy, from the three classic shutters of the treatment of osteoarticular TB;
only chemotherapy is essential [11]. Chemotherapy has been replaced in recent years with
progressively shorter treatment (six months) using more effective drugs [11]. The orthopedic
wrist splint maintained until the disappearance of clinical signs (three to four weeks) and
followed by rehabilitation [11]. The medical treatment of this type of localized TB mainly
consists of anti-bacillary chemotherapy for 12 months, associating isoniazid, rifampicin,
pyrazinamide, and ethambutol. Other than the biopsy, there is very little need for surgery. To date, the treatment options remain debatable. According to Tuli et al., surgical debridement
remains a controversial option for wrist joint TB and, in fact, the optimal duration of treatment
with antitubercular chemotherapy has been an issue of considerable debate. However, if
chemotherapy and debridement are opted for, chemotherapy must be provided prior to any
surgical debridement in order to prevent bony destruction and dissemination of disease [13]. A
prolonged course of nine or 12 months of multiple chemotherapies should be continued [13]. Atypical site, clinical presentation, and delayed biopsy in our case had delayed the diagnosis of
wrist TB. If the wrist TB diagnosis was made early, antituberculous chemotherapy would have
been initiated earlier and would have thus reduced the need for repeated surgical debridement
and improved the clinical outcome. Disclosures Human subjects: Consent was obtained by all participants in this study. Conflicts of interest:
In compliance with the ICMJE uniform disclosure form, all authors declare the following:
Payment/services info: All authors have declared that no financial support was received from
any organization for the submitted work. Financial relationships: All authors have declared
that they have no financial relationships at present or within the previous three years with any
organizations that might have an interest in the submitted work. Other relationships: All
authors have declared that there are no other relationships or activities that could appear to
have influenced the submitted work. Conclusions In conclusion, this case reiterates the point that TB can be a great mimic, which can cause a
possible deep infection if appropriate treatment is not warranted on time. Keeping this in mind,
a clinician should always approach any acute or chronic lesions at the wrist joint with a high
index of suspicion, as TB arthritis is uncommon and easily missed. However, in endemic areas, 2019 Sivasamy et al. Cureus 11(11): e6203. DOI 10.7759/cureus.6203 8 of 9 even if other organisms are grown (which was the main red herring in this case), spinal TB led
to the correct diagnoses. Additional Information 2019 Sivasamy et al. Cureus 11(11): e6203. DOI 10.7759/cureus.6203 References 1. Malaviya AN, Kotwal PP: Arthritis associated with tuberculosis . Best Pract Res Clin
Rheumatol 2003 17:319 343 10 1016/S1521 6942(02)00126 2 1. Malaviya AN, Kotwal PP: Arthritis associated with tuberculosis . Best Pract Res Clin
Rheumatol. 2003, 17:319-343. 10.1016/S1521-6942(02)00126-2 1. Malaviya AN, Kotwal PP: Arthritis associated with tuberculosis . B
Rheumatol. 2003, 17:319-343. 10.1016/S1521-6942(02)00126-2 2. Tseng CC: Tuberculosis arthritis: epidemiology, diagnosis, treatment. J Clin Res Foot Ankle. 2014, 2: 3. Evanchick CC, Davis DE, Harrington TM: Tuberculosis of peripheral joints: an often missed
diagnosis. J Rheumatol. 1986, 13:187-189. 4. Abdulaziz S, Almoallim H, Ibrahim A, et al.: Poncet's disease (reactive arthritis associated
with tuberculosis): retrospective case series and review of literature). Clin Rheumatol. 2012,
31:1521-1528. 10.1007/s10067-012-2042-0 5. Furia JP, Box GG, Lintner DM: Tuberculous arthritis of the knee presenting as a meniscal tear
Am J Orthop (Belle Mead NJ). 1996, 25:138-142. 6. Dhillon MS, Singh P, Sharma S, Gill SS, Nagi ON: Tuberculous osteomyelitis of the cuboid: a
report of four cases. J Foot Ankle Surg. 2000, 39:329-335. 10.1016/S1067-2516(00)80049-6 7. Bozkurt M, Doǧan M, Şeşen H, Turanlı S, Basbozkurt M: Isolated medial cuneiform
tuberculosis: a case report. J Foot Ankle Surg. 2005, 44:60-63. 10.1053/j.jfas.2004.11.011 7. Bozkurt M, Doǧan M, Şeşen H, Turanlı S, Basbozkurt M: Isolated medial cuneiform
tuberculosis: a case report. J Foot Ankle Surg. 2005, 44:60-63. 10.1053/j.jfas.2004.11.011 8. Bajuri MY, Razak KAA: Chronic osteomyelitis of the femur with segmental bone defect:
concepts and treatment. JKIMSU. 2017, 6:127-130. 9. Atan AA, Bajuri MY, Ali AM, et al.: The effects of gentamicin-impregnated collagen sponge
versus gentamicin-impregnated polymethylmethacrylate beads in patients with
osteomyelitis. Asian J Pharm Clin Res. 2018, 11:241-246. 10.22159/ajpcr.2018.v11i12.26910 10. Sbai MA, Benzarti S, Bouzaidi K, Sbei F, Maalla R: A rare localization of tuberculosis of the
wrist: the scapholunate joint. Int J Mycobacteriol. 2015, 4:161-164. 10.1016/j.ijmyco.2015.04.001 11. Wu KC, Tang HM, Yeh KT, Ding DC: Extrapulmonary tuberculosis in the wrist presenting as a
ganglion cyst-like mass: a case report. Tzu Chi Med J. 2012, 24:77-79. 10.1016/j.tcmj.2011.12.005 12. Yao DC, Sartoris DJ: Musculoskeletal tuberculosis. Radiol Clin North Am. 1995, 33:679-689. 13. Tuli SM: Tuberculosis of the Skeletal System. Jaypee Brothers Medical Publishers (P) Ltd.,
New Delhi, India; 2004. 9 of 9
|
https://openalex.org/W3215842193
|
https://jurnal.ugm.ac.id/jpt/article/download/66150/32779
|
Indonesian
| null |
Identifikasi Potensi Wisata di Desa Banjaran Wetan
|
Jurnal pariwisata terapan
| 2,021
|
cc-by-sa
| 6,583
|
Affiliation Affiliation
1Magister Pariwisata Berkelanjutan Sekolah Pascasarjana Universitas Padjadjaran
2Magister Pariwisata Berkelanjutan Sekolah Pascasarjana Universitas Padjadjaran
3Magister Pariwisata Berkelanjutan Sekolah Pascasarjana Universitas Padjadjaran Correspondence Kunkun Kurniawan, Magister Pariwisata Berkelanjutan Sekolah Pascasarjana Universitas Padjadjaran Jl. Dipati Ukur No 53 Kota Bandung. Email: kunkun20001@mail.unpad.co.id 8
Abstract
Banjaran Wetan Village is one of the villages in the Banjaran District, Bandung Regency which does n
yet have a tourism destination. Geographically, Banjaran Wetan Village has the largest area in Banjara
District. One of the natural potentials in the Banjaran Wetan Village area that can be developed into
sustainable tourism destination is Situ Cimeuhmal. Situ Cimeuhmal is located at the foot of the Malab
Mountains with a protected forest area of 46.5 hectares and an altitude of 1,200 meters above sea leve
Stakeholders involved in the management of the Cimeuhmal Situ area consist of PT. Perhutani, BBW
Citarum, PT Jasa Tirta, Community of Children Care for the Universe, and Situ Cimeuhmal Environment
Care Group. The development of Situ Cimeuhmal into a water conservation-based tourism destinatio
is constrained by coordination and role sharing among stakeholders. The purpose of this study is
identify tourism potential in Banjaran Wetan Village and the strategy for developing Situ Cimeuhm
tourism destinations that are in line with the principles of sustainable tourism. The research method w
carried out qualitatively with data collection techniques through observation, interviews, and literatu
studies. The results showed that Banjaran Wetan Village has natural and non-natural tourism potent
that can be developed into a tourism destination. Based on the results of the SWOT analysis, th
ecotourism development strategy of Situ Cimeuhmal is carried out through the use of the attraction
water conservation, local food products, crackers, and coffee as an educational program for tourists
gain new knowledge and experiences and can have an impact on nature conservation, economy, an
socio-culture to the local community. Kawin Cai cultural attractions, coffee festivals, and endorsemen
are important aspects in increasing the promotion of Situ Cimeuhmal as an attractive eco-touris
destination in Banjaran Wetan Village. Keywords: Ecotourism; Water Conservation; Situ Cimeuhmal; Kawin Cai
Article Information: Submitted 27 Maret 2021|Revised 22 July 2021|Accepted 26 July 2021
Copyright © 2021 by the author(s). This article is published by Universitas Gadja
Mada, Indonesia under the Creative Commons Attribution (CC BY 4.0) license. Anyon
may reproduce, distribute, translate, and create derivative works of this article (f
both commercial and noncommercial purposes), subject to full attribution to th
original publication and author(s). The full terms of this license may be seen
http://creativecommons.org/licenses/by/4.0/legalcod Jurnal Pariwisata Terapan Vol. 5., No. 2, 2021
ISSN-2580-1031 (print) ISSN- 2580-104X (online)
https://jurnal.ugm.ac.id/jpt Jurnal Pariwisata Terapan Vol. 5., No. 2, 2021
ISSN-2580-1031 (print) ISSN- 2580-104X (online)
https://jurnal.ugm.ac.id/jpt Pengembangan Ekowisata Situ Cimeuhmal Berbasis
Masyarakat di Desa Banjaran Wetan Kabupaten Bandung
Kunkun kurniawan1, Reiza D Dienaputra2, Cecep Ucu Rahman3 Kunkun kurniawan1, Reiza D Dienaputra2, Cecep Ucu Rahman3 Pendahuluan Desa Banjaran Wetan merupakan desa hasil pemekaran dari Desa Banjaran pada tanggal 25 Juli 1978. Secara wilayah Desa Banjaran Wetan memiliki wilayah 714,93 Ha dan menjadi yang terluas di Kecamatan
Banjaran. Desa Banjaran Wetan memiliki 80 RT, 20 RW dan 6 Dusun. Berdasarkan peringkat status Indeks
Desa Membangun (IDM), Desa Banjaran Wetan termasuk ke dalam kategori Desa Maju. Meskipun Desa
Banjaran Wetan termasuk kategori desa maju, tetapi dari aspek pariwisata masih dikelompokkan
sebagai desa wisata rintisan. Salah satu potensi wisata alam yang menarik untuk dikembangkan menjadi destinasi pariwisata di Desa
Banjaran Wetan adalah Situ Cimeuhmal. Situ Cimeuhmal merupakan danau penampungan air yang
bersumber dari mata air yang berada di bawah kaki pegunungan Malabar kawasan hutan Blok
Cimeuhmal petak 1.b. RPH Banjaran, BKPH Banjaran KPH Bandung Selatan dengan luas kawasan 46,5
hektar. Lokasi strategis Situ Cimeuhmal yang berada di 107037’3.84’BT 705’32.12’LS memiliki fungsi
sebagai daerah tangkapan air. Air yang mengalir dan ditampung di Situ Cimeuhmal berasal dari tiga mata
air, yaitu: Cikahuripan, Balong Mandor dan Kadal Mateng dan memiliki fungsi untuk memenuhi
kebutuhan air bagi penduduk di sekitar Situ Cimeuhmal dan pengairan pertanian yang sekaligus menjadi
hulu dari sungai kecil cikadu dan ujungnya bermuara di Sungai Cisangkuy. Cimeuhmal berasal dari nama
sebuah pohon, yaitu Ki Meuhmal yang dahulu tumbuh di sekitar kawasan Situ Cimeuhmal. Sebelumnya
warga di sekitar menyebut Situ Cimeuhmal dengan nama Situ Ermat. Hal ini berdasarkan penuturan
warga secara turun temurun bahwa di sekitar situ tersebut pernah didiami oleh seorang pegawai jaman
Belanda yang bernama Ermat. Asal usul sejarah yang dimiliki Situ Cimeuhmal memiliki keterkaitan
dengan latar belakang sejarah kehidupan masa lalu. Kawasan yang saat ini berada di wilayah Kecamatan
Banjaran memiliki kaitan dengan catatan sejarah dengan keberadaan Dipati Ukur. Keterkaitan latar
belakang sejarah sebuah tempat (toponimi) menjadi sebuah potensi yang menarik untuk pengembangan
kepariwisataan. Situ Cimeuhmal memiliki lanskap alam yang cukup menarik yakni terletak di ketinggian 1.200 mdpl
dengan latar belakang pegunungan Malabar yang terlihat sangat jelas dan menampakkan pesona
keindahan alami. Terdapat Kawasan pohon pinus yang tumbuh mengelilingi area ini tampak tertata rapi
sehingga pengunjung dapat merasakan suasana yang asri, tenang, dan nyaman. Pohon pinus di sekitar
Situ Cimeuhmal mengeluarkan aroma khas alam pegunungan yang menenangkan hati yang menjadi
salah satu potensi pengembangan destinasi pariwisata yang menampilkan pesona alam dan konservasi
sumber daya air. Kunkun kurniawan, Reiza D Dienaputra, Cecep Ucu Rahman Kunkun kurniawan, Reiza D Dienaputra, Cecep Ucu Rahman Abstract Banjaran Wetan Village is one of the villages in the Banjaran District, Bandung Regency which does not
yet have a tourism destination. Geographically, Banjaran Wetan Village has the largest area in Banjaran
District. One of the natural potentials in the Banjaran Wetan Village area that can be developed into a
sustainable tourism destination is Situ Cimeuhmal. Situ Cimeuhmal is located at the foot of the Malabar
Mountains with a protected forest area of 46.5 hectares and an altitude of 1,200 meters above sea level. Stakeholders involved in the management of the Cimeuhmal Situ area consist of PT. Perhutani, BBWS
Citarum, PT Jasa Tirta, Community of Children Care for the Universe, and Situ Cimeuhmal Environmental
Care Group. The development of Situ Cimeuhmal into a water conservation-based tourism destination
is constrained by coordination and role sharing among stakeholders. The purpose of this study is to
identify tourism potential in Banjaran Wetan Village and the strategy for developing Situ Cimeuhmal
tourism destinations that are in line with the principles of sustainable tourism. The research method was
carried out qualitatively with data collection techniques through observation, interviews, and literature
studies. The results showed that Banjaran Wetan Village has natural and non-natural tourism potential
that can be developed into a tourism destination. Based on the results of the SWOT analysis, the
ecotourism development strategy of Situ Cimeuhmal is carried out through the use of the attraction of
water conservation, local food products, crackers, and coffee as an educational program for tourists to
gain new knowledge and experiences and can have an impact on nature conservation, economy, and
socio-culture to the local community. Kawin Cai cultural attractions, coffee festivals, and endorsements
are important aspects in increasing the promotion of Situ Cimeuhmal as an attractive eco-tourism
destination in Banjaran Wetan Village. Keywords: Ecotourism; Water Conservation; Situ Cimeuhmal; Kawin Cai Article Information: Submitted 27 Maret 2021|Revised 22 July 2021|Accepted 26 July 2021 Copyright © 2021 by the author(s). This article is published by Universitas Gadjah
Mada, Indonesia under the Creative Commons Attribution (CC BY 4.0) license. Anyone
may reproduce, distribute, translate, and create derivative works of this article (for
both commercial and noncommercial purposes), subject to full attribution to the
original publication and author(s). The full terms of this license may be seen at
http://creativecommons.org/licenses/by/4.0/legalcod 84 Pendahuluan Hasil penelusuran di sekitar kawasan Situ terdapat terowongan air yang dibangun pada
jaman Belanda untuk mengalirkan air dari Situ Cimeuhmal menuju sungai Cikadu, sungai Citalugtug yang
bermuara ke Sungai Citanduy di sekitar wilayah Kota Kecamatan Banjaran menjadi objek sejarah
infrastruktur pengelolaan dan pemeliharaan sumber mata air yang melengkapi kawasan ekowisata Situ
Cimeuhmal. Pengembangan wisata yang bertujuan untuk pemeliharaan dan pengelolaan sumber air yang merupakan
bagian dari konsep ekowisata. Konservasi air bersih dari mulai hulu sampai hilir menjadi bagian penting
dalam menjaga kelestarian lingkungan. Hal ini menjadi sebuah ekosistem yang masing-masing subsistem
atau unit memiliki peran dan fungsi saling berkaitan. Pengembangan ekowisata di sebuah destinasi atau
kawasan bertujuan untuk menciptakan dampak positif terkait pelestarian faktor alam, mempertahankan
unsur lokal sosial budaya, dan meningkatkan aspek ekonomis bagi masyarakat sekitar Kawasan tersebut. 85 Jurnal Pariwisata Terapan, Vol. 5., No. 2, 2021 Kunkun kurniawan, Reiza D Dienaputra, Cecep Ucu Rahman Hal ini sejalan dengan hasil penelitian Hermawan (2016) di Desa Nglanggeran yang menyebutkan bahwa
pengembangan ekowisata memberikan dampak sosial dan budaya secara positif. Berdasarkan tinjauan
dari aspek pemangku kepentingan yang terkait dengan kawasan Situ Cimeuhmal, terdapat beberapa
pihak yang menjadi bagian dari pemangku kepentingan. Pada segi kepemilikan lahan, pemangku
kepentingan yang berwenang adalah PT Perhutani sementara lokasi Situ Cimeuhmal berada di Kawasan
Hutan Lindung di bawah pengelolaan Perhutani BKPH Banjaran dan bekerja sama dengan Yayasan
Komunitas Anak Peduli Alam Semesta (KAPAS) tentang perlindungan daerah tangkapan air dan
pengelolaan berdasarkan kerja sama Nomor: 01/SPK/PSDH/BDS/III/2014. Pemangku kepentingan lain
yang terlibat dari aspek operasional dan pemeliharaan yaitu: Balai Besar Wilayah Sungai (BBWS), PT
Perusahaan Jasa Tirta II, dan Kelompok Peduli Lingkungan Situ Cimeuhmal (KPLSC). Pemangku
kepentingan secara adminitasi wilayah yang terlibat adalah Pemerintahan Desa Banjaran Wetan. Kegiatan konservasi sumber daya air Situ Cimeuhmal memiliki potensi untuk dikembangkan menjadi
destinasi pariwisata berbasis konservasi lingkungan alam yang berkelanjutan. Model pariwisata yang
dapat diimplementasikan terkait konservasi lingkungan alam adalah ekowisata (Fennell, 2008). Kegiatan
ekowisata menjadi alat dalam pengembangan bisnis dan ekonomi serta menjaga pengembangan
destinasi pariwisata yang berkelanjutan (Palmer & Chuamuangphan, 2018). Pengembangan ekowisata
bertujuan untuk memberikan dampak positif terhadap pelestarian alam, pemeliharaan budaya lokal, dan
peningkatan kesejahteraan masyarakat di sekitar kawasan ekowisata. Salah satu faktor utama
keberhasilan implementasi pembangunan pariwisata berkelanjutan adalah dukungan dari seluruh
pemangku kepentingan (McComb et al., 2017). Jurnal Pariwisata Terapan, Vol. 5., No. 2, 2021 Metode Model penelitian ini menggunakan metode kualitatif. Teknik pengumpulan data dilakukan melalui
observasi dan wawancara mendalam untuk mendapatkan data primer. Data primer adalah data yang
dikumpulkan secara langsung oleh peneliti dari sumber pertamanya (Suryabrata, 1987). Data primer
yang diperoleh melalui wawancara mendalam, dilakukan kepada Kepala Desa dan Staf Pemerintahan
Desa Banjaran Wetan, Ketua dan anggota Kelompok Peduli Lingkungan Situ Cimeuhmal (KPLSC) serta
Ketua dan anggota Komunitas Anak Peduli Alam Semesta (KAPAS). Wawancara mendalam ini untuk
menggali keterangan, pemikiran dan pengalaman-pengalaman masyarakat terkait keikutsertaannya
dalam pengembangan ekowisata Situ Cimeuhmal. Kombinasi hasil pengamatan langsung dan
wawancara mendalam dalam penelitian kualitatif akan menghasilkan data komprehensif dan bermakna. Data sekunder yang berupa data dokumentasi dan arsip-arsip resmi diperoleh dari sumber yang tidak
langsung (Azhar, 2003). Data sekunder yang diperlukan dalam penelitian ini, diperoleh dari arsip data
Pemerintahan Desa Banjaran Wetan, data PT Perhutani, koran, jurnal, serta beberapa dokumen yang
ada pada penelitian sebelumnya melalui studi literatur. Data primer dan sekunder yang sudah
dikumpulkan, diolah, dan dianalisis secara deskriptif dengan menggunakan model analisis SWOT untuk
mengkaji faktor internal dan faktor eksternal. Kunkun kurniawan, Reiza D Dienaputra, Cecep Ucu Rahman Kunkun kurniawan, Reiza D Dienaputra, Cecep Ucu Rahman ekowisata berbasis masyarakat (Community Based Tourism-CBT) memiliki peranan yang sangat penting
dan mendorong masyarakat memiliki pembagian peran (role sharing) yang lebih besar dan terlibat aktif
dalam kegiatan pariwisata yang berkelanjutan. ekowisata berbasis masyarakat (Community Based Tourism-CBT) memiliki peranan yang sangat penting
dan mendorong masyarakat memiliki pembagian peran (role sharing) yang lebih besar dan terlibat aktif
dalam kegiatan pariwisata yang berkelanjutan. Situ Cimeuhmal sebagai rintisan destinasi pariwisata di Desa Banjaran Wetan, diharapkan dapat
mengerakkan potensi wisata lainnya di Desa Banjaran Wetan. Momentum ini sekaligus menjadi media
untuk membangun kesadaran dan partisipasi masyarakat Desa Banjaran Wetan dalam pengembangan
pariwisata yang berkelanjutan. Potensi yang ada di sekitar masyarakat dapat dikembangkan menjadi
potensi wisata yang menarik. Tersebarnya industri rumah tangga yang berbasis makanan dan minuman
di wilayah Desa Banjaran Wetan dapat dikolaborasikan dengan potensi alam Situ Cimeuhmal untuk
menjadi destinasi pariwisata unggulan. Penelitian ini dilakukan untuk mengetahui potensi apa saja yang
dapat dikembangkan menjadi destinasi pariwisata di Desa Banjaran Wetan, serta bagaimana strategi
pengembangan Situ Cimeuhmal sebagai destinasi ekowisata berbasis sumber daya air yang sejalan
dengan konsep pariwisata berkelanjutan. Pendahuluan Berdasarkan penelitian mengenai ekowisata di Desa
Nglanggeran dikemukakan bahwa faktor-faktor penghambat pengembangan ekowisata adalah SDM
pengelola pariwisata, akses dan fasilitas parkir untuk pengunjung, atraksi, promosi dan produk yang
dihasilkan dari destinasi pariwisata (Jamalina & Wardani, 2017). Beberapa faktor di atas terkait dengan
peran pemerintah, investor, dan masyarakat di sekitar kawasan desa wisata. Hal tersebut menunjukkan
perlunya inisiasi dan pembagian peran (role sharing) yang jelas dalam pengembangan sebuah destinasi
ekowisata. Pembagian peran (role sharing) menjadi permasalahan yang seringkali terjadi dalam
pengembangan sebuah ekowisata. Selain pembagian peran, aspek lain yang berperan penting dalam
pengembangan destinasi wisata adalah mengenai komunikasi antar pemangku kepentingan (Sella &
Yusuf, 2020). Ekowisata berdasarkan tipologi dapat dibedakan menjadi ekowisata berbasis investor, pemerintah, dan
masyarakat (Arida, 2021). Identifikasi ini dilihat dari Inisiasi dan pembagian peran (role sharing) menurut
investasi dan model pengelolaan ekowisata. Setiap tipe ekowisata akan memberikan dampak yang
berbeda dalam pencapaian target dan tujuan. Menurut Arida (2021) permasalahan saat ini adalah model
ekowisata yang melibatkan investor khususnya asing kurang melibatkan dan memberikan dampak
kepada masyarakat sekitarnya. Kondisi ini yang kemudian memunculkan istilah ekowisata semu (pseudo
ecotourism). Rendahnya tingkat pelibatan masyarakat dalam pengembangan ekowisata akan mereduksi
hakikat dan tujuan dari pariwisata berkelanjutan. “Pengembangan pariwisata berbasis masyarakat
(community-based tourism-CBT) merupakan model pembangunan pariwisata yang memberikan peluang
luas kepada masyarakat pedesaan atau lokal untuk berpartisipasi. Community Based Tourism-CBT
merupakan sebuah kegiatan pembangunan pariwisata yang dilakukan sepenuhnya oleh masyarakat. Ide
kegiatan dan pengelolaan dilakukan seluruhnya oleh masyarakat secara partisipatif, dan manfaatnya
dirasakan langsung oleh masyarakat lokal (Sidiq & Resnawaty, 2017)”. Jadi, model pengembangan 86 Jurnal Pariwisata Terapan, Vol. 5., No. 2, 2021 Kunkun kurniawan, Reiza D Dienaputra, Cecep Ucu Rahman Banjaran Wetan. Produk kuliner khas berbasis makanan ringan dari Desa Banjaran Wetan ini dapat
dijadikan potensi wisata edukasi proses pembuatan kerupuk serta oleh-oleh bagi rintisan destinasi
pariwisata di Desa Banjaran Wetan. Banjaran Wetan. Produk kuliner khas berbasis makanan ringan dari Desa Banjaran Wetan ini dapat
dijadikan potensi wisata edukasi proses pembuatan kerupuk serta oleh-oleh bagi rintisan destinasi
pariwisata di Desa Banjaran Wetan. Beberapa lahan pertanian di Desa Banjaran Wetan yang berdekatan dengan kawasan hutan pinus mulai
banyak ditanami pohon kopi dalam 10 tahun terakhir. Banyak petani yang tertarik menanam pohon kopi
karena biji kopi pada saat itu mulai dicari para pemasok kopi. Hasil panen kopi di kawasan ini menjadi
incaran pemasok biji kopi untuk pasar lokal dan ekspor. Hal ini menjadi faktor yang pendorong banyak
petani beralih ke komoditas tanaman kopi yang sebelumnya hanya menanam padi dan singkong yang
menjadi komoditas utamanya. Dampak dari kegiatan petani menanam kopi memberikan peluang baru
bagi tumbuhnya warung kopi yang ada di sekitar Situ Cimeuhmal. Kawasan di sekitar warung-warung
kopi ini menjadi daya tarik bagi pegiat dan pehobi sepeda. Trek yang menanjak menuju Situ Cimeuhmal
menjadi tujuan pengguna sepeda untuk menguji kemampuan fisik serta menjadi tempat istirahat yang
nyaman dengan latar belakang pegunungan Malabar dan pemandangan ke arah kota Kecamatan
Banjaran. Saat ini terdapat 2 merek produk kopi dari kawasan Situ Cimeuhmal, yaitu Kopi Cimeuhmal
dan Kopi Kiara Payung. Potensi wisata alam yang diharapkan menjadi rintisan destinasi pariwisata unggulan di Desa Banjaran
Wetan adalah Situ Cimeuhmal. Pada awalnya Situ Cimeuhmal memiliki peran penting dalam menjaga
pasokan kebutuhan air bersih untuk masyarakat dan sumber pengairan bagi aktifitas pertanian di
sekitarnya sampai dengan tahun 1975. Keadaan Situ Cimeuhmal pada saat itu masih tergenang oleh air
yang jernih dan cukup dalam. Seiring dengan kerusakan hutan (deforestasi) akibat perambahan hutan,
fungsi Situ Cimeuhmal menjadi berkurang akibat sedimentasi yang cukup tinggi. Pada tahun 2003 kondisi
Situ Cimeuhmal terlihat seperti sebuah rawa yang di sekelilingnya hanya ditumbuhi oleh tanaman perdu
dan rerumputan tanpa pohon tegakkan. Potensi sumber air Situ Cimeuhmal mulai terbuka kembali ketika
muncul persoalan krisis air yang menganggu fungsi pertanian di wilayah Desa Banjaran Wetan khususnya
di wilayah RW 12 Kampung Cikamadong kemudian RW 13 dan RW 16 Kampung Kiarapayung. Berdasarkan petunjuk dari para tokoh setempat bahwa di wilayah hutan pernah ada sumber air yang
cukup besar maka upaya penanganan krisis air tersebut mulai dilakukan. Hasil dan Pembahasan Identifikasi Potensi Wisata Di Desa Banjaran Wetan Berdasarkan data demografi, Jumlah penduduk Desa Banjaran Wetan sebanyak 16.961 orang, yaitu Laki-
laki 8.678 orang (51%) dan Perempuan 8.283 orang (49%). Terdiri dari 5.088 KK. Mata pencarian
penduduknya sebanyak 28.78% belum bekerja, 22,36% mengurus rumah tangga, 13,75% pelajar
mahasiswa 13,75%, buruh harian lepas 11,93%, karyawan swasta 8,53% dan sisanya berbagai jenis
pekerjaan lain. Dari aspek kegiatan ekonomi, Desa Banjaran Wetan memiliki 25 industri makanan dan
minuman. Beberapa produk yang sangat populer dari hasil produksi rumah tangga yang berbasis
makanan adalah kerupuk gurileum. Istilah gurileum berasal dari kata gurih dan peleum. Distribusi dari
produk ini tidak saja untuk pasar lokal banjaran tetapi juga pasar regional di Jawa Barat. Selain kerupuk
gurilem, ada produk kerupuk kulit (dorokdok) yang dihasilkan dari produksi rumah tangga dari Desa 87 Jurnal Pariwisata Terapan, Vol. 5., No. 2, 2021 Kunkun kurniawan, Reiza D Dienaputra, Cecep Ucu Rahman Pelaksanaan kegiatan kelompok ini dilakukan dengan berpijak
kepada 3 (tiga) tujuan utama kelompok, yaitu: 1. Menjaga Kualitas dan kuantitas air situ 2. Memanfaatkan air genangan untuk kebutuhan konsumsi dan irigasi sec 3. Mengelola dan menata kawasan situ sebagai area wana wisata edukasi . Mengelola dan menata kawasan situ sebagai area wana wisata edukasi Partisipasi masyarakat oleh kelompok Peduli Lingkungan Situ Cimeuhmal (KPLSC) dalam kegiatan
konservasi sumber daya air Situ Cimeuhmal juga melibatkan sebuah komunitas yang bernama Komunitas
Anak Peduli Alam Semesta (KAPAS). Berdiri pada tahun 2005 sesuai dengan bentuk badan hukum
yayasan dengan nomor 17 tanggal 28 Oktober dan perubahan nomor 16 tanggal 10 September 2018. Isu
strategis yang dijalankan adalah terkait sumber daya air, konservasi lahan, membangun jaring kemitraan
dalam pengelolaan lingkungan hidup, dan pemberdayaan masyarakat mengenai lingkungan hidup. Komunitas Anak Peduli Alam Semesta (KAPAS) memiliki kegiatan dan kontribusi terhadap pengelolaan
kawasan konservasi sumber daya air Situ Cimeuhmal adalah sebagai berikut: kawasan konservasi sumber daya air Situ Cimeuhmal adalah sebagai berikut: 1. Pengadaan sarana pendukung untuk kegiatan pemeliharaan dan penataan kawasan lingkungan Situ
dan sekitarnya (pembuatan Saung, Gazebo, kerja bakti pemeliharaan jalan, pembuatan gorong-
gorong di jalan inspeksi Perhutani, pembuatan sentra pembibitan kelompok, pembuatan rambu-
rambu, pembelian BBM untuk mesin pemotong rumput. . Monitoring dan evaluasi kegiatan (penyelenggaraan pertemuan rutin kelompok 3. Meningkatkan motivasi kelompok dengan stimulus pembiayaan kegiatan swadaya (pengadaan
konsumsi, pembelian bahan awal, dll) 4. Dukungan modal awal untuk kegiatan ekonomi kelompok (warung kelompok, pupuk dan bibit kopi,
pembuatan kolam budidaya ikan, bibit tanaman palawija, dll) 5. Mengupayakan peningkatan kapasitas kelompok melalui keikutsertaan para anggota kelompok pada
pelatihan-pelatihan, loka karya, sosialisasi yang diselenggarakan oleh dinas atau lembaga
pemerintah terkait. 5. Mengupayakan peningkatan kapasitas kelompok melalui keikutsertaan para anggota kelompok pada
pelatihan-pelatihan, loka karya, sosialisasi yang diselenggarakan oleh dinas atau lembaga
pemerintah terkait. Dampak kegiatan yang dilakukan ini memberikan pengaruh terhadap peningkatkan kualitas lingkungan
dan berujung kepada peningkatan kesejahteraan masyarakat yang berpijak kepada 3 (tiga) pilar
keberlanjutan, yaitu: ramah lingkungan (ecological objective), menguntungkan (economical objective),
dan dapat diterapkan oleh kelompok (technical objective). Strategi Pengembangan Situ Cimeuhmal sebagai Destinasi Ekowisata Berbasis Sumber Daya Air
Berkaitan dengan pengembangan ekowisata Situ Cimeuhmal berbasis masyarakat (Community Based
Tourism -CBT), terdapat faktor-faktor yang menjadi pendorong dan penghambat dalam pengembangan
Situ Cimeuhmal sebagai destinasi ekowisata berbasis konservasi air. Kunkun kurniawan, Reiza D Dienaputra, Cecep Ucu Rahman Kondisi yang ada saat itu adalah
Situ Cimeuhmal yang telah berubah dari fungsi aslinya. Hal ini menarik kepedulian masyarakat sekitar
kawasan untuk memperbaikinya. Keterlibatan masyarakat lokal dalam menjaga dan memelihara Situ
Cimeuhmal dinilai sebagai bentuk partisipasi masyarakat dalam perspektif pembangunan berkelanjutan. Secara hakiki hal ini didasari oleh keyakinan bahwa masyarakat itu sendiri yang memahami apa yang
menjadi kebutuhannya. Kegiatan konservasi kawasan Situ Cimeuhmal pada awal mulanya diprakarsai
oleh Bapak Emid Tarmidi (Alm) salah seorang tokoh masyarakat Kampung Kiarapayung yang memiliki
kepedulian, dedikasi, dan motivasi yang tinggi untuk melakukan perbaikan lingkungan di kawasan hutan. Berkat usaha keras yang dilakukan semua pihak, Situ Cimeuhmal dapat kembali kepada fungsi asalnya
sebagai wilayah tangkapan dan sumber daya air. Selain keberadaan Situ Cimeuhmal sebagai pemasok kebutuhan air bersih bagi kebutuhan masyarakat
dan pertanian di sekitarnya, wilayah tangkapan air ini juga memiliki potensi yang dapat dikembangkan
menjadi destinasi ekowisata di wilayah Desa Banjaran Wetan. Berdasarkan tinjauan potensi wisata, Desa
Banjaran Wetan belum memiliki destinasi pariwisata yang dapat diunggulkan sehingga Situ Cimeuhmal 88 Jurnal Pariwisata Terapan, Vol. 5., No. 2, 2021 Kunkun kurniawan, Reiza D Dienaputra, Cecep Ucu Rahman dapat dijadikan sebagai rintisan destinasi pariwisata di wilayah Desa Banjaran Wetan yang berbasis
konservasi sumber daya air. Gambar 1. Peta Penananaman Pohon Situ Cimeuhma
Sumber: KAPAS, 2018 Gambar 1. Peta Penananaman Pohon Situ Cimeuhmal
Sumber: KAPAS, 2018 Gambar 1. Peta Penananaman Pohon Situ Cimeuhmal Sumber: KAPAS, 2018 Sumber: KAPAS, 2018 Kelompok Peduli Lingkungan Situ Cimeuhmal (KPLSC) dan Komunitas Anak Peduli Alam Semesta (KAPAS)
memiliki peran penting dalam mengembalikan fungsi Situ Cimeuhmal yang sempat beberapa waktu lama
kondisinya tidak terawat. Diskusi dan aksi nyata untuk mengembalikan fungsi Situ Cimeuhmal dilakukan
sejak tahun 2004. Kegiatan untuk melakukan penataan dan pemeliharaan lingkungan kawasan Situ
Cimeuhmal dilakukan secara swadaya berdasarkan SK Desa Banjaran Wetan Nomor 520/KEP-
01/01/2011 tanggal 3 Januari 2011. Beberapa kegiatan dan kontribusi yang telah dilakukan oleh
Kelompok Peduli Lingkungan (KPLSC) adalah sebagai berikut: 1. Menyiapkan tenaga kerja secara swadaya untuk kegiatan pemeliharaan dan penataan kawasan
lingkungan Situ Cimeuhmal dan sekitarnya setiap 2 kali dalam sebulan 2. Menyiapkan dan melakukan kordinasi terkait tenaga kerja untuk kegiatan fisik yang dilakukan oleh
pihak BBWS Kelompok Peduli Lingkungan Situ Cimeuhmal (KPLSC) melakukan pemberdayaan ekonomi kepada
masyarakat sekitar melalui kegiatan budidaya tanaman hias, sentra pembibitan tanaman, usaha 89 Jurnal Pariwisata Terapan, Vol. 5., No. 2, 2021 Kunkun kurniawan, Reiza D Dienaputra, Cecep Ucu Rahman budidaya ternak dan ikan terpadu dll. Kunkun kurniawan, Reiza D Dienaputra, Cecep Ucu Rahman Berdasarkan analisis SWOT, faktor
internal dan eksternal yang berpengaruh terhadap pengembangan Situ Cimeuhmal sebagai destinasi
ekowisata berbasis konservasi air adalah sebagai berikut:
F kt
I t
l Strength Strength g
1. Kawasan lingkungan dan situ yang masih asri dan nyaman untuk dikunjungi g
1. Kawasan lingkungan dan situ yang masih asri dan nyaman untuk dikunjungi 2. Terdapat sumber mata air alami sebagai salah satu hulu sungai cisangkuy 3. Situs Sejarah berupa terowongan air yang dibangun pada jaman Belanda
Weaknesses 90 Jurnal Pariwisata Terapan, Vol. 5., No. 2, 2021 Kunkun kurniawan, Reiza D Dienaputra, Cecep Ucu Rahman Kunkun kurniawan, Reiza D Dienaputra, Cecep Ucu Rahman 1. Infrastruktur jalan yang belum ditata dengan baik 2. Informasi dan tata kelola destinasi belum lengkap dan jelas 3. Sinyal Komunikasi yang kurang baik di sekitar kawasan Situ Cimeuhmal
Faktor Eksternal: Opportunities Opportunities 1. Satu satunya daya tarik wisata alam berbentuk situ di sekitar Gunung Malabar yang sangat cocok
untuk dikembangkan sebagai destinasi ekowisata berbasis konservasi air 2. Pemberdayaan masyarakat sekitar melalui pengembangan usaha hulu kopi rakyat dan usaha hilir
kedai kopi untuk memenuhi kebutuhan wisatawan 3. Daya tarik wisata edukasi produksi makanan ringan di wilayah Desa Banjara Threat: 1. Perilaku wisatawan yang tidak ramah lingkungan akan menganggu kelestarian lingkungan
destinasi 2. Konflik pemangku kepentingan dalam pengelolaan destinasi ekowisata Situ Cimeuhmal terkait
profit sharing 3. Pembukaan lahan hutan untuk kepentingan pertanian dan perkebunan Berdasarkan identifikasi faktor yang melibatkan aspek SWOT di atas menunjukkan bahwa strategi yang
dilakukan untuk pengembangan ekowisata Situ Cimeuhmal adalah Strenght – Opportunity:
Optimalisasi ekowisata Situ Cimeuhmal menjadi media edukasi bagi wisatawan terkait konservasi
mata air alami sebagai hulu sungai cisangkuy serta kebutuhan air masyarakat sekitar. Selain itu,
ekowisata ini mendatangkan manfaat bagi pendudul lokal secara ekonomi dan penghasilan baru
terkait kedatangan wisatawan dengan penyediaan produk makanan ringan, kopi lokal dan kedai
kopi. Strenght Opportunity:
Optimalisasi ekowisata Situ Cimeuhmal menjadi media edukasi bagi wisatawan terkait konservasi
mata air alami sebagai hulu sungai cisangkuy serta kebutuhan air masyarakat sekitar. Selain itu,
ekowisata ini mendatangkan manfaat bagi pendudul lokal secara ekonomi dan penghasilan baru
terkait kedatangan wisatawan dengan penyediaan produk makanan ringan, kopi lokal dan kedai
kopi. Weaknesses – Opportunity: Weaknesses – Opportunity: Penataan infrastruktur menuju lokasi dapat dilakukan melalui kerjasama, kordinasi dan
kolaborasi pemerintah setempat dengan masyarakat sekitar. Peningkatan sarana jalan dan
petunjuk yang jelas dan informatif akan memudahkan wisatawan untuk mengunjungi destinasi
ekowisata ini. Penataan infrastruktur menuju lokasi dapat dilakukan melalui kerjasama, kordinasi dan
kolaborasi pemerintah setempat dengan masyarakat sekitar. Peningkatan sarana jalan dan
petunjuk yang jelas dan informatif akan memudahkan wisatawan untuk mengunjungi destinasi
ekowisata ini. Strenght – Threat: Selain pemanfaatan Strenght – Opportunity diperlukan juga strategi peningkatan kualitas aksesibilitas,
amenitas, dan layanan dukungan pariwisata lainnya di kawasan ekowisata Situ Cimeuhmal, yaitu: 1. Aksesibilitas
Peningkatan kualitas aksesibilitas menuju destinasi wisata menjadi prioritas dalam memberikan
kemudahan dan kenyamanan wisatawan. Jalan menuju kawasan hutan pinus perlu diperbaiki dengan
menggunakan konsep dan kualitas jalan yang tidak merusak lingkungan. Fasilitas parkir dibuat di luar
kawasan pohon pinus sekaligus sebagai pintu gerbang destinasi ekowisata Situ Cimeuhmal. Selain
aksesibilitas dari aspek transportasi, diperlukan juga penguatan aksesibilitas dari aspek komunikasi. Saat ini jangkauan komunikasi untuk beberapa provider belum tersedia dengan baik. Akses internet
masih sulit dijangkau baik oleh masyarakat sekitar kawasan ekowisata Situ Cimeuhmal dan
wisatawan yang berkunjung sehingga momen real time berwisata yang sangat penting bagi
wisatawan menjadi kendala. Karakter wisatawan saaat ini membutuhkan kondisi real time ketika
melakukan kegiatan wisata. Momen Post Trip di lokasi wisata menjadi ritual wajib yang dilakukan
oleh wisatawan. 1. Aksesibilitas
Peningkatan kualitas aksesibilitas menuju destinasi wisata menjadi prioritas dalam memberikan
kemudahan dan kenyamanan wisatawan. Jalan menuju kawasan hutan pinus perlu diperbaiki dengan
menggunakan konsep dan kualitas jalan yang tidak merusak lingkungan. Fasilitas parkir dibuat di luar
kawasan pohon pinus sekaligus sebagai pintu gerbang destinasi ekowisata Situ Cimeuhmal. Selain
aksesibilitas dari aspek transportasi, diperlukan juga penguatan aksesibilitas dari aspek komunikasi. Saat ini jangkauan komunikasi untuk beberapa provider belum tersedia dengan baik. Akses internet
masih sulit dijangkau baik oleh masyarakat sekitar kawasan ekowisata Situ Cimeuhmal dan
wisatawan yang berkunjung sehingga momen real time berwisata yang sangat penting bagi
wisatawan menjadi kendala. Karakter wisatawan saaat ini membutuhkan kondisi real time ketika
melakukan kegiatan wisata. Momen Post Trip di lokasi wisata menjadi ritual wajib yang dilakukan
oleh wisatawan. 1. Aksesibilitas
Peningkatan kualitas aksesibilitas menuju destinasi wisata menjadi prioritas dalam memberikan
kemudahan dan kenyamanan wisatawan. Jalan menuju kawasan hutan pinus perlu diperbaiki dengan
menggunakan konsep dan kualitas jalan yang tidak merusak lingkungan. Fasilitas parkir dibuat di luar
kawasan pohon pinus sekaligus sebagai pintu gerbang destinasi ekowisata Situ Cimeuhmal. Selain
aksesibilitas dari aspek transportasi, diperlukan juga penguatan aksesibilitas dari aspek komunikasi. Saat ini jangkauan komunikasi untuk beberapa provider belum tersedia dengan baik. Akses internet
masih sulit dijangkau baik oleh masyarakat sekitar kawasan ekowisata Situ Cimeuhmal dan 2. Amenitas
Destinasi Ekowisata Situ Cimeuhmal belum memiliki fasilitas akomodasi untuk melayani kebutuhan
wisatawan yang berkunjung. Konsep Homestay cocok untuk diterapkan dalam pengembangan
destinasi ekowisata Situ Cimeuhmal yang berkelanjutan. Strenght – Threat: g
Untuk menghindari kerusakan destinasi ekowisata Situ Cimeuhmal perlu dilakukan kordinasi dan
perumusan pedoman tata kelola destinasi dengan mempertimbangkan perlindungan atas
keberlangsungan konservasi sumber mata air alami ini. Untuk menghindari kerusakan destinasi ekowisata Situ Cimeuhmal perlu dilakukan kordinasi dan
perumusan pedoman tata kelola destinasi dengan mempertimbangkan perlindungan atas
keberlangsungan konservasi sumber mata air alami ini. Apabila kelemahan dan ancaman tidak dikelola dengan baik akan menimbulkan potensi
ekowisata ini berkembang ke arah yang akan merusak lingkungan dan selanjutnya potensi
ekowisata ini tidak akan dikunjungi lagi oleh wisatawan. Diperlukan upaya untuk membangun
kesadaran SDM pengelola dan wisatawan akan pentingnya pariwisata yang berkelanjutan. Apabila kelemahan dan ancaman tidak dikelola dengan baik akan menimbulkan potensi
ekowisata ini berkembang ke arah yang akan merusak lingkungan dan selanjutnya potensi
ekowisata ini tidak akan dikunjungi lagi oleh wisatawan. Diperlukan upaya untuk membangun
kesadaran SDM pengelola dan wisatawan akan pentingnya pariwisata yang berkelanjutan. Berdasarkan identifikasi atribut pariwisata secara faktual di lapangan, strategi pengembangan yang
tepat adalah Strenght to Opportunity, yaitu strategi yang dilakukan melalui pemanfaatan Situ Cimeuhmal
sebagai destinasi pariwisata berbasis konservasi sumber daya air yang memiliki kawasan yang asri, 91 Jurnal Pariwisata Terapan, Vol. 5., No. 2, 2021 Kunkun kurniawan, Reiza D Dienaputra, Cecep Ucu Rahman rindang, dan tenang dengan melakukan kegiatan atau event yang dapat mempromosikan ekowisata baru
di bawah kaki Gunung Malabar. Atraksi alam dan ritual event berbasis budaya lokal seperti “Kawin Cai”
tidak cukup untuk menarik wisatawan berkunjung. Diperlukan atraksi wisata lainnya untuk memberikan
pengalaman baru bagi wisatawan ketika berkunjung ke ekowisata Situ Cimeuhmal. Selain event tradisi
“kawin cai” yang rutin diselenggarakan setiap satu tahun sekali untuk tujuan membagikan hasil panen
ikan kepada masyarakat sekitarnya juga diperlukan event menarik lainnya untuk menambah animo
pengunjung ke Ekowisata Situ Cimeuhmal seperti Festival Kopi Gunung Malabar. Peninggalan terowongan air yang yang dibangun pada jaman Belanda sebagai infrastruktur pengelolaan
dan pemeliharaan air dapat dijadikan wisata edukasi berbasis sejarah yang dapat melengkapi daya tarik
ekowisata Situ Cimeuhmal. Kekuatan infrastruktur yang dibangun jaman Belanda menunjukkan kualitas
yang terbaik dan komitmen yang serius dalam pengelolaan dan pemeliharaan sumber daya air. Latar
belakang sejarah lain yang dapat dikembangkan adalah mengenai pengetahuan tentang geo sejarah
pembentukan pegunungan Malabar yang menjadi latar belakang yang menarik ketika wisatawan
berkunjung ke Ekowisata Situ Cimeuhmal. Ada kearifan lokal lain yang menarik ditawarkan bagi
wisatawan adalah aktifitas pandai besi dalam memproduksi alat untuk pertanian. Jurnal Pariwisata Terapan, Vol. 5., No. 2, 2021 Strenght – Threat: Pelibatan masyarakat dalam menyediakan
fasilitas akomodasi bagi wisatawan diharapkan dapat memberikan dampak ekonomi yang bisa
meningkatkan kesejahteraan masyarakat di sekitar destinasi ekowisata Situ Cimeuhmal. Fasilitas lain
yang dapat menjadi potensi peningkatan ekonomi masyarakat sekitar adalah menyediakan kedai
kopi, warung makan dan ruang pamer produk-produk yang dihasilkan oleh masyarakat lokal baik
yang berbentuk produk makanan maupun kerajinan. Salah satu produk khas dari destinasi Ekowisata
Situ Cimeuhmal adalah Kopi yang dihasilkan dari lahan yang ada di sekitar situ. Beberapa petani
mengelola ladang kopi yang tersebar di sekitar kawasan ekowisata Situ Cimeuhmal dan bekerja sama
dengan Kelompok Peduli Lingkungan Situ Cimeuhmal (KPLSC) dan Komunitas Anak Peduli Alam
Semesta (KAPAS) untuk membuat produk kopi dan kedai dengan nama Kopi Cimeuhmal. 2. 2. Amenitas
Destinasi Ekowisata Situ Cimeuhmal belum memiliki fasilitas akomodasi untuk melayani kebutuhan
wisatawan yang berkunjung. Konsep Homestay cocok untuk diterapkan dalam pengembangan
destinasi ekowisata Situ Cimeuhmal yang berkelanjutan. Pelibatan masyarakat dalam menyediakan
fasilitas akomodasi bagi wisatawan diharapkan dapat memberikan dampak ekonomi yang bisa
meningkatkan kesejahteraan masyarakat di sekitar destinasi ekowisata Situ Cimeuhmal. Fasilitas lain
yang dapat menjadi potensi peningkatan ekonomi masyarakat sekitar adalah menyediakan kedai
kopi, warung makan dan ruang pamer produk-produk yang dihasilkan oleh masyarakat lokal baik
yang berbentuk produk makanan maupun kerajinan. Salah satu produk khas dari destinasi Ekowisata
Situ Cimeuhmal adalah Kopi yang dihasilkan dari lahan yang ada di sekitar situ. Beberapa petani
mengelola ladang kopi yang tersebar di sekitar kawasan ekowisata Situ Cimeuhmal dan bekerja sama
dengan Kelompok Peduli Lingkungan Situ Cimeuhmal (KPLSC) dan Komunitas Anak Peduli Alam
Semesta (KAPAS) untuk membuat produk kopi dan kedai dengan nama Kopi Cimeuhmal. 92 Jurnal Pariwisata Terapan, Vol. 5., No. 2, 2021 Kunkun kurniawan, Reiza D Dienaputra, Cecep Ucu Rahman Kunkun kurniawan, Reiza D Dienaputra, Cecep Ucu Rahman Gambar 2. Situ Cimeuhmal
Sumber: Foto Pribadi, 2021. Gambar 2. Situ Cimeuhmal Gambar 2. Situ Cimeuhmal Sumber: Foto Pribadi, 2021. Aspek promosi menjadi salah satu unsur penting dalam pengembangan ekowisata Situ Cimeuhmal. Hal
ini dapat dilakukan melalui promosi yang berbasis offline dan online. Tren penggunaan media online
semakin meningkat sejalan dengan efektifitas dampak yang ditimbulkannya Penggunaan platform media
online menjadi kebutuhan dan tuntutan baru dalam pengembangan destinasi pariwisata. Promosi
pariwisata dapat dilakukan secara below the line dan Above the line. Kunkun kurniawan, Reiza D Dienaputra, Cecep Ucu Rahman Kunkun kurniawan, Reiza D Dienaputra, Cecep Ucu Rahman Strenght – Threat: Kegiatan promosi below the line
banyak melibatkan aktifitas dan interaksi langsung dengan wisatawan sedangkan kegiatan above the line
melibatkan aktifitas dan interaksi melalui media digital seperti IG, Youtube, Facebook, dan Tiktok dengan
menggunakan endorser yang memiliki reputasi baik, berpengaruh dan berkomitmen terhadap
kelestarian alam. 93 Jurnal Pariwisata Terapan, Vol. 5., No. 2, 2021 Kesimpulan Hasil identifikasi potensi wisata yang dapat dikembangkan di Desa Banjaran Wetan adalah potensi
produk kuliner kerupuk gurileum, kerupuk kulit, dan produk kopi beserta ragam olahan lainnya. Ketiga
potensi ini dapat dikembangkan menjadi program wisata edukasi untuk mengetahui proses pembuatan
produk kuliner kerupuk dan pengolahan kopi serta penyajiannya. Wisata edukasi ini menjadi pilihan bagi
wisatawan yang berkunjung ke Desa Banjaran Wetan selain Situ Cimeuhmal sebagai destinasi ekowisata
yang berbasis konservasi sumber daya air. Strategi pengembangan Situ Cimeuhmal menjadi destinasi
ekowisata, adalah melalui peningkatan kualitas dari aspek atraksi, aksesibilitas, dan amenitas. Atraksi
unik baru untuk wisatawan harus disiapkan untuk melengkapi pilihan wisatawan berkunjung, tidak saja
edukasi terkait konservasi air tetapi dapat dikembangkan dari sisi sejarah dan kearifan lokal lain yang
menjadi ciri khas ekowisata Situ Cimeuhmal serta menjadi spot menarik untuk tantangan bersepeda
sambil menikmati alam dan kopi khas lokal. Perbaikan sarana jalan menuju kawasan hutan pinus dan
penyediaan akses komunikasi yang lebih menjangkau area di kawasan Ekowisata Situ Cimeuhmal
menjadi faktor utama dalam pengembangan destinasi pariwisata. Fasilitas akomodasi berbentuk
Homestay memberikan alternatif model kunjungan wisatawan untuk bisa bermalam dan berinteraksi
dengan kehidupan sehari hari masyarakat di kawasan ekowisata Situ Cimeuhmal. Selain itu kegiatan
promosi perlu dioptimalkan baik secara offline maupun online dengan menggunakan endorser public
figure untuk menjangkau efek promosi yang lebih luas dan beragam. Daftar Pustaka Arida, N. S. (2021). Dosen Pariwisata Universitas Udayana Ungkap Ada Pseudo Ekowisata di Bali. Tempo.Com. https://travel.tempo.co Arida, N. S. (2021). Dosen Pariwisata Universitas Udayana Ungkap Ada Pseudo Ekowisata di Bali. Tempo.Com. https://travel.tempo.co Azhar, S. (2003). Metode Penelitian. Yogyakarta: Pustaka Pelajar. Fennell, D. A. (2008). Ecotourism and the myth
193. https://doi.org/10.2167/jost736.0 Jamalina, I. A., & Wardani, D. T. K. (2017). Strategi Pengembangan ekowisata melalui konsep Community Based
Tourism (Cbt) dan manfaat sosial dan ekonomi bagi masyarakat di desa wisata Nglanggeran, Patuk, Gunung
Kidul. Jurnal Ekonomi & Studi Pembangunan, 18(1), 71–85. https://doi.org/10.18196/jesp.18.1.4008 Komunitas Anak Peduli Alam Semesta (KAPAS) (2018). Peta Penanaman Pohon Situ Cim McComb, E. J., Boyd, S., & Boluk, K. (2017). Stakeholder collaboration: A means to the success of rural tourism
destinations? A critical evaluation of the existence of stakeholder collaboration within the Mournes,
Northern
Ireland. Tourism
and
Hospitality
Research,
17(3),
286–297. https://doi.org/10.1177/1467358415583738 Palmer, N. J., & Chuamuangphan, N. (2018). Governance and local participation in ecotourism: community-level
ecotourism stakeholders in Chiang Rai province, Thailand. Journal of Ecotourism, 17(3), 320–337. https://doi.org/10.1080/14724049.2018.1502248 Sella, K., & Yusuf, M. (2020). Identifikasi Peran dan Koordinasi Pemangku Kepentingan Terhadap Pengembangan
Sarana dan Prasarana di Atraksi Wisata Menara Siger , Kabupaten Lampung Selatan. 4(2), 130–146. Sella, K., & Yusuf, M. (2020). Identifikasi Peran dan Koordinasi Pemangku Kepentingan Terhadap Pengembangan
Sarana dan Prasarana di Atraksi Wisata Menara Siger , Kabupaten Lampung Selatan. 4(2), 130–146. Sidiq, A. J., & Resnawaty, R. (2017). Pengembangan desa wisata berbasis partisipasi masyarakat lokal di Desa
Wisata Linggarjati Kuningan, Jawa Barat. Prosiding Penelitian Dan Pengabdian Kepada Masyarakat, 4(1), 38. https://doi.org/10.24198/jppm.v4i1.14208 Sidiq, A. J., & Resnawaty, R. (2017). Pengembangan desa wisata berbasis partisipasi masyarakat lokal di Desa
Wisata Linggarjati Kuningan, Jawa Barat. Prosiding Penelitian Dan Pengabdian Kepada Masyarakat, 4(1), 38. https://doi.org/10.24198/jppm.v4i1.14208
Suryabrata, S. (1987). Metode Penelitian. Jakarta: Rajawali Press. p
g
jpp
Suryabrata, S. (1987). Metode Penelitian. Jakarta: Rajawali Press. Suryabrata, S. (1987). Metode Penelitian. Jakarta: Rajawali Press. 94 Jurnal Pariwisata Terapan, Vol. 5., No. 2, 2021 Kunkun kurniawan, Reiza D Dienaputra, Cecep Ucu Rahman Kunkun kurniawan, Reiza D Dienaputra, Cecep Ucu Rahman
Tabel 1. Daftar Pustaka Perhitungan Bobot IFAS
No
Faktor Internal
Bobot
Rating
BobotxR
ating
Kekuatan (S)
1
Kawasan lingkungan dan situ yang masih asri dan
nyaman
0.20
4
0.80
2
Sumber mata air alami yang menjadi hulu sungai
cisangkuy
0.20
4
0.80
3
Bangunan sejarah berupa terowongan air jaman
Belanda
0.16
3
0.48
Total S
2.08
Kelemahan (W)
1
Infrastruktur jalan yang belum ditata dengan baik
0.16
2
0.32
2
Informasi dan tata kelola destinasi belum lengkap dan
jelas
0.12
1
0.12
3
Sinyal komunikasi yang kurang baik
0.16
1
0.16
Total W
0.60
Total
1
Total EFAS (O-T)
1.48
Sumber: Olahan Peneliti, 2021
Tabel 2. Perhitungan Bobot EFAS
No
Faktor Eksternal
Bobot
Rating
BobotxR
ating
Peluang (O)
1
Satu satunya daya tarik situ di sekitar Gunung Malabar
0.27
4
1.09
2
Pengembangan usaha komoditas kopi dari hulu sampai
hilir
0.23
4
0.91
3
Pengembangan usaha produk makanan ringan
0.09
3
0.27
Total O
2.27
Ancaman (T)
1
Perilaku wisatawan tidak sejalan dengan konsep
konservasi
0.18
1
0.18
2
Konflik kepentingan antar pemangku kepentingan
0.09
2
0.18
3
Penggunaan lahan konservasi untuk fasilitas wisata
0.14
1
0.14
Total T
0.50
Total
1
Total EFAS (O-T)
1.77
Sumber: Olahan Peneliti, 2021 Kunkun kurniawan, Reiza D Dienaputra, Cecep Ucu Rahman
Tabel 1. Perhitungan Bobot IFAS
No
Faktor Internal
Bobot
Rating
BobotxR
ating
Kekuatan (S)
1
Kawasan lingkungan dan situ yang masih asri dan
nyaman
0.20
4
0.80
2
Sumber mata air alami yang menjadi hulu sungai
cisangkuy
0.20
4
0.80
3
Bangunan sejarah berupa terowongan air jaman
Belanda
0.16
3
0.48
Total S
2.08
Kelemahan (W)
1
Infrastruktur jalan yang belum ditata dengan baik
0.16
2
0.32
2
Informasi dan tata kelola destinasi belum lengkap dan
jelas
0.12
1
0.12
3
Sinyal komunikasi yang kurang baik
0.16
1
0.16
Total W
0.60
Total
1
Total EFAS (O-T)
1.48
Sumber: Olahan Peneliti, 2021 Tabel 1. Perhitungan Bobot IFAS Tabel 2. Jurnal Pariwisata Terapan, Vol. 5., No. 2, 2021 Daftar Pustaka Perhitungan Bobot EFAS
No
Faktor Eksternal
Bobot
Rating
BobotxR
ating
Peluang (O)
1
Satu satunya daya tarik situ di sekitar Gunung Malabar
0.27
4
1.09
2
Pengembangan usaha komoditas kopi dari hulu sampai
hilir
0.23
4
0.91
3
Pengembangan usaha produk makanan ringan
0.09
3
0.27
Total O
2.27
Ancaman (T)
1
Perilaku wisatawan tidak sejalan dengan konsep
konservasi
0.18
1
0.18
2
Konflik kepentingan antar pemangku kepentingan
0.09
2
0.18
3
Penggunaan lahan konservasi untuk fasilitas wisata
0.14
1
0.14
Total T
0.50
Total
1
Total EFAS (O-T)
1.77
Sumber: Olahan Peneliti, 2021 95 Jurnal Pariwisata Terapan, Vol. 5., No. 2, 2021 Kunkun kurniawan, Reiza D Dienaputra, Cecep Ucu Rahman Gambar 3. Hasil analisis strategi IFAS EFAS
Sumber: Olahan Peneliti, 2021 Gambar 3. Hasil analisis strategi IFAS EFAS Penataan Penataan
infrastruktur
menuju lokasi dapat dilakukan
melalui kerjasama, kordinasi
dan kolaborasi pemerintah
setempat dengan masyarakat
sekitar. Peningkatan sarana
jalan dan petunjuk yang jelas
dan
informatif
akan
memudahkan
wisatawan
untuk mengunjungi destinasi
ekowisata ini. 2. Pengembangan usaha
komoditas kopi dari hulu
sampai hilir 2. Pengembangan usaha
komoditas kopi dari hulu
sampai hilir 3. Pengembangan usaha
produk makanan ringan Strategi WT
Apabila
kelemahan
dan
ancaman
tidak
dikelola
dengan
baik
akan
menimbulkan
potensi
ekowisata ini berkembang ke
arah yang akan merusak
lingkungan dan selanjutnya
potensi ekowisata ini tidak
akan dikunjungi lagi oleh
wisatawan. Diperlukan upaya
untuk membangun kesadaran
SDM
pengelola
dan
wisatawan akan pentingnya
pariwisata
yang
berkelanjutan. Strategi WT
Apabila
kelemahan
dan
ancaman
tidak
dikelola
dengan
baik
akan
menimbulkan
potensi
ekowisata ini berkembang ke
arah yang akan merusak
lingkungan dan selanjutnya
potensi ekowisata ini tidak
akan dikunjungi lagi oleh
wisatawan. Diperlukan upaya
untuk membangun kesadaran
SDM
pengelola
dan
wisatawan akan pentingnya
pariwisata
yang
berkelanjutan. Gambar 3. Hasil analisis strategi IFAS EFAS Sumber: Olahan Peneliti, 2021 96 Jurnal Pariwisata Terapan, Vol. 5., No. 2, 2021 Kunkun kurniawan, Reiza D Dienaputra, Cecep Ucu Rahman Sumber: Olahan Peneliti, 2021 Strategi WO Strategi SO
Pemanfaatan
Situ
Cimeuhmal
sebagai
destinasi
ekowisata
terkait konservasi mata air alami
hulu
sungai
cisangkuy
serta
kebutuhan air masyarakat sekitar. Selanjutnya membuat event yang
lebih
menarik
untuk
mempromosikan ekowisata Situ
Cimeuhmal. Kemudian membuat
produk wisata sejarah terkait
infrastruktur
pengairan
pada
jaman Belanda di sekitar Situ
Cimeuhmal. Selain
itu,
pemanfaatan
potensi
wisata
edukasi
pembuatan
produksi
makanan
ringan
dan
proses
pengolahan komoditas kopi dari
hulu sampai hilir yang melibatkan
masyarakat di sekitar kawasan
Situ Cimeuhmal g ( )
1. Satu satunya daya tarik
situ di sekitar Gunung
Malabar Tabel 3. Hasil strategi SO IFAS
EFAS Kelemahan (W) 1. Infrastruktur jalan yang
belum ditata dengan baik 2. Informasi dan tata kelola
destinasi belum lengkap
dan jelas 3. Sinyal komunikasi yang
kurang baik Strategi ST 1. Perilaku wisatawan tidak
sejalan dengan konsep
konservasi Sumber: Olahan Peneliti, 2021 97 Jurnal Pariwisata Terapan, Vol. 5., No. 2, 2021
|
https://openalex.org/W2005851086
|
https://escholarship.org/content/qt4n24t7bz/qt4n24t7bz.pdf?t=nqtljl
|
English
| null |
Comparison of Calmodulin Gene Expression in Human Neonatal Melanocytes and Metastatic Melanoma Cell Lines
|
The journal of investigative dermatology/Journal of investigative dermatology
| 1,992
|
cc-by
| 200
|
Permalink Copyright Information
This work is made available under the terms of a Creative Commons
Attribution License, availalbe at
https://creativecommons.org/licenses/by/4.0/ UC Irvine
UC Irvine Previously Published Works
Title
Comparison of Calmodulin Gene Expression in Human Neonatal
Melanocytes and Metastatic Melanoma Cell Lines
Permalink
https://escholarship.org/uc/item/4n24t7bz
Journal
Journal of Investigative Dermatology, 99(6)
ISSN
0022-202X
Authors
Hickie, Robert A
Graham, Mark J
Buckmeier, Julie A
et al.
Publication Date
1992-12-01
DOI
10.1111/1523-1747.ep12614725
Copyright Information
This work is made available under the terms of a Creative Common
Attribution License, availalbe at
https://creativecommons.org/licenses/by/4.0/
Peer reviewed UC Irvine
UC Irvine Previously Published Works
Title
Comparison of Calmodulin Gene Expression in Human Neonatal
Melanocytes and Metastatic Melanoma Cell Lines
Permalink
https://escholarship.org/uc/item/4n24t7bz
Journal
Journal of Investigative Dermatology, 99(6)
ISSN
0022-202X
Authors
Hickie, Robert A
Graham, Mark J
Buckmeier, Julie A
et al. Publication Date
1992-12-01
DOI
10.1111/1523-1747.ep12614725
Copyright Information
This work is made available under the terms of a Creative Common
Attribution License, availalbe at
https://creativecommons.org/licenses/by/4.0/
Peer reviewed Title Title
Comparison of Calmodulin Gene Expression in Human Neonatal
Melanocytes and Metastatic Melanoma Cell Lines Comparison of Calmodulin Gene Expression in Human Neonata
Melanocytes and Metastatic Melanoma Cell Lines Powered by the California Digital Library
University of California eScholarship.org
|
https://openalex.org/W4362469371
|
https://jartel.polinema.ac.id/index.php/jartel/article/download/509/234
|
English
| null |
Monitoring and Controlling System to Artificial Habitat of Antlion (Myrmeleon Sp.) based Wireless Sensor Network
|
Jurnal Jaringan Telekomunikasi
| 2,023
|
cc-by
| 5,447
|
Journal of Telecommunication Network (Jurnal Jaringan Telekomunikasi) Vol. XX, No.1 (2023) Journal of Telecommunication Network (Jurnal Jaringan Telekomunikasi) Vol. XX, No.1 (2023) I. INTRODUCTION support the habitat of farmers who are far from internet access. The addition of a semi-automatic sifting harvest system with a
function to sort the antlions larvae that are ready to harvest. Therefore, a system is made that can monitor and control
temperature and humidity, as well as a harvesting mechanism
in the antlions habitat. Antlion is a type of insect which in the larval stage contains
the active substance sulfonylurea which is useful in the
treatment of Diabetes Mellitus (DM) [1]. The antlions habitat
is in dry sandy areas with a temperature range of 24-30℃ and
humidity of 70-80% so that changes in temperature and
humidity are critical parameters for the chance of antlions
living [2]. However, the monitoring and control of these two
parameters has not been carried out by retreat breeders. In
addition, the harvesting process carried out by farmer still uses
manual sorting by using a sieve in each larval burrow so that it
takes a relatively long time [3]. Based on these problems, an
effort is needed to maintain the availability of antler larvae and
facilitate the harvest process. Therefore, it is necessary to have
a technology-based monitoring and control system to maintain
the survival of the retreat and facilitate the harvest process. Monitoring and Controlling System to Artificial
Habitat of Antlion (Myrmeleon Sp.) based
Wireless Sensor Network Fairuz Rahmadika1, Aad Hariyadi2, Lis Diana Mustafa3
1,2,3 Digital Telecommunication Network Study Program,
Department of Electrical Engineering, State Polytechnic of Malang, Indonesia
2,3 Telecommunication Engineering Study Program,
Department of Electrical Engineering, State Polytechnic of Malang, Indonesia Fairuz Rahmadika1, Aad Hariyadi2, Lis Diana Mustafa3
1,2,3 Digital Telecommunication Network Study Program,
Department of Electrical Engineering, State Polytechnic of Malang, Indonesia
2,3 Telecommunication Engineering Study Program,
Department of Electrical Engineering, State Polytechnic of Malang, Indonesia Fairuz Rahmadika1, Aad Hariyadi2, Lis Diana Mustafa3
1,2,3 Digital Telecommunication Network Study Program,
Department of Electrical Engineering, State Polytechnic of Malang, Indonesia
2,3 Telecommunication Engineering Study Program,
Department of Electrical Engineering, State Polytechnic of Malang, Indonesia 11841160057@student.polinema.ac.id, 2aad.hariyadi@polinema.ac.id, 3lis.diana@polinema.ac.id Abstract—Antlion habitat that breeders create currently does not use a system that can protect the habitat from changes in temperature
and humidity. In addition, when they are about to be sold, breeders filter and sort the antlion one by one. Therefore, a system is made
that can monitor and control changes in temperature and humidity as well as harvesting automatically through the application. This
system is designed using a wireless sensor network with temperature and humidity sensors DHT22, IR MLX90614 with light actuators
and mist makers as well as loadcell sensors to trigger off the sieving motor which is connected to the Firebase database and interface
Flutter aplication. Based on the results of testing the accuracy of temperature, humidity, infrared, and loadcell sensor readings,
respectively, they are 98.07, 97.58%, 98.38%, 98.78% with an average delay 216.97ms can maintain temperature changes in the range
24-30℃ and humidity in the range 70-80% stably. And the average value of harvest time is 30.37 seconds with an average decrease in
sand load of 56.7 grams when the harvesting mechanism is running. Keywords— Antlion, Wireless Sensor Network, DHT22, IRMLX90614, Loadcell A. System design Table I shows the data from the observation of the crab crab
habitat in the breeder with parameters of temperature and
humidity, habitat media, harvest time and size of crab cakes
that are ready to be sold. A. System design Mist Maker OFF
Y
N
Finish Humidity
≤ 70% ? Figure 1 Blok Diagram System
TABLE I
ANTLION OBSERVATION DATA
Temperature
(°C)
Humidity
(%)
Medium
Harvest
Antlion
Size
24 – 30
70 – 80
Dry sand
±3 Week
±4-5mm
Table I shows the data from the observation of the crab crab
habitat in the breeder with parameters of temperature and
humidity, habitat media, harvest time and size of crab cakes
that are ready to be sold. Mikrokontroller
Sender
DHT22
IR MLX90614
Relay
Input
Relay
Relay
Loadcell
Motor 1
Motor 2
Lamp
Cloud
LCD
Output
Smartphone
Router
Node
Sumber AC
220Volt
Power Supply
12Volt
DC Step Down
3.3/5Volt
Relay
Mist Maker
Wire
Wireless
Mikrokontroller
to firebase
Mikrokontroller
Reciever
DC Fan Cloud
Smartphone
Router
Wire
Wireless
Mikrokontroller
to firebase
Mikrokontroller
Reciever Humidity
≤ 80% ? Node Figure 1 Blok Diagram System
TABLE I
ANTLION OBSERVATION DATA
Mikrokontroller
Sender
DHT22
IR MLX90614
Relay
Input
Relay
Relay
Loadcell
Motor 1
Motor 2
Lamp
LCD
Output
Node
Sumber AC
220Volt
Power Supply
12Volt
DC Step Down
3.3/5Volt
Relay
Mist Maker
DC Fan Figure 2 Flowchart Monitoring Sensor The flowchart of the harvesting mechanism system is shown in
Fig. 3. The flowchart of the harvesting mechanism system is shown in
Fig. 3. The flowchart of the harvesting mechanism system is shown in
Fig. 3. Figure 1 Blok Diagram System This sifting system design flowchart has 2 sieving mechanism
parameters, namely harvest and load. Harvest conditions will
determine for active control of motor 1 (opening the shelf base)
and motor 2 (sifter). The load condition is the weight of the
sand that falls into the shelter, the reading of which will
determine the control of motor 2 (sifter) and motor 1 (closing
the rack base). Each of these conditions will be carried out on
the habitat to be harvested or also used to determine the
difference in sand load which will determine the increase in the
population of crabs. The sieving mechanism can be run through
a smartphone or directly through the sieving button on the
habitat. TABLE I
ANTLION OBSERVATION DATA Table I shows the data from the observation of the crab crab
habitat in the breeder with parameters of temperature and
humidity, habitat media, harvest time and size of crab cakes
that are ready to be sold. A. System design The system design contains a block diagram design with
antlion habitat references. The block diagram is shown in Fig. 1. Explanations of the components on the block diagram include: 1. Nodes are 2 retreat habitats with habitat configurations to
the receiver to the microcontroller to firebase and the cloud
so that they can be accessed via smartphones. Utilization of technology related to temperature and
humidity
monitoring
and
control
systems
through
microcontroller devices refers to research [4] and [5], which
uses the DHT22 temperature and humidity sensor and the
MLX90614 IR infrared temperature sensor with outputs active
or deactivated lamp and mist makers actuator that can be
accessed via smartphones for easy monitoring. The system
created works by integrating sensors and actuators with the
ESP8266 microcontroller. The choice of this type of
microcontroller is because it is able to transmit data with a
communication line that was originally directly to the router
using an IP Address, now it can use a MAC Address with a
longer available transmission distance [6][7]. This things to 2. The sender microcontroller is used as a center for
controlling sensors and actuators as well as transmitting
data connected to the receiver microcontroller wirelessly. 3. Inputs include a DHT22 sensor for temperature and
humidity detection, MLX90614 IR sensor for sand surface
temperature detection, and Loadcell sensor for sand load
detection when the sieving system is active. 4. Output includes Relay as an electronic switch to control the
on and off the motor. Mist maker-24 Volt DC fan for
humidity control and lights for temperature control in the
retreat habitat. E-ISSN: 2654-6531 P- ISSN: 2407-0807 25 Journal of Telecommunication Network (Jurnal Jaringan Telekomunikasi) Vol. 13, No.1 (2023) Figure 2 Flowchart Monitoring Sensor
Start
Initiation
Temperature & Humidity, Sand
Temperature, Avrg Temperature, and Time. Avrg Temperature
≤ 24°C ? Avrg Temperature
≥ 30°C ? Lamp OFF
Y
N
Y
N
Lamp ON
Avrg Temperature Calculation =
(Temperature + Sand Temperature) / 2
Humidity
≤ 70% ? Y
Mist Maker ON
N
A
Humidity
≤ 80% ? Mist Maker OFF
Y
N
Finish
A The results of the interview determine the use of a star
topology based on the location of the undur-undur habitat
breeder in Grobyog, Tj. Rejo, Wuluhan, Jember with habitat
data shown in Table I [3], [7], [8]. Humidity
≤ 70% ? Y
Mist Maker ON
N
A
Humidity
≤ 80% ? C. Prototype design The flowchart contains the design flow of the sensor
monitoring system for controlling temperature and humidity as
well as the harvest mechanism system. The sensor monitoring
flow chart is shown in Fig. 2. The prototype design shows the design of the use and
placement of sensors, actuators and microcontrollers in the
artificial turtle habitat. The following is a description of Fig. 4: The following is a description of Fig. 4: The flowchart of the temperature, humidity and feed
monitoring system design has 2 habitat monitoring parameters,
namely the average temperature and humidity. The average
temperature condition is obtained from the calculation of the
average value between the air temperature and the surface
temperature of the sand, the results of which will control the on
and off lights according to the provisions. Humidity is the
humidity of the air in the habitat area, the reading of which will
control the mist maker on and off according to the provisions. Each of these conditions will also run on habitats 1 and 2 which
will then be displayed via the LCD and sent to the server to be
displayed via the smartphone application. 1. A shows the lamp actuator for temperature contro 2. B shows the temperature and humidity sensor, namely
DHT22. 3. C using a sand surface temperature sensor, namely IR
MLX90614. 4. D indicates the weight sensor for sand falling into the
holding container using the loadcell on the LED scale. 5. E shows the ESP8266 microcontroller as the control center
between sensors and actuators. 6. F show the driving motor of the sieving mechanism. 7. G shows the driving motor of the pedestal opening and
closing mechanism. 8. H show mist maker actuator for humidity control. 9. I show LCD. E-ISSN: 2654-6531 P- ISSN: 2407-0807 26 Journal of Telecommunication Network (Jurnal Jaringan Telekomunikasi) Vol. 13, No.1 (2023) Journal of Telecommunication Network (Jurnal Jaringan Telekomunikasi) Vol. 13, No.1 (2023) Figure 5 Implementation of Component Design Figure 3 Flowchart Monitoring Sensor
Start
Initiation
Haverst and load
Haverst ? Sand Load
fit ? Y
Y
N
Motor 1 (opening the slide
shelf) ON
Motor 2 (sieve) ON
Motor 2 (sieve) OFF
Motor 1 (closing the slide
shelf) ON
N
Finish Figure 3 Flowchart Monitoring Sensor
Figure 4 Prototype Design
The implementation of the component design is shown in
Fig 5
Start
Initiation
Haverst and load
Haverst ? Sand Load
fit ? III. RESULTS AND DISCUSSION Based on the planning of the system, testing is carried out to
determine its performance and analyze the results of the design. So that it can find out the results of several test parameters. C. Prototype design Y
Y
N
Motor 1 (opening the slide
shelf) ON
Motor 2 (sieve) ON
Motor 2 (sieve) OFF
Motor 1 (closing the slide
shelf) ON
N
Finish Initiation
Haverst and load Haverst ? Motor 1 (opening the slide
shelf) ON Figure 5 Implementation of Component Design A. Sensor Test Sand Load
fit ? Sensor testing is carried out to determine the accuracy of the
sensor reading value compared to the reading on a standardized
measuring instrument. This test was carried out on three
sensors used in the system, including the DHT22 sensor as a
temperature sensor, IR MLX90614 as a humidity sensor and
loadcell as a load sensor. Sensor test results are shown in the
Table II. TABLE II
DHT22 TEMPERATURE SENSOR ACCURACY TEST RESULTS DHT22 TEMPERATURE SENSOR ACCURACY TEST RESULTS
DHT22 (°C)
Thermometer (°C)
Deviasi Error (%)
26.8
26.2
2.290076336
27.5
26.7
2.996254682
25.4
26.3
3.422053232
28.1
27.5
2.181818182
26.8
26.9
0.383141762
25.7
26.1
1.486988848
26.2
26.9
2.651515152
26
26.4
1.53256705
25.7
26.1
1.509433962
26.3
26.5
0.754716981 Figure 3 Flowchart Monitoring Sensor Figure 3 Flowchart Monitoring Sensor
Figure 4 Prototype Design Based on the results of the temperature sensor test in Table
II, it is known that there is a difference in the temperature value
of the DHT22 reading compared to the value on the UNI-
T333S Thermometer so that the deviation error value is
obtained. Furthermore, the error deviation value is used as a
reference to determine the average error value and accuracy of
the DHT22 sensor through the following calculations: 𝐸𝑟𝑟𝑜𝑟 𝐴𝑣𝑒𝑟𝑎𝑔𝑒= 19.20856619
10
𝑥 100%
= 1.920856619%
𝐴𝑐𝑐𝑢𝑟𝑎𝑐𝑦= 100% −(1.920856619%) = 98.07914338% 𝐸𝑟𝑟𝑜𝑟 𝐴𝑣𝑒𝑟𝑎𝑔𝑒= 19.20856619
10
𝑥 100%
= 1.920856619%
𝐴𝑐𝑐𝑢𝑟𝑎𝑐𝑦= 100% −(1.920856619%) = 98.07914338% Figure 4 Prototype Design The implementation of the component design is shown in
Fig. 5. From the calculation of the temperature sensor accuracy, the
accuracy value is 98.08%. These results indicate that the E-ISSN: 2654-6531 P- ISSN: 2407-0807 27 Journal of Telecommunication Network (Jurnal Jaringan Telekomunikasi) Vol. 13, No.1 (2023) 𝐸𝑟𝑟𝑜𝑟 𝐴𝑣𝑒𝑟𝑎𝑔𝑒= 12.10885147
10
𝑥 100%
= 1.210885147%
𝐴𝑐𝑐𝑢𝑟𝑎𝑐𝑦= 100% −(1.210885147) = 98.78911485% 𝐸𝑟𝑟𝑜𝑟 𝐴𝑣𝑒𝑟𝑎𝑔𝑒= 12.10885147
10
𝑥 100%
= 1.210885147%
𝐴𝑐𝑐𝑢𝑟𝑎𝑐𝑦= 100% −(1.210885147) = 98.78911485% From the calculation of the temperature sensor accuracy,
the accuracy value is 98.38%. These results indicate that the
temperature sensor is of high accuracy because it has a reading
error of below 3% with an accuracy specification on the IR
MLX90614 datasheet of ± 2.56℃ [10]. From the calculation of the load sensor accuracy, the
accuracy value is 98.79%. These results indicate that the load
sensor includes high accuracy because it has a reading error
below 3% [11]. TABLE IV
DHT22 HUMIDITY SENSOR ACCURACY TEST RESULTS
DHT22 (%)
Hygrometer (%)
Error (%)
80.3
78.7
2.033036849
80
78.9
1.394169835
79.9
78.1
2.304737516
82.7
80.1
3.245942572
81.6
80.4
1.492537313
78.2
80.7
3.097893432
77.7
78.6
1.145038168
79.8
77.7
2.702702703
79.2
76.9
2.990897269
77.9
75.1
3.728362184 TABLE IV
DHT22 HUMIDITY SENSOR ACCURACY TEST RESULTS 𝐸𝑟𝑟𝑜𝑟 𝐴𝑣𝑒𝑟𝑎𝑔𝑒= 16.14905225
10
𝑥 100%
= 1.614905225%
𝐴𝑐𝑐𝑢𝑟𝑎𝑐𝑦= 100% −(1.614905225%) = 98,38509477% 𝐸𝑟𝑟𝑜𝑟 𝐴𝑣𝑒𝑟𝑎𝑔𝑒= 16.14905225
10
𝑥 100%
= 1.614905225%
𝐴𝑐𝑐𝑢𝑟𝑎𝑐𝑦= 100% −(1.614905225%) = 98,38509477% 𝐸𝑟𝑟𝑜𝑟 𝐴𝑣𝑒𝑟𝑎𝑔𝑒= 16.14905225
10
𝑥 100%
= 1.614905225%
𝐴𝑐𝑐𝑢𝑟𝑎𝑐𝑦= 100% −(1.614905225%) = 98,38509477% = 2.413531784%
𝐴𝑐𝑐𝑢𝑟𝑎𝑐𝑦= 100% −(2.413531784) = 97.58646822% temperature sensor is of high accuracy because it has a reading
error of below 3% with an accuracy specification on the
DHT22 datasheet of ± 0.5℃ [9]. temperature sensor is of high accuracy because it has a reading
error of below 3% with an accuracy specification on the
DHT22 datasheet of ± 0.5℃ [9]. From the calculation of the humidity sensor accuracy, the
accuracy value is 97.58%. These results indicate that the
humidity sensor includes high accuracy because it has a
reading error below 3% with an accuracy specification on the
DHT22 datasheet of ± 5% [9]. TABLE III TABLE III TABLE III
IR MLX90614 SURFACE TEMPERATURE ACCURACY TEST RESULTS
IRMLX90614 (°C)
Thermogun (°C)
Error (%)
32.59
31.7
2.807570978
25.43
25.2
0.912698413
26.19
26.06
0.49884881
25.9
25.7
0.778210117
27.29
26.9
1.449814126
26.73
25.8
3.604651163
27.15
26.9
0.92936803
25.7
26.2
1.908396947
30.87
31.6
2.310126582
31.3
31.6
0.949367089 TABLE III
IR MLX90614 SURFACE TEMPERATURE ACCURACY TEST RESULTS MLX90614 SURFACE TEMPERATURE ACCURACY TEST RESULTS TABLE V
LOADCELL SENSOR ACCURACY TEST RESULTS
Loadcell (g)
Scales (g)
Error (%)
295.34
300
1.553333333
491.66
500
1.668
998.63
1000
0.137
1481.58
1500
1.228
1981.85
2000
0.9075
2474.59
2500
1.0164
3453.32
3500
1.333714286
3944.81
4000
1.37975
4931.35
5000
1.373
6401.71
6500
1.512153846 TABLE V
LOADCELL SENSOR ACCURACY TEST RESULTS Based on the test results of the temperature sensor in Table
III, it is known that there is a difference in the temperature
value of the MLX90614 IR reading compared to the value of
the Thermogun Infrared GM320 so that the deviation error
value is obtained. Furthermore, the error deviation value is
used as a reference to determine the average error value and
accuracy of the MLX90614 IR sensor through the following
calculations: Based on the load sensor test results in Table V, it is known
that there is a difference in the load value of the Loadcell
reading compared to the value on the Digital Scales so that the
error deviation value is obtained. Furthermore, the error
deviation value is used as a reference to determine the average
error value and accuracy of the Loadcell sensor through the
following calculations: B. Sensor Monitoring Results TABLE VI
MONITORING SENSOR DATA OF HABITAT 1
Time
Habitat 1
Sand
Temperature
(°C)
Temperat
ure
(°C)
∑
Temperature
(°C)
Humidity
(%)
06.00
22.52
23.81
23.165
76.71
07.00
24.04
24.21
24.125
75.61
08.00
25.48
25.71
25.595
72.51
09.00
25.26
25.01
25.135
74.31
10.00
25.06
25.21
25.135
71.11
11.00
25.48
27.71
25.595
68.51
12.00
28.18
27.11
27.645
67.11
13.00
29.2
30.11
29.655
65.71
14.00
30.1
30.61
30.355
64.71
15.00
28.12
28.71
28.415
69.41
16.00
26.94
27.01
26.975
70.21
17.00
25.74
26.11
25.925
73.51
18.00
24.22
24.61
24.415
71.01
19.00
23.78
24.11
23.945
73.21
20.00
24.9
24.31
24.605
74.51
21.00
24.58
23.91
24.245
74.91
22.00
24.72
23.31
24.015
77.01
23.00
24.66
23.71
24.185
76.71
24.00
24.52
24.71
24.615
75.91
01.00
23.94
24.31
24.125
76.11
02.00
24.22
24.81
24.515
80.11
03.00
23.98
24.11
24.045
74.51
04.00
24.4
24.11
24.255
77.81
05.00
24.18
24.01
24.095
76.81
Figure 6 Temperature Change Graph Journal of Telecommunication Network (Jurnal Jaringan Telekomunikasi) Vol. 13, No.1 (2023) Journal of Telecommunication Network (Jurnal Jaringan Telekomunikasi) Vol. 13, No.1 (2023) air temperature are 23.78°C and 24.11°C, respectively,
resulting in an average temperature of 23.945°C which is
<24°C. Based on the results of temperature monitoring, it is
known that the system can run according to the conditions that
have been set, namely the light actuator is active when the
temperature is <24°C and is disabled when >30°C. clearly the temperature changes in habitat 1, see the graph in
Fig. 6. B. Sensor Monitoring Results TABLE VI
MONITORING SENSOR DATA OF HABITAT 1 Time
Habitat 1
Sand
Temperature
(°C)
Temperat
ure
(°C)
∑
Temperature
(°C)
Humidity
(%)
06.00
22.52
23.81
23.165
76.71
07.00
24.04
24.21
24.125
75.61
08.00
25.48
25.71
25.595
72.51
09.00
25.26
25.01
25.135
74.31
10.00
25.06
25.21
25.135
71.11
11.00
25.48
27.71
25.595
68.51
12.00
28.18
27.11
27.645
67.11
13.00
29.2
30.11
29.655
65.71
14.00
30.1
30.61
30.355
64.71
15.00
28.12
28.71
28.415
69.41
16.00
26.94
27.01
26.975
70.21
17.00
25.74
26.11
25.925
73.51
18.00
24.22
24.61
24.415
71.01
19.00
23.78
24.11
23.945
73.21
20.00
24.9
24.31
24.605
74.51
21.00
24.58
23.91
24.245
74.91
22.00
24.72
23.31
24.015
77.01
23.00
24.66
23.71
24.185
76.71
24.00
24.52
24.71
24.615
75.91
01.00
23.94
24.31
24.125
76.11
02.00
24.22
24.81
24.515
80.11
03.00
23.98
24.11
24.045
74.51
04.00
24.4
24.11
24.255
77.81
05.00
24.18
24.01
24.095
76.81 Figure 7 Humidity Change Graph Figure 7 Humidity Change Graph Figure 7 Humidity Change Graph In the graph of changes in habitat humidity 1 shows the
humidity parameters for activating and deactivating the mist
maker-fan actuator. The system is activated at 06.00 with a
humidity reading of 76.71% so that the fan actuator and mist
maker are still in an inactive condition because the humidity
value is between 70-80%. Based on the graph, the mist maker-
fan starts to activate at 11.00 because the humidity value is
68.51% and remains active until 02.00 because the humidity
value has exceeded 80%, which is 80.11%. In accordance with
the results of the humidity parameter, you can turn on the mist
maker and fan according to the conditions set in the system,
namely the mist maker and fan turn on when the humidity is
<70% and turn off when <80%. Figure 6 Temperature Change Graph B. Sensor Monitoring Results The sensor monitoring results display the temperature and
humidity sensor values as well as the surface temperature for
24 hours to determine the active and inactive conditions of the
lamp and mistmaker-fan actuator. Based on Table VI, it is known that there is a change in the
value of air temperature, sand surface temperature and the
average value of both and the value of humidity after
monitoring for 24 hours. The results of this temperature
monitoring are used to determine the temperature value that
can activate or deactivate the lights in each habitat. Meanwhile,
the results of humidity monitoring are used as information on
whether the mistmaker-fan is on or off. The data displayed
shows changes in temperature and humidity values in habitat 1
with a relatively long time span. This is caused by the design
of the habitat that is not closed and the selection of
specifications for lamp actuators and mist makers that are not
in accordance with the shape of the habitat. To find out more Based on the results of the temperature sensor test in Table IV,
it is known that there is a difference in the temperature value of
the DHT22 reading compared to the value on the UNI-T333S
Hygrometer so that the deviation error value is obtained. Furthermore, the error deviation value is used as a reference to
determine the average error value and humidity accuracy of the
DHT22 sensor through the following calculations: 𝐸𝑟𝑟𝑜𝑟 𝐴𝑣𝑒𝑟𝑎𝑔𝑒= 24.13531784
10
𝑥 100% 𝐸𝑟𝑟𝑜𝑟 𝐴𝑣𝑒𝑟𝑎𝑔𝑒= 24.13531784
10
𝑥 100% E-ISSN: 2654-6531 P- ISSN: 2407-0807 28 Journal of Telecommunication Network (Jurna
clearly the temperature changes in habitat 1, see the graph in
Fig. 6. E. Data communication test results Distance
(m)
Mac Address
(Board 1 to Receiver)
IP Address
(Receiver to WiFi)
1
Delivery Success
Connected
5
Delivery Success
Connected
10
Delivery Success
Connected
20
Delivery Success
Failed Connection Lost
30
Delivery Success
Failed Connection Lost
50
Delivery Success
Failed Connection Lost
80
Delivery Success
Failed Connection Lost
90
Delivery Fail
Failed Connection Lost
100
Delivery Fail
Failed Connection Lost Figure 8 Loadcell Sensor Value Graph Fig. 8 shows the results of the graph of measuring the value
of the sand load before and after harvest in accordance with the
data listed in Table VII. Based on these data there are
differences in the value of the sand load for each test. The
average sand load read by the loadcell sensor in each test is
4684.26 grams with an average load reduction of 56.7 grams. The reduction in load is caused when the motor moves when
the harvesting mechanism is active causing vibrations, so that
there is sand spilled around the habitat. Based on Table VIII it can be seen the capacity of the
transmission distance between the two communication lines
through directional or point to point testing. In this system, data
transmission using IP Address communication lines is the
transmission line of the receiver's microcontroller to WiFi, the
maximum stable transmission distance is 10 meters. While the
data transmission system through the MAC Address
communication line is a habitat microcontroller to a receiver
microcontroller, the maximum stable transmission distance is
80 meters. In this case, the MAC Address communication path
on the ESP8266 has a longer transmission capability than the
IP Address so that it can support habitat conditions that are far
from internet access. D. Application Results The results of the application display the flutter application
menu along with system functionality that was previously
connected to the firebase database so that it is integrated into
the microcontroller system [12]. The Flutter application can be
accessed via a smartphone connected to the internet as shown
in Fig. 9. As can observed the dashboard menu displays sensor
data on habitat 1 which consists of the values of surface
temperature, air temperature and humidity as well as an on/off
button for the harvest mechanism. The value of each sensor
will change after ± 10 seconds so that farmers can monitor
changes in temperature and humidity without looking directly
at the location of the crab's habitat. Temperature data will be
collected in the “FORM” menu data log to display a graph of
fluctuations in temperature and humidity values. When the
“ON/OFF” button is pressed, the system will automatically run
the harvesting mechanism and send information to the database
to be forwarded to the microcontroller [13]. After the load
sensor detects the trigger, the information will be sent back to
the database and forwarded to the application as a trigger OFF
button. C. Harvest Mechanism Results The harvesting mechanism is activated when the snails
enter the ready-to-harvest age, which is ± 3 weeks. At harvest
time, the button from the application is pressed so that input 1
is sent to the firebase and read by the microcontroller to start
moving the motor, opening the lid of the habitat mat and the
sieving motor. Sand that is in the habitat will go down to the
side of the sifter when the bottom is open. The sifted sand will
be accommodated into a container equipped with a loadcell
sensor so that the sand load value will be detected as a trigger
to deactivate the base opening and closing motor and the
sieving motor. In this case, the weight of the sand after sieving
must match the weight of the sand before the sieving process
so that the weight of this sand can be used as a parameter for
making the trigger off. The results of the harvest mechanism
include Loadcell load sensor readings with indicators of
deactivation of motor actuators 1 and 2. Figure 6 Temperature Change Graph In the graph of changes in habitat temperature 1 displays
the surface temperature and air temperature parameters from
the IR sensor readings MLX90614 and DHT22. The two
parameters will determine the average temperature value that
functions to activate and deactivate the lamp actuator. The
system is activated at 06.00 with the readings of the surface
temperature of the sand and the air temperature are 22.52°C
and 23.81°C respectively resulting in an average temperature
of 23.165°C which will activate the lamp actuator because the
average temperature value is <24°C. The lamp will be kept
active until 14.00 because after that time there is an increase in
temperature >30°C where the surface temperature of the sand
and the detected air temperature are 30.1°C and 30.61°C,
respectively, resulting in an average temperature of 30.355°C. Furthermore, the lamp actuator will turn on again at 19.00
because the surface temperature of the sand and the detected TABLE VII
HARVEST MECHANISM TEST
Time
Button Status
Loadcell (gram)
Motor
20.47.15
OFF
-2,65
Inactive
20.47.46
ON
4747,10
Active
15.51.34
OFF
-5,71
Inactive
15.52.10
ON
4671,98
Active
16.07.25
OFF
- 2,13
Inactive
16.07.59
ON
4633,7
Active TABLE VII
HARVEST MECHANISM TEST E-ISSN: 2654-6531 P- ISSN: 2407-0807 29 Jaringan Telekomunikasi) Vol. 13, No.1 (2023)
Figure 9 Dashboard menu
E. C. Harvest Mechanism Results Data communication test results
TABLE VIII
TESTING OF TRANSMISTION DISTANCE MAC ADDDRESS AND IP ADDRESS Journal of Telecommunication Network (Jurnal Jaringan Telekomunikasi) Vol. 13, No.1 (2023) Figure 9 Dashboard menu Table VII shows the test data of the harvest mechanism
with load and time parameters. In the first test with a total load
of 4747.10 grams consisting of 500 grams of container load and
4247.10 grams of sand load, the working mechanism time is 31
seconds. In the second test, the total load data was 4671.98
grams consisting of 500 grams of container load with 4171.98
grams of sand so that the mechanism working time was 26
seconds. In the third test, data obtained for a total load of
4633.70 grams consisting of 500 grams of container load with
4133.70 grams of sand load with a mechanism working time of
34 seconds. Based on the test results, it is known that the
average value of harvest time is 30.37 seconds with an average
decrease in sand load of 56.7 grams. Figure 8 Loadcell Sensor Value Graph Figure 9 Dashboard menu E. Data communication test results E. Data communication test results
TABLE VIII
TESTING OF TRANSMISTION DISTANCE MAC ADDDRESS AND IP ADDRESS 𝑎𝑣𝑒𝑟𝑎𝑔𝑒 𝑑𝑒𝑙𝑎𝑦= 𝑡𝑜𝑡𝑎𝑙 𝑑𝑒𝑙𝑎𝑦
𝑡𝑜𝑡𝑎𝑙 𝑝𝑎𝑐𝑘𝑒𝑡 Journal of Telecommunication Network (Jurnal Jaringan Telekomunikasi) Vol. 13, No.1 (2023) Journal of Telecommunication Network (Jurnal Jaringan Telekomunikasi) Vol. 13, No.1 (2023) 𝑎𝑣𝑒𝑟𝑎𝑔𝑒 𝑑𝑒𝑙𝑎𝑦= 4556,41677
21
= 216,9722271 𝑚𝑠
[6] W. Puspitasari and H. Y
Monitoring and Automat 𝑎𝑣𝑒𝑟𝑎𝑔𝑒 𝑑𝑒𝑙𝑎𝑦= 4556,41677
21
= 216,9722271 𝑚𝑠 𝑎𝑣𝑒𝑟𝑎𝑔𝑒 𝑑𝑒𝑙𝑎𝑦= 4556,41677
21
= 216,9722271 𝑚𝑠 [6] W. Puspitasari and H. Y. Perdana R, "Real-Time
Monitoring and Automated Control of Greenhouse
Using
Wireless
Sensor
Network:
Design
and
Implementation," 2018 International Seminar on
Research of Information Technology and Intelligent
Systems
(ISRITI),
2018,
pp. 362-366,
doi:
10.1109/ISRITI.2018.8864377. • Packet loss Packet loss is the number of packets lost during the
transmission process [14]. 𝑝𝑎𝑐𝑘𝑒𝑡 𝑙𝑜𝑠𝑠= ( 𝑝𝑎𝑐𝑘𝑒𝑡 𝑠𝑒𝑛𝑡−𝑝𝑎𝑐𝑘𝑒𝑡 𝑟𝑒𝑐𝑒𝑖𝑣𝑒
𝑝𝑎𝑐𝑘𝑒𝑡 𝑟𝑒𝑐𝑒𝑖𝑣𝑒
) 𝑥 100
𝑝𝑎𝑐𝑘𝑒𝑡 𝑙𝑜𝑠𝑠= H 302 −301
301
I 𝑥 100 = 0.33% 𝑝𝑎𝑐𝑘𝑒𝑡 𝑙𝑜𝑠𝑠= ( 𝑝𝑎𝑐𝑘𝑒𝑡 𝑠𝑒𝑛𝑡−𝑝𝑎𝑐𝑘𝑒𝑡 𝑟𝑒𝑐𝑒𝑖𝑣𝑒
𝑝𝑎𝑐𝑘𝑒𝑡 𝑟𝑒𝑐𝑒𝑖𝑣𝑒
) 𝑥 100
𝑝𝑎𝑐𝑘𝑒𝑡 𝑙𝑜𝑠𝑠= H 302 −301
301
I 𝑥 100 = 0.33% [7] M. Arundina, S. Ngabekti and K. Santoso, "Kesintasan
Undur-Undur pada Media Aklimatisasi," Biosaintifika:
Journal of Biology & Biology Education, vol. 6, no. 1,
pp. 18-23, 2014. Based on testing using the Wireshark application for 7 days
of testing, the average daily delay is 216,9722271 ms which is
included in the good category in the TIPHON standard. Meanwhile, in the packet loss test, the average daily packet loss
is 0.33% which is included in the very good category in the
TIPHON standard[15]. [8] K. Miler, B. E. Yahya and M. Czarnoleski, "Substrate
moisture, particle size and temperature preferences of
trap‐building larvae of sympatric antlions and wormlions
from the rainforest of Borneo," Ecological Entomology,
vol. 44, no. 4, pp. 488-493, 2019. [9] F. Puspasari, T. P. Satya, U. Y. Oktiawati, I. Fahrurrozi
and H. Prisyanti, "Analisis Akurasi Sistem Sensor
DHT22 berbasis Arduino terhadap Thermohygrometer
Standar," JURNAL FISIKA DAN APLIKASINYA, vol. 16, no. 1, pp. 40-45, 2020. IV. CONCLUSION Sensors of air temperature, surface temperature and
humidity with an accuracy of 98.08%, 98.38%, 97.58%,
respectively, can control the temperature range of 24-30℃ and
humidity range of 70-80 quite stable. The loadcell sensor with
an accuracy of 98.78% can regulate the inactivity of the sieving
motor when the sand load value is >4000 grams. The average
load reading is 4684.26 grams with an average time span of
30.37 seconds with a reduction in the sand load per harvest by
an average of 56.7 grams. ESP8266 microcontroller data
communication MAC Address transmission line has a range of
up to 80 m while the IP Address is only 10 m. The average
daily delay obtained is 216.9722271ms in the good category
with an average daily packet loss of 0.33% in the very good
category. [10] I. Inayah, "Analisis Akurasi Sistem Sensor IR
MLX90614 dan Sensor Ultrasonik berbasis Arduino
terhadap Termometer Standar," Jurnal Fisika Unand
(JFU), vol. 10, no. 4, pp. 428-434, 2021. [11] F. Y. Saputra, M. S. Al Amin and Perawati, "Alat
Pengukur Tinggi Badan, Berat Badan, Dan Suhu Badan
Digital Menggunakan Sensor Ultrasonik, Load Cell, Dan
Inframerah Mlx90614," Jurnal Tekno, vol. 19, no. 1, pp. 60-67, 2022. [12] R. Juliana, K. N. VG, R. G and S. P, "Evecurate – A
Smart Event Management App Using Flutter and
Firebase," International Journal of Scientific Research &
Engineering Trends, vol. 7, no. 4, pp. 2519-2524, 2021. • Delay Delay is the transmission time of packets from the
microcontroller to the database and vice versa [14]. The
calculation of the delay value uses the formula for the
difference between the time the packet is received and the time
the packet is sent according to the formula: 𝑑𝑒𝑙𝑎𝑦= (𝑡𝑖𝑚𝑒 𝑟𝑒𝑐𝑒𝑖𝑣𝑒𝑑−𝑡𝑖𝑚𝑒 𝑠𝑒𝑛𝑡) 𝑑𝑒𝑙𝑎𝑦= (𝑡𝑖𝑚𝑒 𝑟𝑒𝑐𝑒𝑖𝑣𝑒𝑑−𝑡𝑖𝑚𝑒 𝑠𝑒𝑛𝑡) 𝑎𝑣𝑒𝑟𝑎𝑔𝑒 𝑑𝑒𝑙𝑎𝑦= 𝑡𝑜𝑡𝑎𝑙 𝑑𝑒𝑙𝑎𝑦
𝑡𝑜𝑡𝑎𝑙 𝑝𝑎𝑐𝑘𝑒𝑡 E-ISSN: 2654-6531 P- ISSN: 2407-0807 30 REFERENCES [13] R. H. Y. Perdana, H. Hudiono and A. F. N. Luqmani,
"Water Leak Detection and Shut-Off System on Water
Distribution Pipe Network Using Wireless Sensor
Network," 2019 International Conference on Advanced
Mechatronics, Intelligent Manufacture and Industrial
Automation (ICAMIMIA), 2019, pp. 297-301, doi:
10.1109/ICAMIMIA47173.2019.9223386. [1] D. H. Susanto, R. W. Kartika, P. H. Heng, A. W. Santoso,
V. M. Lopulalan and A. Wijaya, "Pengaruh Ekstrak
Undur-undur (Myrmeleon sp) terhadap Glukosa Darah
dan Hematokrit pada Tikus Diabetes," Jurnal Ilmiah
Kedokteran Wijaya Kusuma, vol. 9, no. 2, p. 209=2017,
2020. [2] V. Widyawati, Ampuhnya Lintah dan Undur-undur
Tebas Beragam Penyakit, Yogyakarta: Laksana, 2019. [14] P. R. Utami, "Analisis Perbandingan Quality Of Service
Jaringan Internet Berbasis Wireless Pada Layanan
Internet Service Provider (ISP) Indihome dan Fist
Media," Jurnal Ilmiah Teknologi dan Rekayasa, vol. 25,
no. 2, pp. 125-137, 2020. [3] Z. Abiddin, Interviewee, Yusron Herbamart Wuluhan,
Kabupaten Jember. [Interview]. 25 05 2022. [4] A. Wiranto and H. Nurwarsito, "Sistem Monitoring
Pengatur Suhu dan Kelembaban pada Kandang Jangkrik
berbasis Internet of Things (Studi Kasus Budidaya
Jangkrik Perorangan di Kabupaten Blitar)," Jurnal
Pengembangan
Teknologi
Informasi
dan
Ilmu
Komputer, vol. 6, pp. 2673-2680, 2022. [15] Subektiningsih, Renaldi and P. Ferdiansyah, "Analisis
Perbandingan Parameter QoS Standar TIPHON Pada
Jaringan Nirkabel Dalam Penerapan Metode PCQ,"
Explore, vol. 1, no. 57-63, p. 12, 2022. [5] M. Babiuch, P. Foltýnek and P. Smutný, "Using the
ESP32
Microcontroller
for
Data
Processing,"
International Carpathian Control Conference (ICCC)
IEEE, pp. 1-6, 2019. E-ISSN: 2654-6531 P- ISSN: 2407-0807 31
|
https://openalex.org/W2335592662
|
https://zenodo.org/records/2518509/files/article.pdf
|
English
| null |
The Vegetation of West Greenland
|
Journal of ecology
| 1,922
|
public-domain
| 13,858
|
BY R. E. HOLTTUM. BY R. E. HOLTTUM. (With Plates III-V.)
(1) INTRODUCTORY. BY R. E. HOLTTUM. The first description of the vegetation of west Greenland is that contained
in Hans Egede's Perlustration
(4), illustrated by figures of plants collected by
his son Poul Egede, whose herbarium is the earliest existing collection of arctic
plants. Some of the missionary successors of Egede made further
collections
and Rottboll in 1770 published an account of the flora based on their material
The geologist Giesecke made extensive explorations of the west coast during
the years 1806-13, and published a list of plants. J. Vahl visited the whole
coast from south to north during the period 1828-36 and made large collections
of the flora. Some of his work in connection with these has never been pub-
lished, but his material was used by Lange in the preparation of the Conspectu
Florae Groenlandicae
(12), which appeared in 1880, and with its supplements
is the most complete list of Greenland plants. Surgeons on board whaling
vessels and on the various expeditions in search of Franklin collected plants
on the coasts they visited, and lists of these were published by Hooker and
others', but the first general descriptions of the vegetation were made by
Robert Brown (2), who visited various parts of the west coast with Whymper
in 1867, by Berggren (1) (Disko Bay, 1871) and by Hart (5), who accompanied
the Discovery in 1875-6. In 1888 appeared Warming's paper (27), the first
ecological study and still the most complete account of the vegetation of the
country as a whole. This was supplemented as regards the southern districts
by Rosenvinge (25) in 1898, and as regards the Disko neighbourhood by
Porsild (18) in 1902 and in the later papers issued from the Danish Arctic
Station at Godhavn after its foundation by him in 1906. Other papers of eco-
logical interest during this period are by Hartz (accounts of voyages during
1889 to parts of the coast south of Holsteinsborg,
and during 1890 to Disko Bay
and the Waigat), Kruuse (Egedesminde district,
1897) and Vanh6ffen
(Drygalski
Expedition of 1891-3, Umanak fjord). Ostenfeld has written a short account
of the plant life of the extreme north, based on the diary of Wulff,
the Swedish
botanist who died on Rasmussen's second Thule expedition, 1916. There are
also annotated lists of plants collected on various expeditions, but these are
not primarily
of ecological interest2. 1 For complete list of botanical literature up to 1880 see Lauridsen (14).
2 References to a number of these will be found in the various papers by Porsild cited in
the bibliography at the end of this paper. This content downloaded from 128.235.251.160 on Tue, 13 Jan 2015 21:58:25 PM
All use subject to JSTOR Terms and Conditions The Vegetation of West Greenland
Author(s): R. E. Holttum
Source: Journal of Ecology, Vol. 10, No. 1 (May, 1922), pp. 87-108
Published by: British Ecological Society
Stable URL: http://www.jstor.org/stable/2255432 . Accessed: 13/01/2015 21:58 Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp . JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org. British Ecological Society is collaborating with JSTOR to digitize, preserve and extend access to Journal of
Ecology. This content downloaded from 128.235.251.160 on Tue, 13 Jan 2015 21:58:25 PM
All use subject to JSTOR Terms and Conditions 87 THE VEGETATION OF WEST GREENLAND THE VEGETATION OF WEST GREENLAND BY R. E. HOLTTUM. A series of papers entitled " The structure
and biology of arctic flowering
plants" has appeared in the Meddelelser om
GrHnland,
the earliest being Warming's account of the morphology
and biology
of the Ericaceae; other groups dealt with are Saxifragaceae, fresh water
1
L
d
(14) This content downloaded from 128.235.251.160 on Tue, 13 Jan 2015 21:58:25 PM
All use subject to JSTOR Terms and Conditions The Vegetationt
of West Greenland 88 aquatics, Ranunculaceae, Pinguicula, Rosaceae, Cornaceae, Caprifoliaceae
Primulaceae, Caryophyllaceae and Liliaceae. A complete list will be found in
the latest of the series (22). Porsild has in course of publication a list of the
plants growing on and near Disko, with notes on their distribution,
habitat,
seed production and hibernation. He has also numerous important.observa
tions on the occurrence of seedlings. g
The questions of distribution of species and of the origin of the present
flora have been dealt with by several writers. The first important contribution
was by Hooker in 1860. In 1881 Lange criticised Hooker's statistics and con-
clusions, and during the course of the next decade appeared several papers
by Nathorst and Warming (15, 16, 27). Further facts, both floristic and
geological, have since come to light, and Porsild (23) has summarised the
present knowledge of distribution
of species, and the probable conclusions to
be drawn as to the origin of the flora. Geological evidence shows that probably
the whole of Greenland was covered with ice during the glacial period, and
that accordingly the flora must have been completely destroyed; the present
vegetation is therefore due to post-glacial invasion. The nearest land from
which such vegetation could have come is to the north-west,
across Smith
Sound, and it is suggested that all western and circumpolar arctic species
came this way. This accounts for about 36 per cent. of the total number of
species, but a difficulty
arises with regard to the 60 per cent. of southern types. Nathorst supposed that a land bridge existed in post-glacial times connecting
Europe by way of the Faeroes and Iceland with east Greenland, but from a
geological standpoint this appears doubtful, and the number of species oc-
curring commonly in Iceland which are absent in east Greenland is evidence
against it. This content downloaded from 128.235.251.160 on Tue, 13 Jan 2015 21:58:25 PM
All use subject to JSTOR Terms and Conditions (2) TOPOGRAPHY. With the exception of a strip of land of varying width round the coast, the
whole of Greenland is covered with a continuous ice-field. The exposed coastal
region consists mainly of irregular
mountainous land, intersected
by numerous
fjords, at the heads of many of which large iceberg-forming
glaciers reach the
sea. Most of this country consists of gneissic rock, and the lower parts of it,
near the shore, are typically formed of low ice-worn hummocks, with vege-
tation only continuous in the hollows or on old moraines. y
Between the latitudes of 690 and 71? N. there is a change in the character
of the country. Here there are considerable layers of Cretaceous and Tertiary
sedimentary rocks (sandstones and shales, with thin coal seams, containing
the celebrated fossil flora of this part of Greenland) and above them thick
layers of tuff and basalt, bringing the general height of the land up to 3000-
5000 ft. above the sea. These newer rocks weather quite rapidly, and the
appearance of the country is quite different
from that to the north and south
of it. Typically a fairly gentle and even slope rises from the shore, covered
with vegetation at least to the height of the basalt, which then rises in steep
terraces, above a talus of boulders. This shore slope, is broken up by steep
narrow valleys, cut in it by streams formed by melting of snows-above, or
from small glaciers which do not reach the coast. These streams are parti-
cularly vi6lent in spring, and bring down much sediment which is deposited
in fan-shaped deltas, some of considerable size. The whole of the smaller
deltas, and the central regions of-the larger ones, are almost entirely flooded
in the spring, but during the summer their boulder-strewn
surface is baked
dry, except where it is intersected by the stream courses. There is a sandy
beach on the shore almost everywhere in this region, sometimes with low
dunes, and behind the, dune banks small lagoons and salt marshes may occur. The soil on the gneiss rocks, except where there are old moraines, consists
very largely of raw humus formed of remains of the previous vegetation which
has formed a mat on the surface of the rock itself. On the more recent rocks
there is usually a fairly considerable depth of loose soil, but the soil from
weathered tuff is heavy and retentive. (2) TOPOGRAPHY. None of the rocks represented
contains
any quantity of calcium carbonate; nitrogen is also deficient, so that there
must be some organic agent operative in the fixation of atmospheric nitrogen. Doubtless nitrogen-fixing
bacteria are present in the soil, but this matter has
not been investigated Between the latitudes of 690 and 71? N. there is a change in the character
of the country. Here there are considerable layers of Cretaceous and Tertiary
sedimentary rocks (sandstones and shales, with thin coal seams, containing
the celebrated fossil flora of this part of Greenland) and above them thick
layers of tuff and basalt, bringing the general height of the land up to 3000-
5000 ft. above the sea. These newer rocks weather quite rapidly, and the
appearance of the country is quite different
from that to the north and south
of it. Typically a fairly gentle and even slope rises from the shore, covered
with vegetation at least to the height of the basalt, which then rises in steep
terraces, above a talus of boulders. This shore slope, is broken up by steep
narrow valleys, cut in it by streams formed by melting of snows-above, or
from small glaciers which do not reach the coast. These streams are parti-
cularly vi6lent in spring, and bring down much sediment which is deposited
in fan-shaped deltas, some of considerable size. The whole of the smaller
deltas, and the central regions of-the larger ones, are almost entirely flooded
in the spring, but during the summer their boulder-strewn
surface is baked
dry, except where it is intersected by the stream courses. There is a sandy
beach on the shore almost everywhere in this region, sometimes with low
dunes, and behind the, dune banks small lagoons and salt marshes may occur. g
The soil on the gneiss rocks, except where there are old moraines, consists
very largely of raw humus formed of remains of the previous vegetation which
has formed a mat on the surface of the rock itself. On the more recent rocks
there is usually a fairly considerable depth of loose soil, but the soil from
weathered tuff is heavy and retentive. BY R. E. HOLTTUM. Some of the southern plants were undoubtedly introduced by the
Icelandic colonists of the tenth century, but most of them must have been
carried over the sea by natural means. The total number of species and
varieties of vascular plants in the whole of Greenland is 416. This is much
smaller than that of other arctic lands of similar extent, no doubt on account
of the absence of land connection to the south. The following account attempts to summarise the present knowledge of
the vegetation of west Greenland from an ecological standpoint; the writer's
own observations of particular pieces of vegetation have been included in
some cases. He had the privilege of visiting the Danish Arctic Station at
Godhavn on Disko Island in company with Prof. A. C. Seward during the
summer of 1921. The primary object of the expedition was to collect fossil
plants from the well-known localities on Disko and parts of the coast north
of it. Calls made at various places during motor-boat voyages of over 700
miles furnished opportunities for making notes on the vegetation as well as
for collecting herbarium material. The writer is very greatly indebted to the
Director of the Arctic Station, Mr Morten P. Porsild, for much help, both This content downloaded from 128.235.251.160 on Tue, 13 Jan 2015 21:58:25 PM
All use subject to JSTOR Terms and Conditions R. -E. HOLTTUM 89 in identifying
plants collected and in placing his wide knowledge of local
conditions at his visitors' disposal, and also for his very kind hospitality during
two pleasant months. The assistance of A. Erling Porsild on the motor-boat
voyages was also invaluable, both as guide and interpreter
and on account
of his thorough knowledge of the flora. The writer wishes also to express
his very cordial thanks to Prof. Seward for the opportunity
of accompanying
him on the expedition. This content downloaded from 128.235.251.160 on Tue, 13 Jan 2015 21:58:25 PM
All use subject to JSTOR Terms and Conditions (3) CLIMATE. (3) CLIMATE. The mean annual air temperature varies from about 10 C. in the extreme
south of Greenland to about - 80 C. at Upernivik (730 N.), the northernmos
Danish colony. The mean temperatures
for July at the same localities are about
10? and 50 respectively. The total rainfall at Ivigtut, in the extreme south,
is about 130 cm., and at Jakobshavn in Disko Bay only 22 cm. During the
summer there is frequently
low sea fog round the more outlying parts of the
coast, and this has a great influence on the climate of, for instance, the small
islands of the Egedesminde district (11) in the south of Disko Bay. At Cape
Dalton (on the east coast) Hartz and Kruuse (7, p. 343) record that frequent
low sea fog greatly retards the growth of plants near the shore as compared
with those higher up; a marked difference
of temperature was noted, 2? C. at sea level as against 100 at an altitude of a few hundred feet. The differenc
between coast and inland localities as regards climate is illustrated by the
statistics for Godthaab (on coast) and Kornok (inland), both about lat. 640 N. (25, pp. 94-5). The annual rainfall at the former
is 71-4 cm. and at the latter
36-5 cm., while the numbers of days of fog in the year are respectively 57
and 12. Where clear sunny weather prevails during the summer the surface
of dry soil and rocks is raised to 400 C. or more. A black bulb thermometer
in the sun at noon on July 26, 1897, on the south of Disko Bay registered
40.50, the air temperature at the same time being 14.50 (11, p. 367). Porsild
records 480 with black bulb on a "black streak" of Cyanophyceae on gneiss
in Diskofjord (18, p. 210). During the winter snow covers most of the ground, but near the coast
the strong winds leave only a small covering on the exposed shore slopes; in
the valleys it collects and may persist till late in the summer if not exposed
to the direct heat of the sun. The cold dry winds prevent any plants growing
up above the level of the snow surface. (3) CLIMATE. In places well sheltered from the
wind and with a southern exposure, where thick snow collects early, melts
early and is never packed hard, there is a greater protection for the vegetation
in the winter and at the same time the best summer conditions. Here a thick
willow scrub can develop, or a vegetation of herbaceous plants which cannot
stand the more severe conditions of exposed situations. Further inland,
especially on the coasts of some of the fjords, there is greater protection from
wind, and a thick even layer of snow can accumulate over larger areas. In
such situations the greatest development of willow scrub occurs. This content downloaded from 128.235.251.160 on Tue, 13 Jan 2015 21:58:25 PM
All use subject to JSTOR Terms and Conditions The Vegetation
of West Greenland The Vegetation
of West Greenland 90 (2) TOPOGRAPHY. None of the rocks represented
contains
any quantity of calcium carbonate; nitrogen is also deficient, so that there
must be some organic agent operative in the fixation of atmospheric nitrogen
Doubtless nitrogen-fixing
bacteria are present in the soil, but this matter has
not been investigated. This content downloaded from 128.235.251.160 on Tue, 13 Jan 2015 21:58:25 PM
All use subject to JSTOR Terms and Conditions R. E. HOLTTUM R. E. HOLTTUM R. E. HOLTTUM 91 (iii) Heath (Lynghede) (iii) Heath (Lynghede). g
(iv) "Fjaeldmark''";
permanent open vegetation g
(v) Vegetation of fresh water marshes and pools. (v) Vegetation of fresh water marshes and pools. (vi) Shore vegetation. (vi) Shore vegetation. (vii) Nitrophilous vegetation round settlements (vii) Nitrophilous vegetation round settlements. This classification has been adopted, with some additions, by Porsild. It
appears to the present writer, regarding the vegetation from the successional
standpoint, that heath represents the climatic climax over the greater part
of the country, with the probable exception of the birch region in the south,
and of the extreme north where closed vegetation is infrequent. According
to this view the willow scrub must be regarded as a post-climax due to favour-
able local edaphic and topographic conditions, and the "Fjaeldmark" as a
stabilised pre-climax due to unfavourable conditions. A doubtful case is that
of moss bog, which appears to be persistent for a considerable period at least
under certain circumstances; it is not clear whether this will necessarily pass
over into heath, and it will therefore
be described separately. The same remark
applies to the aquatic vegetation of ponds. Salt marsh may possibly in some
cases pass over by silting up into heath, directly or through the stage of fresh
water marsh, but it is not certain that this is the general rule. The highly
nitrogenous soil round settlements bears a characteristic vegetation, chiefly
of grasses, which does not lead to heath. With these exceptions, all the vege-
tation may be regarded as representing
stages in the heath succession. Suc-
cession is naturally very slow, especially in very dry or exposed situations, and
in many cases it is not possible to say whether the heath stage will be reached. The following scheme embodies the above suggestions: The following scheme embodies the above suggestions: A. North of 620 N. Lat. (limit of Betula odorata). (a) Climax: Heath, arising as a result of successions on 1. bare gneiss rock: (i) xerosere (ii) hydrosere; g
(ii) hydrosere; 2. talus slopes of recent rocks; 3. areas of silt brought down by streams. Here the first
stage may be either open vegetation on dry ground or
fresh water marsh vegetation; 4. raised beaches; 5. dunes on sandy shores; 6. moraines. 6. moraines (b) Stabilised pre-climaxes:
Fjaeldmark and (?) Moss bog. 1 The term "fell-field" used by the translators of Warming's (Ecology of Plants (1909) is not
a very suitable English term. This content downloaded from 128.235.251.160 on Tue, 13 Jan 2015 21:58:25 PM
All use subject to JSTOR Terms and Conditions (4) CLASSIFICATION OF VEGETATION TYPES. Warming classified the vegetation of Greenland under the following
headings: Warming classified the vegetation of Greenland under the following
headings: g
(i) The birch scrub of south Greenland (ii) Willow scrub (Pilekrat) and associated herbaceous vegetation (Urte. mark). (ii) Willow scrub (Pilekrat) and associated herbaceous vegetation (Urte. mark). This content downloaded from 128.235.251.160 on Tue, 13 Jan 2015 21:58:25 PM
All use subject to JSTOR Terms and Conditions A. NORTH OF THE REGION OF BETULA ODORATA. (.a) Heath and phases of the heath succession. Heath is the typical vege-
tation over the greater part of the country, varying little in character even
in the extreme south. On Disko it extends to an altitude of over 2000 ft. It
consists essentially of low woody plants, mainly ericaceous, of which the most
important are Empetrum nigrum and Cassiope tetragona,
each being locally
dominant and of about the same range of distribution,
except that Cassiope
occurs at higher altitudes than Empetrum. Both occur on all types of soil. Perhaps the most universally present plant, occasionally also dominant, is
Vaccinium uliginosum v. microphyllum
Lange; Salix glauca also occurs almost
everywhere,
but is never dominant in the heath. Betula nana is frequently
present, and in the drier-
parts may be the dominant phanerogam, the rest of
the vegetation consisting of a carpet of lichens (species of Cladonia, Cetraria,
Alectoria, etc.), but it is rather local in its occurrence without apparent con-
nection with -changes in the soil or other conditions. South of 630 N. Lat. Betula glandulosa replaces B. nana in the heath. Associated ericaceous plants
which are more or less abundant are Phyllodoce caerulea, Ledum palustre v. decumbens Ait., Rhododendron
lapponicum and Loiseleuria procumbens. The
last two plants are rarely met with on the sandstone, though,
both are common
on the gneiss. In many places the vegetation is very mixed, and it is not
possible to name any one plant as the dominant. This is certainly the case
in early stages of the succession, when the ground is not completely covered,
but it is not clear whether any one plant has sufficient
advantage over the
others finally to become dominant in all cases. Certainly one does find almost
pure patches of Cassiope, sometimes in thick tufts, or of trailing matted stems
of Empetrum, but these are usually in rather favourable situations. In the
south Rosenvinge (25, p. 193) records Empetrum as dominant in the heath
near the coast where the climate is moist. Farther inland, where it is drier,
Betula and Vaccinium are dominant; this is no doubt connected with their
deciduous habit. The'associated herbaceous plants differ also, and mosses
tend to disappear. Except on the sandstone, mosses are constantly present
in large numbers and variety of species among the heath plants near the
coast. R. E. HOLTTUM (c) Post-climax: Willow scrub, with associated herbaceous vegetation
as a stage of development. as a stage of development g
(d) Special edaphic types: Nitrophilous vegetation round settlement and Halophilous vegetation of salt marsh. 1 The term "fell-field" used by the translators of Warming's (Ecology of Plants (1909) is not
a very suitable English term. This content downloaded from 128.235.251.160 on Tue, 13 Jan 2015 21:58:25 PM
All use subject to JSTOR Terms and Conditions The Vegetation
of West OXreenlan 92 B. South of 620 N. Lat. (region of Betula odorata)
Climax: Birch-willow
scrub. Stabilised pre-climaxes:
Heath and Grassland. This content downloaded from 128.235.251.160 on Tue, 13 Jan 2015 21:58:25 PM
All use subject to JSTOR Terms and Conditions A. NORTH OF THE REGION OF BETULA ODORATA. On the gneiss the heath occurs as an irregular carpet covering the rock
surface, except in the steepest and most uneven places, where the rock is
exposed; here it is covered almost entirely by species of Gyrophora and
crustaceous lichens, typically black and giving a characteristic
appearance to
the whole landscape. On the more recent rocks of Disko and the Nugsuak
peninsula heath covers the greater part of the permanent shore slopes, though JOURNAL OF ECOLOGY
VOL. X, PLATE III
_~~~~~~~~~~~~~~~~~~~~~~~~~
%0
Ka.r
_C.,._z
't
el
AQ,|_
B
w t.w]
1 E A;~4
o
O!,~~~~~~~~~~~~~~~~~~~~~~~~~u
,A--
i
;'
F~~~~~~~~~~A
g
4;S
=
i
t~~~~~~~~~~~~~~~~
t
S
i
s
- *t,
R"
! _
*; N t~~~~~
vf
;#
*
*
>
s
_i_
r
btf 3~~~~~~~
t *"w
^' s- ld'^ Ms~~~~~~~~~~
s
? s
l X
- t~~~~~~~~
)
l
[ ~~~
~~~
~~~
~~~~~~~
c
idE
ts
'
<
_
z U Z~~~~
HOLTTUM
THE VEGETATION
OF WEST GREENLAND.~~~~~~~~~~M VOL. X, PLATE III JOURNAL OF ECOLOGY TE III
B
o
;'
*; N t~~~
vf
;#
*
*
>
s
_i_
r
btf 3~~~~~~
s~~~~~~~~
s
? s
l X
- t~~~~~~
)
l
[ ~~~
~~~
~~~
~~~~~
c
idE
ts
z U Z~~
M B
o
;' _~~~~~~~~~~~~~~~~~~~~~~~~~
%0
Ka.r
_C.,._z
't
el
AQ,|_
B
w t.w]
1 E A;~4
o
O!,~~~~~~~~~~~~~~~~~~~~~~~~~u
,A--
i
;'
F~~~~~~~~~~A *; N t~~~~
vf
;#
*
*
>
s
_i_
r
btf 3~~~~~~~
s~~~~~~~~~
s
? s
l X
- t~~~~~~~
)
l
[ ~~~
~~~
~~~
~~~~~~
c
idE
ts
z U Z~~~~
M g
4;S
=
i
t
t
S
i
s
- *t,
R"
! _
*; N t~~~~~~
vf
;#
*
*
>
s
_i_
r
btf 3~~~~~~~~
t *"w
^' s- ld'^ Ms~~~~~~~~~~~
s
? s
l X
- t~~~~~~~~~
)
l
[ ~~~
~~~
~~~
~~~~~~~~
c
idE
ts
'
<
_
z U Z~~~~~
HOLTTUM
THE VEGETATION
OF WEST GREENLAND.~~~~~~~~~~M This content downloaded from 128.235.251.160 on Tue, 13 Jan 2015 21:58:25 PM
All use subject to JSTOR Terms and Conditions R. E. HOLTTUM 93 often it is not quite closed, especially at higher altitudes. At Patoot on Aug. 8,
1921, between the heights of 2200 and 2500 feet, A. E. A. NORTH OF THE REGION OF BETULA ODORATA. Porsild and the present
writer observed 67 species of vascular plants, practically all the normal heath
species being represented, though the normal heath dominants were very
restricted,
especially Empetrum. The vegetation was very open, Dryas integri-
folia being among the most abundant species, and forming
large mats which
faced steeply towards the sun. All the plants contained a greater amount of
anthocyan pigment than was normally present at that date in plants at lower
altitudes. Where a small stream flowed down the slope there was a much
more luxuriant vegetation, with a good deal of moss, and it seems probable
that dryness of soil and strong insolation were the factors limiting plant growth,
rather than cold. On the northern
shore of the Nugsuak peninsula, at Ekorgfat
and elsewhere, the snow melts more slowly and insolation is less intense, and
the heath vegetation on the sandstone was more luxuriant, Dryas, for example,
having much broader leaves than usual, and a greater vegetative development. g
g
g
1. Succession on gneiss rock: Xerosere. As above mentioned, the surface
of the rock is usually covered with crustaceous lichens which quite obscure
its natural colour, except close to the sea where the ice-foot forms in winter. The most typical of these lichens are species of Xanthoria and Gyrophora
Starting in small crevices fruticose lichens and xerophytic mosses begin to
form cushions which spread over the surface. Very numerous species may be
present at this stage. One may find tangled mats of Cetraria islandica, Cla-
donia sylvatica, Alectoria jubata, and other lichens of similar habit, or close
Racomitrium cushions fringed by the above lichens, or crusts of Lecanora
tartarea, while Cladonia gracilis is often found climbing up quite steep rock
faces. When these mats are well-established,
higher plants appear in them,
Lycopodium
selago, Pyrola grandiflora
Rad., Vaccinium uliginosum often being
among the first. In shaded places where snow lies late Salix herbacea is very
frequent but may also occur in more exposed situations; it is often accom-
panied by Cassiope hypnoides. In some places small Carices are numerous,
among the commonest being C. rigida, C. alpina and C. scirpioidea. In rather
moist hollows, not wet enough for bog mosses, thick mats of Aulacomnium
turgidum, almost pure or with other mosses, are found. In rock crevices
cushion plants often appear before any other phanerogams are 'present. This content downloaded from 128.235.251.160 on Tue, 13 Jan 2015 21:58:25 PM
All use subject to JSTOR Terms and Conditions A. NORTH OF THE REGION OF BETULA ODORATA. misandra on small tussocks above the water level. Following this the
Eriophorum zone (E. polystachium
and E. scheuchzeri) extends into water
50 cm. deep. With the Eriophorum grow sterile shoots of Chamaenerium
latifolium. Finally comes Hippuris, with submerged mosses. In rock hollows
which have not ponds in them in which Hippuris can grow, Eriophorum
(either
species or both together) almost invariably occurs in shallow water or among
wet bog mosses; or Eriophorum may be absent in less wet places, and Juncus
castaneus the most conspicuous angiosperm, in association with Carices. Kruuse describes ponds as invaded from the west or south-west (direction of
prevailing winds), the zonation being as follows: (1) Empetrum; (2) Poly-
trichum; (3) Sphagnum, with Eriophorum; (4) Hypnum (H. scirpioides and
H. fluitans occurring throughout on the bottom to a depth of at least three
metres). Hippuris vulgaris and Ranunculus hyperboreus
occur in shallow water,
and R. paucistamineus Tausch. to a depth of three metres. ponds by vegetation, implying that they may finally pass over into heath by
accumulation of humus. It seems probable that this may happen, though
slowly, in the case of the smaller ponds, but in the case of larger lakes such
as those described by Porsild in Diskofjord (18, pp. 205-6) there is little
evidence of it. The present writer did not give any special attention to this
matter, but perhaps this is the most suitable place to insert a brief description
of the vegetation of the ponds and hollows in question. The species of aquatic
flowering
plants are all characteristic of temperate regions, and occur in all
ponds which are large enough to be permanent and shallow enough to be
appreciably warmed during the short summer. None of them produce ripe
fruit, and many of them rarely flower,
especially in the north. The commonest
is Hippuris, Menyanthes trifoliata also being very common in the south;
species of Callitriche, Potamogeton,
Sparganium, Myriophyllum
and Ranun-
culus also occur. This co-existence of temperate aquatics with arctic terrestria
vegetation is paralleled in plant remains from the glacial period in temperate
regions. Rikli (24, Taf. 47) describes the zonation round a small pool at
Godhavn. The outermost zone consists of a mat of small Carices, the most
frequent being C. alpina, C. rariflora and C. rigida; here also occur Juncus
castaneus, J. biglumis,
Salix groenlandica,
Tofieldia palustris, Saxifraga stellaris
v. This content downloaded from 128.235.251.160 on Tue, 13 Jan 2015 21:58:25 PM
All use subject to JSTOR Terms and Conditions A. NORTH OF THE REGION OF BETULA ODORATA. The
most typical are Silene acaulis and Diapensia lapponica, but many other
plants characteristic of open vegetation are met with, such as Saxifraga
tricuspidata,
S. aizoon, S. oppositifolia, S. groenlandica, Campanula rotundi-
folia, Viscaria alpina, Potentilla spp., Poa glauca, Cystopteris
fragilis and
Woodsia ilvensis. Many of these plants are found growing in vertical crevices,
especially the ferns. p
y
Hydrosere. In hollows of the rock surface there occur frequently
moss bogs,
or ponds of varying size, with or without aquatic flowering
plants. In the
Egedesminde district Kruuse (11, pp. 384-92) describes the invasion of such This content downloaded from 128.235.251.160 on Tue, 13 Jan 2015 21:58:25 PM
All use subject to JSTOR Terms and Conditions The Vegetation of West Greenlarn 94 ponds by vegetation, implying that they may finally pass over into heath by
accumulation of humus. It seems probable that this may happen, though
slowly, in the case of the smaller ponds, but in the case of larger lakes such
as those described by Porsild in Diskofjord (18, pp. 205-6) there is little
evidence of it. The present writer did not give any special attention to this
matter, but perhaps this is the most suitable place to insert a brief description
of the vegetation of the ponds and hollows in question. The species of aquatic
flowering
plants are all characteristic of temperate regions, and occur in all
ponds which are large enough to be permanent and shallow enough to be
appreciably warmed during the short summer. None of them produce ripe
fruit, and many of them rarely flower,
especially in the north. The commonest
is Hippuris, Menyanthes trifoliata also being very common in the south;
species of Callitriche, Potamogeton,
Sparganium, Myriophyllum
and Ranun-
culus also occur. This co-existence of temperate aquatics with arctic terrestria
vegetation is paralleled in plant remains from the glacial period in temperate
regions. Rikli (24, Taf. 47) describes the zonation round a small pool at
Godhavn. The outermost zone consists of a mat of small Carices, the most
frequent being C. alpina, C. rariflora and C. rigida; here also occur Juncus
castaneus, J. biglumis,
Salix groenlandica,
Tofieldia palustris, Saxifraga stellaris
v. comosa (bearing usually a few flowers at the apex of the inflorescence,
the
rest of which is covered with bulbils) and other plants. Next is a zone of
larger Carices often growing in water, C. rotundata being most frequent, with
C. A. NORTH OF THE REGION OF BETULA ODORATA. lichens and mosses are typically the first plants to appear. 3. Areas of silt brought
down by streams. The deltas at the mouths of
streams in the sandstone districts are the only important areas of this kind. Their vegetation no doubt often originates from plants washed down from
the sides of streams higher up during the spring floods (18, p. 135). Seeds and
bulbils (e.g. of Polygonum viviparum) might well be transported in the same
way. The surface of such deltas consists of sand strewn with boulders of all
sizes, and bears usually an open herbaceous vegetation. On the delta at
Patoot the following plants were noted: Chamaenerium
latifolium,
sometimes
in large clumps covered with very fine purple flowers; Artemisia borealis,
Oxyria digyna, Cerastium alpinum, Silene acaulis, Campanula rotundifolia
Papaver radicatum, Statice maritima v. sibirica, Salix glauca, Trisetum
spicatum, Poa glauca, Calamagrostis purpurascens, Juncus arcticus (locally)
and Equisetum arvense. On the deltas at Kook and other places on the north
coast of the Nugsuak-peninsula Melandrium trflorum, M. apetalum, Les-
querella arctica, Cochlearia officinalis,
Hierochloe alpina, Arctagrostis
latifolia
and Luzula confusa were conspicuous and often comparatively large plants
were present. p
The large delta at Ata was the only one visited by the writer which showed
all stages of succession from the open stony ground subject to violent erosion
of the central area to a true closed heath vegetation. The level of the principal
stream bed at the head of the delta is nearly 200 feet above sea, and the total
width of the delta, which is formed by streams from two adjacent valleys, is
somewhere about three miles. Between the two streams occurs the highest
part, with a vegetation of Cassiope heath. Near the western margin is a flat
area of silt which appears to be on either side of an old water course, which
itself is filled with sand. Here the ground is almost covered with prostrate
Salices (S. glauca or hybrids). The branches of these, which are sometimes
over a metre long, are quite horizontal and in contact with the soil surface. Only very short shoots bearing leaves and inflorescences
grow upwards; such
willows can probably stand fresh deposits of silt, and grow up through it. Violent breaking up of this vegetation is always possible by the formation of
new stream beds which cut right through it, uprooting willows and washing
them away. A. NORTH OF THE REGION OF BETULA ODORATA. comosa (bearing usually a few flowers at the apex of the inflorescence,
the
rest of which is covered with bulbils) and other plants. Next is a zone of
larger Carices often growing in water, C. rotundata being most frequent, with
C. misandra on small tussocks above the water level. Following this the
Eriophorum zone (E. polystachium
and E. scheuchzeri) extends into water
50 cm. deep. With the Eriophorum grow sterile shoots of Chamaenerium
latifolium. Finally comes Hippuris, with submerged mosses. In rock hollows
which have not ponds in them in which Hippuris can grow, Eriophorum
(either
species or both together) almost invariably occurs in shallow water or among
wet bog mosses; or Eriophorum may be absent in less wet places, and Juncus
castaneus the most conspicuous angiosperm, in association with Carices. Kruuse describes ponds as invaded from the west or south-west (direction of
prevailing winds), the zonation being as follows: (1) Empetrum; (2) Poly-
trichum; (3) Sphagnum, with Eriophorum; (4) Hypnum (H. scirpioides and
H. fluitans occurring throughout on the bottom to a depth of at least three
metres). Hippuris vulgaris and Ranunculus hyperboreus
occur in shallow water,
and R. paucistamineus Tausch. to a depth of three metres. 2. Talus slopes of recent rocks. These are frequently too steep and not
sufficiently
stable to allow of the development of a closed vegetation. This is
particularly
the case on the sides of the narrow valleys cut into the sedimentary
rocks; here occur at most only scattered plants of Papaver radicatum, Oxyria
digyna, Cerastium alpinum, etc. Where the slope is more gentle, heath can
develop, the early stages consisting of open vegetation of a great variety of
herbaceous plants, Dryas integrifolia
among the most important, and also low JOURNAL OF ECOLOGY VOL. X, PLATE IV , .,,,,~~3
JA f.s
N~~~~
,
z
"
:
iw , .,,,,~~3
JA f.s
N~~~~
,
z
"
:
iw (b)
Sil.ted-u stea
bdo
det,
At, cls
to (a)
Lov spedn
plant
of
OLTTUMTHE
VEGETATION
O F WiEsTRELAND. |_Il This content downloaded from 128.235.251.160 on Tue, 13 Jan 2015 21:58:25 PM
All use subject to JSTOR Terms and Conditions R. E. HOLTTUM 95 straggling
Salix glauca. Mosses and lichens are not found in any quantity on
the sandstone. Boulder-strewn
slopes below steep terraces of basalt have a
similar first phase of vegetation to that of the raised beach described below,
i.e. This content downloaded from 128.235.251.160 on Tue, 13 Jan 2015 21:58:25 PM
All use subject to JSTOR Terms and Conditions A. NORTH OF THE REGION OF BETULA ODORATA. The situation is very exposed, and the ground may even be free
of snow in the winter. With Salix are various other plants, especially Poly-
gonum viviparum, Pedicularis lanata, Equisetum arvense, with tufts of Poae,
Trisetum and Luzula confusa. The soil is moist almost to the surface every-
where, but the surface itself is hot and dry and in some places baked into a
hard crust. The latter condition prevails especially in parts more remote
from any stream bed; here in the crevices of the uneven surface appear small This content downloaded from 128.235.251.160 on Tue, 13 Jan 2015 21:58:25 PM
All use subject to JSTOR Terms and Conditions The Vegetation of West Greenland 96 dried mosses, but mosses are never abundant on the sandstone, except in
marshes and well-established
heath. Dryas is perhaps the most characteristic
plant of the drier parts which have a fairly permanent soil. Alectoria jubata
is of occasional occurrence,
forming
black patches; other flowering
plants are
Saxifraga tricuspidata,
S. oppositifolia,
Carex rupestris and other Cyperaceae. g
p
pp
p
yp
On the western edge of the delta there is a small lagoon behind the low
dune bank. Round the lagoon is a salt marsh, which gradually changes into
a fresh water marsh on its landward side. When visited by the writer,
perhaps
the most typical plant association here was dominated by Carex stans of
medium height (less than 30 cm.)- with much Equisetum variegatum, less
frequent E. arvense, both quite short, and always some almost prostrate Salix
glauca, with a good deal of moss between the higher plants. Locally Equisetum
variegatum
was dominant over a considerable area, with very fresh green shoots,
the associated plants being as before. On adjacent ground, where the surface
is raised somewhat and drier, E. variegatum
was still dominant but brown and
less upright, and Carex stans almost entirely absent; Polygonum viviparum
and Pedicularis lanata had established themselves. In other parts of the
marshy area were almost pure growths of Eriophorum,rpatches
of each species
alone being easily recognisable from a distance by the different
habit of the
inflorescence. Invading what. appeared to be a silted up-stream course were
very fine plants of Carex stans, with leaves and flowers almost 50 cm. high,
and very long rhizomes. With this were also very vigorous vegetative shoots
of Eriophorum,
much larger than those occurring elsewhere. This content downloaded from 128.235.251.160 on Tue, 13 Jan 2015 21:58:25 PM
All use subject to JSTOR Terms and Conditions A. NORTH OF THE REGION OF BETULA ODORATA. Juncus arcticus
was also pushing its way into the unoccupied ground, its lines of growth being
indicated by the close rows of upright shoots. More raised patches of silt
had prostrate willows growing on them, and it is probable that by progressive
silting the marsh will in time be replaced by heath, through a phase like that
on the drier silted areas described above. Porsild (18, pp. 167-70) describes
similar vegetation near Skansen. Here stream beds in summer are usually
strips of bare sand with no water showing, but the vegetation on either side,
principally of Carex stans, Eriophorum and Equisetum arvense (often pure
patches of each of these species), shows the ground to be very moist below. If
in spring the stream makes a new course for itself, as often happens, the large
Carex and Eriophoru,m
disappear and the area is invaded by heath plants;
traces of former Carex vegetation can sometimes be found in the heath. g
4. Raised beaches. Porsild (18, pp. 172-9) describes the development of
heath vegetation as indicated by a series of raised beaches near Sinigfik
(south Disko). The beaches consisted of rolled shingle of varying size. Species
of Gyrophora
are characteristic of the first phase, with later Sphaerophoron
fragile and Stereocaulon
denudatum. Xerophytic mosses soon appear, Racomt-
trium hypnoides being dominant, growing first in the crevices between the
stones. The mixed mosses finally form cushions as much as one metre long,
resting on bare stones; at the edges of the Racomitrium
cushions occur fruticose R. E. HOLTTUM 97 lichens. The next invaders are xerophilous Polytricha and angiosperms, at
first poorly developed and not flowering. Finally the vegetation becomes
closed by other mosses, particularly species of Dicranum and Oncophorum
and fruticose lichens, especially species of Cladonia, which fill up the spaces
between the other plants. The phanerogams then gain the upper hand and
Racomitrium and lichens occupy less and less of the available surface. In
this case an accumulation of raw humus and the covering of the ground by
mosses, which absorb water and prevent rapid desiccation of the soil, are
necessary before the higher plants can form a sufficiently
vigorous growth to
become dominant. 5. Sand dunes. These are rarely well d?eveloped,
but almost everywhere
in the sandstone districts,
and also where basalt is present, the shore is fringed
by small dunes with characteristic
vegetation. A. NORTH OF THE REGION OF BETULA ODORATA. On the landward side there is
a more or less gradual transition to heath in many cases. Elymus arenarius
v. vtllosus and Honkenya peploides are the only two plants which always occur
on the sand, the latter forming
slightly convex circular patches a metre across;
Mertensia maritima also is often present, but never very abundant. At
Godhavn, on a stretch of basaltic sand, other grasses make their appearance
among the Elymus a little distance back from the sea, Festuca arenaria being
the first of these. Farther from the shore Festuca ovina, Poa alpina, Poa
glauca, Poa pratensis, Trisetum spicatum and others are found, sometimes
almost covering the ground, while Salix glauca soon appears and clothes the
bases of the sand hillocks, which are still crowned with Elymus. Probably
this vegetation would finally develop into heath; at Holsteinsborg Empetrum
was seen to be spreading over some glacial sand by the shore, on which Elymus
also occurred. 6. Moraines. The succession on old moraines will vary according to the
nature of the rock from which they are formed, and with local topographic
and edaphic conditions. If nearly horizontal and of heavy retentive soil they
will be covered with moss bog, but if they are of lighter soil and better drained
heath will probably develop if climatic conditions allow. 6. Moraines. The succession on old moraines will vary according to the
nature of the rock from which they are formed, and with local topographic
and edaphic conditions. If nearly horizontal and of heavy retentive soil they
will be covered with moss bog, but if they are of lighter soil and better drained
heath will probably develop if climatic conditions allow. (b) Stabilised pre-climaxes: (1) "Fjaeldmark." True "Fjaeldmark" is an
open vegetation of isolated stunted flowering
plants, accompanied by mosses
and lichens, which occurs where the climate is not sufficiently
favourable to
permit the development of a closed vegetation of any kind. On Disko it only
occurs at high altitudes, but it is the characteristic
vegetation of the extreme
north. Porsild (18, p. 164) describes a high plateau on Disko, where the ground
was mainly covered with snow; on uncovered patches were Usnea melaxantha
and species of G6yrophora,
also Andreaea petrophila. Of phanerogams the most
frequent was Potentilla emarginata,
with Draba alpina, D. fladnizensis,
Papaver
radicatum, Saxifraga spp. and others. This content downloaded from 128.235.251.160 on Tue, 13 Jan 2015 21:58:25 PM
All use subject to JSTOR Terms and Conditions Journ. of Ecolpgy X The Vegetation of West Greenland The Vegetation of West Greenland 98 As Warming remarks (28, p. 257), the term "Fjaeldmark" covers a very
wide field, which probably should be subdivided, if sufficient
knowledge were
available, into a number of different
vegetation types, corresponding
with
differences
of habitat and climatic conditions. It is obvtious
that in such open
vegetation only those plants occur which can stand the conditions and whose
seeds have been carried there at a time when germination
was possible. Growth
of individuals is exceedingly slow, and its early stages very critical. Though
it is possible that in the course of a long time a stable condition might be
reached, in which the vegetation had attained its maximum possible develop-
ment, probably in practice one cannot decide whether this has occurred. The
number of plants which could exist under the extreme conditions is probably
large, and their actual arrival and establishment
very much a matter of chance. As an example, one may take Warming's (27, pp. 84-6) lists of plants found
on thr.ee nunataks in south Greenland, all 4000-5000 feet above sea. One
species only is common to all three of these. In general, one may say that
almost any species may occur which is found in the first stages of the succession
under similar edaphic conditions at lower altitudes. (2) Moss boys. Porsild (18, p. 226) describes these as occurring on Disko
(i) in gneissic soil in hollows of uneven basalt surfaces; (ii) on horizontal undrained terraces of the trap rocks: here the bog
occurs at even higher altitudes than the heath; g
(iii) on broad foreshores
of the trap rocks as the last phase of vegetation
after the heath: for this there must be a layer of humus thick enough to stop
the water, or a subsoil of gneiss or tuff. Here bog attains its most vigorous
development; (iii) on broad foreshores
of the trap rocks as the last phase of vegetation
after the heath: for this there must be a layer of humus thick enough to stop
the water, or a subsoil of gneiss or tuff. Here bog attains its most vigorous
development; p
(iv) on horizontal moraine soil. They are developed on the gneiss in hollows and on undrained moraine
surfaces, but probably never to such an extent as in case (iii) above. This content downloaded from 128.235.251.160 on Tue, 13 Jan 2015 21:58:25 PM
All use subject to JSTOR Terms and Conditions A. NORTH OF THE REGION OF BETULA ODORATA. This is also the type of vegetation
occurring on nunataks, isolated mountain peaks not covered by the in-
land ice. (b) Stabilised pre-climaxes: (1) "Fjaeldmark." True "Fjaeldmark" is an
open vegetation of isolated stunted flowering
plants, accompanied by mosses
and lichens, which occurs where the climate is not sufficiently
favourable to
permit the development of a closed vegetation of any kind. On Disko it only
occurs at high altitudes, but it is the characteristic
vegetation of the extreme
north. Porsild (18, p. 164) describes a high plateau on Disko, where the ground
was mainly covered with snow; on uncovered patches were Usnea melaxantha
and species of G6yrophora,
also Andreaea petrophila. Of phanerogams the most
frequent was Potentilla emarginata,
with Draba alpina, D. fladnizensis,
Papaver
radicatum, Saxifraga spp. and others. This is also the type of vegetation
occurring on nunataks, isolated mountain peaks not covered by the in-
land ice. 7 Journ. of Ecolpgy X This content downloaded from 128.235.251.160 on Tue, 13 Jan 2015 21:58:25 PM
All use subject to JSTOR Terms and Conditions This content downloaded from 128.235.251.160 on Tue, 13 Jan 2015 21:58:25 PM
All use subject to JSTOR Terms and Conditions 98 The Vegetation of West Greenland The most
extensive example of bog on Disko is at the Mudderbugt on the extreme
eastern point of the island. Here many valleys have their outlet to the sea
and a large flat area of land with a broad sandy shore has been formed. In
places this is bordered with salt marsh and dunes, but most of it is covered
with bog, which has the characteristic
form of isolated hummocks or ridges
covered with heath plants, separated from each other by water or wet moss. The hummocks may be two or even three feet high, they appear to consist of
an accumulation of plant remains and are largely covered with mosses near
the water level. They are presumably kept separated by the flowing off of
drainage water in the spring; as the summer advances the water between the
hummocks evaporates or slowly drains away, and incidentally
forms a breeding
place for swarms of mosqu'itoes which are very trying to the traveller on a
hot day in July. At Skansen the writer noticed a case where the water between
such hummocks drained away almost entirely,
and the intervening
space was
covered with a close crust of liverworts
(Gymnomitrium
etc.), or was becoming This content downloaded from 128.235.251.160 on Tue, 13 Jan 2015 21:58:25 PM
All use subject to JSTOR Terms and Conditions R. E. HOLTTUM 99 overgrown
with phanerogams. At Godhavn in such a crust of small liverworts
a number of seedlings of heath plants were noticed, and also developing bulbils
of Lycopodium selago. Whether the vegetation at Skansen had its origin as a
continuous moss bog is difficult
to say; it seemed to b-e in process of transition
to heath, probably owing to the presence of sand all round it. Rikli (24, Taf. 46) describes these moss bogs in some detail as regards
their probable development. The initial moss cushions may be composed of
numerous species, though according to Porsild the mosses of.moist localities
are less mixed than those in drier situations. Species of Sphagnum occur, but
never dominate in arctic Greenland; in the south thick Sphagnum carpets are
sometimes found. The first phanerogam to appear is often Salix groenlandica
(And.) Lundstr., its horizontal stems half buried in the moss, bearing the
conspicuous deep red-purple
upright inflorescences. Various Carices also occur,
forming compact cushions (C. alpina, C. This content downloaded from 128.235.251.160 on Tue, 13 Jan 2015 21:58:25 PM
All use subject to JSTOR Terms and Conditions The Vegetation of West Greenland The best development 7-2 This content downloaded from 128.235.251.160 on Tue, 13 Jan 2015 21:58:25 PM
All use subject to JSTOR Terms and Conditions The Vegetation of West Greenland 100 of scrub is along the sides of streams, where there is the necessary protection
and more rarely on the borders of lakes. Sometimes also it occurs on well-
established slopes below steep faces of the trap rocks, or in sheltered spots
on the gneiss as at Englishman's harbour, Godhavn. In the latter case the
willows often assume an espalier form if the protection is not good, growing
spread out flat against the rock surface. On Disko this type of vegetation
occurs very little except in Diskofjord and at various localities along the south
coast. On the Nugsuak coasts the present writer saw bushy willows only in
very few localities; the largest were by the bed of a stream near Ata where
the highest shoots were about one metre from the ground, but in no place
did such bushes cover any considerable area. In the vigorous willow scrub in Diskofjord there is little undergrowth,
the
soil consisting of black damp humus completely shaded in the summer. This
condition is modified in the northernmost
recorded area of willow scrub in
west Greenland, namely that described by Porsild (20, pp. 356-60) at Orpik
in Lakse Fjord (about 72? 30' N. Lat.) where an undergrowth
was universally
present consisting of grasses such as Poa pratensis, Trisetum and Hierochloe
also Pyrola grandiflora,
Stellaria longipes, Saxifraga spp., Campanula rotundi-
folia, C. uniflora and other herbaceous plants, with mosses in the more shady
places. "All the herbs are higher and more slender than those occurring on
the heaths, but there is light enough to permit the development of flowers
and fruit; for instance Pyrola is, if possible, larger flowered in the copse than
on the heath. The Vegetation of West Greenland rariflora and others), small species
of Juncus, and occasionally the two common species of Eriophorum (E. poly-
stachium and E. scheuchzeri). Ranunculus nivalis, R. lapponicus and several
species of Pedicularis are frequent. These various plants begin to form, with
the mosses, compact cushions; their dead remains from previous seasons
accumulate, decay being exceedingly slow. On the raised parts of the bog so
formed plants requiring less moist conditions can develop. The first of these
is almost invariably Salix glauca, or a hybrid between this and S. groenlandica,
and this plant remains at the top of the hummock. Later the ordinary heath
plants can gain a footing, and at the Mudderbugt nearly all are commonly
represented. The raised parts of the soil consist essentially of masses of raw
humus penetrated by the roots and stems of living plants. g
The essential characteristic
of the habitat in these cases seems to be wet
undrained soil; the fact that it is undrained is due to underlying
impermeable
soil or rock, or to an accumulated mass of wet peat which is frozen at no
great depth. If there is any considerable depth of soil it is frozen at a depth
of about 60 cm. on Disko; in any case the soil is cold as compared with that
of the heath, or with well-drained slopes in sheltered situations where her-
baceous plants of a less resistant type predominate. Porsild (18, p. 106)
suggests that the origin of the raised parts of the bog vegetation on the
Blaesedal moraine was due to the daily deposition of seeds which would
accumulate on windy -days on the wet edge of the melting snow as it retreated
up the slope. This might account for the observed development of the raised
parts of the vegetation more or less in parallel lines at right angles to the line
of greatest slope of the surface. (This is the case only on gently sloping
surfaces.) (c) Willow scrub (post-climax). As previously mentioned, this vegetation
occurs only in specially favourable situations, the most important factor being
adequate snow cover in winter, affording protection from cold dry winds
which kill off shoots projecting above the snow surface. This content downloaded from 128.235.251.160 on Tue, 13 Jan 2015 21:58:25 PM
All use subject to JSTOR Terms and Conditions The Vegetation of West Greenland There seems
no reason why scrub should not develop on the ground covered by these
plants, and the herbaceous vegetation is therefore
considered as a stage in the
succession to scrub. There are cases, however, quoted by Warming (27, p. 38),
in which this type of vegetation occurs higher up on mountain sides where
there is no willow scrub; climatic or other conditions here cause the arrest
of the succession. The herbaceous vegetation is particularly remarkable in the vicinity of the
"warm" springs of south Disko. These have a fairly constant temperature
at their source of 12-14? C., winter and summer alike. Their course is usually
marked by a luxuriant carpet of Mniobryum albicans, often accompanied by
thick mats of Marchantia. Close to the springs there occur a number of plants
whose normal range of distribution
lies farther to the south; on Disko they
are entirely confined to these localities. With them are plants which are at
or near their northern limit but which can also grow away from the springs
in sheltered sunny places. The most conspicuous of these plants are Archan-
gelica officinalis
and Alchemilla
glomerulans
Buser; the former
is quite confined
to the warm springs, while the latter occurs nowhere else so abundantly as here. g
The most remarkable of these favoured localities from the point of view
of variety and luxuriance of growth of the herbaceous plants is the strip of
land bordering the small bay called Englishman's harbour, lying close to the
entrance to the harbour of Godhavn. The writer visited this place on July 19th,
1921, and some description of it may be appropriate here. The bay lies at
the foot of steep gneiss rocks which rise to a height of perhaps 250 metres. Above the gneiss are steep terraces of basalt, and at the junction of the two
the springs appear, reaching the sea at the western end of the bay; at the
east end is a second smaller spring. The beach is composed of rolled boulders
and shingle among which the only vegetation consisted of a few plants of
Mertensia maritima. The most -considerable flat piece of ground is towards
the east end of the bay where on a wet black peaty soil was a quite closed
vegetation. Alchemilla was everywhere,
and in some places dominant, as was
also Equisetum arvense. The Vegetation of West Greenland In open places pure growths are met with, usually of high,
slender plants of Equisetum arvense, richly branching and as green as the
woodland forms of Europe; or there are small meadows of Poa pratensis,
and along the river there are enormous meadows of Chamaenerium
latifolium
or where a small stream branches off from a bigger, pale green patches of
Mniobryum albicans together with Marchantia polymorpha." Here in the
interior of a fjord, a thick even covering of snow is possible over large areas,
as the winds are never as fierce as on the coast. Even the dwarf birch grows
obliquely upwards, with leaves evenly arranged on all sides of its branches. In the heath "Cassiope sends straight upwards whole bunches of shoots,
closely set with white flowers; Ledum has larger leaves and richer inflores-
cences. And above all these Hierochloe and Arctagrostis
protrude, the latter
attaining a height of 80 cm.... There are far fewer mosses at the bottom be-
tween the heath plants than in the basalt regions; but in the small damp
depressions there occur, in strong contradistinction
to what is the case on the
basalt substratum, quantities of Sphagna-white, grass-green,
wine-red, and
nut-brown." In more southern parts of the coast there are various records of willow
scrub reaching 21 metres in height, but according to Rosenvinge (25, p. 121)
this is exceptional even in the extreme south. Warming (27, pp. 30-33) gives
a description of willows reaching this height near Holsteinsborg (67? N.). This content downloaded from 128.235.251.160 on Tue, 13 Jan 2015 21:58:25 PM
All use subject to JSTOR Terms and Conditions 101 R. E HOLTTUM Amongst the willows grew Archangelica,
which obviously produced ripe seeds,
as young plants were abundant. Other herbaceous plants also occurred, all
taller than when growing in the open. In the soil a number of earthworm
were found; these are probably always present in such situations, but in the
open heath were searched for in vain. Herbaceous vegetation
(Urtemark). In localities similar to those in which
scrub of Salix glauca occurs and often bordering on it, one encounters a more
or less luxuriant growth of herbaceous plants which are not hardy enough
to stand the conditions prevailing in more exposed situations. This content downloaded from 128.235.251.160 on Tue, 13 Jan 2015 21:58:25 PM
All use subject to JSTOR Terms and Conditions The Vegetation of West Greenland The Vegetation of West Greenland 102 able growth
of Epilobium hornemanni
with occasionally
E. lactiflorum,
and
in certain areas numerous
spikes of Habenaria albida and H. hyperborea,
par-
ticularly
of the latter which when in full flower
reaches nearly a foot high. Where the taller plants were not so abundant smaller species were more
noticeable,
particularly
Pinguicula vulgaris and Thalictrum
alpinum; a few
specimens
of Listera cordata
were found. A striking
feature
of the vegetation
was the presence
of numerous
ferns; Aspidium lonchitis,
Dryopteris
linneana
and Cystopteris
fragilis were all abundant. On sloping banks among the rocks where vegetation
had become estab-
lished perhaps the most universally
present
plant was Polygonum
viviparum
though sometimes
nothing
was to be seen but the large leaves of Alchemill
Habenaria albida was very abundant in some places and also Taraxacum
sp.,
with Sibbaldia procumbens
as undergrowth. Among these Dryopteris
sent up
young leaves, and Gnaphalium
norvegicum
showed in places its grey hairy
leaves and black unopened
flower
buds. Grasses never dominate,
but several
species occur, notably Poae, and Phleum alpinum in the wetter
parts. Large
bushy willows were everywhere
among the rocks, but even here the more
prominent
young shoots are killed in the winter,
so that the bushes have
somewhat
the appearance of a trimmed
hedge. g
Porsild (21, p. 255) says that much of the ground
here is never frozen. In
the winter
when the air temperature
outside may be - 300 C. the temperatur
of small pools of water among the vegetation
under the snow may be as much
as 17?. Only a few rocks project through
the snow which otherwise
covers
everything,
and "up through
the snow layer in the warmest
places chimney
like holes are formed,
through
which warm and damp air as from
a greenhous
streams
to meet one." Earthworms
occur in the soil. In the paper just referred
to Porsild describes
a second locality, 0sterli
a slope beneath masses of tuff
on the east side of 0sterdal, behind the Arctic
Station; and in the 0sterdal,
a little higher up, occur more warm springs
with attendant
carpets of Mniobryum
albicans and much ground
covered
with
Alchemilla. On the 9sterli low bushes of Salix in many parts cover the slope
almost to the top, but in other places there is an almost purely herbaceou
vegetation. This content downloaded from 128.235.251.160 on Tue, 13 Jan 2015 21:58:25 PM
All use subject to JSTOR Terms and Conditions The Vegetation of West Greenland There were also numerous plants of Archangelica
with inflorescence
shoots a metre high, though only just in flower. Luzula
parvijiora also is a very characteristic
plant both here and by the warm springs
in the 0sterdal near the Arctic Station. In other places there was a consider- The Vegetation of West Greenland Dupontia fisheri was dominant here, with much Carex stans, and with these
were associated the species of Eriophorum and Equisetum, and Salix glauca. On a small adjoining area Carex glareosa was dominant, with C. ursina, and
thick patches of Ranunculus hyperboreus. (Microcoleus sp.). All round the edge of the water, and even in it at low tide,
was a thick mat of Glyceria vilfoidea, which in such situations is always quite
sterile. A little farther from the lagoon the Glyceria mat was invaded by
small Carices, C. ursina and C. salina v. subspathacea; in places C. ursina
became dominant, accompanied by Stellaria humifusa, a little Glyceria and
mosses. The most luxuriant vegetation occurred along the small water courses
which drained the marsh. Where the plants grew more thinly, on wet mud
near the dune bank, Triglochin palustre was exceedingly abundant. It was
mostly sterile, but obvious.from
the dark olive green of its tiny leaves, which
were only about 5 cm. high. On each plant was already formed
a comparatively
large underground
winter bud. Not far inland from the lagoon the salt marsh
proper passed gradually into the previously described fresh water marsh. Dupontia fisheri was dominant here, with much Carex stans, and with these
were associated the species of Eriophorum and Equisetum, and Salix glauca. On a small adjoining area Carex glareosa was dominant, with C. ursina, and
thick patches of Ranunculus hyperboreus. The most characteristic
salt marsh plants in this part of Greenland appear
to be Glyceria vilfoidea, Carex glareosa, C. ursina, C. salina and Stellaria
humifusa. Dupontia fisheri is a high arctic type, and has its southern limit
on north Disko. (2) Nitrophilous vegetation
round settlements. In the neighbourhood
of all
settlements a good deal of nitrogenous animal refuse accumulates, and the
vegetation responds to this in a remarkable way. Certain grasses, notably
Alopecurus alpinus, develop to an enormously greater extent than under
normal circumstances
and form a close vegetation of thick tussocks. In most
places where there are long-established settlements Alopecurus has entirely
the upper hand and itself covers nearly all the ground, though in some places,
as often under bird cliffs,
Glyceria distans forms similar tussocks. Other plants
which commonly occur are Cochlearia officinalis and Cerastium alpinum. The Vegetation of West Greenland Round the Arctic Station it was very noticeable that many other grasses
responded to the same stimulus, Phippsia algida and Festuca ovina for example
growing into tufts many times larger than on poorer sandy soil. The sites of old
settlements,
long uninhabited,
are still marked by the presence of thick growths
of Alopecurus. Hartz (6, p. 58) records a number of introduced weeds round
the settlements of Christianshaab and Ritenbenk, many of them growing
vigorously and flowering. Among these were Polygonum convolvulus,
Stellaria
media, Galium aparine, Chenopodium
album, Sinapis arvensis and Poa annua. The Vegetation of West Greenland Here at the end of August Epilobium
angustifolium
comes into
flower,
and careful
search among the larger plants reveals small specimens
of
Botrychium
lunaria. Almost
pure patches of Gnaphalium
norvegicum
are found y
p
p
p
g
At both 0sterli and Englishman's
harbour
Porsild has placed notices
asking
the natives not to pull up any of the plants or gather turf for fuel, and this
request has been respected,
so that these interesting
patches of vegetation
are
being carefully
preserved. g
(d) Special edaphicformations. (1) Salt marsh. This is very restricted
in
its occurrence,
and is usually confined
to localities where lagoons are formed
behind sand banks, as on the delta at Ata. The bottom of the Ata lagoon is
formed
of muddy silt held together
by a continuous
crust of a blue-green
alga VOL. X, PLATE V JOURNAL OF ECOLOGY (a) Heath on gneiss, Godhavn, with Cassiofie tetragona
(below) and Emizpetrum
nigrumn
(on right). Climax type. (a) Heath on gneiss, Godhavn, with Cassiofie tetragona
(below) and Emizpetrum
nigrumn
(on right). Climax type. ....... - z
-
(b) Bog near Godhavn, with hummocks covered by Salixglauca. (b) Bog near Godhavn, with hummocks covered by Salixglauca. HOLTTUM THE VEGETATION
OF WEST
GREENLAND. This content downloaded from 128.235.251.160 on Tue, 13 Jan 2015 21:58:25 PM
All use subject to JSTOR Terms and Conditions This content downloaded from 128.235.251.160 on Tue, 13 Jan 2015 21:58:25 PM
All use subject to JSTOR Terms and Conditions R. E. HOLTTUM 103 (Microcoleus sp.). All round the edge of the water, and even in it at low tide,
was a thick mat of Glyceria vilfoidea, which in such situations is always quite
sterile. A little farther from the lagoon the Glyceria mat was invaded by
small Carices, C. ursina and C. salina v. subspathacea; in places C. ursina
became dominant, accompanied by Stellaria humifusa, a little Glyceria and
mosses. The most luxuriant vegetation occurred along the small water courses
which drained the marsh. Where the plants grew more thinly, on wet mud
near the dune bank, Triglochin palustre was exceedingly abundant. It was
mostly sterile, but obvious.from
the dark olive green of its tiny leaves, which
were only about 5 cm. high. On each plant was already formed
a comparatively
large underground
winter bud. Not far inland from the lagoon the salt marsh
proper passed gradually into the previously described fresh water marsh. This content downloaded from 128.235.251.160 on Tue, 13 Jan 2015 21:58:25 PM
All use subject to JSTOR Terms and Conditions B. SOUTH OF THE REGION OF BETULA ODORA TA. South of about 620 N. Lat. birch-willow
scrub occupies considerable areas,
and must probably be considered as the climatic climax vegetation. Heath
and grassland are also very extensive, and are probably in many cases
stabilised pre-climaxes due to local conditions of soil, altitude, and exposure. This content downloaded from 128.235.251.160 on Tue, 13 Jan 2015 21:58:25 PM
All use subject to JSTOR Terms and Conditions 04
Th
f The Vegetation of West Greenland 104 The scrub, consisting of Betula odorata Bechst. v. tortuosa Rgl. with the
associated bushy species Salix glauca, Alnus ovata (Schr.) v. repens Lange
(occurring also in Salix scrub as far north as 680) and Sorbus americana, is
usually very thick, and rarely up to a man's height. On level ground especially,
where there is less shelter, the upper surface of the bushy layer is very even,
and no doubt determined by wind and snow-covering. In the most sheltered
inland valleys only do larger tree-growths
of birch occur. In Tasermiut fjord
some specimens attain 6-7 metres in height, and Sorbus nearly 4 metres. The birches never grow straight upright with one main trunk; several branches
appear about the soil level, and some of these are usually prostrate at the
base. Here also, as in the willow scrub farther north, where the birches are
not too thick, grasses and other plants occur as undergrowth. Juniperus
communis v. nana Hartm. is common in this region, but does not usually
form bushy growths; specimens as old as 400 years have been recorded. In
connection with the natural vegetation of this region it is interesting
to note
that seeds of Picea excelsa were sown in Agdluitsok fjord (60? 30' N.) in 1846,
and in 1886 three specimens collected had only reached one metre in height. p
y
g
In this region willow scrub occurs in moister and less sunny places than
birch scrub, though Salix glauca is usually present in the latter. Though it
may attain 21 metres, the scrub is normally not more than half that height
Often the bushes grow very thickly, and the ground vegetation is of mosses
only (Hylocomium
etc.), but where they are more open herbaceous flowerin
plants occur, grasses often predominating,
especially Anthoxanthum
odoratum
Airaflexuosa, Phleum alpinum and species of Poa and Festuca. Willow scrub
here may be found up to 1400 feet above sea-level. This content downloaded from 128.235.251.160 on Tue, 13 Jan 2015 21:58:25 PM
All use subject to JSTOR Terms and Conditions (5) GENERAL CONSIDERATIONS WITH REGARD TO THE
CONDITIONS OF PLANT GROWTH. The principal factor controlling
the nature of the vegetation, especially in
the south, is the oceanic climate, with its short and comparatively cool and
damp summer. This causes the vegetation over a long stretch of coast to be
very uniform in character, and leads'to the accumulation of raw humus,
giving heath as the climatic climax; this is further accentuated by the non-
calcareous nature of the rocks from which the soil is formed. The July iso-
therm of 100 C., which is approximately coincident with the northern tree
limit, just touches the south of Greenland, and here alone is anything in the
nature of tree growth possible. The only important arborescent species is
Betula odorata, which is characteristic also of the tree limit in the oceanic
climates of Iceland and Scandinavia. In favourable situations, however,
though no tree growth is possible, considerable vegetative development during
the short summer may take place, even as far north as 730; in more exposed
situations the amount of growth in a single summer is very restricted. g
g
y
In such exposed situations the xeromorphic character of most of the
plants is very marked. This is particularly the case with the evergreen woody
plants of the heath such as Empetrurm
and Cassiope. These must be protected
from loss of water by transpiration when the activity of their roots is com-
pletely suspended owing to the freezing
of the ground; hence' the need for their
narrow leaves with recurved margins and protective hairs. In summer these
characters are probably of service in the drier habitats, as for example on the
sandstone, but in most places the presence in the heath of larger leaved plants
such as Salix glauca, whose transpiration must be considerable, would seem
to indicate that such extreme protection against drought is not essential at
this season. Salix and Betula are of course deciduous, and so avoid the
possibility of severe loss of water in the winter or spring when the ground is
still frozen or very cold. Rosenvinge (25, p. 193) remarks that Empetrum
in south Greenland does not grow in situations which are exposed to the
winter " foehn," which may melt the snow cover from above the heath plants
while the ground is still frozen. The root development of Salix glauca is often
very considerable, a specimen collected at Atanikerdluk possessing a main
root nine metres long. B. SOUTH OF THE REGION OF BETULA ODORA TA. y
p
On sheltered slopes a herbaceous vegetation, mainly of grasses, is probably
an early stage in the development of scrub. The number of species occurring
is said to be considerable, with no definite dominant. In drier soil on less
sheltered plains Anthoxanthum
odoratum
and Aira flexuosa are the character-
istic species, with a few other plants, including Leontodon autumnale and
Taraxacum sp. Such grassland may occur up to 1200 feet above sea-level, and
provides pasture for the sheep and cattle which in this part of Greenland
alone can be reared. In Tasermiut fjord Nardus stricta replaces Anthoxanthum j
The character of the heath and fell-field
in the south is much the same as
farther north, but the distribution of these types of vegetation is more re-
stricted. Species like Papaver radicatum, Silene acaulis and Dryas integrifoli
are here confined to the most unfavourable situations, the competition else-
where being too severe. g
The nitrophilous vegetation round the settlements is also the same, with
the addition of other species, particularly Rumex acetosa and Matricaria
inodora. As in the north, a similar vegetation is found to persist on the sites
of old habitations. This is especially interesting
in the case of the ruins of
the former Icelandic col-onies,
some of them dating from the tenth century, R. E. HOLTTUM 10oo and probably uninhabited for at least 400 years. Here still grow plants which
must have been introduced by the Norsemen, such as Polygonum aviculare,
Stellaria media, Capsella bursa pastoris, Poa annua, Vicia cracca, Achillea
millefolium. These plants never grow beyond the limits of the old settlements
It is possible that some of the plants now common in south Greenland may have
been introduced in the same way, and Rosenvinge suggests that this may be
the case with Leontodon autumnale and Anthoxanthum
odoratum. This content downloaded from 128.235.251.160 on Tue, 13 Jan 2015 21:58:25 PM
All use subject to JSTOR Terms and Conditions (5) GENERAL CONSIDERATIONS WITH REGARD TO THE
CONDITIONS OF PLANT GROWTH. The ground is always frozen at a depth of at most a
metre almost everywhere, so that long roots are horizontal. Those of the The Vegetationt
of West Greenland 106 evergreen
plants are not so extensive,
but some of the herbaceous
plants, such
as Papaver and species of Potentilla,
have very long roots, corresponding
with
their relatively
greater transpiration. The extensive development
of fresh
green flowering
shoots of Chamaenerium
latifolium
in numerous
apparently
arid localities shows that in many places where heath may be formed
there
is no need for xeromorphy
during
the summer. p y
g
For ability to withstand
low temperature
the internal
organisation
of the
plant must be responsible. It may be that the formation
of anthocyan
pig-
ments in the leaves of practically
all plants, both in spring and autumn, is
of value in this connection. The habit of growth
of the heath plants afford
protection
against desiccation
by winter
winds rather than against low tem-
perature
alone. Mycorrhizal
fungi are present
in the roots of most of the common
plants
Ectotropic mycorrhiza
occurs in the case of Betula, Salix, Polygonum
and
Dryas; endotropic
in the Ericaceous plants and Diapensia. This may be
connected
with the fact that decomposition
of humus in the soil is exceedingl
slow, and available nitrogenous
compounds
correspondingly
limited. It is
characteristic
of all the cushion plants, for example, that dead leaves remain
attached to the short stems for many years before
they are decomposed. yy
y
p
Little work has been done on the germination
of seeds in the arctic under
natural conditions,
though this is a matter of very great importance
in the
study of the development
of vegetation. Porsild has made numerous
observa
tions on the ripening
of seeds and the occurrence
of seedlings,
but these are
not yet published. Hart (5, pp. 74-5, 306, 342) observed no seedlings at
Discovery Bay (81? 42' N. Lat.) and concluded that no ripe seeds are pro-
duced, and that spreading
of plants was exclusively
vegetative,
or at most
due to the chance importation
of ripe seed from farther
south. He grew
seedlings
of wheat, peas, mustard
etc. under a glass shelter
at a temperatur
of about 330 F. Some of the wheat seeds had been exposed to four successive
winters
at that latitude before
being sown, but were quite uninjured. This content downloaded from 128.235.251.160 on Tue, 13 Jan 2015 21:58:25 PM
All use subject to JSTOR Terms and Conditions BIBLIOGRAPHY. (1) Berggren, S. "Bidrag till Kannedom om Fanerogamflora
ved Diskobugten och Auleit-
sivikfjorden
paa Gronlands Vestkyst." Ofv. Kgl. Vet.-Ak. Forh. No. 7, Stockholm, 1871. (1) Berggren, S. "Bidrag till Kannedom om Fanerogamflora
ved Diskobugten och Auleit-
sivikfjorden
paa Gronlands Vestkyst." Ofv. Kgl. Vet.-Ak. Forh. No. 7, Stockholm, 1871. (2) Brown, R. "Florula Discoana." Tr. Bot. Soc. Edinb. 9, 1868. j
p
y
g
(2) Brown, R. "Florula Discoana." Tr. Bot. Soc. Edinb. 9, 1868. (3) Cleve, A. "Zum Pflanzenleben in schwedischen Hochgebirgen." Bihang till Kgl. Svensk
Vet.-Ak. Handl. Bd. 26, Afd. 3, 1900. (4) Egede, H. Det gamle Gronlands ny Perlustration,
Copenhagen, 1729. (5) Hart, H. C. "On the Botany of the British Polar Expedition of 4875-6." Journ. Bot. N.S. 9, 1880. (6) Hartz, N. "Botaniske Rejseberetning fra Vestgronland, 1889 og 1890." Meddelelser
om
Gronland,Heft 15, 1898. ,
(7) Hartz, N. and Kruuse, C. " The vegetation of N.E. Greenland, 690 25' to 750 N." Medd. om Giyinl. Heft 30, 1911. (8) Hesselman, H. " Om Mykorrhizabildringar
hos arktiska Vhxter." Bihang till Kgl. Svensk
Vet.-Ak. Handl. Bd. 26, Afd. 3, 1900. (9) Hooker, J. D. "Outlines of the distribution
of Arctic Plants." Tr. Linn. Soc. 22, 1860. (10) Kjellman, F. R. "Aus dem Leben der Polarpflanzen." In Nordenskjold's Studien und
Forschungen
veranlas8t
durch meine Reise im hohen Norden, deutsche Ausgabe, Leipzig, 1885. g
g
g
(11) Kruuse, C. "Vegetationen i Egedesminde Skjaergaard." Medd. om Gronl. Heft 14, 1898. (12) Lange, J. M. "Conspectus Florae Groenlandicae." Medd. om GrH)nl. Heft 3, 1880. g
(13) Lange, J. M. "Studiei uiber Gronlands Flora." Engler's Jahrb. Bd. 1, 1881. g
g
(14) Lauridsen, P. "Bibliographia Groenlandica." Medd. om Gr#6nl. Heft 13, 1890 g
(15) Nathorst, A. G. " Beitrage der Polarforschung
zur Pflanzengeographie
der Vorzeit. " In
Nordenskjold's Studien und Forschungen,
1885. j
g
(16) Nathorst, A. G. "Kritische Bemerkungen iber die Geschichte der Vegetation Grbnlands."
Engler's Jahrb. Bd. 14, 1892. g
(17) Ostenfeld, C. H. " Plante og Dyreliv paa Gr0nlands Nordkyst. " In Rasmussen's Gryunlan
langs Polar-havet,
Copenhagen, 1919. g
p
g
(18) Porsild, M. P. "Bidrag til en Skildring af Vegetationen paa 0en Disko." Medd. om
GrAnl. Heft 25, 1902. (19) Porsild, M. P. "The plant-life of Hare Island off the coast of West Greenland Arbejder
fra den Danske Arktiske Station paa Disko, No. 3)." Medd. om Gr6nl. Heft 47, 1910. (20) Porsild, M. P. (5) GENERAL CONSIDERATIONS WITH REGARD TO THE
CONDITIONS OF PLANT GROWTH. Appa-
rently ripe seeds from herbarium
specimens
collected by Hart would not
germinate,
but the trial was not made till four years later. g
y
Kjellman (10, p. 480) observes
that in northern
Siberia 85 out of the 150
species collected
on the voyage of the Vega had ripe seeds, and of the other
some were probably
only found too early in the year for the fruits
to be mature p
y
y
y
y
Cleve (3, pp. 101-3) in a paper on the alpine vegetation
of Sweden includes
a short section on this matter. He concludes
that a large proportion
of arctic
and alpine plants produce ripe seeds, but that it is usually necessary
for these
to rest at least over one winter
before
they are capable of germination. y
p
g
Porsild (19, p. 273) says that seeds of many of the plants of west Greenland
mature under the snow, and that if gathered in the autumn they are not
properly
ripe. Some of the plants which are beyond the normal
northern
limit
of their distribution
produce no ripe seeds, and this is also the case for most R. E. HOLTTUM 107 of the aquatic plants. The present occurrence of these types may represent
patches of a once continuous distribution in a somewhat warmer climate
subsequent to the glacial period. The most important agent in the dispersal of seeds is the wind blowing
over the surface of the snow and sea ice in winter. Porsild gives a list of plants
not occurring on Hare Island (19, p. 268); these plants are all distributed by
the wind in winter, and have not reached Hare Island because strong tidal
currents prevent the formation of continuous sea ice between it and Disko. Plants which are completely covered by snow in winter cannot have their
seeds dispersed in this way; they are mostly of southern type which only
form seeds, if at all, quite at the end of the season, so that they have little
chance of spreading more than short distances, whereas the localities in which
they can flourish are often few and far between. Another important factor
in seed dispersal is the water from melting snow in the spring. This content downloaded from 128.235.251.160 on Tue, 13 Jan 2015 21:58:25 PM
All use subject to JSTOR Terms and Conditions BIBLIOGRAPHY. "Vascular plants of West Greenland between 710 and 730 N. (ibid. No. 6)."
Medd. om Grvnl. Heft 50, 1912. This content downloaded from 128.235.251.160 on Tue, 13 Jan 2015 21:58:25 PM
All use subject to JSTOR Terms and Conditions This content downloaded from 128.235.251.160 on Tue, 13 Jan 2015 21:58:25 PM
All use subject to JSTOR Terms and Conditions This content downloaded from 128.235.251.160 on Tue, 13 Jan 2015 21:58:25 PM
All use subject to JSTOR Terms and Conditions The Vegetation of West Green land The Vegetation of West Green land 108 (21) Porsild, M. P. "Naturfredning
i dansk Gr0nland (ibid. No. 8)." Medd. om GrCqnl. left 51,
1915. (22) Porsild, M . P. "Liliales. (The structure
and biology of arctic flowering
plants, 14.)" Medd. om. Gr?nl. Heft 37, 1920. (23) Porsild, M. P. "Plantevaext." In Grinland i Tohundredaaret
for Hans Egedes Landing,
Copenhagen, 1921. (24) Rikli, M. "Vegetationsbilder
aus D5,nisch-Westgr6nland."
In Karsten und Schenk, Vege-
tatiomsbilder,
vii Reihe, Heft 8, 1910. (25) Rosenvinge, L. K. "Det sydligste Gr0nlands Vegetation." Medd. om Gr nl. Heft 15, 1898. (25) Rosenvinge, L. K. "Det sydligste Gr0 (26) Vanhoffen, E. Chapters on Vegetation in Drygalski's Gr6nland-Expedition
der Gesell-
schatftfiur
Erdkunde zu Berlin, Bd. 2, 1897. (27) Warming, E. "Om Gr0nlands Vegetation." Medd. om Gr6nl. Heft 12 (also in Engler's
Jahrb. Bd. 10), 1888. (28) Warming, E. (Ecology of Plants, Oxford, 1909.
|
https://openalex.org/W3189490001
|
http://pure-oai.bham.ac.uk/ws/files/145446222/d1ea00010a.pdf
|
English
| null |
Insights into HONO sources from observations during a solar eclipse
|
Environmental science. Atmospheres
| 2,021
|
cc-by
| 8,750
|
University of Birmingham
Insights into HONO sources from observations
during a solar eclipse
Singh, Ajit; Crilley, Leigh; Pope, Francis; Bloss, William
DOI:
10.1039/d1ea00010a
License:
Creative Commons: Attribution (CC BY)
Document Version
Publisher's PDF, also known as Version of record
Citation for published version (Harvard):
Singh, A, Crilley, L, Pope, F & Bloss, W 2021, 'Insights into HONO sources from observations during a solar
eclipse', Environmental Science: Atmospheres. https://doi.org/10.1039/d1ea00010a University of Birmingham Insights into HONO sources from observations
during a solar eclipse
Singh, Ajit; Crilley, Leigh; Pope, Francis; Bloss, William DOI:
10.1039/d1ea00010a
License:
Creative Commons: Attribution (CC BY) Document Version
Publisher's PDF, also known as Version of record Citation for published version (Harvard):
Singh, A, Crilley, L, Pope, F & Bloss, W 2021, 'Insights into HONO sources from observations during a solar
eclipse', Environmental Science: Atmospheres. https://doi.org/10.1039/d1ea00010a Citation for published version (Harvard):
Singh, A, Crilley, L, Pope, F & Bloss, W 2021, 'Insights into HONO sources from observations during a solar
eclipse', Environmental Science: Atmospheres. https://doi.org/10.1039/d1ea00010a Link to publication on Research at Birmingham portal Download date: 24. Oct. 2024 General rights
l
li General rights
Unless a licence is specified above, all rights (including copyright and moral rights) in this document are retained by the authors and/or the
copyright holders. The express permission of the copyright holder must be obtained for any use of this material other than for purposes
permitted by law. eely distribute the URL that is used to identify this publication. •Users may freely distribute the URL that is used to identify this publication. Users may freely distribute the URL that is used to identify this publication. •Users may download and/or print one copy of the publication from the University of Birmingham research portal for the purpose of private
study or non-commercial research. study o
o
co
e c a
esea c
•User may use extracts from the document in line with the concept of ‘fair dealing’ under the Copyright, Designs and P
•Users may not further distribute the material nor use it for the purposes of commercial gain. Where a licence is displayed above, please note the terms and conditions of the licence govern your use of this document. Where a licence is displayed above, please note the terms and conditions of the licence govern your use of th When citing, please reference the published version. Environmental Science:
Atmospheres Cite this: DOI: 10.1039/d1ea00010a
Received 13th February 2021
Accepted 29th June 2021
DOI: 10.1039/d1ea00010a
rsc.li/esatmospheres
Atmosphere
PAPER
s Article. Published on 20 July 2021. Downloaded on 8/11/2021 7:19:11 PM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. View Article Online
View Journal aSchool of Geography, Earth and Environmental Sciences, University of Birmingham,
Birmingham, B15 2TT, UK. E-mail: W.J.Bloss@bham.ac.uk; a.singh.2@bham.ac.uk;
F.Pope@bham.ac.uk
bDepartment of Chemistry, York University, Toronto, Canada. E-mail: lcrilley@yorku.
ca
† Electronic
supplementary
information
(ESI)
available.
See
DOI:
10.1039/d1ea00010a Insights into HONO sources from observations
during a solar eclipse† Open Access Article. Published on 20 July 2021. Downloaded on 8/11/2021 7:19:11 PM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. Ajit Singh,
a Leigh R. Crilley,
b Francis D. Pope
a and William J. Bloss
*a Nitrous acid (HONO) is a major, and often the dominant, precursor to primary OH radical production in the
daytime boundary layer, driving the removal of many primary pollutants and formation of secondary species
such as ozone and many aerosol components. A number of photochemical HONO production
mechanisms have been proposed, alongside homogeneous gas-phase reactions, to account for field
observations of daytime HONO. The range of production mechanisms show varying dependencies upon
precursor species such as NO2, available surfaces for heterogeneous reactions, and dark/photoenhanced
aspects. Here, we exploit measurements of HONO and related species during a near-total solar eclipse
as a natural perturbation to the atmospheric photochemistry to assess the characteristics of the
production mechanisms occurring at an urban background location. Little variation in HONO abundance
was observed in response to changing light levels during the eclipse, pointing to relatively balanced
photochemical source and (well-understood) sink terms. We employ a series of simple kinetic
simulations to explore the consistency of different potential source mechanisms with the observations,
finding evidence for a dominant role for photochemical processing of traffic-derived NO2 upon surfaces
producing HONO, alongside indications of a smaller contribution from direct vehicular emissions. Other
mechanisms involving dark heterogeneous reactions were not, in isolation, consistent with the
observations. The critical role of NO2, ultimately derived overwhelmingly from local road traffic
emissions at this location, points to significant future reductions in daytime HONO production with
vehicle fleet evolution and reduction of tailpipe emissions. Received 13th February 2021
Accepted 29th June 2021 Environmental signicance Nitrous acid (HONO) is a major, oen the dominant, boundary layer precursor to the key daytime atmospheric oxidant OH; however sources of HONO are poorly
understood, with a number of candidate formation mechanisms advanced. Here, we use a natural perturbation – a solar eclipse – to identify the photochemical
factors governing HONO formation, from their variation with this large-scale change in solar intensity. We demonstrate that HONO is not controlled by a single
source, but that photoenhanced production (i.e. reactions accelerated by sunlight) are a major component of HONO production, suggesting a dominant role for
photochemical processing of traffic-derived NO2 upon surfaces producing HONO, alongside a smaller contribution from direct vehicular emissions. In both
cases, association with (anthropogenically derived) NO2 indicates that HONO formation may fall with future vehicle eet evolution and tailpipe emissions
reduction. photolysis of ozone and subsequent reaction of electronically
excited oxygen atoms with water vapour, a series of recent eld
campaigns have shown that the photolysis of nitrous acid
(HONO) is a major/the dominant primary OH precursor in the
continental boundary layer.3,4 photolysis of ozone and subsequent reaction of electronically
excited oxygen atoms with water vapour, a series of recent eld
campaigns have shown that the photolysis of nitrous acid
(HONO) is a major/the dominant primary OH precursor in the
continental boundary layer.3,4 Take down policy down policy
the University of Birmingham exercises care and attention in making items available there are rare occasions when an it
ded in error or has been deemed to be commercially or otherwise sensitive. this is the case for this document, please contact UBIRA@lists.bham.ac.uk providing details and we will remove access
ely and investigate. If you believe that this is the case for this document, please contact UBIRA@lists.bham.ac.uk providing details and we
the work immediately and investigate. Download date: 24. Oct. 2024 Open Access Article. Published on 20 July 2021. Downloaded on 8/11/2021 7:19:11 PM.
This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. (R3) 2NO2 + H2O / HONO + HNO3 Primary sources of HONO include vehicle emissions,9,10
biomass burning,11–13 microbial activities in soils14,15 and asso-
ciated with biocrusts.16 Laboratory studies have shown that
photoenhanced HONO production can occur following NO2
uptake to surfaces including soot,17 aromatic species,18 humic
acids19,20 and TiO2.21 Photoenhanced NO2 heterogeneous reac-
tions on ‘urban grime’ on building surfaces22 have also been
shown to generate appreciable levels of HONO. Further poten-
tial sources are photolysis of nitric acid,23 nitrophenols24 and of
particulate nitrate.25–27 Many eld and laboratory studies have been carried out to
investigate the sources of HONO in different environments, e.g. urban, rural, coastal, forest and vehicle tunnels.5,9,28–34 In
a recent study, Tong, et al.35 performed measurements at urban
and suburban environments in Beijing during winter, and
observed that direct emission and homogenous gas phase
sources made a larger contribution in urban areas, while
heterogeneous sources were suggested to be more signicant
away from urban centres. A recurrent challenge remains
quantitatively reconciling daytime HONO concentrations in
many urban environments with the current known sources.5–7 3
Approach to exploring the
anticipated response of HONO during
the solar eclipse Here, we take advantage of a near-total solar eclipse (90%
attenuation) as a natural short-term perturbation to atmo-
spheric photochemistry to explore the chemical processes
affecting HONO abundance. The reduction in photolysis
frequencies during an eclipse is effectively uniform over a large
area (relative to the chemical lifetime and hence spatial foot-
print of HONO, NOx and related species), of relevance to the
challenges
of
heterogeneity.36,37
Comparison
of
observed
temporal behaviour during the eclipse therefore provides
a unique test of our understanding of HONO photochemistry. We report the temporal variation of HONO, NOx, O3 and aerosol
characteristics during the eclipse, and use these measurements
to explore the nature of sources of HONO which are consistent
with their behaviour. HONO abundance is expected to respond to rapid changes in
light conditions (i.e. with and without the shading condition of
the solar eclipse), to an extent dependent upon the dominant
HONO source and sink mechanisms, and their respective vari-
ation with solar radiation levels. To explore the variation in
HONO (and other species) abundance anticipated across the
eclipse timeframe (ca. 2 hours) a series of simple simulations
were performed representing different scenarios (mechanisms)
for potential atmospheric processes/emissions that may form
and remove HONO. We do not attempt to reproduce the
detailed gas- or heterogeneous chemistry using a formal model
or chemical mechanism – as many of the species that would be
required were not measured (notably, no VOC data were avail-
able, precluding modelling HOx sources or sinks). However,
with some assumptions (below), over the limited duration of the
eclipse event, the time evolution of HONO can be calculated
from its initial abundance, the time variation of various
possible source term(s), and its chemical removal which is
dominated ($95%, see below) by photolysis. This allows us to
explore the response of potential HONO source(s), para-
meterised in terms of the measured variation of sunlight, NOx
abundance and aerosol particle parameters, by comparison 1
Introduction Atmospheric chemical processing in the sunlit troposphere is
driven primarily by the OH radical, which initiates the removal
of most organic compounds, and drives the formation of ozone,
and secondary organic and inorganic aerosols.1,2 While the
dominant primary source of OH in the free troposphere is the HONO + hn (l < 440 nm) / NO + OH
(R1) (R1) HONO is formed from the slow homogeneous reaction
between OH and NO, and in this sense acts as a photolabile
reservoir for OH (and NO) with (R1) and (R2) forming a null
cycle; however observed levels of HONO during daytime are
typically one order of magnitude higher than (R1) and (R2) in
isolation predict, indicating the presence of additional HONO
formation mechanisms or emissions (and hence net OH Environ. Sci.: Atmos. © 2021 The Author(s). Published by the Royal Society of Chemistry Environ. Sci.: Atmos. View Article Online View Article Online Paper Environmental Science: Atmospheres Paper where gas phase HONO is sampled, using a stripping coil, into
an acidic solution and is derivatived into an azo dye. The light
absorption at 550 nm of the azo dye is then measured with
a spectrometer using an optical path length of 2.4 m. The
LOPAP was operated and calibrated according to the standard
procedures,39 with a sampling height of 3 m above ground level,
and data acquired at 5 min time resolution. Baseline (zero)
measurements, obtained by sampling zero air, were taken at
frequent intervals (8 hours). The detection limit (2s) under the
instrument operating conditions of these measurements was
determined to be 6 pptV. In addition to the LOPAP, co-located
measurements of NO, NO2 and NOx (Thermo Scientic 42c-
Mo
convertor
for
NO2
measurement,
hence
potentially
responding to NOy interferences), and ozone (Thermo Scientic
49i) were performed. An optical particle spectrophotometer (TSI
3330) measured the particle number size distribution, and
converted from particle number concentration (PNC) to total
surface area (TSA) via the TSI AIM soware (version 9), which
assumes particle sphericity. NOx, O3 and particle data were
obtained at a 1 min time resolution. Meteorological data (rela-
tive humidity, solar intensity, air temperature, and wind speed
and direction) was obtained at 1 min time resolution from
a nearby weather station operated by the University of
Birmingham/UK Met Office, located within 100 m of the EROS
site. Environ. Sci.: Atmos. 1
Introduction At the University of Birmingham site, the eclipse (20th
March 2015) occurred from 08:25 to 10:40 GMT (rst/last
contact) with the maximum coverage (89% of the sun's face)
occurring at 9:31.40 production).5–7 The dark formation of HONO through the
heterogeneous hydrolysis of nitrogen dioxide ((R3) below) has
been known for a number of years;8 however recent laboratory
and eld students have identied a number of additional
candidate HONO formation mechanisms and/or emission
sources. OH + NO + M / HONO + M
(R2)
2NO2 + H2O / HONO + HNO3
(R3) OH + NO + M / HONO + M
(R2)
2NO2 + H2O / HONO + HNO3
(R3) (R2) © 2021 The Author(s). Published by the Royal Society of Chemistry 2
Observations during a solar eclipse HONO, NOx, O3 and particle number concentration were
measured at an urban background monitoring station situated
(52.45 N, 1.93 W) within the Edgbaston campus of the
University of Birmingham, UK. Possible anthropogenic emis-
sion sources located nearby include a suburban rail line (north-
west, ca. 90 m away), and a suburban road (east, 125 m away). HONO was measured using a long-path absorption photometer
(LOPAP).38 Briey, the LOPAP is a wet chemical technique, © 2021 The Author(s). Published by the Royal Society of Chemistry Environmental Science: Atmospheres
View Article Online View Article Online Environmental Science: Atmospheres Paper with the observed HONO variation over the course of the
eclipse. maintain the observed levels (Section 4.2 below) and would, to
an extent, offset the HONO + OH reaction. We also discount the
proposed,
but
uncertain,
H2O þ NO*
2mechanism.44–46
A
number of assumptions regarding advection/mixing are made:
we assume air mass homogeneity over the timescale of the
eclipse
period,
and
neglect vertical
mixing. The former
assumption is likely justied considering the background
nature of the measurement site and limited timespan of the
analysis; recent measurements of vertical proles of HONO in
urban areas47 point to a broadly uniform prole near to the
ground (<200 m), with some increase near ground level attrib-
uted to traffic source mixing and ground surface production –
although the measurement period spans the early morning
time when boundary layer breakup/enhanced vertical mixing
may occur. Finally, each potential HONO source mechanism is
considered in isolation, while in reality multiple mechanisms
likely occur in parallel. These and other assumptions are
considered further in the discussion, below. p
The attenuation of photolysis frequencies was approximated
using values obtained from the TUV model41 for clear-sky
conditions, with the reduction in j values modelled as a 1-p
sine wave. We note that this is an approximation, both to the
actual photolysis frequencies and the geometric coverage of the
solar disk, but one which is acceptable in the context of the
analysis which follows. The calculated jNO2 values prior to and
subsequent to the eclipse were consistent with those that would
be estimated using photo-stationary steady state (NOx/O3
abundance, neglecting HO2/RO2 reactions) to within 5.6 1.5%
(pre-eclipse), 21 1.9% (during the eclipse) and 6.4 2.0%
(post-eclipse). The different scenarios examined are described
in Table 1. Open Access Article. Published on 20 July 2021. Downloaded on 8/11/2021 7:19:11 PM.
This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. In each case, the initial HONO concentration [HONO]t¼0,
and the source mechanism scaling constant C (Table 1) was
determined from the measured values of HONO, NOx, jNO2 and
total aerosol surface area (TSA) as appropriate. Both [HONO]t¼0
and C were optimised independently, for each source scenario,
by minimising the sum of squares of residuals between calcu-
lated and observed HONO concentrations for the two 1 hour
periods perior to and immediately following the eclipse period
(from 07:30–08:30 and 10:40–11:40 respectively). We adopt the
pre/post eclipse period to allow prediction of hypothetical, no-
eclipse HONO concentrations; we test the sensitivity of this
time period selection by repeating the tting including both the
pre/post and eclipse periods (ESI, Fig. S1†); while the optimised
parameters vary slightly the conclusions regarding which
mechanism(s) are consistent/inconsistent with the data are
unchanged). All other species were set to their actual observed
levels for the relevant point in time. It is important to note that
we do not assume HONO is in steady state/equilibrium with its
production and removal terms. For each source scenario, we
repeat the analysis under two conditions – rstly for the actual
(eclipse inuenced) reduction in solar intensity (photolysis d[HONO]/dt ¼ Sx jHONO [HONO]
(1) (1) d[HONO]/dt ¼ Sx jHONO [HONO] The different HONO sources considered, Sx, are listed in
Table 1. In each case HONO concentrations were calculated by
numerically evaluating the differential equations for HONO
formation/removal, using time-dependent observed values of
the input parameters (jNO2, [NO2] etc), on a 60 second timestep. A number of assumptions are inherent in this approach: eqn
(1) neglects HONO sinks other than photolysis (i.e., reaction
with OH). Using OH measurements performed near to the
measurement site during a previous campaign,42,43 the rate of
loss of HONO due to reaction with OH is calculated to be 1.1–
5.0% (using [OH] of 2 106 to 9 106 mol cm3) of that due to
photolysis, at the lowest photolysis frequency corresponding to
the eclipse maximum, hence this approximation is reasonable. Eqn (1) also neglects homogeneous sources of HONO – from the
(slow) termolecular OH + NO chemistry, which we estimate
would contribute <7% of the HONO production necessary to The different HONO sources considered, Sx, are listed in
Table 1. Environ. Sci.: Atmos. 2
Observations during a solar eclipse In each case, the calculation was initiated at 07:30
GMT, prior to the start of the eclipse, and the subsequent
evolution of HONO calculated according to the various alter-
native mechanisms presented below. In general, each case may
be represented via eqn (1), where Sx is the source term for
HONO formation in each scenario: Open Access Article. Published on 20 July 2021. Downloaded on 8/11/2021 7:19:11 PM.
This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. During the eclipse, weather conditions were calm (wind speed
of 0.5 0.4 m s1) with a mean relative humidity of 76 6%
and temperature of 6.7 1.7 C, which are typical for the time of
year in the UK. In particular, the usual rise in temperature with
daytime was observed, which may have begun the process of
boundary layer break-up, although there was little evidence for
change in NOx or PM levels associated with this. Relatively little
change in HONO concentration was observed during the eclipse
period (Fig. 1a), suggesting that closely balanced photochemical
source and photochemical sink mechanisms may be domi-
nating HONO abundance. Ozone and NOx were observed to be
tightly coupled during eclipse period (08:25 am to 10:40 am
GMT, Fig. 1b), and to follow their expected photochemistry,48
with NO and O3 levels falling and NO2 rising. During the eclipse,
little variation was observed in the particle TSA (Fig. 1c), which
was found to be reasonably well correlated with the HONO
concentration (R2 ¼ 0.63), possibly suggesting a common
source or dilution process. Vehicle emissions have been iden-
tied as a signicant source of HONO in urban environments,
where a low HONO/NOx ratio is commonly used as a proxy
parameter to evaluate HONO emissions from traffic.3,49,50 In the
present study, high HONO/NO2 (1.9–4.1%) and HONO/NOx
(1.0–1.8%) ratios were observed before, during and aer the
eclipse. Previous empirically determined emission ratios of
HONO/NOx based on tunnel and car exhaust studies are
between 0.3% and 1.0%,9,51 with recent evaluations at the high
end of this range, e.g. 1.24 0.35%50 and 0.72–1.01%33 (deter-
mined in Birmingham). Thus the measured HONO/NOx ratio is
at the top end of/in excess of the ratio of expected for vehicle
emissions – suggesting that these are not the sole source of
HONO at this site. Direct emission of HONO from vehicles is another potential
source in urban areas.9,33,50,51 In an urban area, particle total
surface area (TSA) can be used as a reasonably conserved tracer
for vehicle emissions.54 Case 5 (Fig. 2E) considers a scenario
where the HONO source scales with the TSA; however this also
does not match the measured HONO levels during the eclipse,
implying that a source scaled to particle TSA alone – or direct
vehicular emission of HONO in isolation – cannot explain the
observed HONO concentrations in this environment. Open Access Article. Published on 20 July 2021. Downloaded on 8/11/2021 7:19:11 PM.
This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. Similar
results were observed when particle number concentration was
used as a metric of traffic particulate matter emissions, as
opposed to TSA (not shown). However, HONO production
proportional to the product of jNO2 and TSA was a better match
to observed (Case 6, Fig. 2F). Open Access Article. Published on 20 July 2021. Downloaded on 8/11/2021 7:19:11 PM.
This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. In each case HONO concentrations were calculated by
numerically evaluating the differential equations for HONO
formation/removal, using time-dependent observed values of
the input parameters (jNO2, [NO2] etc), on a 60 second timestep. A number of assumptions are inherent in this approach: eqn
(1) neglects HONO sinks other than photolysis (i.e., reaction
with OH). Using OH measurements performed near to the
measurement site during a previous campaign,42,43 the rate of
loss of HONO due to reaction with OH is calculated to be 1.1–
5.0% (using [OH] of 2 106 to 9 106 mol cm3) of that due to
photolysis, at the lowest photolysis frequency corresponding to
the eclipse maximum, hence this approximation is reasonable. Eqn (1) also neglects homogeneous sources of HONO – from the
(slow) termolecular OH + NO chemistry, which we estimate
would contribute <7% of the HONO production necessary to Table 1
Summary of HONO production scenarios and optimised parameter values (HONO source strength determined in ppb s1 and NO2
mixing ratio in ppb)
Cases Potential sources
HONO source strength (S)/ppb
s1
Scale factor (C)
Initial HONO
([HONO]t¼0)/
ppb
Fit RMS
residual/
ppb
Case 1 No source term
S1 ¼ 0
N/A
1.2
0.325
Case 2 Photolysis-related
S2 ¼ C jNO2
7.2 102 ppb
1.04
0.118
Case 3 NO2-related
S3 ¼ C [NO2]
1.57 105 s1
0.71
0.067
Case 4 Source(s) related to photolysis and NO2 S4 ¼ C jNO2 [NO2]
3.99 103
1.0
0.096
Case 5 Source(s) related to aerosol surface area
(TSA)
S5 ¼ C TSA
1.65 106 ppb s1 (mm2 cm3)1 0.68
0.078
Case 6 Source(s) related to photolysis and TSA S6 ¼ C jNO2 TSA
4.42 104 ppb (mm2 cm3)1
0.97
0.090
Case 7 Source(s) related to NO2 and TSA
S7 ¼ C [NO2] TSA
7.61 108 s1 (mm2 cm3)1
0.67
0.105
Case 8 Source(s) related to photolysis, NO2 and
TSA
S8 ¼ C jNO2 [NO2] TSA
2.31 105 (mm2 cm3)1
0.94
0.075 Table 1
Summary of HONO production scenarios and optimised parameter values (HONO source strength determined in ppb s1 and NO2
mixing ratio in ppb) © 2021 The Author(s). Published by the Royal Society of Chemistry © 2021 The Author(s). Published by the Royal Society of Chemistry Environ. Sci.: Atmos. 4
Results & discussion Previous work in urban areas has suggested a daytime source
of HONO that scaled with NO2 levels and sunlight, specically
jNO2.7,52 Case 2 considers a HONO source term which is directly
proportional to jNO2. The predicted HONO behaviour broadly
matches that observed, indicating that in reality, the predomi-
nant source is likely to scale with jNO2 or some similar measure
of solar insolation. Dark conversion of NO2 to HONO upon
moist surfaces is a long-established HONO formation mecha-
nism (e.g. Finlayson-Pitts, et al.53), and case 3 considers a HONO
source term which is directly proportional to NO2 – however this
scenario shows a poor t of predicted to observed HONO levels,
indicating that the predominant source does not scale with
[NO2] in isolation. Photosensitized conversion of NO2 on
surfaces containing organics has been identied as a signicant
daytime HONO source in a number of studies.18,19 To explore the
effect of such reactions, a HONO source which scaled with the
product of jNO2 and the NO2 mixing ratio was explored in case 4
(Fig. 2D), where a signicantly better t was observed, indi-
cating that a source term which scaled with jNO2 [NO2] could
account
for
the
majority
of
HONO
production
in
this
environment. Environ. Sci.: Atmos. Open Access Article. Published on 20 July 2021. Downloaded on 8/11/2021 7:19:11 PM.
This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. View Article Online Paper
ew Article Online Paper
rticle Online Environmental Science: Atmospheres Paper from these measurements (5 105 mol cm3), a steady state
HONO level of 0.04 ppb is obtained, approximately a factor of 15
lower than that observed mid-eclipse, indicating that OH + NO
is not responsible for HONO formation in isolation, and rather
forms a small contributor to the total HONO source. The role of
OH + NO as a secondary HONO production term, alongside
other, larger mechanisms, is considered further below. frequencies), and secondly, for comparison, a hypothetical
scenario where no eclipse occurred and clear-sky photolysis (as
derived from the changing SZA) applied throughout the 2 hour
time period. The resulting predicted and observed HONO
concentrations are shown in Fig. 2, and discussed below. 4.2
Evaluation of HONO source scenarios Cases 7 and 8 (Fig. 2G and H respectively) represent HONO
formation via NO2 conversion on the surface of aerosol parti-
cles, without (case 7/Fig. 2G) and with (case 8/Fig. 2H) photo-
enhancement respectively. These represent secondary chem-
ical (photochemical) sources involving aerosol particles. In
these cases, the HONO production rate is assumed to scale with
TSA [NO2] and jNO2 TSA [NO2], respectively. Case 7 did
not reproduce the observed behaviour, indicating that such
heterogeneous aerosol conversion of NO2 to HONO, without
photoenhancement, could not, in isolation, account for HONO
production in this environment. Case 8 showed signicantly
closer reproduction of the measured HONO evolution under the
eclipse conditions, giving the best t (lowest RMS residual) of
all the scenarios considered, and indicating that a mechanism
dependent upon solar insolation (jNO2), NO2 abundance and Fig. 2 compares the observed (red line/points) and simulated
(actual, eclipse condition – solid blue line; hypothetical no
eclipse condition – dashed blue line) time evolution of the
HONO abundance for each scenario considered. Case 1 (Fig. 2A)
considers a scenario where no HONO source exists; this clearly
diverges from the observed behaviour, indicating that addi-
tional HONO sources are in fact present. Note that the t
(minimised RMS residual) is affected by inclusion of the second
tting period, aer the eclipse – but in which there is little
sensitivity to the initial HONO level, due to the rapid photolytic
decay. The reaction of OH radicals with NO forms HONO (via
(R2)); no OH data were available during these experiments, but
OH levels may be estimated from prior observations at this
location (Heard, et al.42 – taking the mean OH level for 09:30 © 2021 The Author(s). Published by the Royal Society of Chemistry View Article Online Paper Environmental Science: Atmospheres ace area is able to reproduce the observed HONO
cross the eclipse The similarit of eclipse and non
– as observed for the actual HONO. While we obtain the opt
t for case 8 in comparison to (e g ) case 2 the geom
series of (a) measured HONO and solar intensity, (b) NOx and ozone, and (c) particle total surface area (TSA) and HONO/NO2 Open Access Article. Published on 20 July 2021. Downloaded on 8/11/2021 7:19:11 PM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. Fig. 4.2
Evaluation of HONO source scenarios 1
Time series of (a) measured HONO and solar intensity, (b) NOx and ozone, and (c) particle total surface area (TSA) and HONO/NO2 ratio. – as observed for the actual HONO. While we obtain the optimal
t for case 8, in comparison to (e.g.) case 2, the geometric
surface area of ground surfaces signicantly exceeds that of – as observed for the actual HONO. While we obtain the optimal
t for case 8, in comparison to (e.g.) case 2, the geometric
surface area of ground surfaces signicantly exceeds that of aerosol surface area is able to reproduce the observed HONO
behaviour across the eclipse. The similarity of eclipse- and non-
eclipse simulations reects the balanced photolytic dependence aerosol surface area is able to reproduce the observed HONO
behaviour across the eclipse. The similarity of eclipse- and non-
eclipse simulations reects the balanced photolytic dependence © 2021 The Author(s). Published by the Royal Society of Chemistry Environ. Sci.: Atmos. View Article Online View Article Online Environmental Science: Atmospheres Environmental Science: Atmospheres Comparison between measured (red points) and calculated HONO mixing ratio during the solar eclipse, using actual eclipse ph
ncies (solid blue line) and (hypothetical) non-eclipse photolysis frequencies (dotted blue line). Shading indicates eclipse duration f
contact. Source scenarios (panels A–H) as defined in Table 1. onmental Science: Atmospheres Open Access Article. Published on 20 July 2021. Downloaded on 8/11/2021 7:19:11 PM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. Fig. 2
Comparison between measured (red points) and calculated HONO mixing ratio during the solar eclipse, using actual eclipse photolysis
frequencies (solid blue line) and (hypothetical) non-eclipse photolysis frequencies (dotted blue line). Shading indicates eclipse duration from first
to last contact. Source scenarios (panels A–H) as defined in Table 1. © 2021 The Author(s). Published by the Royal Society of Chemistry Environ. Sci.: Atmos. View Article Online Paper Paper Environmental Science: Atmospheres son between observed HONO mixing ratios, and HONOconserved (the HONO abundance calculated after accounting for
der eclipse and (hypothetical) non-eclipse scenarios, (b) variation in HONOconserved/NOx ratio under actual (eclipse) and Open Access Article. Published on 20 July 2021. Downloaded on 8/11/2021 7:19:11 PM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. Fig. 3
(a) Comparison between observed HONO mixing ratios, and HONOconserved (the HONO abundance calculated after accounting for
photolytic loss) under eclipse and (hypothetical) non-eclipse scenarios, (b) variation in HONOconserved/NOx ratio under actual (eclipse) and
hypothetical (non-eclipse) conditions. Fig. 3
(a) Comparison between observed HONO mixing ratios, and HONOconserved (the HONO abundance calculated after accounting for
photolytic loss) under eclipse and (hypothetical) non-eclipse scenarios, (b) variation in HONOconserved/NOx ratio under actual (eclipse) and
hypothetical (non-eclipse) conditions. aerosol particles (by at least a factor of 16-fold, assuming an
aerosol surface area of 300 mm2 cm3 and boundary layer height
of 200 m, which may be representative of the measurement
location early in the morning – low winds, under an eclipse
condition), and hence it may be likely that ground surface-
mediated
photoenhanced
conversion
of
NO2
dominated
HONO formation at this location, potentially augmented by
contributions from aerosol surfaces. wintertime reported previously by Heard, et al.42 resulted in
a slight improvement to the statistical t – although increasing
degrees of freedom would be expected to improve agreement. © 2021 The Author(s). Published by the Royal Society of Chemistry There are no conicts to declare. There are no conicts to declare. Acknowledgements We thank the University of Birmingham for supporting Ajit
Singh through the Elite Scholarship Scheme. The authors would
also like to acknowledge UK Natural Environmental Research
Council project SNAABL (Sources of Nitrous Acid in the Atmo-
spheric Boundary Layer; NE/M013545/1) for funding this
research. Open Access Article. Published on 20 July 2021. Downloaded on 8/11/2021 7:19:11 PM.
This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. (2) where the [HONO]conserved estimates the total HONO concen-
tration that has been formed/emitted up to each point in time,
derived from the observations aer accounting for photolysis. HONO removal through reaction with OH is neglected, as
above. Cn and Cn1 represent the observed HONO at time tn and
tn1, while Dt is the time difference. A modest difference in
HONOconserved between the eclipse and without-eclipse condi-
tions can be seen (Fig. 3a). Concentrating on the actual (eclipse)
condition, the conserved HONO shows an approximately
monotonic rise, indicating an approximately constant emission
rate. This is at variance with traffic-related emissions, which
would be expected to show the characteristic “rush hour”
behaviour,
i.e. peaking
at
8–9
am. Fig. 3b
shows
the
HONOconserved/NOx ratio – an approximately constant ratio
would be expected if direct emissions dominated production (of
both species), assuming NOx is conserved, however in fact
signicant variation with time is observed (the chemical NOx
lifetime is estimated to be 17.5 hours with respect to OH + NO2,
using the measured OH from Heard, et al.,42 indicating that NOx
can be considered to be approximately conserved on our time-
scale of 3.5 hours). The lack of a traffic/rush-hour pattern in the
HONOconserved and HONOconserved/NOx ratio further supports
the inference that direct vehicular/combustion emissions, in
isolation, are not the dominant source of HONO at this location,
with some additional, photochemical/heterogeneous atmo-
spheric chemical term is required to account for the observed
behaviour. However, the results from scenarios 1–8, which
point to involvement of NO2 in all HONO production mecha-
nisms with a degree of consistency with the observed behaviour,
reinforce the ultimate importance of vehicle emissions (as the
overall source of NOx). Environmental Science: Atmospheres As
a further evaluation of the performance of scenario 8, we
applied this analysis approach to data from the days preceding
and following the eclipse event (i.e. 19th and 21st March
respectively); the simulation was able to satisfactorily reproduce
the observed HONO levels (ESI, Fig. S2†). We note that in cases
3, 4, 7 and 8 the NOx would vary due to changing PSS in the
(hypothetical) non-eclipse case. This approximation has the
effect of biasing the non-eclipse (hypothetical) simulations high
(as the real NO2 would be lower, in the absence of the eclipse-
derived photolysis attenuation). In reality, it is likely that a combination of mechanisms
occur in parallel, and addition of the OH + NO reaction, with OH
levels assumed equal to those measured42 and scaled to the
relative solar intensity expected during the eclipse (which may
underestimate OH, given the importance of non-photolytic
sources
such
as
alkene
ozonolysis
at
this
location
in Environ. Sci.: Atmos. View Article Online Environmental Science: Atmospheres Paper proportionality, by comparing the observed and predicted
HONO levels across the eclipse perturbation. Source terms
lacking a photolytic component – proportional to NO2 concen-
tration, aerosol particle number or aerosol particle surface area
– were not consistent with the observed behaviour. Source terms
with an insolation dependence, here represented as jNO2, better
replicated the measurements, with strong evidence for the
involvement of NO2, and the best agreement (in isolation) ob-
tained for a combination of jNO2, NO2 concentration and aerosol
surface area. Both the observed source behaviour, and the
temporal
variation
in
the
HONOconserved
levels
and
the
HONOconserved/NOx ratio, were inconsistent with primary traffic
emissions dominating HONO production in this urban back-
ground location – although of course road traffic is ultimately
the dominant emission source for NOx, and hence the key
HONO precursor NO2 identied here. Accordingly, future
vehicle eet evolution, which is expected to lead to signicantly
lower urban NOx levels through improved emissions technolo-
gies and penetration of electric vehicles, will also lead to lower
HONO concentrations, and hence lower local OH radical
production and local atmospheric oxidation in urban boundary
layer locations such as that considered here. References 1 B. W. LaFranchi, G. M. Wolfe, J. A. Thornton, S. A. Harrold,
E. C. Browne, K. E. Min, P. J. Wooldridge, J. B. Gilman,
W. C. Kuster, P. D. Goldan, J. A. de Gouw, M. McKay,
A. H. Goldstein, X. Ren, J. Mao and R. C. Cohen, Closing
the peroxy acetyl nitrate budget: observations of acyl peroxy
nitrates (PAN, PPN, and MPAN) during BEARPEX 2007,
Atmos. Chem. Phys., 2009, 9, 7623–7641, DOI: 10.5194/acp-
9-7623-2009. Environ. Sci.: Atmos. 4.3
Insight from temporal variation of HONO production A further analysis was undertaken to derive the cumulative
HONO production with time, i.e. to remove the effect of the loss
of HONO through photolysis under both the real (eclipse) and
hypothetical (non-eclipse) conditions. This quantity – hypo-
thetically conserved HONO, [HONO]conserved, was calculated
using eqn (2): ½HONOconserved ¼ Cn þ
X
t¼n ðn1Þ
jHONO ðCn þ Cn1Þ Dt
2
(2) ½HONOconserved ¼ Cn þ
X
t¼n ðn1Þ
jHONO ðCn þ Cn1Þ Dt
2
(2) 5
Concluding remarks Phys., 2014, 14, 2805–2822, DOI: 10.5194/acp-
14-2805-2014. 15 R. Oswald, T. Behrendt, M. Ermel, D. Wu, H. Su, Y. Cheng,
C. Breuninger,
A. Moravek,
E. Mougin,
C. Delon,
B. Loubet, A. Pommerening-R¨oser, M. S¨orgel, U. P¨oschl,
T. Hoffmann, M. O. Andreae, F. X. Meixner and I. Trebs,
HONO emissions from soil bacteria as a major source of
atmospheric reactive nitrogen, Science, 2013, 341, 1233–
1235, DOI: 10.1126/science.1242266. M. Haeffelin and J. F. Doussin, Study of the unknown
HONO daytime source at a European suburban site during
the
MEGAPOLI
summer
and
winter
eld
campaigns,
Atmos. Chem. Phys., 2014, 14, 2805–2822, DOI: 10.5194/acp-
14-2805-2014. 16 B. Weber, D. Wu, A. Tamm, N. Ruckteschler, E. Rodr´ıguez-
Caballero, J. Steinkamp, H. Meusel, W. Elbert, T. Behrendt,
M. S¨orgel, Y. Cheng, P. J. Crutzen, H. Su and U. P¨oschl,
Biological soil crusts accelerate the nitrogen cycle through
large NO and HONO emissions in drylands, Proc. Natl. Acad. Sci. U. S. A., 2015, 112, 15384–15389, DOI: 10.1073/
pnas.1515818112. 7 J. D. Lee, L. K. Whalley, D. E. Heard, D. Stone, R. E. Dunmore,
J. F. Hamilton, D. E. Young, J. D. Allan, S. Laufs and
J. Kleffmann, Detailed budget analysis of HONO in central
London reveals a missing daytime source, Atmos. Chem. Phys., 2016, 16, 2747–2764, DOI: 10.5194/acp-16-2747-2016. 8 B. J. Finlayson-Pitts, L. M. Wingen, A. L. Sumner, D. Syomin
and K. A. Ramazan, The heterogeneous hydrolysis of NO2 in
laboratory systems and in outdoor and indoor atmospheres:
an integrated mechanism, Phys. Chem. Chem. Phys., 2003, 5,
223–242, DOI: 10.1039/b208564J. 17 M. E. Monge, B. D'Anna, L. Mazri, A. Giroir-Fendler,
M. Ammann, D. Donaldson and C. George, Light changes
the atmospheric reactivity of soot, Proc. Natl. Acad. Sci. U. S. A., 2010, 107, 6605–6609, DOI: 10.1073/pnas.0908341107. 9 R. Kurtenbach, K. H. Becker, J. A. G. Gomes, J. Kleffmann,
J. C. L¨orzer, M. Spittler, P. Wiesen, R. Ackermann, A. Geyer
and
U. Platt,
Investigations
of
emissions
and
heterogeneous formation of HONO in a road traffic tunnel,
Atmos. Environ., 2001, 35, 3385–3394, DOI: 10.1016/s1352-
2310(01)00138-8. 18 C. George, R. S. Strekowski, J. Kleffmann, K. Stemmler and
M. Ammann, Photoenhanced uptake of gaseous NO2 on
solid organic compounds: a photochemical source of
HONO?,
Faraday
Discuss.,
2005,
130,
195–210,
DOI:
10.1039/b417888m. 10 B. Rappengl¨uck, G. Lubertino, S. Alvarez, J. Golovko,
B. Czader and L. 5
Concluding remarks Ackermann, Radical precursors and
related species from traffic as observed and modeled at an
urban highway junction, J. Air Waste Manage. Assoc., 2013,
63, 1270–1286, DOI: 10.1080/10962247.2013.822438. 19 K. Stemmler, M. Ammann, C. Donders, J. Kleffmann and
C. George, Photosensitized reduction of nitrogen dioxide
on humic acid as a source of nitrous acid, Nature, 2006,
440, 195–198, DOI: 10.1038/nature04603. 20 S. Laufs, M. Cazaunau, P. Stella, R. Kurtenbach, P. Cellier,
A. Mellouki, B. Loubet and J. Kleffmann, Diurnal uxes of
HONO above a crop rotation, Atmos. Chem. Phys., 2017, 17,
6907–6923, DOI: 10.5194/acp-17-6907-2017. 11 R. J. Yokelson, J. D. Crounse, P. F. DeCarlo, T. Karl,
S. Urbanski,
E. Atlas,
T. Campos,
Y. Shinozuka,
V. Kapustin, A. D. Clarke, A. Weinheimer, D. J. Knapp,
D. D. Montzka, J. Holloway, P. Weibring, F. Flocke,
W. Zheng, D. Toohey, P. O. Wennberg, C. Wiedinmyer,
L. Mauldin, A. Fried, D. Richter, J. Walega, J. L. Jimenez,
K. Adachi, P. R. Buseck, S. R. Hall and R. Shetter,
Emissions from biomass burning in the Yucatan, Atmos. Chem. Phys., 2009, 9, 5785–5812, DOI: 10.5194/acp-9-5785-
2009. 21 J. M. Langridge, R. J. Gustafsson, P. T. Griffiths, R. A. Cox,
R. M. Lambert and R. L. Jones, Solar driven nitrous acid
formation
on
building
material
surfaces
containing
titanium dioxide: a concern for air quality in urban areas?,
Atmos. Environ.,
2009,
43,
5128–5131,
DOI:
10.1016/
j.atmosenv.2009.06.046. 22 A. M. Baergen
and
D. J. Donaldson,
Photochemical
renoxication of nitric acid on real urban grime, Environ. Sci. Technol., 2013, 47, 815–820, DOI: 10.1021/es3037862. 12 J. M. Roberts, P. Veres, C. Warneke, J. A. Neuman,
R. A. Washenfelder,
S. S. Brown,
M. Baasandorj,
J. B. Burkholder, I. R. Burling, T. J. Johnson, R. J. Yokelson
and J. de Gouw, Measurement of HONO, HNCO, and other
inorganic acids by negative-ion proton-transfer chemical-
ionization mass spectrometry (NI-PT-CIMS): application to
biomass burning emissions, Atmos. Meas. Tech., 2010, 3,
981–990, DOI: 10.5194/amt-3-981-2010. 23 X. Zhou, H. Gao, Y. He, G. Huang, S. B. Bertman, K. Civerolo
and J. Schwab, Nitric acid photolysis on surfaces in low-NOx
environments:
signicant
atmospheric
implications,
Geophys. Res. Lett.,
2003,
30,
2217,
DOI:
10.1029/
2003gl018620. 24 I. Bejan, Y. A. El Aal, I. Barnes, T. Benter, B. Bohn, P. Wiesen
and J. Kleffmann, The photolysis of ortho-nitrophenols:
a new gas phase source of HONO, Phys. Chem. Chem. Phys.,
2006, 8, 2028–2035, DOI: 10.1039/b516590c. 13 N. Theys, R. Volkamer, J. F. M¨uller, K. 5
Concluding remarks HONO, NOx, O3 and particle number size distributions were
measured at an urban background location over the course of
a near-total solar eclipse in Birmingham, UK. The observed NO,
NO2 and O3 responded as anticipated from the well-understood
atmospheric photochemistry. HONO levels showed limited
variation across the eclipse period; given HONO removal is
overwhelmingly dominated by photolysis, this observation
indicates that photolytic sources predominate in producing
HONO at this location. A simplied approach was adopted to
explore which factors were required within the HONO source
mechanism,
assuming
direct
and
multiplicative 2 X. Ren, H. Gao, X. Zhou, J. D. Crounse, P. O. Wennberg,
E. C. Browne, B. W. LaFranchi, R. C. Cohen, M. McKay,
A. H. Goldstein and J. Mao, Measurement of atmospheric
nitrous acid at Bodgett Forest during BEARPEX2007, Atmos. Chem. Phys., 2010, 10, 6283–6294, DOI: 10.5194/acp-10-
6283-2010. 3 J. Kleffmann, Daytime sources of nitrous acid (HONO) in the
atmospheric boundary layer, ChemPhysChem, 2007, 8, 1137–
1144, DOI: 10.1002/cphc.200700016. 4 F. Spataro and A. Ianniello, Sources of atmospheric nitrous
acid: state of the science, current research needs, and © 2021 The Author(s). Published by the Royal Society of Chemistry View Article Online Paper Paper Environmental Science: Atmospheres future prospects, J. Air Waste Manage. Assoc., 2014, 64, 1232–
1250, DOI: 10.1080/10962247.2014.952846. M. Van Roozendael, Global nitrous acid emissions and
levels of regional oxidants enhanced by wildres, Nat. Geosci., 2020, 13, 681–686, DOI: 10.1038/s41561-020-0637-7. M. Van Roozendael, Global nitrous acid emissions and
levels of regional oxidants enhanced by wildres, Nat. Geosci., 2020, 13, 681–686, DOI: 10.1038/s41561-020-0637-7. 5 J. Kleffmann, T. Gavriloaiei, A. Hofzumahaus, F. Holland, 14 H. Su, Y. Cheng, R. Oswald, T. Behrendt, I. Trebs,
F. X. Meixner, M. O. Andreae, P. Cheng, Y. Zhang and
U. P¨oschl, Soil nitrite as a source of atmospheric HONO
and OH radicals, Science, 2011, 333, 1616–1618, DOI:
10.1126/science.1207687. R. Koppmann,
L. Rupp,
E. Schlosser,
M. Siese
and
A. Wahner, Daytime formation of nitrous acid: a major
source of OH radicals in a forest, Geophys. Res. Lett., 2005,
32, L05818, DOI: 10.1029/2005gl022524. g
6 V. Michoud, A. Colomb, A. Borbon, K. Miet, M. Beekmann,
M. Camredon, B. Aumont, S. Perrier, P. Zapf, G. Siour,
W. Ait-Helal, C. Af, A. Kukui, M. Furger, J. C. Dupont,
M. Haeffelin and J. F. Doussin, Study of the unknown
HONO daytime source at a European suburban site during
the
MEGAPOLI
summer
and
winter
eld
campaigns,
Atmos. Chem. Environmental Science: Atmospheres Paper 25 L. Wang, L. Wen, C. Xu, J. Chen, X. Wang, L. Yang, W. Wang,
X. Yang, X. Sui, L. Yao and Q. Zhang, HONO and its potential
source particulate nitrite at an urban site in North China
during the cold season, Sci. Total Environ., 2015, 538, 93–
101, DOI: 10.1016/j.scitotenv.2015.08.032. R. Woodward-Massey, C. Ye, D. E. Heard, S. Tong, S. Hou,
Y. Sun, J. Xu, L. Wei and P. Fu, Insights into air pollution
chemistry
and
sulphate
formation
from
nitrous
acid
(HONO) measurements during haze events in Beijing,
Faraday
Discuss.,
2021,
226,
223–238,
DOI:
10.1039/
d0fd00100g. 26 C. Ye, X. Zhou, D. Pu, J. Stutz, J. Festa, M. Spolaor, C. Tsai,
C. Cantrell, R. L. Mauldin, T. Campos, A. Weinheimer,
R. S. Hornbrook, E. C. Apel, A. Guenther, L. Kaser,
B. Yuan,
T. Karl,
J. Haggerty, S. Hall, K. Ullmann,
J. N. Smith, J. Ortega and C. Knote, Rapid cycling of
reactive nitrogen in the marine boundary layer, Nature,
2016, 532, 489–491, DOI: 10.1038/nature17195. 35 S. Tong, S. Hou, Y. Zhang, B. Chu, Y. Liu, H. He, P. Zhao and
M. Ge, Exploring the nitrous acid (HONO) formation
mechanism
in
winter
Beijing:
direct
emissions
and
heterogeneous production in urban and suburban areas,
Faraday
Discuss.,
2016,
189,
213–230,
DOI:
10.1039/
c5fd00163c. 27 C. Reed, M. J. Evans, L. R. Crilley, W. J. Bloss, T. Sherwen,
K. A. Read, J. D. Lee and L. J. Carpenter, Evidence for
renoxication
in
the
tropical
marine
boundary
layer,
Atmos. Chem. Phys., 2017, 17, 4081–4092, DOI: 10.5194/acp-
17-4081-2017. 36 L. R. Crilley, L. Kramer, F. D. Pope, L. K. Whalley, D. R. Cryer,
D. E. Heard, J. D. Lee, C. Reed and W. Bloss, On the
interpretation
of
in
situ
HONO
observations
via
photochemical steady state, Faraday Discuss., 2016, 189,
191–212, DOI: 10.1039/c5fd00224a. 28 J. Stutz, B. Alicke, R. Ackermann, A. Geyer, S. Wang,
A. B. White, E. J. Williams, C. W. Spicer and J. D. Fast,
Relative humidity dependence of HONO chemistry in
urban areas, J. Geophys. Res.: Atmos., 2004, 109, D03307,
DOI: 10.1029/2003jd004135. 37 B. H. Lee, E. C. Wood, S. C. Herndon, B. L. Lefer, W. T. Luke,
W. H. Brune, D. D. Nelson, M. S. Zahniser and J. W. Munger,
Urban measurements of atmospheric nitrous acid: a caveat
on the interpretation of the HONO photostationary state, J. Geophys. Res.:
Atmos., 2013, 118, 12,274–12,281, DOI:
10.1002/2013jd020341. 29 X. Li,
T. Environmental Science: Atmospheres Brauers,
R. H¨aseler,
B. Bohn,
H. Fuchs,
A. Hofzumahaus, F. Holland, S. Lou, K. D. Lu, F. Rohrer,
M. Hu, L. M. Zeng, Y. H. Zhang, R. M. Garland, H. Su,
A. Nowak, A. Wiedensohler, N. Takegawa, M. Shao and
A. Wahner, Exploring the atmospheric chemistry of nitrous
acid (HONO) at a rural site in Southern China, Atmos. Chem. Phys., 2012, 12, 1497–1513, DOI: 10.5194/acp-12-
1497-2012. 38 J. Heland, J. Kleffmann, R. Kurtenbach and P. Wiesen, A New
Instrument To Measure Gaseous Nitrous Acid (HONO) in the
Atmosphere, Environ. Sci. Technol., 2001, 35, 3207–3212,
DOI: 10.1021/es000303t. 39 J. Kleffmann and P. Wiesen, Technical Note: Quantication
of
interferences
of
wet
chemical
HONO
LOPAP
measurements under simulated polar conditions, Atmos. Chem. Phys., 2008, 8, 6813–6822, DOI: 10.5194/acp-8-6813-
2008. 30 J. Kleffmann,
R. Kurtenbach,
J. L¨orzer,
P. Wiesen,
N. Kalthoff,
B. Vogel
and
H. Vogel,
Measured
and
simulated vertical proles of nitrous acid—Part I: Field
measurements, Atmos. Environ., 2003, 37, 2949–2955, DOI:
10.1016/S1352-2310(03)00242-5. 40 E. Hanna, J. Penman, T. J´onsson, G. R. Bigg, H. Bj¨ornsson,
S. Sj´ur+arson, M. A. Hansen, J. Cappelen and R. G. Bryant,
Meteorological effects of the solar eclipse of 20 March
2015: analysis of UK Met Office automatic weather station
data and comparison with automatic weather station data
from the Faroes and Iceland, Philos. Trans. R. Soc., A, 2016,
374, 20150212, DOI: 10.1098/rsta.2015.0212. 31 K. Acker, A. Febo, S. Trick, C. Perrino, P. Bruno, P. Wiesen,
D. M¨oller, W. Wieprecht, R. Auel, M. Giusto, A. Geyer,
U. Platt and I. Allegrini, Nitrous acid in the urban area of
Rome, Atmos. Environ., 2006, 40, 3123–3133, DOI: 10.1016/
j.atmosenv.2006.01.028. 41 S. Madronich, S. Flocke, J. Zeng, I. Petropavlovskikh and
J. Lee-Taylor, Tropospheric Ultraviolet and Visible (TUV)
Radiation
Model,
National
Center
for
Atmospheric
Research, Boulder, CO, USA, 2003. 32 H. Meusel, U. Kuhn, A. Reiffs, C. Mallik, H. Harder,
M. Martinez,
J. Schuladen,
B. Bohn,
U. Parchatka,
J. N. Crowley,
H. Fischer,
L. Tomsche,
A. Novelli,
T. Hoffmann,
R. H. H. Janssen,
O. Hartogensis,
M. Pikridas, M. Vrekoussis, E. Bourtsoukidis, B. Weber,
J. Lelieveld, J. Williams, U. P¨oschl, Y. Cheng and H. Su,
Daytime formation of nitrous acid at a coastal remote site
in
Cyprus
indicating
a
common
ground
source
of
atmospheric HONO and NO, Atmos. Chem. Phys., 2016, 16,
14475–14493, DOI: 10.5194/acp-16-14475-2016. 42 D. E. Heard, L. J. Carpenter, D. J. Creasey, J. R. Hopkins,
J. D. Lee, A. C. 5
Concluding remarks J. Zarzana, N. Kille,
L. Clarisse,
I. De
Smedt,
C. Lerot,
H. Finkenzeller,
F. Hendrick, T. K. Koenig, C. F. Lee, C. Knote, H. Yu and © 2021 The Author(s). Published by the Royal Society of Chemistry Environ. Sci.: Atmos. View Article Online Environmental Science: Atmospheres Environmental Science: Atmospheres Lewis, M. J. Pilling, P. W. Seakins,
N. Carslaw and K. M. Emmerson, High levels of the
hydroxyl radical in the winter urban troposphere, Geophys. Res. Lett., 2004, 31, DOI: 10.1029/2004gl020544. 43 K. M. Emmerson, N. Carslaw and M. J. Pilling, Urban
Atmospheric Chemistry During the PUMA Campaign 2:
Radical Budgets for OH, HO2 and RO2, J. Atmos. Chem.,
2005, 52, 165–183, DOI: 10.1007/s10874-005-1323-2. 33 L. J. Kramer, L. R. Crilley, T. J. Adams, S. M. Ball, F. D. Pope
and W. J. Bloss, Nitrous acid (HONO) emissions under real-
world driving conditions from vehicles in a UK road tunnel,
Atmos. Chem. Phys., 2020, 20, 5231–5248, DOI: 10.5194/acp-
20-5231-2020. 44 S. Li, J. Matthews and A. Sinha, Atmospheric hydroxyl radical
production from electronically excited NO2 and H2O, Science,
2008, 319, 1657–1660, DOI: 10.1126/science.1151443. 45 S. Carr, D. E. Heard and M. A. Blitz, Comment on
“Atmospheric
Hydroxyl
Radical
Production
from 34 W. J. Bloss, L. Kramer, L. R. Crilley, T. Vu, R. M. Harrison,
Z. Shi, J. D. Lee, F. A. Squires, L. K. Whalley, E. Slater, © 2021 The Author(s). Published by the Royal Society of Chemistry Environ. Sci.: Atmos. View Article Online Paper Paper © 2021 The Author(s). Published by the Royal Society of Chemistry Environmental Science: Atmospheres Electronically Excited NO2 and H2O”, Science, 2009, 324, 336,
DOI: 10.1126/science.1166669. with a Mixed Fleet, J. Air Waste Manage. Assoc., 2017, 67,
797–805, DOI: 10.1080/10962247.2017.1293573. with a Mixed Fleet, J. Air Waste Manage. Assoc., 2017, 67,
797–805, DOI: 10.1080/10962247.2017.1293573. 51 51 T. W. Kirchstetter,
R. A. Harley
and
D. Littlejohn,
Measurement of nitrous acid in motor vehicle exhaust,
Environ. Sci. Technol., 1996, 30, 2843–2849, DOI: 10.1021/
es960135y. 46 T. J. Dillon and J. N. Crowley, Reactive quenching of
electronically excited NO2* and NO3* by H2O as potential
sources of atmospheric HOx radicals, Atmos. Chem. Phys.,
2018, 18, 14005–14015, DOI: 10.5194/acp-18-14005-2018. 52 K. Wong, C. Tsai, B. Lefer, C. Haman, N. Grossberg,
W. Brune, X. Ren, W. Luke and J. Stutz, Daytime HONO
vertical gradients during SHARP 2009 in Houston, TX,
Atmos. Chem. Phys., 2012, 12, 635–652, DOI: 10.5194/acp-
12-635-2012. 47 F. Meng, M. Qin, K. Tang, J. Duan, W. Fang, S. Liang, K. Ye,
P. Xie, Y. Sun, C. Xie, C. Ye, P. Fu, J. Liu and W. Liu, High-
resolution vertical distribution and sources of HONO and
NO2 in the nocturnal boundary layer in urban Beijing,
China, Atmos. Chem. Phys., 2020, 20, 5071–5092, DOI:
10.5194/acp-20-5071-2020. Open Access Article. Published on 20 July 2021. Downloaded on 8/11/2021 7:19:11 PM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. 53 B. Finlayson-Pitts, L. Wingen, A. Sumner, D. Syomin and
K. Ramazan, The heterogeneous hydrolysis of NO2 in
laboratory
systems
and
in
outdoor
and
indoor
atmospheres: an integrated mechanism, Phys. Chem. Chem. Phys., 2003, 5, 223–242, DOI: 10.1039/b208564j. 48 P. A. Leighton, Photochemistry of air pollution, Academic
Press, New York, 1960. 49 Y. Elshorbany, B. Steil, C. Br¨uhl and J. Lelieveld, Impact of
HONO on global atmospheric chemistry calculated with an
empirical parameterization in the EMAC model, Atmos. Chem. Phys., 2012, 12, 9977–10000, DOI: 10.5194/acp-12-
9977-2012. 54 G. Lv, C.-L. Song, S.-Z. Pan, J.-H. Gao and X.-F. Cao,
Comparison
of
number,
surface
area
and
volume
distributions of particles emitted from a multipoint port
fuel injection car and a gasoline direct injection car,
Atmos. Pollut. Res.,
2014,
5,
753–758,
DOI:
10.5094/
apr.2014.084. 50 Y. Liang, Q. Zha, W. Wang, L. Cui, K. H. Lui, K. F. Ho,
Z. Wang, S.-c. Lee and T. Wang, Revisiting Nitrous Acid
(HONO) Emission from On-road Vehicles: A Tunnel Study © 2021 The Author(s). Published by the Royal Society of Chemistry Environ. Sci.: Atmos.
|
https://openalex.org/W4311674548
|
https://journals.plos.org/plosone/article/file?id=10.1371/journal.pone.0278908&type=printable
|
English
| null |
Cognitive adverse effects of epilepsy and its predictors attending outpatient department of South Gondar zone hospitals, Amhara Region, Ethiopia 2020 /2021
|
PloS one
| 2,022
|
cc-by
| 6,368
|
PLOS ONE RESEARCH ARTICLE Background Epilepsy is the most common neurologic disorder which is further complicated by neurobe-
havioral co-morbidities, cognitive impairment, psychiatric disorders, and social problems. However, assessments of cognitive status of epileptic patients are far too low during clinical
visits. This calls for early neuropsychological assessment soon after the diagnosis of epi-
lepsy for a better treatment plan and outcome for epileptic patients. Editor: Muhammad Junaid Farrukh, UCSI
University, MALAYSIA
Received: April 28, 2022
Accepted: November 28, 2022
Published: December 9, 2022
Copyright: © 2022 Asnakew et al. This is an open
access article distributed under the terms of the
Creative Commons Attribution License, which
permits unrestricted use, distribution, and
reproduction in any medium, provided the original
author and source are credited. Editor: Muhammad Junaid Farrukh, UCSI
University, MALAYSIA Accepted: November 28, 2022
Published: December 9, 2022 Objective Copyright: © 2022 Asnakew et al. This is an open
access article distributed under the terms of the
Creative Commons Attribution License, which
permits unrestricted use, distribution, and
reproduction in any medium, provided the original
author and source are credited. This study aimed to assess the cognitive adverse effects of epilepsy and its predictors
attending outpatient departments of South Gondar Zone hospitals Amhara region Ethiopia
2020/2021. OPEN ACCESS Citation: Asnakew S, Legas G, Belete A, Admasu
FT, Yitbarek GY, Aytenew TM, et al. (2022)
Cognitive adverse effects of epilepsy and its
predictors attending outpatient department of
South Gondar zone hospitals, Amhara Region,
Ethiopia 2020 /2021. PLoS ONE 17(12): e0278908. https://doi.org/10.1371/journal.pone.0278908 Cognitive adverse effects of epilepsy and its
predictors attending outpatient department
of South Gondar zone hospitals, Amhara
Region, Ethiopia 2020 /2021 Sintayehu AsnakewID1*, Getasew Legas1, Amsalu Belete1, Fitalew Tadele Admasu2,
Getachew Yideg Yitbarek2, Tigabu Munye Aytenew3, Biruk Demise4, Eshetie
Molla Alemu4, Muluken Adela Alemu5, Wubet Alebachew Bayih6, Dejen Getaneh Feleke6,
Ermias Sisay Chanie6, Binyam Munye Birhane6, Demewoz Kefale6 a1111111111
a1111111111
a1111111111
a1111111111
a1111111111 1 Department of Psychiatry, School of Medicine, College of Health Science, Debre Tabor University, Debre
Tabor, Ethiopia, 2 Department of Biomedical Sciences, College of Health Science, Debre Tabor University,
Debre Tabor, Ethiopia, 3 Department of Nursing, College of Health Science Debre Tabor University, Debre
Tabor, Ethiopia, 4 Departments of Social and Population Health, College of Health Science, Debre Tabor
University, Debre Tabor, Ethiopia, 5 Departemnt of Pharmacy, College of Health Science, Debre Tabor
University, Debre Tabor, Ethiopia, 6 Department of Pediatrics and Child Health and Neonatal Nursing,
College of Health Science, Debre Tabor University, Debre Tabor, Ethiopia * sintie579@gmail.com * sintie579@gmail.com * sintie579@gmail.com PLOS ONE PLOS ONE Conclusions Prevalence of cognitive impairment in this study was high. Regular neuropsychiatric assess-
ment of patients with epilepsy should be encouraged especially for those participants with
longer durations of illness, who are rural residents, who take combined Phenobarbital and
Phenytoin, participants who had a history of head injury, depression, and anxiety. Results Prevalence of cognitive impairment in this study was 69.2% (95%CI; 65.4, 73.1). Rural resi-
dents (AOR = 4.16,95%CI, 1.99,8.67), respondents who couldn’t read and write (AOR =
2.62, 95%CI; 1.24, 5.5,) longer duration of seizure disorder (AOR = 4.59,95%CI;
2.01,10.52), taking combined Phenobarbital and Phenytoin (AOR = 4.69,95%CI;
1.88,11.69), having history of head injury (AOR = 3.29,95%CI;1.30,8.32), having depression
(AOR = 4.76,95%CI;2.83,7.98), and anxiety (AOR = 3.11,95%CI; 1.58,6.12) were signifi-
cantly associated with cognitive impairment. Data Availability Statement: All relevant data are
within the article. A multi-center institutional-based cross-sectional study was conducted. A total of 509
respondents were included with a response rate of 93.9%. Previously adapted pretested
structured questionnaire was used containing, socio-demographic, clinical, and seizure
related factors. Mini-Mental State Examination (MMSE) was used to measure cognitive
impairment. A systematic random sampling technique was applied. Data were entered into
Epi data version 4.4.2 then exported to SPSS version 24 for analysis. Descriptive statistics,
bivariable and multivariable binary logistic regressions with odds ratios and 95% confidence Funding: The fund for this research work was
obtained from Debre Tabor University. The funders
had no role in study design, data collection and
analysis, decision to publish, or preparation of the
manuscript. Competing interests: The authors have declared
that no competing interests exist. 1 / 13 PLOS ONE | https://doi.org/10.1371/journal.pone.0278908
December 9, 2022 PLOS ONE Cognitive adverse effects of epilepsy and its predictors interval were employed. The level of significance of association was determined at a p-value
< 0.05. interval were employed. The level of significance of association was determined at a p-value
< 0.05. PLOS ONE | https://doi.org/10.1371/journal.pone.0278908
December 9, 2022 Introduction Epilepsy is a neurological condition characterized by recurrent seizures [1, 2] which account
for 1% of the global burden of disease [3]. It affects at least 100 million people worldwide, spe-
cifically in childhood and adolescence [4] and 1 in 26 of the population may develop seizures
at some point in their lives [5]. The vast majority of them live in low- and middle-income
countries [6] and its burden in low-income countries is more than twice that found in high-
income countries because of the higher risk factors such as poverty, higher rate of infectious
diseases, and brain injuries [7]. A community-based study done in 5 African countries including Kenya, Tanzania, Uganda,
Ghana, and South Africa showed the prevalence of epilepsy ranging between 7 and 15 per
1000 people [8]. Likewise, in a study done on older children in rural Kenya, the adjusted preva-
lence estimates of a lifetime and active epilepsy were 41/1000 and 11/1000, respectively [9]. Moreover, very high prevalence rates of epilepsy have been found in the Zay society Ethiopia
with a prevalence of 29.5/1000 [10]. Although seizures are the most common clinical manifestation of epilepsies, individuals
with epilepsy are at risk of numerous health problems, include cognitive problems, mental
health conditions, including depression, anxiety, and somatic co-morbidities. For many indi-
viduals with epilepsy, the co-morbidities are more burdensome than the seizures that cognitive
abnormalities are among the most common and troublesome [11]. Cognitive impairment is a significant cognitive decline from a previous level of perfor-
mance or functioning in one or more domains cogitation (complex attention, executive func-
tion, learning, and memory, language, perceptual-motor, or social cognition) [12, 13]. Cognitive impairment is a frequent feature of different types of seizure disorder. Nearly
70% of patients with TLE have problems in declarative memory function, which represents the
most common cognitive impairment in this group. Impairment of executive function and low
intelligence levels are also quite often observed in about 30% of the patients with TLE [14]. In 2 / 13 PLOS ONE Cognitive adverse effects of epilepsy and its predictors newly diagnosed and untreated epileptic patients, cognitive problems are already present in
more than 50% of patients [13]. Likewise, cognitive impairment has been reported in around
30%–40% of epileptic patients [15]. In another study, between 20–50% of patients with epilepsy have memory impairment [16]. Study design, period, and setting A multi-centered institutional-based cross-sectional study was conducted at South Gondar
Zone hospitals from December 2020 to January 2021. Debre Tabor is the capital city of the
South Gondar zone which is located about 666km, North of Addis Ababa. There are 8 hospi-
tals in this zone which include Debre Tabor comprehensive specialized hospital, Andabet,
Mekane-Eyesus, Addis Zemen, Ebnat, Nefas Mewucha, Dr. Ambachew Memorial, and Simada
primary hospitals. There were about 1101 patients with epilepsy who visit the outpatient
department of South Gondar Zone hospitals per month. Introduction In a study done on newly onset untreated epileptic patients, 49.4% of the participants experi-
enced impairment in attention or executive function, 47.8% impairment in episodic memory,
and 39.3% subjective deficits in memory, and 35.2% subjective deficits in attention [17]. Research done in the USA using MMSE among participants living with epilepsy cognitive
impairment was found in all domains of cognitive function, including 39.5% with impairment
in visual memory, 23.7% each in attention and executive function, 18.4% in visuospatial skills,
and 15.8% for both verbal memory and language [18]. Magnitude of cognitive impairment
among patients with epilepsy has been evident with studies conducted in Indonesia 69.2%
[19], India 36% [20], Burkina Faso 61.8% [21], South-Eastern Nigeria 19.6% [22], Pakistan
39.5% [23] and Ethiopia 26.92% [24]. Cognitive impairment further complicated to difficulty
in performing daily day-to-day activities, such individuals may complain of impaired atten-
tion, word-finding difficulty, verbal fluency difficulty, forgetfulness, and psychomotor slowing
[25]. Thus, epileptic patients with cognitive impairment will have poor quality of life [26]. A variety of factors contribute to cognitive adverse effects (CAE) of epilepsy, including clin-
ical factors (history of medical illness, previous history of mental illness, depression, and anxi-
ety brain insult), seizure-related factors (seizure frequency, duration of seizure disorder,
seizure type and age at seizure onset), and antiepileptic medications, and substance-related
medications [16, 27–29]. Studies of the cognitive sequelae of epilepsy in Africans including
Ethiopia are generally lacking. Therefore, this study will add a body of knowledge about the magnitude and factors associ-
ated with cognitive impairment in patients living with epilepsy. It is also important to provide
baseline information for policy-makers and health care managers to integrate mental health
services with the primary health care system to screen and manage cognitive impairment
among patients with epilepsy. Thus, this study was intended to assess the cognitive adverse effects and associated factors
among epileptic patients attending South Gondar Zone Hospitals Amhara region, Ethiopia
2020/2021. Sample size determination The sample size was determined by using a single population proportion formula. Taking into
account the following assumption, the proportion (p) is 26.92%; (taken from research con-
ducted on cognitive impairments of epilepsy at Black-Lion neurology clinic, Addis Ababa,
Ethiopia [24]), the margin of error (0.05); level of confidence (95%), and non-response rate
10% giving the sample size of 334. 3 / 13 PLOS ONE | https://doi.org/10.1371/journal.pone.0278908
December 9, 2022 PLOS ONE Cognitive adverse effects of epilepsy and its predictors We also calculated the sample size based on factors by taking four variables including duration
of seizure disorder, seizure frequency, type of medication, and age at seizures onset giving the
sample size of 447, 389,369,542 respectively. Taking the larger sample size, the final size to be 542. Study participants and sampling procedures This research was conducted in eight (8) South Gondar Zone governmental hospitals among
patients living with epilepsy including Debre Tabor comprehensive specialized Hospital,
Andabet, Estie, Addis Zemen, Ebnat, Lay Gaynt, Dr.Ambachew Memorial, and Simada pri-
mary hospitals. All epileptic patients attending outpatients departments were the source popu-
lations and those age 18 years and above were included. Those patients with epilepsy who had
a co morbid intellectual disability and acutely sick were excluded. A systematic random sam-
pling technique was applied. The sample size was distributed to each hospital proportional to
the numbers of epileptic patients attending the outpatient department per month in each of
the hospitals. Then interval k was calculated and the first participant was determined by the
lottery method. Thus, K = N/n. . .1101/542 ~ 2. Finally, the participants were interviewed every 2 intervals Thus, K = N/n. . .1101/542 ~ 2. Finally, the participants were interviewed every 2 intervals
Of all invited (542) participants, twenty five (25) of the eligible participants refused to par-
ticipate and eight (8) of the questionnaires were discarded because of incomplete data. Finally,
509 participants completed the questionnaires with a response rate of 93.9% (Fig 1). Study variables The dependent variable was cognitive impairment which was measured using MMSE as a
dichotomous variable (Yes/No). Independent variables include socio-demographic factors
(age, gender, marital status, educational status resident), clinical variables (history of medical
illness, history of mental illness, history of head injury, depression, and anxiety), seizure-
related factors (seizure frequency, duration of seizure disorder, seizure type, age at seizure
onset), medication and substance-related factors (AEDs, substance intake). PLOS ONE | https://doi.org/10.1371/journal.pone.0278908
December 9, 2022 Data collection tools Data was collected through face-to-face interviews by using a previously adapted standard
questionnaire with the Amharic version of the tool. Cognitive adverse effects of epilepsy were
measured using MMSE which had sensitivity and specificity of 87% and 82%, respectively and
the MMSE has shown a high degree of correlation with a variety of gold standards tools includ-
ing MoCA and DSM clinical diagnosis and had enter data reliability 0.89 (Cronbach α) [30]. Cognitive impairment. According to the Mini-Mental State Examination tool (MMSE),
individuals living with epilepsy who scored less than 25 out of the total score of 30 were con-
sidered to have a deficit in cognition [31]. Depression and anxiety. Depression and anxiety were screened using HADS in which
those individuals who scored > = 8 on HADS were categorized as having depression and anxi-
ety [32]. Current use: using at least one of a specific substance for non-medical purposes within the
last three months (alcohol, khat, tobacco, others). Ever use of a substance: using at least one of any specific substance for a non-medical pur-
pose at least once in the lifetime (alcohol, khat, tobacco, others) [33]. Medical illness. To examine a history of medical illness, respondents were asked: ‘Did
you have any medical illness (DM, HTN, HIV/AIDS, etc.?)’ and responses were yes/no. Intellectual disability. It has been screened using DSM-V which is defined by significant
limitations in both intellectual functioning (reasoning, learning, and problem solving) and in PLOS ONE | https://doi.org/10.1371/journal.pone.0278908
December 9, 2022 4 / 13 PLOS ONE Cognitive adverse effects of epilepsy and its predictors Fig 1. Sampling procedure for the cognitive adverse effect of epilepsy and its predictors attending South Gondar hospitals Amhara Ethiopia 2020/2021. https://doi.org/10.1371/journal.pone.0278908.g001 Fig 1. Sampling procedure for the cognitive adverse effect of epilepsy and its predictors attending South Gondar hospitals Amhara Ethiopia 2020/2021. https://doi org/10 1371/journal pone 0278908 g001 gnitive adverse effect of epilepsy and its predictors attending South Gondar hospitals Amhara Ethiopia 2020/2021. 908.g001 Fig 1. Sampling procedure for the cognitive adverse effect of epilepsy and its predictors attending South Gondar hospitals Amhara Ethiopia 2020/2021. https://doi.org/10.1371/journal.pone.0278908.g001 https://doi.org/10.1371/journal.pone.0278908.g001 https://doi.org/10.1371/journal.pone.0278908.g001 adaptive behaviour (conceptual, social, and practical skills) and it has been screened using
DSM-V [34]. PLOS ONE | https://doi.org/10.1371/journal.pone.0278908
December 9, 2022 Data quality control and data collectors Investigators selected 24 Bsc psychiatry professionals working in the eight hospitals as the data
collector three for each hospital. The training was given to the data collectors and supervisors
on how to properly utilize the data collection tools. The questionnaire was translated into
Amharic and then back into English to check its consistency. A week before the actual data collection, pretest was conducted on 5% (28) of samples from
Mekane-Eyesus hospital to check the clarity of the instrument that the data obtained from this
was not included in the main analysis part. Based on the finding from the pretest, the question-
naire was revised and adapted especially on the structured questionnaire. Once the partici-
pants agreed to participate; they were given pieces of information and signed the informed
consent. Supervision was held regularly during data collection. The collected data were checked
daily for completeness and consistency. 5 / 13 PLOS ONE | https://doi.org/10.1371/journal.pone.0278908
December 9, 2022 PLOS ONE Cognitive adverse effects of epilepsy and its predictors Data processing and analysis Data were entered into Epi data version 4.4.2 and then export to SPSS version 24 for further
statistical analysis. Frequency tables and diagrams were used for presenting the descriptive
results. Bivariable analysis was used to look for the association between predictors and dependent
variables and multivariable logistic regression analysis was done to control for confounding
and identify the most important associated variables. The strength of associations was indi-
cated by AOR with a 95% confidence interval. P-value < 0.05 was considered statistically
significant. Hosmer and Lemeshow’s test (p = 0.208) was used to check model fitness. Multi-co linearity
was checked to see the correlation among the independent variables by using variance inflation
factor and tolerance. In this case, the value of variance inflation factor was <10 and tolerance
was greater than 0.1 which indicated that there was no dependency between independent
variables. Clinical factors In this study, about 89(17.5%) of them had a history of medical illness, 72(14.1%) had a history
of head injury, 305 (59.9%) of them had mild to severe depression and 220(43.2%) of them
were anxious. Socio-demographic characteristics In this study, a total of 509 individuals were included with a response rate of 93.9%. The major-
ity of the respondents were males 312(61.3%) and in the age group between 18–29 years, Most
of them were orthodox followers 341(67%), and Amhara by ethnicity 485(95.2%). About 172
(33.8%) of them couldn’t read and write, and nearly half of them were the rural residents 251
(49.3%) (Table 1). Ethical consideration The ethical clearance was obtained from the ethical review committee of Debre Tabor Univer-
sity, and a permission letter was obtained from each hospital. We received written informed
consent from study participants and confidentiality was maintained by omitting personal
identifiers. The purpose of the study, direct and indirect advantages of being included in the study
were explained to the participants. This form indicated that participation was voluntary and
that clients had the right to withdraw from completing the questionnaire at any time they
wish. Participants were also informed that there was no expectation of any benefits for them asso-
ciated with participating in the study but those who had cognitive impairment on the MMSE
scale got appropriate intervention timely. Substance-related factors In this study majority of the respondents had ever use 370(72.7%) and the current use of 248
(48.7%) alcohol respectively (Fig 2). PLOS ONE Table 1. Socio-demographic characteristics of patents with epilepsy attending outpatient department of South Gondar zone hospitals Amhara, Ethiopia, 2021
(n = 509). Characteristics
Category
Frequency
Percent
Sex
Male
312
61.3
Female
197
38.7
Age
18–29
295
58.0
30–39
99
19.4
40–49
49
9.6
> = 50
66
13
Ethnicity
Amhara
485
95.2
Tigray
13
2.6
Oromo
11
2.2
Religion
Orthodox
341
67
Muslim
85
16.7
Protestant
14
2.8
Catholic
69
13.5
Residence
Rural
251
49.3
Urban
258
50.7
Educational status
Unable to read and write
173
34
primary (1–8)
58
11.4
High school (9–12)
128
25.1
College and above
150
29.5
https://doi org/10 1371/journal pone 0278908 t001 https://doi.org/10.1371/journal.pone.0278908.t001 Seizure and medication-related factors The majority of participants were diagnosed for general tonic-clonic seizure 191(37.5%) and
were treated with either Phenobarbital or Phenytoin 213(41.9%) (Table 2). 6 / 13 PLOS ONE | https://doi.org/10.1371/journal.pone.0278908
December 9, 2022 Cognitive adverse effects of epilepsy and its predictors PLOS ONE | https://doi.org/10.1371/journal.pone.0278908
December 9, 2022 PLOS ONE Table 2. Seizure and medication-related factors of cognitive impairment in patients with epilepsy attending outpatient department of south Gondar zone hospitals
Amhara Ethiopia 2021(n = 509). Characteristics
Category
Frequency
Percent
Age at seizure onset
<10
278
54.6
10–19
91
17.9
20–29
45
8.8
30–39
95
18.7
Duration of the seizure disorder
<10
315
61.9
10–29
136
26.7
> = 30
58
11.4
Frequency of seizures
Daily to every other day
146
28.7
Weekly to every other week
45
8.8
Once in three to four weeks
116
22.8
Once in the past 1–6 months
100
19.6
6–11 months ago
65
12.8
1–4 years ago
37
7.3
Type of seizure
Simple partial
96
18.9
Focal with secondary
generalization
66
13
Complex partial
109
21.4
GTCs
191
37.5
Atonic
28
5.5
Myoclonic
19
3.7
Type of AEDs
Carbamazepine
79
15.5
Phenobarbital /Phenytoin
213
41.9
Na+ Valporate
72
14.1
Phenobarbital +Phenytoin
145
28.5
https://doi org/10 1371/journal pone 0278908 t002 tors of cognitive impairment in patients with epilepsy attending outpatient department of south Gondar zone hospitals https://doi.org/10.1371/journal.pone.0278908.t002 95%CI; 1.30, 8.32), depression (AOR = 4.76, 95%CI; 2.83, 7.98), and anxiety (AOR = 3.11, 95%
CI; 1.58, 6.12) as compared with their counterparts. Moreover, participants who were rural
residents (AOR = 4.16, 95%CI; 1.99, 8.67) and could not read and write (AOR = 2.62, 95%CI;
1.24, 5.51) scored lower on the MMSE test (Table 3). 95%CI; 1.30, 8.32), depression (AOR = 4.76, 95%CI; 2.83, 7.98), and anxiety (AOR = 3.11, 95%
CI; 1.58, 6.12) as compared with their counterparts. Moreover, participants who were rural
residents (AOR = 4.16, 95%CI; 1.99, 8.67) and could not read and write (AOR = 2.62, 95%CI;
1.24, 5.51) scored lower on the MMSE test (Table 3). Prevalence and associated factors of cognitive impairment In this study magnitude of cognitive impairment among epileptic patients was 69.2% (95CI;
65.4, 73.1). To determine the association of independent variables with cognitive impairment, bivari-
able, and multivariable binary logistic regression analyses was carried out. On the bivariate analysis respondents who are rural residents, unable to read and write,
duration of seizure disorder> = 30 years, respondents who took combined Phenobarbital and
Phenytoin, respondents who had a history of medical illness, head injury, depression, and anx-
iety were significantly associated with cognitive impairment at p-value <0.05. These variables were taken to multivariable analysis to control confounding effects. In mul-
tivariable analysis, rural residents, respondents who could not read and write, participants
who lived with the seizure disorder for > = 30 years, and who had a history of head injury, par-
ticipants who took combined Phenobarbital and Phenytoin, having depression and anxiety
were significantly associated with cognitive impairment. When controlling for other variables, the odds of developing cognitive impairment among
epileptic patients were 4.59 times higher among those participants who had seizure disorders>
= 30years as compared with those who had <10 years duration of seizure disorder
(AOR = 4.59, 95%CI; 2.01,10.52). Participants who took combined Phenobarbital and Phenyt-
oin were more affected by cognitive impairment as compared with those participants who
took Carbamazepine (AOR = 2.03, 95%CI; 1.21, 4.32). The likelihood of developing cognitive
impairment was greater among participants who had a history of head injury (AOR = 3.29, 7 / 13 PLOS ONE | https://doi.org/10.1371/journal.pone.0278908
December 9, 2022 Cognitive adverse effects of epilepsy and its predictors PLOS ONE | https://doi.org/10.1371/journal.pone.0278908
December 9, 2022 Discussion Most people with epilepsy in low and middle income countries do not seek medical treatment
for their epilepsy and all types of epilepsy frequently experience cognitive and emotional diffi-
culties. This study aimed to assess the magnitude of cognitive impairment in epileptic patients
and was investigated in relation to the socio-demographic, and other factors which greatly
affect the higher executive brain function. This study showed that the magnitude of cognitive
impairment was found to be 69.2% (95%CI; 65.4, 73.1). Factors, including living in a rural
area, not being able to read or write, having a history of head injury, having a seizure disorder
for > = 30 years, using combined phenobarbital and phenytoin, having depression and anxiety
were significantly associated with cognitive impairment. The result of the current study was in line with the study carried out in Indonesia 69.2%
[19] but higher than the studies done in the USA 39.5% [18], India 36% [20], Slovakia 37%
[35], Burkina Faso 61.8% [21], South-Eastern Nigeria 19.6% [22], Pakistan 39.5% [23] and
Ethiopia 26.92% [24]. The difference might be the variation in the tool, socio-cultural differ-
ences, the ages of the participants included, and sample size. For example, in Indian and PLOS ONE | https://doi.org/10.1371/journal.pone.0278908
December 9, 2022 8 / 13 PLOS ONE Cognitive adverse effects of epilepsy and its predictors Fig 2. Substance-related factors of cognitive impairment in patients with epilepsy attending outpatient department of south Gondar zone hospitals
Amhara Ethiopia 2020 /2021(n = 509). Fig 2. Substance-related factors of cognitive impairment in patients with epilepsy attending outpatient department of south Gondar zone hospitals
Amhara Ethiopia 2020 /2021(n = 509). Fig 2. Substance-related factors of cognitive impairment in patients with epilepsy attending outpatient department of south Gondar zone hospitals
Amhara Ethiopia 2020 /2021(n = 509). https://doi org/10 1371/journal pone 0278908 g002 https://doi.org/10.1371/journal.pone.0278908.g002 https://doi.org/10.1371/journal.pone.0278908.g002 https://doi.org/10.1371/journal.pone.0278908.g002 Nigerian studies the sample size was 100 and 102 respectively which were much lower than the
current study. Moreover, in the current study MMSE was used but CSID was utilized in the,
Nigeria study. In the contrary, the current study was lower than the Tunisians study (100% of cases had
cognitive problems) [36]. The reason could be the difference in the type of seizure disorder
included in the studies. In the Tunisian study, the patients were only those with temporal lobe
epilepsy but in the current study, all types of seizure disorders were included. PLOS ONE | https://doi.org/10.1371/journal.pone.0278908
December 9, 2022 PLOS ONE Table 3. Factors associated with cognitive impairment among patients with epilepsy attending outpatient department of South Gondar hospitals 2020(n = 509). Characteristics
Category
Cognitive impairment
COR(95%CI)
AOR(95%CI)
Yes
No
Residence
Rural
202
49
2.97(1.99,4.42)
4.16(1.99,8.67)
Urban
150
108
1
1
Educational status
Unable to read & write
154
19
4.05(2.26,7.28)
2.62(1.24,5.51)
Primary school
45
13
1.73(0.86,3.50)
0.43(0.16,1.18)
Secondary school
53
75
0.35(0.22,0.58)
0.28(0.14,0.55)
college and above
100
50
1
1
Duration of seizure disorder
<10
207
108
1
1
10–29
98
38
1.35(0.87,2.09)
1.51(0.51,4.44)
> = 30
47
11
2.23(1.11,4.47)
4.59(2.01,10.52)
Type of AEDs
Carbamazepine
61
18
1
1
Phenobarbital /Phenytoin
139
74
0.55(0.31,1.01)
0.26(0.58,2.73)
Na+ Valporate
25
47
0.16(0.08,0.32)
1.97(0.70,5.50)
Phenobarbital +Phenytoin
127
18
2.08(1.01,4.28)
4.69(1.88,11.69)
History of head injury
Yes
63
9
4.06(1.89,8.69)
3.29(1.30,8.32)
No
289
148
1
1
History of medical illness
Yes
79
10
4.25(2.14,8.46)
0.82(0.29,2.25)
No
273
147
1
1
Depression
Yes
256
49
5.88(3.89,8.87)
4.76(2.83,7.98)
No
96
108
1
1
Anxiety
Yes
191
29
5.24(3.32,8.25)
3.11(1.58,6.12)
No
161
128
1
1
Note that: Hosmer Lemshow test -0.208, Tolerance >0.1, variance inflation factor<10
https://doi.org/10.1371/journal.pone.0278908.t003 contributes for the development of cognitive impairment compared with those who had no
history of head injury. Participants who were rural residents and could not write and read
scored poor cognition in this test. This might be because the MMSE examination is affected by
the educational status that overestimates the cognitive impairment in the lower schooling pop-
ulation [44]. Moreover, people who lived in rural areas and those who couldn’t read and write
might not get enough information about the treatment of epilepsy so that they become late in
seeking help. Therefore, living with untreated seizures for a longer duration exposes poor cog-
nitive performance. These calls for neuropsychological evaluation and measuring the cognitive status of patients
with epilepsy early in the course of the disease. This also indicated the necessity of regular
screening of the cognitive side effects of antiepileptic drugs and co morbid disorders in the
course of the disease and long term therapy using screening tools. Discussion On the independent predictors of cognitive impairment, rural residents, participants who
could not read and write, longer duration of illness, taking combined Phenobarbital and Phe-
nytoin, having a history of head injury, those participants who had depression and anxiety
scored lower on the MMSE test compared with their counterparts. Participants who had > =
30 years duration of illness were more affected to develop cognitive impairment compared
with <10 years duration of illness. This was supported by Merkans (Ethiopia) [24], and USA
[37] studies. Likewise, those participants who took combined Phenobarbital and Phenytoin
were highly affected by cognitive impairment. This was in agreement with the previous study
[38]. The current study also revealed that participants who were positive for depression scored
lower on the neuropsychiatric test(MMSE) which was supported by the previous work [39]. Studies showed that depression decrease the volume of bilateral hippocampi, alters the cortical
thickness and a reduction of neuronal cell density in the frontal lobe. These results illustrated
the negative effects of depression on the cognitive function of epileptic patients [40–42]. Higher anxiety was also negatively associated with cognitive function which was supported
by the USA [18], and Slovakia studies [35]. Likewise, respondents with a history of head injury
scored lower on MMSE which was supported by the study conducted in Australia [43]. This
could be, head injury is the risk for seizure disorder i.e. leads frequent seizure which in turn PLOS ONE | https://doi.org/10.1371/journal.pone.0278908
December 9, 2022 9 / 13 Cognitive adverse effects of epilepsy and its predictors Acknowledgments The authors acknowledge Debre Tabor University for reviewing and approval of ethical issues. We extend our gratitude to data collectors, supervisors, and study participants for their time
and effort. Conclusion The prevalence of cognitive impairment was found to be high. Longer duration of seizure
disorder, taking combined Phenobarbital and Phenytoin, having a history of head injury,
respondents who were rural residents, who couldn’t write and read, those participants who
had depression and anxiety were greatly affected by cognitive impairment compared with
their counterparts. Therefore, emphasis should be given to those rural residents, with his-
tory of head injury, couldn’t read and write, on increasing the availability of second-genera-
tion AEDs and avoidance of routine prescriptions of combined old generation medications
like phenytoin and phenobarbital at a time. Similarly, communities’ health education
regarding the treatment of epilepsy is crucial, especially for those rural residents and
participants who couldn’t read and write. Moreover, integration for routine screening of
patients for depression and anxiety is recommended for early prevention of cognitive
impairment. Limitations People with epilepsy may have recall bias that they may face trouble in recalling the onset and
the duration of the illness. Additionally, social desirability bias may be a problem since data
collection method was face to face interview which force interviewees to give socially accept-
able responses especially in case of substance related questions. Moreover, the MMSE is affected by educational status i.e. participants with lower educa-
tional status might show lower on the test and vice versa. PLOS ONE | https://doi.org/10.1371/journal.pone.0278908
December 9, 2022 10 / 13 PLOS ONE Cognitive adverse effects of epilepsy and its predictors Author Contributions Conceptualization: Amsalu Belete, Fitalew Tadele Admasu, Getachew Yideg Yitbarek, Tigabu
Munye Aytenew, Muluken Adela Alemu, Wubet Alebachew Bayih, Ermias Sisay Chanie,
Demewoz Kefale. References 1. McLin WM, de Boer HM. Public perceptions about epilepsy. Epilepsia. 1995; 36(10):957. https://doi. org/10.1111/j.1528-1157.1995.tb00952.x PMID: 7555958 2. Berhanu S, Alemu S, Asmera J, Prevett M. Primary care treatment of epilepsy in rural Ethiopia. Ethio-
pian Journal of Health Development. 2002; 16(3):235–40. 3. Diseases OMdlSPfN, Organization WH, Epilepsy GCa, Diseases PfN, Neuroscience, Epilepsy IBf,
et al. Atlas: Epilepsy Care in the World: World Health Organization; 2005. 4. Reynolds EH. The ILAE/IBE/WHO Global Campaign against epilepsy: bringing epilepsy “Out of the
Shadows”. Epilepsy & Behavior. 2000; 1(4):S3–S8. 5. Jacobs M, Jensen FE. Introduction to institute of medicine report: epilepsy across the spectrum: pro-
moting health and understanding. Epilepsy currents. 2012; 12(6):243. https://doi.org/10.5698/1535-
7511-12.6.243 PMID: 23447725 6. Ngugi AK, Bottomley C, Kleinschmidt I, Sander JW, Newton CR. Estimation of the burden of active and
life-time epilepsy: a meta-analytic approach. Epilepsia. 2010; 51(5):883–90. https://doi.org/10.1111/j. 1528-1167.2009.02481.x PMID: 20067507 7. Newton CR, Garcia HH. Epilepsy in poor regions of the world. The Lancet. 2012; 380(9848):1193–201. https://doi.org/10.1016/S0140-6736(12)61381-6 PMID: 23021288 8. Ngugi AK, Bottomley C, Kleinschmidt I, Wagner RG, Kakooza-Mwesige A, Ae-Ngibise K, et al. Preva-
lence of active convulsive epilepsy in sub-Saharan Africa and associated risk factors: cross-sectional
and case-control studies. The Lancet Neurology. 2013; 12(3):253–63. https://doi.org/10.1016/S1474-
4422(13)70003-6 PMID: 23375964 9. Mung’ala-Odera V, White S, Meehan R, Otieno G, Njuguna P, Mturi N, et al. Prevalence, incidence and
risk factors of epilepsy in older children in rural Kenya. Seizure. 2008; 17(5):396–404. https://doi.org/10. 1016/j.seizure.2007.11.028 PMID: 18249012 10. Almu S, Tadesse Z, Cooper P, Hackett R. The prevalence of epilepsy in the Zay Society, Ethiopia—an
area of high prevalence. Seizure. 2006; 15(3):211–3. https://doi.org/10.1016/j.seizure.2006.01.004
PMID: 16488161 11. England M, Liverman C, Schultz A, Strawbridge L. Committee on the Public Health Dimensions of the
Epilepsies; Board on Health Sciences Policy; Institute of Medicine. Epilepsy Across the Spectrum: Pro-
moting Health and Understanding. The National Academies Press (US), March; 2012. 12. Association AP. Desk reference to the diagnostic criteria from DSM-5®: American Psychiatric Pub;
2014. 13. Hermann BP, Seidenberg M, Dow C, Jones J, Rutecki P, Bhattacharya A, et al. Cognitive prognosis in
chronic temporal lobe epilepsy. Annals of neurology. 2006; 60(1):80–7. https://doi.org/10.1002/ana. 20872 PMID: 16802302 14. Helmstaedter C, Kockelmann E. Cognitive outcomes in patients with chronic temporal lobe epilepsy. Epilepsia. 2006; 47:96–8. https://doi.org/10.1111/j.1528-1167.2006.00702.x PMID: 17105474 15. Byard RW, Blumbergs P, Rutty G, Sperhake J, Banner J, Krous HF. Data curation: Fitalew Tadele Admasu, Ermias Sisay Chanie. Data curation: Fitalew Tadele Admasu, Ermias Sisay Chanie. Formal analysis: Getasew Legas, Amsalu Belete, Fitalew Tadele Admasu, Tigabu Munye Ayte-
new, Biruk Demise, Eshetie Molla Alemu, Wubet Alebachew Bayih, Dejen Getaneh Feleke. Funding acquisition: Amsalu Belete, Fitalew Tadele Admasu, Ermias Sisay Chanie. Investigation: Muluken Adela Alemu, Ermias Sisay Chanie. Investigation: Muluken Adela Alemu, Ermias Sisay Chanie. Methodology: Sintayehu Asnakew, Getasew Legas, Getachew Yideg Yitbarek, Biruk Demise,
Muluken Adela Alemu, Binyam Munye Birhane. Project administration: Binyam Munye Birhane. Resources: Getachew Yideg Yitbarek. Resources: Getachew Yideg Yitbarek. Software: Getasew Legas, Getachew Yideg Yitbarek, Biruk Demise, Binyam Munye Birhane. Supervision: Sintayehu Asnakew, Getasew Legas, Tigabu Munye Aytenew, Binyam Munye
Birhane, Demewoz Kefale. Validation: Tigabu Munye Aytenew, Eshetie Molla Alemu, Wubet Alebachew Bayih, Dejen
Getaneh Feleke, Demewoz Kefale. Writing – original draft: Biruk Demise, Eshetie Molla Alemu, Wubet Alebachew Bayih,
Dejen Getaneh Feleke, Demewoz Kefale. Writing – review & editing: Biruk Demise, Eshetie Molla Alemu, Wubet Alebachew Bayih,
Dejen Getaneh Feleke, Demewoz Kefale. 11 / 13 PLOS ONE | https://doi.org/10.1371/journal.pone.0278908
December 9, 2022 PLOS ONE Cognitive adverse effects of epilepsy and its predictors References Clinical neuropsychology and neocortical epilepsies. Advances in neurology. 2000;
84:457–62. PMID: 11091888 29. Jokeit H, Ebner A. Effects of chronic epilepsy on intellectual functions. Progress in brain research. 135:
Elsevier; 2002. p. 455–63. 30. Gugssa SA, Davey G, Ejigu AA, Metaferia G, Medhin G, Kelkile T. Population norms for the mini-mental
state examination in Ethiopia. Ethiop Med J. 2011; 49(3):239–47. PMID: 21991757 31. Folstein MF, Folstein SE, McHugh PR. “Mini-mental state”: a practical method for grading the cognitive
state of patients for the clinician. Journal of psychiatric research. 1975; 12(3):189–98. 32. Zigmond AS, Snaith RP. The hospital anxiety and depression scale. Acta psychiatrica scandinavica. 1983; 67(6):361–70. https://doi.org/10.1111/j.1600-0447.1983.tb09716.x PMID: 6880820 33. Humeniuk R, Ali R, Babor TF, Farrell M, Formigoni ML, Jittiwutikarn J, et al. Validation of the alcohol,
smoking and substance involvement screening test (ASSIST). Addiction. 2008; 103(6):1039–47. https://doi.org/10.1111/j.1360-0443.2007.02114.x PMID: 18373724 34. Guha M. Diagnostic and statistical manual of mental disorders: DSM-5. Reference Reviews. 2014. 35. Jarčusˇkova´ D, Palusˇna´ M, Gazda J, Feketeova´ E, Gdovinova´ Z. Which clinical and neuropsychological
factors are responsible for cognitive impairment in patients with epilepsy? International Journal of Public
Health. 2020; 65(6):947–56. https://doi.org/10.1007/s00038-020-01401-7 PMID: 32533220 36. Nouha F, Sawsan D, Salma S, Olfa H, Hanen H, Mariem D, et al. Cognitive impairment in patients with
temporal lobe epilepsy. J Neuropsychiatry. 2018; 2(2):5. 37. Blakemore C, Ettlinger G, Falconer M. Cognitive abilities in relation to frequency of seizures and neuro-
pathology of the temporal lobes in man. Journal of neurology, neurosurgery, and psychiatry. 1966; 29
(3):268. https://doi.org/10.1136/jnnp.29.3.268 PMID: 5937644 38. Singh P, Pandey AK. Psychiatric morbidity in epilepsy. Int J Res Med Sci. 2017; 5(10):4267. 39. Martin RC, Griffith HR, Faught E, Gilliam F, Mackey M, Vogtle L. Cognitive functioning in community
dwelling older adults with chronic partial epilepsy. Epilepsia. 2005; 46(2):298–303. https://doi.org/10. 1111/j.0013-9580.2005.02104.x PMID: 15679511 40. Sheline YI, Gado MH, Kraemer HC. Untreated depression and hippocampal volume loss. American
journal of psychiatry. 2003; 160(8):1516–8. https://doi.org/10.1176/appi.ajp.160.8.1516 PMID:
12900317 41. Cotter D, Mackay D, Chana G, Beasley C, Landau S, Everall IP. Reduced neuronal size and glial cell
density in area 9 of the dorsolateral prefrontal cortex in subjects with major depressive disorder. Cere-
bral cortex. 2002; 12(4):386–94. https://doi.org/10.1093/cercor/12.4.386 PMID: 11884354 42. Cotter DR, Pariante CM, Everall IP. Glial cell abnormalities in major psychiatric disorders: the evidence
and implications. Brain research bulletin. 2001; 55(5):585–95. https://doi.org/10.1016/s0361-9230(01)
00527-5 PMID: 11576755 43. References Lack of evidence for a causal rela-
tionship between hypoxic-ischemic encephalopathy and subdural hemorrhage in fetal life, infancy, and
early childhood. Pediatric and Developmental Pathology. 2007; 10(5):348–50. https://doi.org/10.2350/
06-08-0154.1 PMID: 17929988 16. Aldenkamp AP. Cognitive impairment in epilepsy: state of affairs and clinical relevance. Seizure-Euro-
pean Journal of Epilepsy. 2006; 15(4):219–20. 17. Witt J-A, Helmstaedter C. Should cognition be screened in new-onset epilepsies? A study in 247
untreated patients. Journal of neurology. 2012; 259(8):1727–31. https://doi.org/10.1007/s00415-012-
6526-2 PMID: 22580844 18. Miller LA, Galioto R, Tremont G, Davis J, Bryant K, Roth J, et al. Cognitive impairment in older adults
with epilepsy: characterization and risk factor analysis. Epilepsy & Behavior. 2016; 56:113–7. https://
doi.org/10.1016/j.yebeh.2016.01.011 PMID: 26859320 19. Handayani S, Harun Y, Mukhlisa M, Bahar E. Factors that influence cognitive function in epilepsy
patients at neurology clinic Mohammad Hoesin Hospital Palembang. Journal of the Neurological Sci-
ences. 2019; 405:71. 20. Kumar V, Vatsala M. Cross sectional study to determine the cognitive impairment among epilepsy
patients. Int J Res Med Sci. 2019; 7:1465–71. 21. Dabilgou AA, Drave´ A, Adeline MJ, Kyelem KAN, Napon C, Millogo A, et al. Cognitive Disorders in
Patients with Epilepsy Attending at Neurology Outpatient Clinics. A Multicenter Prospective Cross-Sec-
tional Study from Burkina Faso. 2019. 12 / 13 PLOS ONE | https://doi.org/10.1371/journal.pone.0278908
December 9, 2022 December 9, 2022 PLOS ONE Cognitive adverse effects of epilepsy and its predictors 22. Arinzechi EO, Ogunrin OA, Nwosu CM, Nwani PO, Enwereji KO, Asomugha LA, et al. A community-
based case–control study of prevalence and pattern of cognitive impairments in patients with epilepsy
residing in South-Eastern Nigeria. Journal of neurosciences in rural practice. 2016; 7(3):405. https://doi. org/10.4103/0976-3147.181488 PMID: 27365959 23. Malik M, Hussain A, Malik S, Hashmi A. Cognition and Memory Impairment Among Patients of Epilepsy
in Pakistan-The Role of Conventional and Newer Anti-Epileptics. 24. Merkena M. Prevalence of cognitive adverse outcomes in epileptic outpatients. J Neurol Stroke. 2016;
4(5):00155. 25. England MJ, Liverman CT, Schultz AM, Strawbridge LM. Epilepsy across the spectrum: Promoting
health and understanding.: A summary of the Institute of Medicine report. Epilepsy & Behavior. 2012;
25(2):266–76. https://doi.org/10.1016/j.yebeh.2012.06.016 PMID: 23041175 26. DiIorio CK, Bamps YA, Edwards AL, Escoffery C, Thompson NJ, Begley CE, et al. The prevention
research centers’ managing epilepsy well network. Epilepsy & Behavior. 2010; 19(3):218–24. https://
doi.org/10.1016/j.yebeh.2010.07.027 PMID: 20869323 27. Dodrill CB. Neuropsychological effects of seizures. Epilepsy & Behavior. 2004; 5:21–4. https://doi.org/
10.1016/j.yebeh.2003.11.004 PMID: 14725843 28. Jones-Gotman M. 44.
Tombaugh TN, McIntyre NJ. The mini-mental state examination: a comprehensive review. Journal of
the American Geriatrics Society. 1992; 40(9):922–35. https://doi.org/10.1111/j.1532-5415.1992.
tb01992.x PMID: 1512391 References Semple BD, Zamani A, Rayner G, Shultz SR, Jones NC. Affective, neurocognitive and psychosocial
disorders associated with traumatic brain injury and post-traumatic epilepsy. Neurobiology of disease. 2019; 123:27–41. https://doi.org/10.1016/j.nbd.2018.07.018 PMID: 30059725 44. Tombaugh TN, McIntyre NJ. The mini-mental state examination: a comprehensive review. Journal of
the American Geriatrics Society. 1992; 40(9):922–35. https://doi.org/10.1111/j.1532-5415.1992. tb01992.x PMID: 1512391 13 / 13 PLOS ONE | https://doi.org/10.1371/journal.pone.0278908
December 9, 2022 December 9, 2022
|
https://openalex.org/W4220930662
|
https://dergipark.org.tr/tr/download/article-file/2227697
|
English
| null |
A Study on the Determination of Mechanic Harvest Properties of Some Sweet Cherry Varieties
|
Black sea journal of agriculture
| 2,022
|
cc-by
| 4,850
|
A STUDY ON THE DETERMINATION OF MECHANIC HARVEST
PROPERTIES OF SOME SWEET CHERRY VARIETIES
Ersen OKUR1*, Selçuk ARIN1
1Tekirdag Namık Kemal University, Agricultural Faculty, Department of Biosystem Engineering, 59030, Tekirdağ, Turkey Ersen OKUR1*, Selçuk ARIN1 1Tekirdag Namık Kemal University, Agricultural Faculty, Department of Biosystem Engineering, 59030, Tekirdağ, Turkey Abstract: Cherry is a hand-harvested fruit due to various difficulties and constraints. This condition results in the use of labor at a high
rate. Turkey's annual cherry production is at the level of 724 thousand tons. Various tools or machines that have been developed for
cherry harvest have the potential to contribute greatly to the production in this area. The share of the labor required for harvesting in
cherry production in the total labor requirement is around 70%. In this research, it was aimed to collect the necessary data to
mechanize the cherry harvest by determining the physico-mechanical properties of cherry fruit. As a result of the present study,
several physical, biological, and mechanical properties of four sweet cherry variety (0900 Ziraat, Starks Gold, Merton Late, Lambert)
were determined and compared in terms of fruit mass, net fruit weight, tensile force, weight, thickness, length, width, sphericity,
surface area, volume of fruit, and also weight, width, length, sphericity of seed, tensile force of stalk, stalk length, and weight. Tensile
force of fruit, tensile force of stalk and weight of the fruit of 0900 Ziraat variety were found 2.579 N, 7.041 N, 9.592 g, respectively. After the evaluation of the obtained data, it was determined that all four cherry varieties examined were suitable for mechanical
harvesting. However, the most suitable variety for mechanical harvesting was found as ‘0900 Ziraat’. Keywords: Sweet cherry, Fruit harvesting, Mechanization, Tensile force Keywords: Sweet cherry, Fruit harvesting, Mechanization, Tensile force
*Corresponding author: Tekirdag Namık Kemal University, Agricultural Faculty, Department of Biosystem Engineering, 59030, Tekirdağ, Turkey
E mail: eokur@nku.edu.tr (E. OKUR)
Ersen OKUR
https://orcid.org/0000-0003-1933-7642
Received: February 02, 2022
Accepted: March 11, 2022
Published: April 01, 2022
Selçuk ARIN
https://orcid.org/0000-0002-6053-5950
Cite as: Okur E, Arın S. 2022. A study on the determination of mechanic harvest properties of some sweet cherry varieties. BSJ Agri, 5(2): 137-142. Keywords: Sweet cherry, Fruit harvesting, Mechanization, Tensile force 2021). ‘0900 Ziraat’ variety is the cherry variety with the
highest production volume in Turkey and is an important
export product (Erogul, 2018). Black Sea Journal of Agriculture
doi: 10.47115/bsagriculture.1066907 Open Access Journal
e-ISSN: 2618 – 6578 Research Article
Volume 5 - Issue 2: 137-142 / April 2022 This work is licensed under Creative Commons Attribution 4.0 International License BSJ Agri / Ersen OKUR and Selçuk ARIN
This work is licensed under Creative Commons Attribution 4.0 International License 1. Introduction 69% of the exported cherries were sold to EU
member countries (Anonymous, 2020; Anonymous, BSJ Agri / Ersen OKUR and Selçuk ARIN 137 Black Sea Journal of Agriculture and
Gökduman,
2020). Kocabiyik
et
al. (2009)
determined the human energy cost, work success, and
some physico-mechanical properties of fruits in apple,
peach, apricot, cherry, and plum harvest. As a result, they
determined the highest human energy input for cherry
harvest. Pırlak and Güleryüz (2000) reported that the
fruit harvest is 100-250 times higher than the grain
harvest in terms of labor and approximately 40 times
higher in terms of production costs. (2009) Merton Late, and Lambert) were used in the study
(Figure 1). Merton Late, and Lambert) were used in the study
(Figure 1). Merton Late, and Lambert) were used in the study
(Figure 1). Figure 1. Photos of garden and cherry varieties used in
the experiments. Erdoğan (1988) evaluated the human labor needs in
horticultural
agriculture
in
terms
of
harvest
mechanization. According to the results of the study,
human labor needs in strawberry and cherry harvesting
were found to be higher than other fruits. Pırlak and
Güleryüz (2000) examined the mechanical harvesting of
fruit species and concluded that the human labor
required for manual cherry harvesting is high. This
process constitutes 40-80% of the working time in
production. Figure 1. Photos of garden and cherry varieties used in
the experiments. For each cherry variety, measurements and calculations
were made on a total of 103 samples from randomly
selected trees, in order to determine the properties of
cherries for mechanical harvesting and to determine
their suitability for mechanical harvesting. These are
weight, thickness, width, height, sphericity, surface area,
volume, density, net fruit weight, tensile force of the fruit
from the stalk, seed weight, width, thickness, height,
sphericity, surface area, and tensile force of the fruit
stalks from the branch, weight, length, thickness, and
number of the stalks. In addition, pH measurements and
color analyzes were performed for the examined
varieties (Vursavuş et al., 2006; Kocabıyık et al., 2009;
Pérez-Sánchez et al., 2010; İkinci and Bolat, 2015; Sarısu
et al., 2019). Knowing the biological properties of agricultural
products are necessary and important in the design,
construction, operation, control, determination of yields,
analysis and evaluation of the quality of the products. Knowing these properties is beneficial not only for
engineers but also for food scientists, processors, plant-
breeding designers and specialists (Mohsenin, 1986). 1. Introduction Cherry (Purunus avium L.) belongs to Rosaceae family,
Prunoideae subfamily, Prunus genus in botany. The
origin of the cherry is known as the region between the
Southern Caucasus, the Caspian Sea, and Northeast
Anatolia. It spread east and west from these gene centers
and covered a large area on Earth. Accordingly, our
country is one of the origin centers of cherries (Başkaya,
2011). There are around 1500 cherry variety in the
world, and this number is increasing day by day with
ongoing breeding studies. Besides, the same varieties
were named with different names, and different varieties
were named with the same names in terms of regions
(Çakaryıldırım, 2003). Table 1. The highest cherry produced countries in the
world (Anonymous, 2022) Table 1. The highest cherry produced countries in the
world (Anonymous, 2022)
Rankings
Country
Production
(metric ton)
Cultivated
Area (ha)
1
Turkey
724944
82729
2
USA
294900
34400
3
Chile
255471
39645
4
Uzbekistan
185068
12718
5
Iran
164080
24033 Cherry cultivated areas and the amount of cherry
produced accordingly increase every year in Turkey. Our
production, which was 230000 tons from 29000 ha in
2000, increased from 67046 ha to 417905 tonnes in
2010 and from 82729 ha to 724944 tonnes in 2020
(Anonymous, 2022). Harvesting of cherry fruit is still mostly done by hand
both in the world and in Turkey. Therefore, harvesting is
one of the areas where the labor requirement per unit
area is high in fruit growing and hand-harvesting
constitutes 30-60% of the total production cost (Moser,
1989; Gezer, 2001). When the situation of mechanical
fruit harvesting in Turkey is examined, it is seen that the
process with the highest labor requirement per unit area
is harvesting. When we look at the share of the labor
required for harvest in the total labor requirement, sour
cherry, and cherry production ranks first with 70%
(Gezer, 2001). Like cherries, sour cherry harvest costs
account for 30-60% of the total production cost (Yılmaz According to 2020 World cherry production data, Turkey
ranks first in terms of cherry production and cultivated
area (Table 1; Anonymous, 2022). Of the 664224 tons of
cherries produced in Turkey in 2019, 80508 tons were
exported. This amount constitutes 12.1% of the total
production. 2.2. Statistical Analysis Analysis of variance (ANOVA) was applied to the data
obtained from the study by using the statistical package
program (Statistica, 1999), and the differences between
the group means were determined with the Duncan
Multiple Comparison Test (Genç and Soysal, 2018). 2.3. The Tensile Force of the Fruit from the Stalk (FT-f)
and the Stalk from the Branch (FT-s) 1. Introduction For
these reasons, studies have been carried out by many
researchers to determine the physico-mechanical and
biological properties of different fruits. Research has
been carried out for many fruits such as apricot (Altıkat
and Temiz, 2019), apple (Polat et al., 2020), plum (Alnıak
Sezer and Çetin, 2021), black berry (Çalışır and Aydın,
2004), sour cherry and cherry (Özgüven et al., 2001;
Vursavuş et al., 2006; Kocabıyık et al., 2009; Pérez-
Sánchez et al., 2010; Göksel and Aksoy, 2014; İkinci and
Bolat, 2015; Taşova and Güzel, 2017; Krumov and
Christov, 2018; Sarısu et al., 2019). However, it should
not be forgotten that there are many variety of each fruit
and that even the same variety vary according to the
climate, soil, and cultivation method of the region in
which they are grown (Altikat and Temiz, 2019). Due to
that, it is important to conduct studies on the samples of
fruit variety in different regions, to have a large amount
of data on the characteristics of that fruit variety, and to
create a more reliable infrastructure for studies on
mechanical harvesting. In this study, it was aimed to
determine the physico-mechanical properties of four
cherry variety grown in Tekirdağ, to collect the data
necessary for the mechanization of cherry harvest and to
determine the suitability of these variety for mechanical
harvesting. The thickness, width, height of the fruits, and the length
and thickness of the fruit stalks were measured with a
digital caliper (Mitutoyo) with an accuracy of 0.01 mm,
and the weights of the fruits and stalks were measured
with a precision balance (AND - GF 600) with an accuracy
of 0.001 g. The tensile force of the fruit (from the stalk)
and the tensile force of the fruit stalk (from the branch)
were made with a 1 gr precision hand dynamometer
(Lutron FG 5020), pH measurements were analyzed by
pH meters (Hanna Instruments pH 211), and color
measurements were obtained from a colorimeter
(HunterLab D25LT - Reston, VA). Sphericity, net fruit
weight, surface area, and density values were calculated
from the obtained measurement results. Additionally, net
fruit weight was calculated by subtracting the weight of
the fruit stalk and seed from the total fruit weight. 2.1. Cherry Varieties To determine the tensile force of the fruit (from the
stalk), an equipment was made on the tip of the hand
dynamometer to pull the fruit from the stalk. With the
help of this equipment, the samples were pulled in the
direction of the fruit stalk axis and the tensile forces (FT-f) 3.1. pH Values The data obtained by measuring the parameters of fruit,
seeds, and stalks of four different cherry varieties were
statistically analyzed and the results were summarized in
Table 4, 5, and 6. The juices of the fruits were squeezed, and pH
measurements were made in three replications for each
cherry variety, and the averages are given in Table 2. Numerically the highest pH was measured in Lambert,
and the lowest pH was measured in 0900 Ziraat variety. The results are in line with the work done by Vursavuş et
al. (2006), Göksel and Aksoy (2014), Sarısu et al. (2019),
Eroğul and Özmen (2020). 3.3. Fruit Parameters Measurements Measured and calculated fruit parameters of investigated
cherry variety are shown in Table 4. When the tensile forces of the fruits pulling from the
stalk were examined, the highest value (2.579 N) was
measured in 0900 Ziraat variety, the lowest value (1.530
N) was measured in Lambert variety and the difference
was statistically significant (P < 0.05). When the weights
of the fruits were examined, all variety were found to be
statistically different from each other, the highest value
was found in 0900 Ziraat with 9.592 g, and the lowest
value was determined in Lambert variety with 3.684 g. BSJ Agri / Ersen OKUR and Selçuk ARIN 2.1. Cherry Varieties BSJ Agri / Ersen OKUR and Selçuk ARIN
The cherry fruits used for the research were obtained
from the trees in the orchard of Tekirdağ Viticulture
Research Institute (40.970860 N, 27.472279 E). Four
different cherry varieties (0900 Ziraat, Starks Gold, 138 Black Sea Journal of Agriculture Table 2. pH analysis results of cherry varieties
Varieties
pH
0900 Ziraat
3.86
Starks Gold
3.95
Merton Late
3.97
Lambert
4.24 Table 2. pH analysis results of cherry varieties were determined as N (Newton). After the fruits were
plucked from their stalks, the remaining stalks were
pulled from the branch with the help of an equipment
and the tensile force (FT-s) was measured. Fruits and
stalks which tensile forces were measured were
numbered for further measurements. 2.4. Fruit Sizes and Sphericity The size measurements of the fruits were made with a
digital caliper and the sphericity values were calculated
by placing the values in Equation 1 given below (Moser,
1989; Vursavuş et al., 2006; Yılmaz and Gökduman,
2020): Table 3. Color analysis results of cherry varieties
Varieties
L*
a*
b*
0900 Ziraat
18.25
9.25
2.66
Starks Gold
60.32
0.55
28.19
Merton Late
15.86
6.93
1.28
Lambert
13.83
1.56
0.23 Table 3. Color analysis results of cherry varieties ф = (𝑥𝑦𝑧)1 3
⁄ /𝑧 (1)
where;
ф = Sphericity (%)
x: Height of fruit (mm)
y: Thickness of the fruit (mm)
z: Width of the fruit (mm) (1) ф = (𝑥𝑦𝑧)1 3
⁄ /𝑧 ф = (𝑥𝑦𝑧)1 3
⁄ /𝑧 Findings for the variety ‘Ziraat 0900’ are in agreement
with Vursavuş et al. (2006). There are differences
between the values measured in the study and the values
obtained by Göksel and Aksoy (2014), Sarısu et al. (2019), Eroğul and Özmen (2020). It is thought that these
differences
may
be
caused
by
cultivation
and
environmental conditions. y: Thickness of the fruit (mm) 3.2. Color Measurements L* (lightness), a* (redness), b* (yellowness) values were
measured
with
HunterLab
colorimeter. The
measurements were made from randomly selected 3
samples for each variety and their mean values are
summarized in Table 3. According to the results of the study, the variety with the
highest brightness (L*) and yellowness (b*) values was
‘Starks Gold’, and the variety with the highest redness
(a*) value was ‘0900 Ziraat’. Table 4. Results of investigated fruit parameters
Fruit parameters
Cherry varieties
0900 Ziraat
Starks Gold
Merton Late
Lambert
Tensile force of fruit (from stalk) (N)
2.579±0.117a
2.187±0.872b
2.246±0.664b
1.530±0.803c
Weight (g)
9.592±1.095a
6.000±0.948c
7.130±1.592b
3.684±0.732d
Width (mm)
26.351±1.384a
22.526±1.474c
23.424±1.885b
15.917±1.133d
Thickness (mm)
23.213±1.201a
19.858±1.236b
20.143±1.719b
17.557±1.496c
Length (mm)
24.957±1.121a
21.267±1.096b
20.850±1.550c
16.467±1.208d
Sphericity (%)
0.939±0.022b
0.939±0.022b
0.912±0.025c
1.042±0.032a
Surface area (cm2)
19.228±1.712a
14.049±1.514b
14.402±2.188b
8.672±1.260c
Volume (cm3)
9.850±1.544a
5.728±1.066c
6.966±1.621b
3.528±0.760d
Density (g/cm3)
0.984±0.097c
1.061±0.122a
1.027±0.059b
1.063±0.181a
Net fruit weight (without seed) (g)
9.223±1.090a
5.623±0.927c
6.794±1.567b
3.403±0.717d
a-dMean values with different superscripts in the same row indicate a significant difference (P<0.05), n=103. Table 4. Results of investigated fruit parameters Table 4. Results of investigated fruit parameters Table 4. Results of investigated fruit parameters The lengths of the cherry stalks
were determined, and there was no statistical difference
between Merton Late, 0900 Ziraat, and Lambert varieties. Starks Gold variety was found to have the shortest stalk
length statistically. Similarly, as a result of stalk thickness
measurements, the lowest value was found in Starks Gold
variety. Merton Late and Lambert were measured as the
varieties with the highest stalk thicknesses. The obtained
results are similar to other studies on the subject (İkinci
and Bolat, 2015; Sarısu et al., 2019). For 0900 Ziraat variety, the tensile force of the fruit while
pulling from the branch was found to be 2.579 N, and the
tensile force of the fruit stalk from the branch was found
to be 7.041 N. Therefore, if the cherry is picked by pulling
from the fruit, the fruit will be separated from the stalk
first, and the stalk will remain on the branch. When it is
desired to collect the cherries without a stalk, they can be
picked by applying a lower force, but since the upper part
of the fruit whose stalk is broken off will be opened, the
endurance time will be reduced. In addition, when
stalkless picking is preferred, pulling down from the top
of the fruit rather than squeezing from the sides will
cause less damage. Table 4. Results of investigated fruit parameters BSJ Agri / Ersen OKUR and Selçuk ARIN 139 Black Sea Journal of Agriculture Black Sea Journal of Agriculture Table 5. Results of investigated seed parameters
Seed parameters
Cherry varieties
Ziraat 0900
Starks Gold
Merton Late
Lambert
Weight (g)
0.369±0.041a
0.377±0.042a
0.336±0.046b
0.284±0.044c
Width (mm)
11.126±0.421a
10.332±0.440c
10.581±0.429b
9.473±0.506d
Thickness (mm)
7.153±0.323b
6.988±0.291c
7.360±0.373a
6.823±0.390d
Length (mm)
9.240±0.342a
8.901±0.385b
8.976±0.372b
8.300±0.353c
Sphericity (%)
0.810±0.023c
0.834±0.022b
0.837±0.029b
0.856±0.024a
Surface area (cm2)
2.547±0.155a
2.330±0.159c
2.462±0.147b
2.067±0.188d
a-dMean values with different superscripts in the same row indicate a significant difference (P<0.05), n= 103. Table 5. Results of investigated seed parameters Table 6. Results of investigated stalk parameters
Stalk parameters
Cherry varieties
Ziraat 0900
Starks Gold
Merton Late
Lambert
Tensile force of stalk (from branch) (N)
7.041±2.793ab
6.355±2.391b
7.669±2.577a
7.522±2.567a
Number of stalks in cluster (unit)
1.320±0.546c
2.350±0.667a
1.728±0.795b
1.728±0.795b
Weight (g)
0.317±0.024b
0.106±0.016c
0.176±0.060a
0.176±0.063a
Length (mm)
55.680±5.549a
41.630±4.381b
54.534±5.917a
54.534±6.105a
Thickness (mm)
1.113±0.143b
1.196±0.086c
1.156±0.163a
1.164±0.157a
a-dMean values with different superscripts in the same row indicate a significant difference (P<0.05), n= 103. Table 6. Results of investigated stalk parameters Table 6. Results of investigated stalk parameters were found in Starks Gold varieties. Measured and
calculated values are summarized in Table 6. The sphericity values of the fruits were calculated, the
highest value (104%) was calculated for Lambert variety,
the lowest value (91%) was calculated for Merton Late
variety and the difference was statistically significant
(P<0.05). The surface area values of the fruits were
calculated, the highest value (19.228 cm2) was found for
0900 Ziraat variety, the lowest value (8.672 cm2) was
found for Lambert variety and the difference was
statistically significant (P<0.05). When the net fruit
weights were examined, all variety were found to be
statistically different from each other, the highest value
was found in 0900 Ziraat with 9.223 g, and the lowest
value was determined in Lambert variety with 3.403 g. The results regarding fruit sizes are consistent with the
findings obtained in previous studies (Vursavuş et al.,
2006; Delice et al., 2012; Sarısu et al., 2019; Eroğul and
Özmen, 2020). The number of cherry stalks in cluster was evaluated,
and there was no statistical difference between Merton
Late and Lambert variety (P>0.05). The highest value
was measured in Starks Gold with 2.350 units, and the
lowest value was measured in 0900 Ziraat variety with
1.320 units (P<0.05). 3.4. Fruit Seed Parameters Measurements When fruit seed weights were examined, the highest
values were measured in 0900 Ziraat and Starks Gold, the
lowest value was obtained from in Lambert variety and
the differences were found to be statistically significant. The measurement results were given in Table 5. In
addition, the sphericity values of the fruit seeds were
calculated, the highest value (85%) was in Lambert
variety, the lowest value (81%) was in 0900 Ziraat
variety and the differences compared to other variety
was found to be statistically significant (P<0.05). The
results are in line with the values measured by İkinci and
Bolat (2015). 3.6. Ratio of Fruit Mass to Tensile Force (M/FT) Ratio of fruit mass to tensile force (M/FT) results Tensile force of fruit (from stalk)
FT-f
N
2.579
2.187
2.246
1.530
Tensile force of stalk (from branch)
FT-s
N
7.041
6.355
7.669
7.522
Weight
M
g
9.592
6.000
7.13
3.684
According to tensile force of fruit
(from stalk)
M / FT-f
3.719
2.744
3.175
2.408
According to tensile force of stalk
(from branch)
M / FT-s
1.362
0.944
0.930
0.490
Fi
2 M/F
l
f
h
h
i
i Figure 2. M/FT-f values for the cherry varieties. harvest. As a result of the evaluation and analysis of the
obtained data, it was determined that all four cherry
varieties were suitable for mechanical harvesting. However, it was determined that the most suitable 3.6. Ratio of Fruit Mass to Tensile Force (M/FT) 3.6. Ratio of Fruit Mass to Tensile Force (M/FT)
Moser (1989) stated that if the ratio of fruit mass to
tensile force (M/FT) is equal or greater than 1, the fruit is
machine harvestable. The relationship between the
tensile force of the product and the mass of the product is
very important in the design of the harvesting units of
the harvesters, especially in terms of the selection of the
harvesting method. Table 7 summarizes the fruit mass to
tensile force ratio results of the research. The results indicates that all variety were suitable for
machine harvesting (M/FT>1) in the case of harvesting
the fruit without a stalk, and 0900 Ziraat variety was
suitable for harvesting with a stalk. When the results
were examined, it was determined that 0900 Ziraat was
the most suitable variety for machine harvesting in the
case of picking cherries with or without stalks. The
results found are in agreement with similar studies
(Kocabıyık et al., 2009). Between the examined varieties,
least suitable variety for machine harvesting is Lambert
variety (Figure 2). BSJ Agri / Ersen OKUR and Selçuk ARIN
141
Table 7. Ratio of fruit mass to tensile force (M/FT) results
Symbol
Unit
Cherry varieties
0900 Ziraat
Starks Gold
Merton Late
Lambert
Tensile force of fruit (from stalk)
FT-f
N
2.579
2.187
2.246
1.530
Tensile force of stalk (from branch)
FT-s
N
7.041
6.355
7.669
7.522
Weight
M
g
9.592
6.000
7.13
3.684
According to tensile force of fruit
(from stalk)
M / FT-f
3.719
2.744
3.175
2.408
According to tensile force of stalk
(from branch)
M / FT-s
1.362
0.944
0.930
0.490
Figure 2. M/FT-f values for the cherry varieties. 4. Conclusion
In the research, the physico-mechanical properties of
four cherry varieties grown in Tekirdag were determined
and the data were measured to mechanize the cherry
harvest. As a result of the evaluation and analysis of the
obtained data, it was determined that all four cherry
varieties were suitable for mechanical harvesting. However, it was determined that the most suitable Table 7. BSJ Agri / Ersen OKUR and Selçuk ARIN 3.5. Fruit Stalk Parameters If cherries are to be picked with a stalk, they should be
picked by pulling from the stalk, not from the fruit. Since
the thickness of the stalks is an average of 1.157 mm, a
design should be made to hold the fruit tightly so that the
stalk does not slip while pulling. In addition, the stalk
length, which was found to be 51.59 mm on average, is The tensile force of the cherry stalks from the branch was
measured, and there was no statistical difference
between the 0900 Ziraat variety and the other examined
varieties. However, the highest values were found in
Merton Late and Lambert varieties, and the lowest values BSJ Agri / Ersen OKUR and Selçuk ARIN 140 Black Sea Journal of Agriculture sufficient for the handle to be held by an apparatus. Since the difference between the tensile force values of
the fruits from the stalks is statistically significant, it will
be beneficial for the tensile force to be adjustable rather
than fixed in the designs to be made. The difference
between the weight, width, thickness, height, and volume
values of the cherries was also found to be statistically
significant. For this reason, the system should be
designed to be able to change and adjust according to
different properties of fruit varieties when necessary. According to the results of the research conducted by
Krumov and Christov (2018), cherry fruits should be
transported with appropriate methods and the variety
are suitable for mechanized harvesting. In the study
conducted by Peterson and Wolford (2001) on
mechanization, they stated that the cherries harvested by
the machine they developed were damaged only 2-6%
more than the traditionally harvested cherries, and the
ratio of marketable cherries were found as 85-92%. Eroğul and Özmen (2020) reported that 0900 Ziraat
variety is the most suitable variety after storage, shelf-life
properties and stands out in terms of some quality References Moser E. 1989. Bağ bahçe sebze ve endüstri kültürlerinde
mekanizasyon uygulamaları (Çeviri: Tuncer İK, Özgüven F). Türkiye Zirai Donatım Kurumu (T.Z.D.K.) Mesleki Yayınları,
Yayın No: 52, Ankara, Türkey, pp: 98. Alnıak Sezer S, Çetin M. 2021. Determination of some physical
and mechanical properties of plum fruit in different harvest
periods. Anadolu J Agr Sci, 36: 73-79. Alnıak Sezer S, Çetin M. 2021. Determination of some physical
and mechanical properties of plum fruit in different harvest
periods. Anadolu J Agr Sci, 36: 73-79. Altıkat S, Temiz Ş. 2019. Physico-mechanical and some
chemical properties of apricot varieties in Iğdır province. YYU
J Agr Sci, 29 (3): 373-381. Özgüven F, Vursavuş K, Özgüven AI. 2001. Determination of
some parameters related to mechanical harvesting of sour
cherry and similar fruits under the laboratory conditions. Fruit, nut, and vegetable production engineering. Proceeding
of the 6th International Symposium, September 11-14, 2001,
Posdam, Germany, pp: 111-115. ́́́ Anonymous. 2020. Kiraz değerlendirme raporu, 2020. URL:
https://www.tarimorman.gov.tr/BUGEM/Belgeler/MİLLİ%2
0TARIM/Ürün%20Masaları%20Ürün%20Değerlendirme%2
0Raporları%20yayımlandı/Kiraz%20Değerlendirme%20Rap
oru.pdf (access date: April 10, 2021). Pérez-Sánchez R, Gómez-Sánchez MA, Morales-Corts MR. 2010. Description and quality evaluation of sweet cherries cultured
in Spain. J Food Qual, 33(4): 490-506. Anonymous. 2021. List of Exporters for the Selected Product in
2019. Trade
Map. URL: Anonymous. 2021. List of Exporters for the Selected Product in
2019. Trade
Map. URL:
https://www.trademap.org/Country_SelProduct.aspx?nvpm
=1%7c%7c%7c%7c%7c080929%7c%7c%7c6%7c1%7c1%
7c2%7c1%7c1%7c2%7c1%7c1%7c1 (access date: April 10,
2021). Peterson DL, Wolford SD. 2001. Mechanical harvester for fresh
market quality stemless sweet cherries. Transactions of the
ASAE 2001, 44(3): 481-485. https://www.trademap.org/Country_SelProduct.aspx?nvpm
=1%7c%7c%7c%7c%7c080929%7c%7c%7c6%7c1%7c1%
7c2%7c1%7c1%7c2%7c1%7c1%7c1 (access date: April 10,
2021) Pırlak L, Güleryüz M. 2000. Meyve türlerinin mekanik yolla
hasatı. Tarımsal Mekanizasyon 19. Ulusal Kongresi Bildiri
Kitabı, Haziran 1-2, 2000, Erzurum, Türkey, pp: 253-258. Anonymous. 2022. Dünya kiraz üretim değerleri. URL:
http://www.fao.org/faostat/en/?#data/QC
(access
date:
January 30, 2022). Polat M, Mertoğlu K, Eskimez İ. 2020. Elmada bazı özelliklerin
birlikte ele alınabilme potansiyelleri: Pinova örneği. Ziraat
Müh, 370: 115-125. Başkaya Z. 2011. The geographical foundations of cherry
farming in Turkey. Doğu Coğrafya Derg, 16(26): 45-71. Sarısu HC, Demirtaş İ, Aksu M, Altındal M. 2019. Some physical
and chemical properties of new sweet cherry davraz. Harran
Tarım ve Gıda Bil Derg, 23(4): 391-399. Çakaryıldırım N. 2003. Kiraz. TEAE Bakış, 3(12): 1-4. Çalışır S, Aydın C. 2004. Some physico-mechanic properties of
cherry laurel (Prunus lauracerasus L.) fruits. J Food Eng,
65(1): 145-150. Statistica. 1999. Statistica for the windows operating system,
Stat Soft, Inc., Tulsa, OK, US. Acknowledgments Mohsenin NM. 1986. Physical properties of plant and animal
materials: Structure, physical characteristics and mechanical
properties. Gordon and Breach Science Publishers, Inc., New
York, US, pp: 891. This article has been produced from PhD thesis of Ersen
Okur. This article has been produced from PhD thesis of Ersen
Okur. Black Sea Journal of Agriculture variety for mechanical harvesting was ‘0900 Ziraat’. However, it would be beneficial to conduct similar
studies in different regions for different cherry varieties,
and thus to collect more results and high diversity of data
on mechanization. Genç S, Soysal Mİ. 2018. Parametric and nonparametric post
hoc tests. BSJ Eng Sci, 1(1): 18-27. hoc tests. BSJ Eng Sci, 1(1): 18 27. Gezer İ. 2001. Türkiye’de mekanik meyve hasadının durumu. Tarımsal Mekanizasyon 20. Ulusal Kongresi, September 13- 15, 2001, Şanlıurfa, Türkey, pp: 251-256. 15, 2001, Şanlıurfa, Türkey, pp: 251-256. Göksel Z, Aksoy U. 2014. Physico-chemical characteristics of
some table sweet cherry varieties. Turkish J Agric Natural Sci,
Special Issue 2: 1856-1862. Conflict of Interest The author declared that there is no conflict of interest. Krumov S, Christov N. 2018. Mechanical properties of the fruits
of some perspective sweet cherry cultivars. Scientific Papers,
Series B, Horticul, 62: 21-24. Author Contributions E.O.: initiated the research idea, developed, organized,
analyzed and interpreted the data and wrote the
manuscript. S.A.: supervised the research, suggested the
research methods, structured the paper and edited the
manuscript. İkinci A, Bolat İ. 2015. Investigation of the performance of some
sweet cherry varieties grown in the GAP region. Harran
Tarım ve Gıda Bil Derg, 19(2): 54-65. Kocabıyık H, Kavdır İ, Özpınar S. 2009. Technical and
economical analysis of some fruits hand harvested in
Çanakkale, and determination of some propeties of fruits
related with mechanical harvest. J Tekirdag Agri Fac, 6(1):
45-53. 4. Conclusion In the research, the physico-mechanical properties of
four cherry varieties grown in Tekirdag were determined
and the data were measured to mechanize the cherry BSJ Agri / Ersen OKUR and Selçuk ARIN 141 Black Sea Journal of Agriculture References Delice A, Ekinci N, Özdüven FF, Gür E. 2012. Determinations of
factors that effect on quality properties of 0900 Ziraat cherry
variety in Lapseki. J Tekirdag Agri Fac, 9(3): 27-34. Taşova M, Güzel M. 2017. Determination physico-mechanic
properties and color values of İstanbul variety cherry
(Prunus cerasus L.). Gaziosmanpasa J Sci Res, 6(Special
Issue): 73-80. Erdoğan D. 1988. Bahçe bitkileri tarımında insan işgücü
ihtiyaçlarının
hasat
mekanizasyonu
açısından
değerlendirilmesi. Ankara Üniv Ziraat Fak Yıllığı, 38, Ankara,
Turkey, pp: Vursavuş K, Kelebek H, Selli S. 2006. A study on some chemical
and physico-mechanic properties of three sweet cherry
varieties (Prunus avium L.) in Turkey. J Food Eng, 74: 568-
575. Eroğul D, Özmen B. 2020. Determination of storage and shelf
life of sweet cherry varieties produced in early region. Ege
Ü niv Ziraat Fak Derg, 57(3): 359-366. Yılmaz D, Gökduman ME. 2020. Effects of different harvesting
times on physico-mechanical properties of sour cherry
(Prunus cerasus L.). Erwerbs-Obstbau, 62: 335-340. Erogul D. 2018. An overview of sweet cherry fruit cultivation in
Turkey. Trends in Horticul, 1: 1-4. BSJ Agri / Ersen OKUR and Selçuk ARIN 142
|
https://openalex.org/W4390675483
|
https://bmcpulmmed.biomedcentral.com/counter/pdf/10.1186/s12890-023-02820-x
|
English
| null |
Asymmetrical high-flow nasal cannula performs similarly to standard interface in patients with acute hypoxemic post-extubation respiratory failure: a pilot study
|
BMC pulmonary medicine
| 2,024
|
cc-by
| 7,633
|
Abstract Background Standard high-flow nasal cannula (HFNC) is a respiratory support device widely used to manage post-
extubation hypoxemic acute respiratory failure (hARF) due to greater comfort, oxygenation, alveolar recruitment,
humidification, and reduction of dead space, as compared to conventional oxygen therapy. On the contrary, the
effects of the new asymmetrical HFNC interface (Optiflow® Duet system (Fisher & Paykel, Healthcare, Auckland, New
Zealand) is still under discussion. Our aim is investigating whether the use of asymmetrical HFNC interface presents
any relevant difference, compared with the standard configuration, on lung aeration (as assessed by end-expiratory
lung impedance (EELI) measured by electrical impedance tomography (EIT)), diaphragm ultrasound thickening
fraction (TFdi) and excursion (DE), ventilatory efficiency (estimated by corrected minute ventilation (MV)), gas
exchange, dyspnea, and comfort. Methods Pilot physiological crossover randomized controlled study enrolling 20 adults admitted to the Intensive
Care unit, invasively ventilated for at least 24 h, and developing post-extubation hARF, i.e., PaO2/set FiO2 < 300 mmHg
during Venturi mask (VM) within 120 min after extubation. Each HFNC configuration was applied in a randomized
60 min sequence at a flow rate of 60 L/min. Results Global EELI, TFdi, DE, ventilatory efficiency, gas exchange and dyspnea were not significantly different, while
comfort was greater during asymmetrical HFNC support, as compared to standard interface (10 [7–10] and 8 [7–9],
p-value 0.044). Conclusions In post-extubation hARF, the use of the asymmetrical HFNC, as compared to standard HFNC interface,
slightly improved patient comfort without affecting lung aeration, diaphragm activity, ventilatory efficiency, dyspnea
and gas exchange. Clinical trial number ClinicalTrial.gov. Registration number: NCT05838326 (01/05/2023). *Correspondence:
Nicolò Sella
nico.sella@hotmail.it.com Full list of author information is available at the end of the article Full list of author information is available at the end of the article © The Author(s) 2024. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use,
sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and
the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this
article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. BMC Pulmonary Medicine BMC Pulmonary Medicine Boscolo et al. BMC Pulmonary Medicine (2024) 24:21
https://doi.org/10.1186/s12890-023-02820-x Open Access Asymmetrical high-flow nasal cannula
performs similarly to standard interface
in patients with acute hypoxemic post-
extubation respiratory failure: a pilot study Annalisa Boscolo1,2,3, Tommaso Pettenuzzo2, Francesco Zarantonello2, Nicolò Sella2*, Elisa Pistollato1,
Alessandro De Cassai2, Sabrina Congedi1, Irene Paiusco1, Giacomo Bertoldo1, Silvia Crociani1, Francesca Toma1,
Giulia Mormando4, Giulia Lorenzoni5, Dario Gregori5 and Paolo Navalesi1,2 © The Author(s) 2024. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use,
sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and
the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this
article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included
in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will
need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The
Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available
in this article, unless otherwise stated in a credit line to the data. Abstract If material is not included
in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will
need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The
Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available
in this article, unless otherwise stated in a credit line to the data. Materials and methodsh This pilot physiological crossover randomized controlled
study included all consecutive adult patients, admitted to
the intensive care unit (ICU) of the University Hospital
of Padua (Italy) between May 8th and June 10th 2023,
undergoing invasive mechanical ventilation for at least
24 h and experiencing post-extubation hARF, defined as
arterial partial pressure of oxygen (PaO2) to set inspira
tory fraction of oxygen (FiO2) ratio < 300 mmHg during
VenturiMask (VM) support [8], within 120 min after
extubation. Exclusion criteria were: (i) long-term oxy
gen therapy, (ii) need for rescue noninvasive ventilation
after extubation (based on predefined criteria [19]), (iii)
chronic pulmonary disease, (iv) moderate-severe cardiac
failure, (v) severe acute respiratory syndrome coronavi
rus-2 infection, (vi) pregnancy, (vii) presence of trache
ostomy, (viii) contraindications to EIT [20] or HFNC
interface [4, 21], and (ix) requiring nasogastric tubes for
mandatory clinical reasons, i.e., delayed gastric emptying,
upper abdominal surgery. f
Indeed, they have been tested as first-line treatment
for avoiding intubation in patients experiencing hARF or
exacerbations of chronic obstructive pulmonary disease
(COPD), for preventing re-intubation, especially in non
surgical patients at low or moderate risk of extubation
failure or in post-operative patients at low or high risk of
pulmonary complications, and for preoxygenation during
endotracheal intubation [5, 7–16]. Recently, a new HFNC interface using asymmetrical
prongs was approved for clinical practice [17, 18]. Unlike
standard nasal cannulas with equally sized prongs, the
asymmetrical prongs deliver different flow rates between
the two nostrils [17, 18]. Bench studies have demon
strated that the asymmetrical configuration resulted
in higher positive end-expiratory pressure (PEEP) and
accelerated clearance of the anatomical dead space [2,
17]. As compared to the conventional HFNC interface,
Slobod and colleagues recently found the asymmetri
cal interface to be associated only with reduced minute
ventilation and work of breathing in 10 non-intubated
patients with mild-to-moderate hARF, likely attributable
to the enhanced carbon dioxide (CO2) clearance from the
upper airway [18]. Introduction effect of the asymmetrical HFNC interface to prevent
extubation failure. In recent years, high-flow nasal cannula (HFNC) oxy
gen therapy has become popular among intensivists to
manage patients with hypoxemic acute respiratory fail
ure (hARF) [1]. According to the last European Respira
tory Society task force, HFNC is a valuable intervention
for improving lung aeration, oxygenation and alveolar
recruitment in different populations, such as post-oper
ative patients and nonsurgical subjects at risk of extuba
tion failure or pulmonary complications [1–3].i We designed this pilot study for investigating whether
in patients developing hARF early after extubation, the
use of asymmetrical HFNC interface presents any rel
evant difference, compared with the standard configu
ration, on lung aeration - as assessed by end-expiratory
lung impedance (EELI) measured by electrical impedance
tomography (EIT) -, on diaphragm ultrasound thicken
ing fraction (TFdi) and excursion (DE) -, ventilatory effi
ciency - estimated by corrected minute ventilation (MV)
-, gas exchange, dyspnea, and comfort. p
y
p
[
]
HFNC delivers up to 60 L/min of warmed humidified
gas, with an inspired fraction of oxygen (FiO2) ranging
from 21 to 100% [4, 5]. HFNC promotes naso-pharyn
geal dead space washout, leading to a decrease of min
ute ventilation and diaphragm activity, and may increase
to some extent the end-expiratory lung volume (EELV)
consequent to a variable rise of end-expiratory airway
pressure [2, 5, 6]. Finally, by delivering warm and well
humidified gas, HFNC may facilitate the clearance of tra
cheobronchial secretions. Overall, HFNC has the poten
tial to improve oxygenation and patient comfort, while
increasing EELV and reducing inspiratory effort [5, 7, 8]. Indeed, they have been tested as first-line treatment
for avoiding intubation in patients experiencing hARF or
exacerbations of chronic obstructive pulmonary disease
(COPD), for preventing re-intubation, especially in non
surgical patients at low or moderate risk of extubation
failure or in post-operative patients at low or high risk of
pulmonary complications, and for preoxygenation during
endotracheal intubation [5, 7–16]. HFNC delivers up to 60 L/min of warmed humidified
gas, with an inspired fraction of oxygen (FiO2) ranging
from 21 to 100% [4, 5]. HFNC promotes naso-pharyn
geal dead space washout, leading to a decrease of min
ute ventilation and diaphragm activity, and may increase
to some extent the end-expiratory lung volume (EELV)
consequent to a variable rise of end-expiratory airway
pressure [2, 5, 6]. Introduction Finally, by delivering warm and well
humidified gas, HFNC may facilitate the clearance of tra
cheobronchial secretions. Overall, HFNC has the poten
tial to improve oxygenation and patient comfort, while
increasing EELV and reducing inspiratory effort [5, 7, 8].i (2024) 24:21 Page 2 of 11 Boscolo et al. BMC Pulmonary Medicine New & noteworthy The asymmetrical high-flow nasal cannula oxygen therapy (Optiflow® Duet system (Fisher &
Paykel, Healthcare, Auckland, New Zealand) provides greater comfort as compared to standard interface; while their
performance in term of lung aeration, diaphragm activity, ventilatory efficiency, dyspnea, and gas exchange is similar. Keywords High-flow nasal cannula, High flow nasal oxygen, High flow nasal therapy, Asymmetrical cannula,
Asymmetrical cannula, DUET Diaphragm ultrasound Diaphragm ultrasound evaluation was performed at the
bedside, during quiet breathing, with the patient in a
semi-recumbent position, by two trained intensivists (AB
and TP) [19], using a 4–12 MHz linear array transducer
(Mindray M9, North America, NJ, USA), placed perpen
dicular to the right chest wall between the 9th and 10th
intercostal spaces (at the level of apposition) after the EIT
evaluation, as previously published [19, 31].h Either with conventional HFNC or asymmetrical
device, the set FiO2 was titrated to maintain a peripheral
oxygen saturation between 92% and 98%, the gas flow
rate was set at 50–60 L/min (AIRVO 2, Fisher&Paykel
Healthcare, New Zealand), based on the size of the
nostril, and the temperature of the heated humidi
fier (Fisher&Paykel Healthcare, New Zealand) was set
at 37° C (absolute humidity delivered 44 mgH2O/L) for
the entire study period. Each step was 60 min long and a
10-min ‘wash-out’ phase with VM support was required
before each step. The standard and asymmetrical inter
faces were identically sized, i.e., small, medium, or large,
according to the distance between the patient’s nostrils,
as recommended by the manufacturers [21]. The diaphragm thickness was measured at both end-
expiration and inspiration, and TFdi was calculated as
the average of three respiratory cycles, according to the
formula: TFdi (%) = (inspiratory thickness-expiratory
thickness)/expiratory thickness*100 [32]. Diaphragm
ultrasound assessment was performed only on the right
side due to the lower interobserver reproducibility on the
left side [19, 33, 34]. The intra- and inter-observer agree
ment between the two observers was previously pub
lished [19]. DE was measured with a low frequency curved array
probe (2–5 MHz) positioned just below the costal arch
at the midclavicular line and by angling the ultrasound
beam as much as possible cranially and perpendicular
to the diaphragmatic dome. The diaphragm is identi
fied as a bright line that covers the liver and the spleen. During inspiration, the diaphragm should move toward
the probe [32, 35]. Excursion is quantified in M-mode,
with the M-line placed perpendicularly to the direction
of motion, and as the mean of three quiet breathings [32,
36]. Baseline demographic and clinical data were collected
from electronic health records. Diaphragm ultrasound During the last 10 min of
each phase (i.e., MV, standard HFNC, and asymmetrical
HFNC) the following variables were collected: respira
tory rate, pH, arterial oxygen saturation (SaO2), PaO2/set
FiO2, arterial pressure of carbon dioxide (PaCO2) and
FiO2, systolic blood pressure, diastolic blood pressure,
comfort, dyspnea, and EIT and ultrasound variables. Comfort and dyspnea were evaluated using a numeri
cal scale (NRS) (ranging between 0 and 10) and the Borg
scale, respectively [7, 25]. All patients were blinded to the
novelty of the asymmetrical interface. EIT acute respiratory distress (RR ≥ 35/min, agitation, recruit
ment of accessory muscles, and peripheral oxygen satu
ration < 90%); (2) life-threatening cardiac arrhythmias; (3)
copious secretions [22–24]. After meeting the inclusion criteria, a 16-electrode EIT
belt was placed around the chest, as previously described
[20, 26]. The following EIT parameters were recorded
during the last 10 min of each step and before ultra
sound assessment: (i) the average global tidal volume
(VT) and the percentage of VT distributed to non-depen
dent and dependent lung regions (VTglob, VTnon-dep,
and VTdep, respectively); (ii) the MV and the corrected
MV, calculated as [(VTglob x PaCO2)/40 mmHg]* respi
ratory rate per minute− 1, where 40 mmHg is the ideal
value of PaCO2 [27, 28]; (iii) the global and regional
changes in EELI (estimating EELV) during the VM and
in each HFNC phase (ΔEELIglob, ΔEELInon-dep, and
ΔEELIdep, respectively); (iv) the global inhomogeneity
index (GI) and the regional ventilation delay (RVD) [29,
30]. The study was approved by the Institutional Ethics
Committee of Padua University Hospital (reference num
ber: AOP2949) and registered on ClinicalTrial.gov (reg
istration number NCT05838326, 01/05/2023). The study
was carried out according to the Declaration of Helsinki
and written informed consent was obtained from all
patients. The asymmetrical HFNC interfaces were kindly
provided by Fisher & Paykel Healthcare (New Zealand)
only for research purposes, without any economic inter
ests. The industry was not involved in any phase of the
study. Randomization Randomization was performed within 120 min after
extubation, immediately after validation of the oxygen
ation criteria, defined as PaO2/set FiO2 < 300 mmHg
during VM support. According to a web-based blocked
random sequence, all patients received HFNC therapy
through the asymmetrical interface Optiflow® Duet sys
tem (Fisher & Paykel, Healthcare, Auckland, New Zea
land) and through the conventional interface. Materials and methodsh Attending ICU physicians identified patients as
ready to undergo the first spontaneous breathing trial
when they met the following predefined criteria in a
daily screening, as previously described [22–24]: (1)
PaO2/FiO2 ≥ 200 mmHg with PEEP ≤ 8 cmH2O and
FiO2 ≤ 0.4; (2) respiratory rate (RR) ≤ 30/min (during pres
sure support ventilation); (3) a cooperative cognitive state
(Richmond agitation-sedation scale between 0 and − 1);
and (4) hemodynamic stability (heart rate < 140 beats
min−1 and mean arterial pressure > 60 mmHg with nor
epinephrine < 0.1 mcg/kg/min or dobutamine < 5 mcg/
kg/min and without epinephrine). After a 30-minute
spontaneous breathing trial, the patient was extubated
only in the absence of any of these criteria: (1) signs of However, no study has, in so far, focused on patients
developing hARF after extubation and on assessing the Boscolo et al. BMC Pulmonary Medicine (2024) 24:21 Boscolo et al. BMC Pulmonary Medicine Page 3 of 11 Statistical analysis Continuous data are presented as median and interquar
tile range [IQR]. Being a pilot study, a sample size was
not calculated a priori. Comparisons between differ
ent interfaces were analyzed using the Wilcoxon signed Boscolo et al. BMC Pulmonary Medicine (2024) 24:21 Page 4 of 11 Boscolo et al. BMC Pulmonary Medicine R (version 4.0.3, R Foundation for Statistical Computing,
Vienna, Austria). R (version 4.0.3, R Foundation for Statistical Computing,
Vienna, Austria). rank test for paired data and all p-values were adjusted
by Benjamini and Hochberg method. Missing data was
omitted from the analysis. Subset analyses were per
formed according to the improvement on lung aeration
during the asymmetrical support. All statistical tests
were two-tailed and statistical significance was defined
by p < 0.05. Analyses were conducted using Prism (ver
sion 5.0; GraphPad Software, Inc, La Jolla, CA, USA) and rank test for paired data and all p-values were adjusted
by Benjamini and Hochberg method. Missing data was
omitted from the analysis. Subset analyses were per
formed according to the improvement on lung aeration
during the asymmetrical support. All statistical tests
were two-tailed and statistical significance was defined
by p < 0.05. Analyses were conducted using Prism (ver
sion 5.0; GraphPad Software, Inc, La Jolla, CA, USA) and Continuous variables are expressed as median, with interquartile range [IQR], and categorical variables are expressed as absolute values (%)
Abbreviations: N patient number, Tot Total, BMI body mass index, SAPS II simplified acute physiology score II at ICU admission, ICU intensive care unit, IMV invasive
mechanical ventilation, LOS length of stay, PaO2 partial arterial pressure of oxygen, FiO2 inspiratory fraction of oxygen Asymmetrical HFNC vs. venturi mask Asymmetrical HFNC vs. venturi mask Asymmetrical HFNC oxygen therapy decreased cor
rected MV and respiratory rate, as compared to VM
(p = 0.050 and p = 0.007, respectively); while SaO2 and
PaO2/set FiO2 were higher during asymmetrical HFNC,
as compared to VM (p = 0.006 and p = 0.001, respectively)
(Table 2). Finally, both comfort and dyspnea were not
different as compared to VM (p = 0.104) (Table 3). Addi
tional data on asymmetrical HFNC oxygen therapy, as
compared to VM, are reported in the Supplementary
materials (Table S3). Results As shown in Fig. 1, we evaluated for enrollment 35
patients and excluded 14 patients not meeting inclu
sion criteria. One patient was dropped out because of Fig. 1 CONSORT flow diagram for crossover trials. *Patients not meeting inclusion criteria: rescue noninvasive ventilation after extubation N = 5, chronic
pulmonary disease N = 2, tracheostomy N = 2, mandatory nasogastric tube N = 5. Abbreviations: HFNC high flow nasal cannula, N number, EIT electrical
impedance tomography Fig. 1 CONSORT flow diagram for crossover trials. *Patients not meeting inclusion criteria: rescue noninvasive ventilation after extubation N = 5, chronic
pulmonary disease N = 2, tracheostomy N = 2, mandatory nasogastric tube N = 5. Abbreviations: HFNC high flow nasal cannula, N number, EIT electrical
impedance tomography Boscolo et al. BMC Pulmonary Medicine (2024) 24:2 Boscolo et al. BMC Pulmonary Medicine Page 5 of 11 Furthermore, TFdi (p = 0.910) and DE (p = 0.891) were not
different in the two HFNC phases (Fig. 3 and S2). Furthermore, TFdi (p = 0.910) and DE (p = 0.891) were not
different in the two HFNC phases (Fig. 3 and S2). inadequate EIT recordings, leaving 20 patients eligible
for analysis. Patients’ characteristics are reported in Table 1. Median
age was 65 [55–76] years and seven (35%) were female. The most frequent etiology for admission to the ICU
was elective abdominal surgery (35%), followed by sep
tic shock (15%) and trauma (15%) (Table 1). Extubation
occurred after the first spontaneous breathing trial in 17
out of 20 (85%) patients, and after the second attempt in
3 (15%) patients. During VM support, PaO2/FiO2 was
195 [177–259] mmHg. The total duration of invasive
mechanical ventilation was 30 [27–84] hours, and ICU
stay was 2 [2–5] days. One patient (5%) died during the
ICU stay (Table 1). variables are expressed as median, with interquartile range [IQR], and categorical variables are expressed as absolute values (%) Gas exchange, dyspnea, and comfort Gas exchange, dyspnea, and comfort As shown in Table 3, respiratory rate, SaO2, PaO2/set
FiO2, gas exchange and hemodynamic parameters were
not different between asymmetrical and standard HFNC. Comfort, but not dyspnea, was higher during asym
metrical HFNC, compared to the conventional interface
(p = 0.044 and p = 0.763, respectively) (Table 3). s: N patient number, Tot Total, BMI body mass index, SAPS II simplified acute physiology score II at ICU admission, ICU intensive care u
ventilation, LOS length of stay, PaO2 partial arterial pressure of oxygen, FiO2 inspiratory fraction of oxygen Lung aeration and diaphragm activityf As shown in Fig. 2 and S1, no differences were found in
the percent change of global EELI (p = 0.159) and its dis
tribution in dependent (p = 0.364) and not-dependent
(p = 0.836) lung regions when passing from VM and
asymmetrical or standard HFNC. Also, MV and cor
rected MV were similar between asymmetrical and stan
dard HFNC, as well as global VT, GI and RVD (Table 2). Table 1 Characteristics of the study population
N
Gender
Size
cannula
Age
(years)
BMI
(kg/m2)
SAPS II
ICU admission
IMV
(hours)
ICU LOS
(days)
PaO2/FiO2
(Venturi
mask)
1
F
S
52
27
21
Trauma
84
7
173
2
M
L
84
25
52
Septic shock
131
5
263
3
F
M
77
26
47
Septic shock
281
14
169
4
M
L
54
29
49
Septic shock
320
14
145
5
F
M
53
21
34
Abdominal surgery
30
2
283
6
M
M
60
27
23
Neurosurgery
30
3
176
7
M
L
73
26
40
Abdominal surgery
25
2
279
8
F
M
77
27
40
Gastrointestinal bleeding
40
14
213
9
F
M
73
21
48
Abdominal surgery
28
2
184
10
M
L
65
26
38
Trauma
27
2
194
11
M
L
59
27
18
Endocrinological surgery
27
2
219
12
M
L
85
28
34
Neurosurgery
28
2
186
13
F
S
69
22
34
Abdominal surgery
28
2
245
14
M
L
68
27
38
Abdominal surgery
28
2
196
15
M
M
33
41
40
Trauma
29
2
183
16
M
M
52
28
26
Gastrointestinal bleeding
288
4
225
17
M
M
79
27
40
Thoracic surgery
83
4
180
18
F
M
64
26
29
Abdominal surgery
25
2
290
19
M
L
63
29
45
Otorhinolaryngological
surgery
36
3
163
20
M
M
60
33
18
Abdominal surgery
25
2
272
Total
20
F
(35%)
-
65
[55–76]
27
[26–29]
38
[27–44]
-
30
[27–84]
2
[2–5]
195
[177–259]
C
ti
i bl
d
di
ith i t
til
[IQR]
d
t
i
l
i bl
d
b
l t
l
(%) Page 6 of 11 Boscolo et al. BMC Pulmonary Medicine y
pedance tomography during standard and asymmetrical HFNC oxygen therapy. Variables are expressed a
tional data are reported in Fig. S1. Abbreviations: HFNC high-flow nasal cannula, DE diaphragmatic excursion, TFdi diaphragmatic thickening fraction, VT tidal volume, dep dependent, non-dep non-
dependent, EELI delta end-expiratory lung impedance (measured as percent change from Venturi mask), GI global inhomogeneity index, RVD regional ventilation
delay, MV minute ventilation, VM Venturi mask, N number Lung aeration and diaphragm activityf A: global lung aeration; B: lung aeration in dependent area; C: lung aera
significant, HFNC high-flow nasal cannula, dep dependent, non-dep non-dependent, EELI end-expiratory
rom VM) Fig. 2 Electrical impedance tomography during standard and asymmetrical HFNC oxygen therapy. Variables are expressed as median, with an interquar
tile range [IQR]. Additional data are reported in Fig. S1. A: global lung aeration; B: lung aeration in dependent area; C: lung aeration in non-dependent area. Abbreviations: ns not significant, HFNC high-flow nasal cannula, dep dependent, non-dep non-dependent, EELI end-expiratory lung impedance (measured
as percent change from VM) Fig. 2 Electrical impedance tomography during standard and asymmetrical HFNC oxygen therapy. Variables are expressed as median, with an interquar
tile range [IQR]. Additional data are reported in Fig. S1. A: global lung aeration; B: lung aeration in dependent area; C: lung aeration in non-dependent area. Abbreviations: ns not significant, HFNC high-flow nasal cannula, dep dependent, non-dep non-dependent, EELI end-expiratory lung impedance (measured
as percent change from VM) Boscolo et al. BMC Pulmonary Medicine Boscolo et al. BMC Pulmonary Medicine (2024) 24:21 Page 7 of 11 Page 7 of 11 Table 2 Electrical impedance tomography parameters
Variable
Standard HFNCa
(N = 20)
Asymmetrical
HFNCb
(N = 20)
Adjusted p-value
(a-b)
Venturi maskc
(n = 20)
Adjusted
p-value
(b-c)
MV (change from VM), %
−13 [−26, 1]
−12 [−18, −2]
0.731
–
0.145*
Corrected MV
(change from VM), %
−10 [−21, 3]
−11 [−19, −0.10]
0.992
–
0.050*
Global VT, (change from VM), %
106 [87–130]
100 [94–113]
0.654
–
0.749
VT non−dep, %
57 [31–66]
58 [41–66]
0.731
53 [42–60]
0.749
VT dep, %
51 [37–68]
51 [39–57]
0.731
48 [41–59]
0.936
Ratio VT non−dep/dep
1.04 [0.60–1.49]
1.12 [0.71–1.53]
0.843
1.10 [0.80–1.50]
0.749
GI index
48 [38–51]
49 [45–65]
0.654
49 [42–73]
0.749
Global RVD, %
11 [8–17]
12 [8–18]
0.336
11 [7–16]
0.936
RVD non−dep, %
14 [11–20]
14 [9–24]
0.336
13 [9–21]
0.749
RVD dep, %
9 [6–13]
9 [7–13]
0.654
9 [6–11]
0.936
Variables are expressed as median, with an interquartile range [IQR]. *p-value a-c: 0.018 and 0.006, respectively. Lung aeration and diaphragm activityf a = Standard HFNC; b = Asymmetrical HFNC;
c = Venturi mask
Abbreviations: HFNC high-flow nasal cannula, DE diaphragmatic excursion, TFdi diaphragmatic thickening fraction, VT tidal volume, dep dependent, non-dep non-
d
d
d l
d
l
d
(
d
h
f
k)
l b l
h
d
l
l Abbreviations: HFNC high-flow nasal cannula, DE diaphragmatic excursion, TFdi diaphragmatic thickening fraction, VT tidal volume, dep dependent, non-dep non-
dependent, EELI delta end-expiratory lung impedance (measured as percent change from Venturi mask), GI global inhomogeneity index, RVD regional ventilation
delay, MV minute ventilation, VM Venturi mask, N number Abbreviations: HFNC high-flow nasal cannula, DE diaphragmatic excursion, TFdi diaphragmatic thickening fraction, VT tidal volume, dep dependent, non-dep non-
dependent, EELI delta end-expiratory lung impedance (measured as percent change from Venturi mask), GI global inhomogeneity index, RVD regional ventilation
delay, MV minute ventilation, VM Venturi mask, N number Fig. 3 Diaphragm ultrasound evaluation during standard and asymmetrical HFNC oxygen therapy. Variables are expressed as median, with an inter
quartile range [IQR]. Additional data are reported in Fig. S2. A: TFdi; B: diaphragm excursion. Abbreviations: HFNC high-flow nasal cannula, TFdi diaphragm
thickening fraction Fig. 3 Diaphragm ultrasound evaluation during standard and asymmetrical HFNC oxygen therapy. Variables are expressed as median, with an inter
quartile range [IQR]. Additional data are reported in Fig. S2. A: TFdi; B: diaphragm excursion. Abbreviations: HFNC high-flow nasal cannula, TFdi diaphragm
thickening fraction Fig. 3 Diaphragm ultrasound evaluation during standard and asymmetrical HFNC oxygen therapy. Variables are expressed as median, with an inter
quartile range [IQR]. Additional data are reported in Fig. S2. A: TFdi; B: diaphragm excursion. Abbreviations: HFNC high-flow nasal cannula, TFdi diaphragm
thickening fraction Boscolo et al. BMC Pulmonary Medicine (2024) 24:21 Boscolo et al. Lung aeration and diaphragm activityf BMC Pulmonary Medicine Page 8 of 11 Page 8 of 11 Table 3 Gas exchange, hemodynamic parameters, dyspnea, and
comfort
Variable
Standard
HFNCa
(N = 20)
Asym
metrical
HFNCb
(N = 20)
Ad
justed
p-value
(a-b)
Venturi
maskc
(n = 20)
Ad
justed
p-value
(b-c)
Set FiO2, %
40 [40–40] 40 [38–40] 0.763
40
[40–40]
0.104
Respiratory
rate*min− 1
14 [10–16] 14 [11–18] 0.173
18
[13–19]
0.007
pH
7.46
[7.43–7.48]
7.46
[7.42–7.50]
0.763
7.43
[7.41–7.46]
0.350
SaO2, %
97 [96–98] 97 [97–98] 0.763
96
[94–97]
0.006
PaO2/set FiO2,
mmHg
237
[211–322]
247
[225–300]
0.173
195
[177–259]
0.001
PaCO2, mmHg
42 [39–44] 42 [39–45] 0.763
41
[39–43]
0.492
SBP, mmHg
135
[122–152]
138
[122–158]
0.763
135
[126–164]
0.236
DBP, mmHg
63 [55–72] 64 [56–71] 0.763
65
[59–72]
0.452
Comfort
(range 0–10)
8 [7–9]
10 [7–10]
0.044
8 [6–10]
0.104
Dyspnea
(range 0–10)
0 [0–1]
0 [0–1]
0.763
0 [0–2]
0.104
Variables are expressed as median, with an interquartile range [IQR]. a = Standard HFNC; b = Asymmetrical HFNC; c = Venturi mask
Abbreviations: HFNC high-flow nasal cannula, PaO2 arterial partial pressure
of oxygen, PaCO2 arterial partial pressure of carbon dioxide, FiO2 inspiratory
fraction of oxygen, SBP systolic blood pressure, DBP diastolic blood pressure,
SaO2 arterial oxygen saturation, N number led to a significant improvement of the ‘PEEP effect’ [17,
37, 38]. On the contrary, in 10 ICU patients affected by
mild-to-moderate hARF, Slobod et al. showed that the
asymmetrical HFNC interface did not affect alveolar
recruitment, dorsal fraction of ventilation and end-expi
ratory lung impedance, thus suggesting no major effect
on alveolar aeration [18]. Our results are in keeping with
those findings despite uneven populations. Indeed, in
our cohort of adults experiencing post-extubation hARF,
EELI was similar between asymmetrical and standard
configuration, without any difference between ventral
and dorsal aeration. However, our results on lung aera
tion may be limited because, first, we did not measure
EELV directly with computed tomography [39], but only
through a derived EIT parameter (i.e., EELI) and, second,
because we cannot exclude that some patients breathe
with their mouths open, which may decrease the ‘PEEP
effect’ associated with HFNC oxygen therapy [11, 40]. Subset analysis Additional analyses were performed considering only
patients (14, 70%) improving lung aeration using stan
dard HFNCs, as shown in the Supplementary materials
(Tables S4 and S5). Once again, the asymmetrical inter
face, as compared to the standard configuration, shows
higher patient comfort (p = 0.016). Lung aeration and diaphragm activityf However, all above mentioned bench studies, describ
ing an increased ‘PEEP effect’ during the asymmetrical
HFNC, suffer an important limitation worthy of discus
sion, such as collecting data from ‘normal’ upper airway
models, and not accounting for anatomical abnormalities
that may affect nasal flow distribution and ‘PEEP effect’
[17, 37, 38]. Likewise, data on the role of asymmetrical nostrils in
reducing the patients’ work of breathing are still conflict
ing. Interestingly, Slobod et al. found that the inspiratory
esophageal pressure-time product was slightly reduced
with the asymmetrical HFNC, in comparison with the
standard interface [18]. Since we cannot exclude that the
presence of the esophageal catheter, useful for measuring
the esophageal pressure-time product or the diaphragm
electrical activity, may affect either the dead space clear
ance or the ‘PEEP effect’, we decided to remove any
nasogastric tube before protocol initiation [37, 41]. As
an alternative, we decided to explore the patient inspira
tory effort by ultrasound assessment, a less invasive tech
nique, with easy learning and high reproducibility [11, 40,
42]. Based on our findings, both TFdi and DE were simi
lar between different interfaces. Discussion In this pilot physiological study, randomizing 20 ICU
patients with post-extubation hARF, the use of the asym
metrical HFNC, despite showing similar performances in
terms of lung aeration, TFdi, DE, ventilatory efficiency,
and gas exchange, was associated with improved patient
comfort, compared to standard HFNC interface. f
Furthermore, standard HFNC support has been shown
to reduce dead space and to improve CO2-washout in
mixed populations (i.e., hypoxemic ICU patients, hyper
capnic COPD subjects ect) [1]. In keeping with those
previous findings, Tatkov et al. showed an increased
CO2-washout, and Slobod et al. observed an increased
ventilatory efficiency during asymmetrical HFNC sup
port, as compared to the standard interface [17, 18]. According to our findings, the asymmetrical HFNC
performed similarly to the standard interface, probably
due to an important heterogeneity in patients’ baseline
characteristics that may affect the comparison with the
above-mentioned study [18]. Indeed, in our cohort only 4 While conventional HFNC has been shown to gener
ate a ‘PEEP effect’, promoting alveolar recruitment and
improving oxygenation, with a flow-dependent increase
in global EELI, a valid surrogate of alveolar recruitment,
clinical evidences on the potential benefits of the use of
asymmetrical prongs on lung aeration are still lacking [2,
5–8, 18]. Recent bench studies, collecting data from anatomi
cally ‘correct’ three-dimensional upper airway models,
showed that an increase in asymmetrical nare occlusion Page 9 of 11 Boscolo et al. BMC Pulmonary Medicine (2024) 24:21 Boscolo et al. BMC Pulmonary Medicine non-dep
non-dependent
NRS
numerical scale
PaO2
arterial partial pressure of oxygen
PaCO2
arterial pressure of carbon dioxide
PEEP
positive end-expiratory pressure
RVD
regional ventilation delay
RR
respiratory rate
SAPS II
simplified acute physiology score II PaO2
SaO2
arterial oxygen saturation
SBP
systolic blood pressure
Tot
total
VT
tidal volume
TFdi
diaphragm ultrasound thickening fraction
VM
Venturi mask out of 20 (20%) patients were intubated for ‘primary’ ARF
or pneumonia (i.e., patient n. 1, 3, 4, 17), while Slobod et
al. enrolled 6 out of 10 (60%) patients with ‘primary’ ARF
[18]. Finally, despite the absence of relevant differences
between the standard and asymmetrical interface, our
results suggest greater comfort during the asymmetri
cal HFNC interface, favoring their application routinely. The reasons why our patients reported greater comfort
during the asymmetrical interface are not entirely clear. Data availability
h d
d The datasets used and/or analysed during the current study are available for
the corresponding author on reasonable request. Funding
f
d No funding was received for the present study. The asymmetrical HFNC
interfaces were kindly provided by Fisher & Paykel Healthcare (New Zealand)
only for research purposes, without any economic interests. The industry was
not involved in any phase of the study. ff
In conclusion, in acute post-extubation hARF, the use
of the asymmetrical HFNC, as compared to standard
HFNC, improved patient comfort slightly, despite similar
performances in terms of lung aeration, TFdi, DE, venti
latory efficiency, dyspnea and gas exchange. Acknowledgements We are indebted to all ICU colleagues who made this work possible (Eugenio
Serra MD, Arianna Peralta MD, Luisa Muraro MD, Paolo Persona MD, Enrico
Petranzan MD). Furthermore, the authors thank Kirstin Elisabeth Rose, MD, for
proofreading the article and editing English. Author contributions AB, TP, FZ, NS and PN substantially contributed to the study design, data
interpretation and the writing of the manuscript. GL, EP, ADC, SC, FT, IP, GB,
SC and GM contributed to data collection, interpretation, and the writing
of the manuscript. GL, ADC and DG conceived, performed and guaranteed
the accuracy of data analysis. AB, ZF, NS, DG and PT critically revised the
manuscript for important intellectual content. All authors read and approved
the final manuscript. Supplementary Information The online version contains supplementary material available at https://doi. org/10.1186/s12890-023-02820-x. f
Our study has some limitations. First, during our trial,
it was not possible to control the potential impact of
spontaneous patient movements on EIT recordings [26],
although we marked the initial EIT belt position, as pre
viously described [26]. Second, we cannot exclude that
the absence of a ‘PEEP effect’ may be due to the potential
impact of mouth breathing during HFNC, as previously
described [11, 40]. Third, we enrolled patients with a
median invasive mechanical ventilation of 30 h. So doing,
the effect on alveolar recruitment and CO2 clearance in
case of longer endotracheal intubation remains unclear
and further studies are required to clarify this issue. Finally, due to the explorative nature of our investigation
we cannot exclude that our study could be underpowered
to assess any difference between the asymmetrical and
standard interface. In fact, our sample size could not be
enough to measure a possible effect on lung aeration and
diaphragm activity. In addition, further studies are nec
essary to explore the impact of asymmetrical HFNCs in
different clinical settings, with different patient selection. Supplementary Material 2: Additional data on diaphragm ultrasound
evaluation, EIT and gas exchanges Abbreviations Abbreviations
BMI
body mass index
CO2
carbon dioxide
COPD
chronic obstructive pulmonary disease
DBP
diastolic blood pressure
DE
diaphragm excursion
dep
dependent
EELI
end-expiratory lung impedance
EIT
electrical impedance tomography
EELV
end-expiratory lung volume
GI
global inhomogeneity index
HFNC
high-flow nasal cannula
hARF
hypoxemic acute respiratory failure
ICU
intensive care unit
IMV
invasive mechanical ventilation
IQR
interquartile range
LOS
length of stay
MV
minute ventilation
N
number Competing interests PN research lab received grants/research equipment by Draeger, Mindray,
Intersurgical SPA and Gilead. PN receives royalties from Intersurgical SPA
for Helmet Next invention. He also received speaking fees from Getinge,
Intersurgical SPA, Gilead, MSD, Draeger. NS and FZ received speaking fees from
Getinge. The other authors have no other competing interests to declare. Discussion However, our results seem to be promising for realizing
further studies investigating the impact of asymmetrical
cannulas on lung aeration and diaphragm activity in dif
ferent clinical settings, with different patient selection. Supplementary Information Ethics approval and consent to participate The study was approved by the Institutional Ethics Committee of Padua
University Hospital (reference number: AOP2949) and written informed
consent was obtained from all patients. References 20. Sella N, Pettenuzzo T, Zarantonello F, Andreatta G, De Cassai A, Schiavolin
C, et al. Electrical impedance tomography: a compass for the safe route to
optimal PEEP. Respir Med. 2021;187:106555. 1. Oczkowski S, Ergan B, Bos L, Chatwin M, Ferrer M, Gregoretti C, et al. ERS clini
cal practice guidelines: high-flow nasal cannula in acute Respiratory Failure. Eur Respir J. 2022;59(4):2101574. 1. Oczkowski S, Ergan B, Bos L, Chatwin M, Ferrer M, Gregoretti C, et al. ERS clini
cal practice guidelines: high-flow nasal cannula in acute Respiratory Failure. Eur Respir J. 2022;59(4):2101574. 21. Park S. High-flow nasal cannula for Respiratory Failure in adult patients. Acute
Crit Care. 2021;36(4):275–85. p
2. Pinkham M, Tatkov S. Effect of flow and cannula size on generated pressure
during nasal high flow. Crit Care Lond Engl. 2020;24(1):248. p
2. Pinkham M, Tatkov S. Effect of flow and cannula size on generated pressure
during nasal high flow. Crit Care Lond Engl. 2020;24(1):248. 22. Boles JM, Bion J, Connors A, Herridge M, Marsh B, Melot C, et al. Weaning
from mechanical ventilation. Eur Respir J. 2007;29(5):1033–56. 3. Goligher EC, Dres M, Patel BK, Sahetya SK, Beitler JR, Telias I, et al. Lung-
and diaphragm-protective ventilation. Am J Respir Crit Care Med. 2020;202(7):950–61. 3. Goligher EC, Dres M, Patel BK, Sahetya SK, Beitler JR, Telias I, et al. Lung-
and diaphragm-protective ventilation. Am J Respir Crit Care Med. 2020;202(7):950–61. 23. Dres M, Dubé BP, Mayaux J, Delemazure J, Reuter D, Brochard L, et al. Coexistence and impact of Limb muscle and diaphragm weakness at Time
of Liberation from Mechanical Ventilation in Medical Intensive Care Unit
patients. Am J Respir Crit Care Med. 2017;195(1):57–66. 4. Roca O, Hernández G, Díaz-Lobato S, Carratalá JM, Gutiérrez RM, Masclans JR,
et al. Current evidence for the effectiveness of heated and humidified high
flow nasal cannula supportive therapy in adult patients with Respiratory
Failure. Crit Care Lond Engl. 2016;20(1):109. 24. Dubé BP, Dres M, Mayaux J, Demiri S, Similowski T, Demoule A. Ultrasound
evaluation of diaphragm function in mechanically ventilated patients:
comparison to phrenic stimulation and prognostic implications. Thorax. 2017;72(9):811–8. 5. Mauri T, Alban L, Turrini C, Cambiaghi B, Carlesso E, Taccone P, et al. Optimum
support by high-flow nasal cannula in acute hypoxemic Respiratory Failure:
effects of increasing flow rates. Intensive Care Med. 2017;43(10):1453–63.l 25. Vitacca M, Bianchi L, Zanotti E, Vianello A, Barbano L, Porta R, et al. References Assess
ment of physiologic variables and subjective comfort under different levels
of pressure support ventilation. Chest. 2004;126(3):851–9. 6. Lee JH, Rehder KJ, Williford L, Cheifetz IM, Turner DA. Use of high flow nasal
cannula in critically ill infants, children, and adults: a critical review of the
literature. Intensive Care Med. 2013;39(2):247–57. 26. Zarantonello F, Sella N, Pettenuzzo T, Andreatta G, Calore A, Dotto D, et al. Early physiologic effects of Prone Positioning in COVID-19 Acute Respiratory
Distress Syndrome. Anesthesiology. 2022;137(3):327–39. 7. Maggiore SM, Idone FA, Vaschetto R, Festa R, Cataldo A, Antonicelli F, et al. Nasal high-flow versus Venturi mask oxygen therapy after extubation. Effects
on oxygenation, comfort, and clinical outcome. Am J Respir Crit Care Med. 2014;190(3):282–8. 27. Fusina F, Albani F, Bertelli M, Cavallo E, Crisci S, Caserta R, et al. Corrected
Minute Ventilation is Associated with Mortality in ARDS caused by COVID-19. Respir Care. 2021;66(4):619–25. 8. Maggiore SM, Jaber S, Grieco DL, Mancebo J, Zakynthinos S, Demoule A, et al. High-Flow Versus VenturiMask Oxygen Therapy to Prevent Re-intubation in
Hypoxemic patients after Extubation: a Multicenter, Randomized Clinical Trial. Am J Respir Crit Care Med. 2022;206(12):1452–62. 28. Zheng M. Dead space ventilation-related indices: bedside tools to evaluate
the ventilation and perfusion relationship in patients with acute respiratory
distress syndrome. Crit Care Lond Engl. 2023;27(1):46. 9. Arman PD, Varn MN, Povian S, Davis A, Uchakin P, Bhar A, WEANING, AND
EXTUBATION [Internet]. Effects of Direct Extubation to High-Flow Nasal
Cannula Compared to Standard Nasal Cannula in Patients in the Intensive
Care Unit. In: A53 CRITICAL CARE: PROBLEMS RELATED TO INTUBATION,. American Thoracic Society; 2017 [cited 2022 Aug 26]. p. A1887–A1887. (American Thoracic Society International Conference Abstracts). Available
from: https://www.atsjournals.org/doi/abs/https://doi.org/10.1164/ajrccm-
conference.2017.195.1_MeetingAbstracts.A1887. 29. Zhao Z, Möller K, Steinmann D, Frerichs I, Guttmann J. Evaluation of an
electrical impedance tomography-based Global Inhomogeneity Index for
pulmonary ventilation distribution. Intensive Care Med. 2009;35(11):1900–6. 30. Hsu YL, Tien AJ, Chang MY, Chang HT, Möller K, Frerichs I, et al. Regional ven
tilation redistribution measured by electrical impedance tomography during
spontaneous breathing trial with automatic tube compensation. Physiol
Meas. 2017;38(6):1193–203. 31. Bernardinello N, Cocconcelli E, Boscolo A, Castelli G, Sella N, Giraudo C, et al. Prevalence of diaphragm dysfunction in patients with interstitial lung Disease
(ILD): the role of diaphragmatic ultrasound. Respir Med. 2023;216:107293. 10. Cho JY, Kim HS, Kang H, Kim SH, Choe KH, Lee KM, et al. Received: 22 August 2023 / Accepted: 18 December 2023 19. Boscolo A, Sella N, Pettenuzzo T, Pistollato E, Calabrese F, Gregori D, et al. Diaphragm dysfunction predicts weaning outcome after bilateral lung trans
plant. Anesthesiology. 2023. https://doi.org/10.1097/ALN.000000000004729. Author details
1D
f 1Department of Medicine (DIMED), University of Padua, Padua, Italy Page 10 of 11 Page 10 of 11 Boscolo et al. BMC Pulmonary Medicine (2024) 24:21 Boscolo et al. BMC Pulmonary Medicine (2024) 24:21 Boscolo et al. BMC Pulmonary Medicine (2024) 24:21 2Institute of Anaesthesia and Intensive Care, Padua University Hospital, 13,
Giustiniani Street, Padua 35128, Italy a systematic review and network meta-analysis. Intensive Care Med. 2022;48(2):137–47. a systematic review and network meta-analysis. Intensive Care Med. 2022;48(2):137–47. 16. Thille AW, Monseau G, Coudroy R, Nay MA, Gacouin A, Decavèle M, et al. Non-invasive ventilation versus high-flow nasal oxygen for postextubation
Respiratory Failure in ICU: a post-hoc analysis of a randomized clinical trial. Crit Care Lond Engl. 2021;25(1):221. 3Thoracic Surgery and Lung Transplant Unit, Department of Cardiac,
Thoracic, Vascular Sciences, and Public Health, University of Padua, Padua,
Italy 4Emergency Department, Padua University Hospital, Padua, Italy
5Unit of Biostatistics, Epidemiology, and Public Health, Department
of Cardiac, Vascular Sciences, and Public Health, University of Padua,
Thoracic, Padua, Italy 17. Taktov S, Rees M, Gulley A, van den Lgt H, Nilius GN. Asymmetrical nasal
high flow ventilation improves clearance of CO2 from the anatomical
dead space and increases positive airway pressure. J Appl Physiol (1985). 2023;134(2):365–77. 18. Slobod D, Spinelli E, Crotti S, Lissoni A, Galazzi A, Grasselli G, et al. Effects of an
asymmetrical high flow nasal cannula interface in hypoxemic patients. Crit
Care Lond Engl. 2023;27(1):145. Received: 22 August 2023 / Accepted: 18 December 2023 References J Bras Pneumol Publicacao of Soc Bras Pneumol E Tisilogia. 2020;46(6):e20200064. clinical uses. J Bras Pneumol Publicacao of Soc Bras Pneumol E Tisilogia. 2020;46(6):e20200064. 37. Brüning J, Hildebrandt T, Heppt W, Schmidt N, Lamecker H, Szengel A, et al. Characterization of the airflow within an average geometry of the Healthy
Human Nasal Cavity. Sci Rep. 2020;10(1):3755. 38. Narang I, Carberry JC, Butler JE, Gandevia SC, Chiang AKI, Eckert DJ. Physi
ological responses and perceived comfort to high-flow nasal cannula therapy
in awake adults: effects of flow magnitude and temperature. J Appl Physiol. 2021;131(6):1772–82. 41. Parke RL, McGuinness SP. Pressures delivered by nasal high flow oxygen dur
ing all phases of the respiratory cycle. Respir Care. 2013;58(10):1621–4.
42. Cammarota G, Sguazzotti I, Zanoni M, Messina A, Colombo D, Vignazia GL, et
al. Diaphragmatic Ultrasound Assessment in subjects with Acute Hypercap
nic Respiratory Failure admitted to the Emergency Department. Respir Care.
2019;64(12):1469–77. References Comparison of
Postextubation Outcomes Associated with High-Flow Nasal Cannula vs. Conventional Oxygen Therapy in patients at high risk of Reintubation: a
Randomized Clinical Trial. J Korean Med Sci. 2020;35(25):e194. 32. Haaksma ME, Smit JM, Boussuges A, Demoule A, Dres M, Ferrari G, et al. EXpert consensus on Diaphragm UltraSonography in the critically ill (EXO
DUS): a Delphi consensus statement on the measurement of diaphragm
ultrasound-derived parameters in a critical care setting. Crit Care Lond Engl. 2022;26(1):99. 11. Longhini F, Pisani L, Lungu R, Comellini V, Bruni A, Garofalo E, et al. High-Flow
Oxygen Therapy after Noninvasive Ventilation interruption in patients recov
ering from Hypercapnic Acute Respiratory Failure: a physiological crossover
trial. Crit Care Med. 2019;47(6):e506–11. 33. Cammarota G, Rossi E, Vitali L, Simonte R, Sannipoli T, Anniciello F, et al. Effect
of awake prone position on diaphragmatic thickening fraction in patients
assisted by noninvasive ventilation for hypoxemic acute Respiratory Failure
related to novel coronavirus Disease. Crit Care Lond Engl. 2021;25(1):305. 12. Ferrer M, Sellarés J, Valencia M, Carrillo A, Gonzalez G, Badia JR, et al. Non-
invasive ventilation after extubation in hypercapnic patients with chronic
respiratory disorders: randomised controlled trial. Lancet Lond Engl. 2009;374(9695):1082–8. 34. Cammarota G, Simonte R, Longhini F, Spadaro S, Vetrugno L, De Robertis E. Advanced Point-of-care Bedside Monitoring for Acute Respiratory Failure. Anesthesiology. 2023;138(3):317–34. 13. Fernandez R, Subira C, Frutos-Vivar F, Rialp G, Laborda C, Masclans JR, et al. High-flow nasal cannula to prevent postextubation Respiratory Failure in
high-risk non-hypercapnic patients: a randomized multicenter trial. Ann
Intensive Care. 2017;7(1):47. 35. Tuinman PR, Jonkman AH, Dres M, Shi ZH, Goligher EC, Goffi A, et al. Respira
tory muscle ultrasonography: methodology, basic and advanced principles
and clinical applications in ICU and ED patients-a narrative review. Intensive
Care Med. 2020;46(4):594–605. 14. Ranieri VM, Tonetti T, Navalesi P, Nava S, Antonelli M, Pesenti A, et al. High-Flow nasal oxygen for severe hypoxemia: oxygenation response
and outcome in patients with COVID-19. Am J Respir Crit Care Med. 2022;205(4):431–9. 36. Santana PV, Cardenas LZ, de Albuquerque ALP, de Carvalho CRR, Caruso
P. Diaphragmatic ultrasound: a review of its methodological aspects and 15. Fernando SM, Tran A, Sadeghirad B, Burns KEA, Fan E, Brodie D, et al. Noninvasive respiratory support following extubation in critically ill adults: Boscolo et al. BMC Pulmonary Medicine (2024) 24:21 Boscolo et al. BMC Pulmonary Medicine Page 11 of 11 Page 11 of 11 (2024) 24:21 clinical uses. 42. Cammarota G, Sguazzotti I, Zanoni M, Messina A, Colombo D, Vignazia GL, et
al. Diaphragmatic Ultrasound Assessment in subjects with Acute Hypercap
nic Respiratory Failure admitted to the Emergency Department. Respir Care.
2019;64(12):1469–77. 41. Parke RL, McGuinness SP. Pressures delivered by nasal high flow oxygen dur
ing all phases of the respiratory cycle. Respir Care. 2013;58(10):1621–4. Publisher’s Note
S
i
N
i 39. Gattinoni L, Caironi P, Valenza F, Carlesso E. The role of CT-scan studies for the
diagnosis and therapy of acute respiratory distress syndrome. Clin Chest Med. 2006;27(4):559–70. Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations. 40. Garofalo E, Bruni A, Pelaia C, Cammarota G, Murabito P, Biamonte E, et al. Evaluation of a New Interface Combining High-Flow Nasal Cannula and CPAP. Respir Care. 2019;64(10):1231–9.
|
https://openalex.org/W4319312936
|
https://www.frontiersin.org/articles/10.3389/fpsyg.2023.1089386/pdf
|
English
| null |
Parenting style and Chinese preschool children’s pre-academic skills: A moderated mediation model of approaches to learning and family socioeconomic status
|
Frontiers in psychology
| 2,023
|
cc-by
| 8,490
|
© 2023 Xia. This is an open-access article
distributed under the terms of the Creative
Commons Attribution License (CC BY). The
use, distribution or reproduction in other
forums is permitted, provided the original
author(s) and the copyright owner(s) are
credited and that the original publication in this
journal is cited, in accordance with accepted
academic practice. No use, distribution or
reproduction is permitted which does not
comply with these terms. Parenting style and Chinese
preschool children’s pre-academic
skills: A moderated mediation
model of approaches to learning
and family socioeconomic status Xiaoying Xia 1,2* 1 School of Early Childhood Education, Shanghai Normal University Tianhua College, Shanghai, China,
2 University of the Pacific, Stockton, CA, United States This study examined the mediating role of children’s approaches to learning (ATL)
in parenting style and Chinese preschoolers’ pre-academic skills (i.e., literacy and
numeracy) as well as the moderating role of family socioeconomic status (SES)
in the mediating process. Participants were 307 children aged five to six years old
from four public kindergartens in Shanghai, China. Parents provided demographic
information and reported their parenting style (i.e., authoritative, authoritarian, and
permissive) and teachers rated each child’s ATL and pre-academic skills. Results
indicated that: (1) authoritative parenting positively related to children’s pre-academic
skills while no significant relationships were identified either for authoritarian or
permissive parenting with pre-academic skills; (2) children’s ATL partially mediated
the relationship between authoritative parenting and children’s pre-academic skills;
(3) family SES moderated the relationship between children’s ATL and pre-academic
skills. Specifically, children’s ATL was more strongly related to pre-academic skills
for children from low SES families as compared to their high-SES peers. These
findings contribute to the understanding of the effects of parenting styles on Chinese
children’s early academic achievement and underscore the importance of ATL to
children’s pre-academic skills, especially for low-SES children. CITATION
Xia X (2023) Parenting style and Chinese
preschool children’s pre-academic skills: A
moderated mediation model of approaches to
learning and family socioeconomic status. Front. Psychol. 14:1089386. doi: 10.3389/fpsyg.2023.1089386 © 2023 Xia. This is an open-access article
distributed under the terms of the Creative
Commons Attribution License (CC BY). The
use, distribution or reproduction in other
forums is permitted, provided the original
author(s) and the copyright owner(s) are
credited and that the original publication in this
journal is cited, in accordance with accepted
academic practice. No use, distribution or
reproduction is permitted which does not
comply with these terms. parenting style, approaches to learning, pre-academic skills, family SES, moderated
mediation TYPE Original Research
PUBLISHED 06 February 2023
DOI 10.3389/fpsyg.2023.1089386 TYPE Original Research
PUBLISHED 06 February 2023
DOI 10.3389/fpsyg.2023.1089386 TYPE Original Research
PUBLISHED 06 February 2023
DOI 10.3389/fpsyg.2023.1089386 parenting style, approaches to learning, pre-academic skills, family SES, moderated
mediation KEYWORDS 1. Introduction Transitioning from preschools to primary grades represents a critical life milestone for young
children as they are required to be academically and social–emotionally ready to function well in
formal schooling (Hair et al., 2006). Children’s pre-academic skills, including emergent literacy and
numeracy skills, are essential ingredients of school readiness (Jeon et al., 2020). Research suggests
that children’s early academic skills at school entry are consistent and strong predictors of later
academic achievement (Duncan et al., 2007; Davies et al., 2016). However, the development gap in
children’s academic achievement has appeared as early as kindergarten, and studies have shown that
the academic skills of children from low-income families are significantly lower than that of their
peers from high-income families (Zhang et al., 2019). In order to narrow the gap in children’s
academic readiness for school, it is necessary to explore the factors that affect the development of
children’s early academic skills.l Parenting style has reoccurred in the literature as influential for young children’s development,
nevertheless most studies have sampled Western populations and relatively little is known about the Frontiers in Psychology Frontiers in Psychology 01 frontiersin.org frontiersin.org 10.3389/fpsyg.2023.1089386 Xia associations between parenting style and preschool children’s
pre-academic skills in the non-Western contexts, like China. In addition,
despite the evidence for the link between parenting styles and children’s
academic achievement, relatively little empirical work has been
conducted to uncover the mechanism underlying this relationship. Some scholars have speculated that the pathway from parenting to
academic achievement can be accounted for by “academic enablers”
(DiPerna, 2006), including children’s motivation, engagement, and
learning behaviors. These attitudes and behaviors in learning are
conceptualized as approaches to learning (ATL; Li-Grining et al., 2010). ATL has been documented as a significant contributor to children’s
academic achievement in many studies (e.g., Schmerse, 2020), however,
empirical work examining the mediating role of ATL in the pathway
from family factors to children’s academic achievement is relatively
limited. Moreover, prior research suggests that children’s ATL varies
across diverse socioeconomic status (SES) families (Buek, 2019). However, the question remains unclear whether ATL functions similarly
or differently across children of different SES families. If low-SES
children reap more benefits from ATL, promoting children’s ATL would
be an effective way to mitigate the SES-based achievement gap. If
high-SES children reap more benefits from ATL, ATL would be viewed
as an important factor for the widening achievement gap. 1.1. Parenting style and children’s
pre-academic skills Parenting style is conceptualized as a constellation of attitudes and
behaviors that parents display in rearing their children (Darling and
Steinberg, 1993). In Baumrind’s (1967) seminal work, three types of
parenting style were proposed: authoritarian, authoritative, and
permissive. Authoritarian parents, characterized by high demandingness
and low responsiveness, have high expectations for children’s obedience
and respect for parental authority and show limited responsiveness to
children’s needs. Authoritative parents, characterized by high
responsiveness and high demandingness, tend to set rules for children
and rely on reasoning and negotiations with children to enforce rules. Permissive parents are high on responsiveness but low on
demandingness. They set few limits on their children’s behaviors and
show high responsiveness to children’s needs and wants. Parenting style constitutes an environmental factor for children’s
development and may play an indirect role via children’s personal
characteristics (Xie and Li, 2022). DiPerna (2006) proposed that
“academic enablers” such as motivation, academic engagement and
learning behaviors are important contributors to children’s academic
achievement. Children’s approaches to learning (ATL) refers to positive
learning attitudes and behaviors that children display in learning
activities, including persistence, sustained focus, initiative, planning,
and cooperative engagement in group learning (McCWayne et al., 2004). As ATL is closely involved in the process of learning, it has been
consistently shown as an important predictor of children’s learning
outcomes. For example, McCWayne et al. (2004) sampled US Head Start
children and found ATL significantly predicted preschool children’s
language and math skills. A longitudinal study by Vitiello and Greenfield
(2017) found that preschool children with higher levels of ATL
performed better in math and reading than those with lower ATL when
were in the first grade of primary school. Research sampling Chinese
preschool children has found similar conclusions. For example, Zhang
and Zhou (2018) found that children’s ATL such as self-control,
persistence, curiosity and interest significantly predicted preschool
children’s language development and math skills. Based on bioecological systems theory, parent–child interactions are
of critical importance to children’s acquisition of early academic skills
(Bronfenbrenner and Morris, 2006). A large body of empirical research
suggests that authoritative parenting positively and authoritarian and
permissive parenting negatively associated with children’s academic
achievement [see a review by Pinquart (2016)]. However, most of the
extant research has been conducted in the West and studies sampling
non-Western populations report inconsistent conclusions. 1. Introduction Therefore, this
study aimed to investigate whether parenting style was associated with
Chinese children’s early academic skills and to further explore whether
this relation was mediated by children’s ATL and whether family SES
moderated the mediating process. studies regarding parenting style have primarily focused on primary and
secondary school students (Xie et al., 2022), and research sampling
preschool children has mostly focused on the effect of parenting styles
on children’s emotional and social development (Lin et al., 2022). Furthermore, children of different age groups have different
developmental needs. There is some evidence that specific parenting
dimensions have differential effects on children’s academic performance
across developmental stages. For example, Zhu and Liu (2019) found
that authoritative parenting has a stronger positive effect on 6–9 years
old children’s reading score, while permissive parenting has a stronger
positive effect on 10–15 years old adolescents’ reading performance. f
Research on the effects of parenting style on Chinese preschool
children’s academic achievement is relatively limited and the few extant
studies tend to be less consistent. For example, Ren and Edwards (2017)
found a positive correlation between authoritative parenting and
preschool children’s early academic skills, while Li et al. (2016) failed to
find a significant correlation between the two variables. Another study
sampling Chinese American preschool children also reported an
insignificant link between parenting style (i.e., authoritative and
authoritarian) and academic achievement (Chen et al., 2015). They
believed that parenting style functioned more by providing emotional
family atmosphere, and have minimal effects on children’s academic
performance (Chen et al., 2015). In view of the limited research in the
Chinese context, the contribution of different parenting dimensions to
preschool
children’s
pre-academic
skills
requires
further
empirical research. frontiersin.org Frontiers in Psychology 1.1. Parenting style and children’s
pre-academic skills For example,
authoritarian parenting was identified to be positively related with
Chinese adolescents’ academic achievement (Leung et al., 1998) and
permissive parenting was reported as more beneficial for Spanish
adolescents’ school outcomes than authoritative parenting (Garcia and
Gracia, 2009). As parenting styles could be interpreted differently across
cultures, the effect of parenting style on children’s school outcomes
might be different accordingly. In the Chinese context, the existing ATL is a developmental characteristic and parenting style is
foundational for the formation of ATL in children as it provides an
emotional environment for parent–child daily interactions. Western
studies have found that supportive parenting positively predicts
preschool children’s learning interest and persistence, while corporal
punishment and authoritarian parenting can damage children’s learning
motivation, which in turn affects their persistence in tasks and learning Frontiers in Psychology 02 10.3389/fpsyg.2023.1089386 10.3389/fpsyg.2023.1089386 Xia strategies (Martin et al., 2013). Research on Chinese preschool children
also found that parenting styles affect the development of preschool
children’s ATL. Specifically, authoritative parenting style positively
predicted, while authoritarian parenting style negatively predicted
children’s ATL (Ye et al., 2021). Based on the literature above, ATL might
be an important mechanism by which parenting styles affect children’s
early academic performance. skills as well as the moderating role of family SES in the mediating
process. The hypothesized model is presented in Figure 1. To test the
model, the following hypotheses were proposed: (1) authoritative
parenting positively related to children’s pre-academic skills (Hypothesis
1); (2) authoritarian and permissive parenting negatively related to
children’s pre-academic skills (Hypothesis 2); (3) children’s approaches
to learning mediated the relationship between parenting style and
children’s pre-academic skills (Hypothesis 3); (4) family SES moderated
the relationship between children’s approaches to learning and
pre-academic skills (Hypothesis 4). 2.1. Participants In this study, a clustered random sampling method was used to elicit
participants from four public kindergartens in Shanghai, China,
including one model kindergarten, one first-class kindergarten and two
second-class kindergartens. In general, model kindergartens and first-
class kindergartens have more children from higher SES families while
second-class kindergartens include more children from lower SES
children. Chinese public kindergartens provide services for three to six
years old children and the present study focused on five to six years old
children in senior classes of kindergartens. A total of 341 questionnaires
were distributed among the target population and 324 respondents
agreed to participate in the study and returned the family demographic
and parenting questionnaires, resulting in a return rate of 95%. List-wise
deletion method was used to delete cases with missing values. Finally, a
sample of 307 parents and their children from 11 classrooms of the four
kindergartens were involved in this study. Twenty two teachers of the
child participants were invited to rate children’s school readiness based
on daily observations.hi The child participants in this study ranged from five to six years old
(Mean = 5.53, SD = 0.50) with five years olds accounting for 46.9%
(n = 144) and 6 years olds accounting for 53.1% (n = 163). The
proportions of boys and girls were balanced, among which 154 were
girls, accounting for 50.2% (n = 154). The majority children in this study
were singletons, accounting for 83.4% (n = 256). As regards parent
participants, the mean age is 35.5 years (SD = 3.6) with two thirds
respondents being mothers (65.1%, n = 200). Approximately 84%
(n = 258) of the population obtained a college degree or above, suggesting
that the majority participants were relatively well-educated. The vast
majority of families are intact families, accounting for 97.7% (n = 300). 2. Materials and methods Family income, parents’ educational levels and occupational prestige
are important indicators of family socioeconomic status (SES) and have
been consistently documented to relate to children’s academic
achievement (Conger and Donnellan, 2007). Children from low-SES
families tend to score lower on academic skills than their higher-SES
peers (Ip et al., 2016). The academic achievement gaps between high and
low-SES children have emerged as early as preschool years. Prior
research has suggested that ATL can work as a protective factor for
children in the context of adversity. However, Luthar et al. (2000)
claimed that children’s early ATL skills can serve as resilience for
children who experience risk factors and they argued that high-risk
children benefited more from ATL than low-risk children. Also, there is
some empirical evidence in the literature that ATL may function as a
source of resilience for children facing risks. For example, Razza et al. (2015) investigated children from low-income families in the
United States and claimed that children’s ATL is particularly beneficial
for improving the academic skills of at-risk children. In the study, they
primarily sampled low-income children and did not investigate the
relationship between ATL and children’s academic achievement across
diverse family SES. Shah et al. (2018) sampled 6,200 U.S. preschool
children and explored the interaction between family SES and children’s
curiosity in shaping children’s academic achievement. They found that
children’s curiosity positively predicted their reading and math
performance, and the association was stronger among children from
lower SES families. As children’s curiosity constitutes an important
element of ATL, it can be hypothesized that family SES would moderate
the relationship between ATL and pre-academic skills. In view of this,
the moderating role of family SES in the relationship between children’s
ATL and their pre-academic skills will be explored among Chinese
preschoolers in the present study. Frontiers in Psychology 2.2.3. Children’s approaches to learning According to the statistics reported by Shanghai Municipal Statistics
Bureau (2015), the average per capita disposable income for each
household was RMB 3,975 ($576) a month. In terms of family economic
status, 31.3% (n = 96) families earned a monthly income below $1,450;
53.3% (n = 163) families reported a monthly income ranging from
$1,450 to $4,350; 15.7% (n = 48) families had a monthly income higher
than $4,350. According to the statistics reported by Shanghai Municipal Statistics
Bureau (2015), the average per capita disposable income for each
household was RMB 3,975 ($576) a month. In terms of family economic
status, 31.3% (n = 96) families earned a monthly income below $1,450;
53.3% (n = 163) families reported a monthly income ranging from
$1,450 to $4,350; 15.7% (n = 48) families had a monthly income higher
than $4,350. Children’s learning behavior was measured with the approaches to
learning sub-domain of the social competence subscale from the Early
Development Instrument (EDI; Janus and Offord, 2007). The EDI
consists of 103 items and assesses children’s early development in
physical health, language and cognitive development, social competence,
emotional maturity, and communication skills and general knowledge. The Chinese version of the EDI has been validated in Chinese preschool
children and presented desirable psychometrical soundness (Ip et al.,
2013). The ATL subdomain of the EDI contains 9 items and measures
children’s learning behaviors such as working independently, following
directions and class routines, and adjusting to changes. Sample items
include “follows directions” and “works independently.” Questions in
the ATL measure were answered on a 3-point scale, ranging from often
or very true, sometimes or somewhat true, to never or not true. For data
analysis purposes, all responses on 3-point Likert-scale items were
coded 0, 5, and 10 (Janus and Offord, 2007). The average score in this
domain was calculated to represent children’s ATL. Higher scores
indicate stronger ATL for children. The Cronbach alpha coefficient was
0.95 for the measure of ATL in the current sample. 2.2.2. Parenting style Parenting style was measured with the Parenting Styles and
Dimensions Questionnaire (PSDQ; Robinson et al., 2001). The PSDQ is
comprised of 32 items and uses a 5-point Likert format (1 = never,
2 = once in a while, 3 = about half of the time, 4 = very often, 5 = always)
to assess how frequent parents display certain attitudes and behaviors in
rearing their children. There are three dimensions in the PSDQ:
authoritative parenting (15 items), authoritarian parenting (12 items),
and permissive parenting (5 items). The authoritative parenting
evaluates how frequent parents demonstrate warmth and acceptance,
reasoning and induction, and democratic participation in parent–child
interactions. The sample items include, “take child’s desire into account
before asking him/her to do something.” The authoritarian parenting
evaluates the frequency of parents’ use of physical coercion, verbal
hostility, and non-reasoning and punitive strategies. The sample items
include, “guide child by punishment more than by reason.” The
permissive parenting measures how frequent parents show
permissiveness in rearing children. The sample items include, “I state
punishments to my child and do not actually do them.” The mean of
each parenting dimension was calculated by averaging all the items in
that specific dimension. Higher scores indicated a higher frequency of
a specific type of parenting. In this sample, the Cronbach’s alphas for
authoritative, authoritarian, and permissive parenting were 0.87, 0.76,
and 0.62, respectively. 2.2.1. Family socioeconomic status In the present study, family monthly income, parents’ educational
levels, and occupational prestige were included as indicators for
family SES. The parent participants reported diverse levels of family
monthly income, ranging from $427 and below (coded as 1), $427–
853 (coded as 2), $853–1,422 (coded as 3), $1422–4,266 (coded as 4),
to $4266–7,110 (coded as 5), and to $7,110 and above (coded as 6). In addition, parents’ educational levels were assessed on a 5-point
scale, from junior middle school and below (coded as 1), secondary
school (coded as 2), three year college (coded as 3), four year college
(coded as 4) to graduate programs and above (coded as 5). Also,
parents’ occupation was evaluated on a 3-point scale and ranged
from farmers, migrant workers, service workers (coded as 1), foreign
enterprise workers, general enterprise workers, and freelancers
(coded as 2) to civil servants, professional and technical personnel,
enterprise management personnel, cultural, educational and
scientific personnel (coded as 3). A composite score of the family SES
was computed using principal component analysis based on the
following formula: Family SES = (β1*Z parental education+β2*Z
parental occupation+β3* Z family income)/εf, with β1-3 as the factor
loadings and εf as the eigenvalue for the first factor (OECD, 2012). In general, higher SES scores represent higher family SES in the
present study. 1.4. The present study FIGURE 1
The moderated mediation model. FIGURE 1
The moderated mediation model. As reviewed above, previous work has primarily investigated the
direct effect of parenting styles on the academic achievement among
middle school and primary school-age children in the Western context. However, the effects of parenting styles on preschool children’s
pre-academic skills received less attention. In addition, the mechanism
underlying the relationship between parenting styles and children’s
academic performance was not adequately addressed. ATL is
documented to closely relate to children’s pre-academic skills, but less
work is conducted to explore whether this relationship varies across
diverse SES families. In view of this, this study considers both family
variables and child variables to test a moderated mediation model by
examining the mediating role of ATL in parenting style (i.e.,
authoritative, authoritarian, and permissive) and children’s pre-academic Frontiers in Psychology 03 frontiersin.org 10.3389/fpsyg.2023.1089386 Xia 2.2.4. Children’s pre-academic skills Children’s pre-academic skills were measured using items from the
language and cognitive development subscale of the EDI (Janus and
Offord, 2007). The language and cognitive development subscale
consists of 26 items and assesses children’s basic and advanced literacy
skills, basic numeracy skills and interest in literacy/numeracy and
memory. It is noted that the subdomain of interest in literacy/numeracy
and memory was not included for further analysis as it was not
applicable for the measure of pre-academic skills. Finally, 21 items in the
language and cognitive development subscale were used to measure
children’s pre-academic skills. Basic literacy assesses children’s basic
abilities to recognize written words and to participate in literacy-
oriented activities. Sample items include “knows how to handle a book”
and “is able to write his or her name in Chinese.” Advanced literacy
assesses children’s emergent reading and writing. Sample items include
“is able to read simple words” and “is interested in writing voluntarily.”
Basic numeracy assesses children’s number recognition and counting,
and basic mathematical concepts. Sample items include “sort and classify
objects by a common characteristic” and “recognize geometrical shapes.”
All answers to the questions in these measures were scored on a 2-point
scale: “yes” if a child possesses a skill and “no” if he or she does not
(Janus and Offord, 2007). For data analysis purposes, all responses on
binary items were coded 0 or 10.The mean of all items in the basic
literacy, advanced literacy and basic numeracy subdomains was
calculated to represent children’s pre-academic skills. The Cronbach
alpha coefficient was 0.83 for the measure of pre-academic skills. Frontiers in Psychology frontiersin.org 2.3. Data collection After obtaining approval from the IRB in the university, the
investigator contacted the principals of the potential kindergartens and
finally obtained permission from four kindergartens in Shanghai, China. Parent participants were elicited by distributing letters when they came
to school to pick up their children. If they were willing to participate,
they signed an informed consent form and give permission for the
investigator to access their children’s data. One parent of the child was
asked to fill out a survey questionnaire containing items about their 04 Frontiers in Psychology 10.3389/fpsyg.2023.1089386 10.3389/fpsyg.2023.1089386 Xia family background information and parenting style. The parent surveys
were collected by the investigator the next day. Subsequent to collecting
parent data, teachers, who were familiar with the child participants, were
invited to rate children’s approaches to learning and pre-academic skills
based on their daily observations. To reduce measurement errors, all the
teacher participants received two rounds of measurement training based
on the training materials provided by Offord Centre for Child Studies
(i.e., the developer of the EDI). Each participating teacher was given a
teachers’ guide on the use of the EDI assessment forms in which the
assessment criteria for all the items were clarified and children’s
exemplary performance and behaviors were also presented in the
training sessions. Teachers rated children’s ATL and pre-academic skills
based on their daily observations and personal interactions
with children. (r = −0.12, p < 0.05) and permissive parenting negatively associated with
children’s ATL (r = −0.12, p < 0.05). Children’s ATL was highly linked
with their pre-academic skills (r = 0.51, p < 0.001). In addition, family
SES positively related to authoritative parenting (r = 0.27, p < 0.001), but
negatively related to authoritarian (r = −0.15, p < 0.01) and permissive
parenting (r = −0.13, p < 0.05). These results suggest that parents from
higher SES families reported higher levels of authoritative parenting but
lower levels of authoritarian and permissive parenting. In addition,
family SES positively related to children’s early academic skills (r = 0.24,
p < 0.01) and no significant relationship was identified for children’s ATL
with family SES. A series of t-tests were conducted to determine whether parenting
style, children’s approaches to learning and pre-academic skills vary by
children’s gender, age, and only-child status as these variables were
evidenced to be related to either parenting style or children’s
development outcomes. 2.4. Data analysis At first, means, standard deviations and bivariate correlations of the
main variables were calculated using SPSS 26.0. Second, a sequential
multiple regression was conducted to determine the relationship
between parenting style on children’s pre-academic skills while
controlling for children’s age, gender, with siblings or not, and family
SES. The β value (standardized regression coefficient) was referred to
determine the effect of each type of parenting style on pre-academic
skills. A statistically significant β value would indicate the potential
influence of specific type of parenting. Third, Model 4 and Model 14 in
PROCESS Macro (Hayes, 2017; Version 3.3) were employed to test the
mediating role of ATL and the moderated mediating role of family SES
in parenting styles and children’s pre-academic skills, respectively. The
effects are assumed as statistically significant when the lower limit (Boot
LLCI) and upper limit (Boot ULCI) of bootstrap results in 95%
confidence interval are either below or above zero (Hayes, 2017). 2.3. Data collection Statistics showed that there were significant
gender differences in children’s ATL. Girls scored significantly higher
than boys (t = −3.40, p < 0.01) in ATL. Also, age difference in children’s
pre-academic skills was identified and 6-year-old children were
significantly higher than that of 5-year-old children (t = −2.12, p < 0.05). There are significant differences in the authoritative and permissive
parenting between only child and children with siblings. Parents of only
child reported higher authoritative parenting then those with more than
one children (t = 3.15, p < 0.01), whereas parents with more than one
children reported higher levels of permissive parenting than those with
one children (t = −2.02, p < 0.05). In view of the effects of children’s
gender, age, only child status on the main research variables, they were
used as control variables in the follow-up analysis. 3.2. Parenting style and children’s
pre-academic skills To test the relationship between three types of parenting styles and
children’s pre-academic skills, a sequential multiple regression with
children’s pre-academic skills as the dependent variable, and the three
types of parenting styles as independent variables was conducted after
controlling for child and family demographic variables. Multiple
regression analysis (see Table 2) showed that demographic factors such
as children’s gender, age, and whether they were only child or not and
family SES altogether, explained 8.6% of the variation in children’s
pre-academic skills (ΔR2 = 0.086, F [4,302] =7.07, p < 0.001). Family SES
positively related to children’s pre-academic skills (β = 0.25, p < 0.01), and
children from high-SES families scored higher in pre-academic skills 3.4. The moderated mediating model TABLE 3 Total effect, direct effect, and indirect effect. TABLE 3 Total effect, direct effect, and indirect effect. Effects
Boot (SE)
Boot LLCI
Boot ULCI
Total effect
0.24
0.06
0.13
0.35
Direct effect
0.13
0.05
0.03
0.23
Indirect effect
0.11
0.04
0.05
0.19
n = 307. In order to examine the moderating effect of family SES in the
mediation model, Model 14 in PROCESS Macro (Hayes, 2017; Version
3.3) was used to test the hypothesized moderated mediation model with
children’s gender, age, and only child status as control variables,
pre-academic skills as the outcome variable, authoritative parenting as
the predictor, ATL as the mediating variable, and family SES as the
moderating variable. The interaction between family SES and children’s
ATL significantly contributed to children’s early academic skills
(β = −0.05, t = −3.26, p < 0.01). The moderated mediation index was
−0.01, and the 95% Bootstrap confidence interval was [−0.019, −0.003],
excluding 0. These results suggest that the moderated mediation model
was validated, supporting hypothesis 4 proposed in this study. Family
SES moderated the effect of ATL on children’s pre-academic skills, and
then moderated the mediating effect of ATL on authoritative parenting
and children’s pre-academic skills. than children from low-SES families. Three types of parenting styles
explained a 3.6% variation in children’s pre-academic skills (ΔR2 = 0.036,
F [3,299] = 4.13, p < 0.01). Moreover, among the three types of parenting
dimensions, authoritative parenting significantly and positively
associated with children’s early academic skills (β = 0.19, p < 0.01), while
authoritarian parenting and permissive parenting did not significantly
relate to children’s pre-academic skills with control variables included. Thus, hypothesis 1 was supported, and hypothesis 2 was rejected in the
present study. To further probe the interactions between family SES and children’s
ATL, conditional regression lines of children’s pre-academic skills on
ATL were plotted for children at one standard deviation below, the
mean, and one standard deviation above for family SES (Aiken and West,
1991), labeled as low, medium, and high SES in Figure 2. Regardless of
children’ family SES levels, children’s ATL all positively related to their
pre-academic skills with a stronger relationship for children from low
SES families. Among the three levels of family SES, ATL exerted the
largest mediating effect for low SES families [β = 0.13, Boot CI 95% (0.05,
0.22)]. The mediating effect of children’s approaches to learning is 3.1. Descriptive analysis Zero-order correlations, means, and standard deviations of the main
variables were presented in Table 1. Correlational analyzes showed that
authoritative parenting positively related to children’s ATL (r = 0.20,
p < 0.01), and pre-academic skills (r = 0.23, p < 0.001). Authoritarian
parenting was negatively related to children’s pre-academic skills TABLE 1 Means, standard deviations, and zero-order correlations of main variables. Variables
2
3
4
5
6
1. Family SES
0.27***
−0.15**
−0.13*
0.04
0.24***
2. Authoritative parenting
–
−0.33***
−0.32***
0.20**
0.23***
3. Authoritarian parenting
–
0.44***
−0.10
−0.12*
4. Permissive parenting
–
−0.12*
−0.07
5. Approaches to learning
–
0.51***
6. Pre-academic skills
–
M
3.98
2.00
2.36
8.32
8.20
SD
0.52
0.42
0.65
2.17
1.64
n = 307, *p < 0.05, **p < 0.01, ***p < 0.001. TABLE 1 Means, standard deviations, and zero-order correlations of main variables. 1 Means, standard deviations, and zero-order correlations of main variables. 05 05 Frontiers in Psychology 10.3389/fpsyg.2023.1089386 Xia B
SE B
β
ΔR2
F
t
p
Step 1
0.086***
7.07
0.000
Child gender
0.35
0.18
0.11
1.94
0.053
Child age
0.45
0.18
0.14*
2.47
0.014
Only child status
0.18
0.26
0.04
0.68
0.498
Family SES
0.14
0.03
0.25***
4.24
0.000
Step 2
0.036**
4.13
0.003
Authoritative
0.61
0.19
0.19**
3.66
0.000
Authoritarian
−0.14
0.25
−0.04
0.57
0.572
Permissive
0.06
0.16
0.02
0.37
0.710
n = 307, ***p < 0.001; **p < 0.01, *p < 0.05. 4. Discussion
4.1. Parenting style and children’s
pre-academic skills This study found that authoritative parenting positively related to
children’s pre-academic skills after controlling for child and family
demographics. This finding is in line with previous literature that
authoritative parenting positively contributed to preschool children’s
academic skills and language development (Bingham et al., 2017). Also,
a meta-analysis, primarily focused on parenting style of Chinese
adolescents, reported that positive parenting behaviors such as
emotional warmth, autonomy and support positively linked with
adolescents’ academic achievement (Xie et al., 2022). According to self-
determination theory (SDT; Deci and Ryan, 1985), children’s intrinsic
motivation for learning stems from the satisfaction of their psychological
needs for autonomy, competence and belonging. In the daily
interactions, authoritative parents create affective and secure family
environment by showing support, respect, understanding and caring for
children, which in turn may motivate children to participate more
actively in learning activities and acquire better pre-academic skills. Results from the present study showed that children’s ATL partially
mediated the relationship between authoritative parenting and children’s
pre-academic skills. Authoritative parenting can not only directly related
to children’s early academic skills, but also indirectly related to
pre-academic skills via ATL. This result supports Hill and Palacios’
(2020) study with Early Childhood Longitudinal Study-Kindergarten
Cohort (ECLS-K) 2011 longitudinal data, which found that parents’
warmth and support helps promote children’s ATL, which in turn
facilitates children’s acquisition of reading skills. Authoritative parenting
characterized by warmth and support, reasoning guidance, and
democratic participation helps to create a positive and democratic
family atmosphere, which is conducive to children’s development of
ATL. And enhanced ATL directly affects children’s learning process and
promotes the acquisition of academic skills. Although correlation analysis indicated that authoritarian parenting
negatively correlated with children’s pre-academic skills, its effect on
pre-academic skills was not identified after controlling for demographic
variables and other types of parenting style (i.e., authoritative and
permissive parenting). This result is in contrast to many studies conducted
in the West. For example, Tamis-LeMonda et al. (2004) reported that
authoritarian behaviors such as interference and harsh punishment could
damage parent–child relationships, thereby negatively affecting children’s
development of pre-academic skills. Previous research sampling Chinese
populations has mostly focused on adolescents and reported a negative
link between authoritarian parenting and academic achievement (Chen
et al., 2015; Xie et al., 2022). 3.3. Approaches to learning as a mediator Model 4 in PROCESS Macro (Hayes, 2017; Version 3.3) was
used to test the mediating role of children’s ATL in the relationship
between authoritative parenting and pre-academic skills. Authoritative
parenting
positively
predicted
children’s
pre-academic skills (β = 0.24, t = 4.29, p < 0.001) and children’s ATL
(β = 0.23, t = 4.04, p < 0.001). Also, children’s ATL positively related
to their early academic skills (β = 0.47, t = 9.20, p < 0.001). After
adding the mediating variable of ATL to the model, the predictive
effect of authoritative parenting on children’s academic skills is still
significant (β = 0.13, t = 2.64, p < 0.01), which indicates that
authoritative parenting can not only directly predict children’s early
academic skills, but also indirectly predict children’s early academic
skills through ATL. Children’s ATL plays a partial mediating role
in authoritative parenting and children’s early academic skills,
supporting research hypothesis 3. The direct effect, indirect effect
and total effect of authoritative parenting on children’s early
academic skills were shown in Table 3. The total effect was 0.24 and
the indirect effect was 0.11. Altogether, the indirect effect
accounted for 45.83% of the total effect. This result suggests that
ATL mediated 45.83% of the effect of authoritative parenting on
children’s pre-academic skills. 0
1
2
3
4
5
6
7
8
9
10
Low Approaches to Learning High Approaches to Learning
Children's Pre-academic Skills
Low SES
Medium SES
High SES
FIGURE 2
The moderating role of family SES on approaches to learning and
pre-academic skills. Frontiers in Psychology 06 frontiersin.org 10.3389/fpsyg.2023.1089386 10.3389/fpsyg.2023.1089386 Xia relatively weaker for medium SES families [β = 0.10, Boot CI 95% (0.04,
0.17)] and the weakest for high-SES families [β = 0.07, Boot CI 95%
(0.03, 0.13)]. support, tend to allow children to explore their environment, encourage
young children to solve problems independently and involve children in
the process of decision making. All of these behaviors are conducive to the
development of ATL (Jeon and Neppl, 2019). In contrast, restrictive and
harsh parenting behaviors will inhibit children’s exploratory behavior and
hinder the development of positive learning behaviors (Larzelere et al.,
2013). Moreover, in this study ATL was found to positively relate to
pre-academic skills. This result is consistent with prior literature that
preschool children’s ATL significantly predicted their early language and
math skills (Zhang and Zhou, 2018). 4. Discussion
4.1. Parenting style and children’s
pre-academic skills Nonetheless, a recent study focusing on
Chinese preschoolers revealed that the sub-dimensions of authoritarian
parenting (i.e., verbal hostility and non-reasoning/punitive strategies) had
differential effects on children’s school outcomes (Li and Xie, 2017). Specifically, they found negative effects of parents’ non-reasoning/
punitive strategies but positive effects of parents’ verbal hostility on
children’s language, literacy, numeracy and science skills. As a global
parenting construct of authoritarian parenting was adopted in the present
study, the conflicting effects among the various sub-dimensions of
authoritarian parenting might hinder the possibility to identify significant
effects. In addition, permissive parenting was not found to relate to
children’s pre-academic skills. It should be noted that the low internal
consistency reliability of permissive parenting in this study may limit its
ability to exert significant effects on children’s pre-academic skills. 4.3. The moderating role of family
socioeconomic status The relationship between ATL and children’s early academic skills
was moderated by family SES in this study. Although in different SES
families, children with higher levels of ATL display better academic
skills, the relationship was stronger for children from lower SES
families. It can be seen that ATL is an important protective factor for
low-SES children’s the development of academic skills. This result
supports a prior study by Shah et al. (2018), who found that family SES
moderated the relationship between curiosity and preschool children’s
reading and math skills, and this relationship was stronger for low-SES
children. Also, Razza et al. (2010) claimed that ATL acted as a protective
factor for economically disadvantaged children who are at risk in
school readiness measures. The greater effects of ATL on children’s
pre-academic skills might be because children from low SES families
have less access to home educational resources and the drive for
academic achievement is closely related to children’s ATL (Deci and
Ryan, 1985). In the context of scarce educational resources, the way how
children interact with the environment and peers and the initiative and
activeness they display in learning activities play a more important role
in children’s acquisition of academic skills. High levels of approaches to
learning can help children cope with the disadvantages associated with
low SES family environment. 3.3. Approaches to learning as a mediator McDermott et al.’s (2014) longitudinal
study of preschool children also showed that children with lower levels of
motivation and focused attention in preschools performed worse in
different academic subjects when they were in the first grade of primary
schools. It is possible that children with higher levels of ATL can maintain
focus, follow instructions, and inhibit certain behaviors in learning settings
so that they can make better use of learning opportunities and develop
stronger academic skills (Schmerse, 2020). In contrast, children with poor
ATL are easily distracted and lack perseverance in tasks, which will hinder
the development of children’s early academic skills (Beisly et al., 2020). Frontiers in Psychology frontiersin.org 4.2. The mediating role of approaches to
learning As expected, this study found that authoritative parenting positively
related to ATL, which in turn positively associated with children’s
pre-academic skills. Authoritative parents, characterized by autonomy and Frontiers in Psychology 07 frontiersin.org 10.3389/fpsyg.2023.1089386 10.3389/fpsyg.2023.1089386 Xia Bingham, G., Jeon, H., Kwon, K., and Lim, C. (2017). Parenting styles and home literacy
opportunities: associations with children’s oral language skills. Infant Child Dev. 26:e2020.
doi: 10.1002/icd.2020 Ethics statement The studies involving human participants were reviewed and
approved by University of the Pacific. Written informed consent to
participate in this study was provided by the participants’ legal guardian/
next of kin. Author contributions XX collected and analyzed the data, and wrote the
manuscript independently. 5. Implications based on different levels of family SES. As the participants were all from
one economically developed city of China (i.e., Shanghai), the
conclusions cannot be generalized to other Chinese regions. Future
research can increase sample size to include children from both
economically developed and backward regions in China to replicate this
study. Last but not the least, this study does not distinguish between
mothers and fathers, and whether maternal and paternal parenting
styles have different effects on children’s ATL and early academic skills
also deserves further investigation. This study contributes to the current literature by examining the
mechanism underlying the relationship between parenting style and
Chinese children’s pre-academic skills. The results of the study have
several important implications. At first, we found that authoritative
parenting made a positive contribution to children’s early academic skills. Parents are encouraged to adopt authoritative parenting and tend to
be responsive to children’s feelings or needs, respect children’s opinions,
and give explanations for obeying certain rules and consider children’s
wishes in family affairs. The respect, understanding, and care provided
by parents helps create benevolent family atmosphere and facilitate
children’s development of early academic skills. Second, authoritative
parenting positively related to children’s pre-academic skills via ATL and
this mediating effect was stronger for low-SES children. As ATL positively
related to children’s pre-academic skills, parents should attach importance
to ATL and try to cultivate desirable learning behaviors when they are
involved in children’s education. At present, many parents put more
emphasis on children’s academic achievement than on children’s
ATL. The preschool years constitute a critical period for children’s
development of ATL. Thus, parents should adopt positive and democratic
parenting behaviors to enhance children’s attention, persistence,
independence and other good learning behaviors, thereby promoting the
acquisition of early academic skills. Kindergartens can also improve
teachers’ knowledge and understanding of ATL and conduct workshops
and specialized trainings for teachers on how to stimulate children’s
curiosity in activities, develop children’s persistence in tasks, and guide
children to actively explore and solve problems independently. Third, for
economically underprivileged children, parents need to enhance
communications with their children and try to adopt authoritative
parenting, which is beneficial for the development of ATL and
pre-academic skills as well. Data availability statement The original contributions presented in the study are included in the
article/supplementary material, further inquiries can be directed to the
corresponding author. 6. Limitations and suggestions for
further studies The author declares that the research was conducted in the absence
of any commercial or financial relationships that could be construed as
a potential conflict of interest. Despite its contributions to the literature, several limitations need
to be acknowledged in this study. At first, as parenting style was assessed
only through parental self-reports, parental bias might be created as a
result of social desirability. More accurate information can be obtained
by direct observations of parent–child interactions in the family context. Second, this study used a cross-sectional design to measure ATL and
children’s pre-academic skills. In the future, longitudinal designs could
be used to further validate the mediating role of ATL in family variables
and children’s academic outcomes. Third, in order to include diverse SES
populations, the study elicited parents from four public kindergartens Bronfenbrenner, U., and Morris, P. A. (2006). “The bioecological model of human
development,” in Handbook of Child Psychology. Vol. 1: Theoretical Models of Human
Development. eds. W. R. Damon and M. Lerner. 6th ed (Hoboken, NJ: John Wiley & Sons,
Inc), 793–828. Publisher’s note All claims expressed in this article are solely those of the authors and
do not necessarily represent those of their affiliated organizations, or
those of the publisher, the editors and the reviewers. Any product that
may be evaluated in this article, or claim that may be made by its
manufacturer, is not guaranteed or endorsed by the publisher. Beisly, A. H., Kwon, K. A., and Jeon, S. (2020). Executive function and learning
behaviours: associations with academic skills among pre-schoolers. Early Child Dev. Care
190, 2469–2483. doi: 10.1080/03004430.2019.1585347 Baumrind, D. (1967). Child care practices anteceding three patterns of preschool
behaviour. Genet. Psychol. Monogr. 75, 43–88. References Tracing children’s
approaches to learning through head start, kindergarten, and first grade: different pathways
to different outcomes. J. Educ. Psychol. 106, 200–213. doi: 10.1037/a0033547 Deci, E. L., and Ryan, R. M. (1985). Intrinsic Motivation and Self-Determination in
Human Behaviour. New York: Plenum. OECD (2012). PISA 2009 Technical Report. Paris: OECD. DiPerna, J. C. (2006). Academic enablers and student achievement: Implications for
assessment and intervention services in the schools. Psychol. Sch. 43, 7–17. doi: 10. 1002/
pits.20125 Pinquart, M. (2016). Associations of parenting styles and dimensions with academic
achievement in children and adolescents: a meta-analysis. Educ. Psychol. Rev. 28, 475–493. doi: 10.1007/s10648-015-9338-y Duncan, G. J., Dowsett, C. J., Claessens, A., Magnuson, K., Huston, A. C., Klebanov, P.,
et al. (2007). School readiness and later achievement. Dev. Psychol. 43, 1428–1446. doi:
10.1037/0012-1649.43.6.1428 Razza, R. A., Martin, A., and Brooks-Gunn, J. (2010). Associations among family
environment, sustained attention, and school readiness for low-income children. Dev. Psychol. 46, 1528–1542. doi: 10.1037/a0020389 Garcia, F., and Gracia, E. (2009). Is always authoritative the optimum parenting style?. Evidence from Spanish families. Adolescence 44, 101–131. Razza, R., Martin, A., and Brooks-Gunn, J. (2015). Are approaches to learning in
kindergarten associated with academic and social competence similarly? Child Youth Care
Forum 44, 757–776. doi: 10.1007/s10566-015-9307-0 Hair, E., Halle, T., Terry-Humen, E., Lavelle, B., and Calkins, J. (2006). Children's school
readiness in the ECLS-K: predictions to academic, health, and social outcomes in first
grade. Early Child. Res. Q. 21, 431–454. doi: 10.1016/j.ecresq.2006.09.005 Ren, L., and Edwards, C. P. (2017). Chinese parents’ expectations and child pre-academic
skills: the indirect role of parenting and social competence. Early Educ. Dev. 28, 1052–1071. doi: 10.1080/10409289.2017.1319784 Hayes, A. F. (2017). Introduction to Mediation, Moderation, and Conditional Process
Analysis: A Regression-Based Approach. New York: Guilford Publications. Robinson, C. C., Mandelco, B., Olsen, S. F., and Hart, C. H. (2001). “The parenting styles
and dimensions questionnaire (PSDQ)” in Handbook of Family Measurement Techniques. Vol. 3: Instruments and Index. eds. B. F. Perlmutter, J. Touliatos and G. W. Holden
(Thousand Oaks, CA: Sage), 319–321. Hill, T., and Palacios, N. (2020). The influence of parental warmth and stress on reading
through approaches to learning: racial/ethnic variation. Infant Child Dev. 30:e2210. doi:
10.1002/icd.2210 Ip, P., Li, S., Rao, N., Ng, S., Lau, W., and Chow, C. B. (2013). Validation study of the
Chinese early development instrument (CEDI). BMC Pediatr. 13:146. doi:
10.1186/1471-2431-13-146 Schmerse, D. (2020). References Bingham, G., Jeon, H., Kwon, K., and Lim, C. (2017). Parenting styles and home literacy
opportunities: associations with children’s oral language skills. Infant Child Dev. 26:e2020. doi: 10.1002/icd.2020 Aiken, L. S.., and West, S. G. (1991). Multiple regression: Testing and interpreting
interactions. Sage Publications, Inc. Baumrind, D. (1967). Child care practices anteceding three patterns of preschool
behaviour. Genet. Psychol. Monogr. 75, 43–88. Baumrind, D. (1967). Child care practices anteceding three patterns of preschool
behaviour. Genet. Psychol. Monogr. 75, 43–88. Beisly, A. H., Kwon, K. A., and Jeon, S. (2020). Executive function and learning
behaviours: associations with academic skills among pre-schoolers. Early Child Dev. Care
190, 2469–2483. doi: 10.1080/03004430.2019.1585347 08 Frontiers in Psychology frontiersin.org Xia 10.3389/fpsyg.2023.1089386 Buek, K. (2019). Early growth trajectories of children’s ATL: the contribution
of parent and family characteristics. Psychol. Sch. 56, 1053–1072. doi: 10.1002/
pits.22224 Lin, X., Liao, Y., and Li, H. (2022). Parenting styles and social competence in Chinese
pre-schoolers: a moderated mediation model of singleton and self-regulation. Early Educ. Dev. 33, 437–451. doi: 10.1080/10409289.2021.1940643 Luthar, S. S., Cicchetti, D., and Becker, B. (2000). The construct of resilience: a critical
evaluation and guidelines for future work. Child Dev. 71, 543–562. doi:
10.1111/1467-8624.00164 Chen, S. H., Main, A., Zhou, Q., Bunge, S. A., Lau, N., and Chu, K. (2015). Effortful
control and early academic achievement of Chinese American children in immigrant
families. Early Child. Res. Q. 30, 45–56. doi: 10.1016/j.ecresq.2014.08.004 Martin, A., Ryan, R. M., and Brooks-Gunn, J. (2013). Longitudinal associations among
interest, persistence, supportive parenting, and achievement in early childhood. Early
Child. Res. Q. 28, 658–667. doi: 10.1016/j.ecresq.2013.05.003 Conger, R. D., and Donnellan, M. B. (2007). An interactionist perspective on the
socioeconomic context of human development. Annu. Rev. Psychol. 58, 175–199. doi:
10.1146/annurev.psych.58.110405.085551 McCWayne, C. M., Fantuzzo, J. W., and McDermott, P. A. (2004). Preschool
competency in context: an investigation of the unique contribution of child
competencies to early academic success. Dev. Psychol. 40, 633–645. doi:
10.1037/0012-1649.40.4.633 Darling, N., and Steinberg, L. (1993). Parenting style as context: an integrative model. Psychol. Bull. 113, 487–496. doi: 10.1037/0033-2909.113.3.487 Davies, S., Janus, M., Duku, E., and Gaskin, A. (2016). Using the early development
instrument to examine cognitive and non-cognitive school readiness and elementary
student achievement. Early Child. Res. Q. 35, 63–75. doi: 10.1016/j.ecresq.2015.10.002 McDermott, P. A., Rikoon, S. H., and Fantuzzo, J. W. (2014). References Preschool quality effects on learning behaviour and later
achievement in Germany: moderation by socioeconomic status. Child Dev. 91, 2237–2254. doi: 10.1111/cdev.13357 Ip, P., Rao, N., Bacon-Shone, J., Li, S. L., Ho, F. K. W., Chow, C. B., et al. (2016). Socioeconomic gradients in school readiness of Chinese preschool children: the mediating
role of family processes and kindergarten quality. Early Child. Res. Q. 35, 111–123. doi:
10.1016/j.ecresq.2015.10.005 Shah, P. E., Weeks, H. M., Richards, B., and Kaciroti, N. (2018). Early childhood curiosity
and kindergarten reading and math academic achievement. Pediatr. Res. 84, 380–386. doi:
10.1038/s41390-018-0039-3 Shanghai Municipal Statistics Bureau (2015). Shanghai Statistical Yearbook 2015,
Shanghai. China: China Statistics Press. Janus, M., and Offord, D. R. (2007). Development and psychometric properties of the
early development instrument (EDI): a measure of children’s school readiness. Can. J. Behav. Sci. 39, 1–22. doi: 10.1037/cjbs2007001 Tamis-Lemonda, C., Shannon, J., Cabrera, N., and Lamb, M. (2004). Fathers and mothers
play with their 2-and 3-year-olds: contributions to language and cognitive development. Child Dev. 75, 1806–1820. doi: 10.1111/j.1467-8624.2004.00818.x Jeon, S., and Neppl, T. K. (2019). Economic pressure, parent positivity, positive parenting, and
child social competence. J. Child Fam. Stud. 28, 1402–1412. doi: 10.1007/s10826-019-01372-1 Vitiello, V. E., and Greenfield, D. B. (2017). Executive functions and approaches to
learning in predicting school readiness. J. Appl. Dev. Psychol. 53, 1–9. doi: 10.1016/j. appdev.2017.08.004 Jeon, H. J., Peterson, C. A., Luze, G., Carta, J. J., and Langill, C. C. (2020). Associations
between parental involvement and school readiness for children enrolled in head start and
other early education programs. Child Youth Serv. Rev. 118:105353. doi: 10.1016/j. childyouth.2020.105353 Xie, S., and Li, H. (2022). Self-regulation mediates the relations between family factors
and preschool readiness. Early Child. Res. Q. 59, 32–43. doi: 10.1016/j.ecresq.2021.10.005 Larzelere, R. E., Morris, A. S., and Harrist, A. W. (Eds.) (2013). Authoritative Parenting:
Synthesizing Nurturance and Discipline for Optimal Child Development. Washington, D. C.:
American Psychological Association. Xie, Y., Shi, Z., Yin, L., and Lan, L. (2022). A meta-analysis of the relationships between
Chinese parenting styles and Children’s academic achievement (Chinese). Psychol. Dev. Educ. 38, 366–379. Leung, K., Lau, S., and Lam, W. L. (1998). Parenting styles and academic achievement:
a cross-cultural study. Merrill-Palmer Q. 44, 157–172. Ye, P., Wang, C., and Lin, C. (2021). The relationship between family socioeconomic
status and children’s learning quality: the mediating role of parental rearing style (Chinese). Stud. Presch. Educ. 7, 17–28. References Li, X., and Xie, J. (2017). Parenting styles of Chinese families and children’s social-
emotional and cognitive developmental outcomes. Eur. Early Child. Educ. Res. J. 25,
637–650. doi: 10.1080/1350293X.2017.1331077 Zhang, X., Hu, B. Y., Ren, L., and Zhang, L. (2019). Family socioeconomic status and
Chinese children’s early academic development: examining child-level mechanisms. Contemp. Educ. Psychol. 59:101792. doi: 10.1016/j.cedpsych.2019.101792 Li, Y., Xu, L., Liu, L., Lv, Y., Wang, Y., and Huntsinger, C. S. (2016). Can preschool
socioeconomic composition moderate relationships between family environment and
Chinese children’s early academic and social outcomes? Child Youth Serv. Rev. 60, 1–10. doi: 10.1016/j.childyouth.2015.11.009 Zhang, L., and Zhou, J. (2018). The development of approaches to learning and the
prediction for early achievement in language and mathematics (Chinese). Global Educ. 47,
113–128. Li-Grining, C. P., Votruba-Drzal, E., Maldonado-Carreño, C., and Haas, K. (2010). Children’s early approaches to learning and academic trajectories through fifth grade. Dev. Psychol. 46, 1062–1077. doi: 10.1037/a0020066 Zhu, M., and Liu, J. (2019). Effects of parenting style on children’s academic ability
(Chinese). Sociol. Stud. Child. 6, 43–62. 09 frontiersin.org Frontiers in Psychology
|
https://openalex.org/W3198206636
|
https://www.nature.com/articles/s41388-021-02001-2.pdf
|
English
| null |
The role of reciprocal fusions in MLL-r acute leukemia: studying the chromosomal translocation t(4;11)
|
Oncogene
| 2,021
|
cc-by
| 11,009
|
INTRODUCTION failed with the full-human counterpart to convincingly create
leukemia with the MLL-AF4 fusion gene alone. Our study from
2010 demonstrated that the AF4-MLL fusion protein is indis-
pensable for leukemia onset, as the onset of leukemia was
observed only with AF4-MLL or both fusion genes, but never with
MLL-AF4 alone [5]. Expression of AF4-MLL alone caused B-/T-type
leukemia, but only in one-third of the transplanted mice, which
may indicate that other transcription factors (e.g., RUNX1) were
somehow complementing functions deriving from the missing
MLL-AF4 allele to drive leukemia [6]. MLL-r leukemia is diagnosed in 5–10% of all acute leukemia
patients, and the spectrum of MLL fusion partners has increased
over the last 30 years of research to more than 100 [1]. The most
frequent translocation in proB ALL is t(4;11)(q21;q23) which
represents overall about 57% of all cases. In this particular
translocation, the two genes MLL (KMT2A) and AF4 (AFF1) are fused
in a balanced recombination event to cause the generation of the
two fusion genes MLL-AF4 and AF4-MLL, respectively. Most of the
identified breakpoints of leukemia patients cluster to MLL introns
9–11 (~81%), and AF4 introns 3 and 4 (~86%), indicating that these
regions of both genes are the preferred hotspots for the
illegitimate recombination event. Of note, breakpoints within
MLL intron 11 are most frequently found in infant ALL, and appear
to change the biology of the reciprocal AF4-MLL fusion protein by
disrupting of the first PHD finger of the PHD domain. This changes
the binding properties of CYP33 (PPIE, Peptidyl-prolyl cis-trans
isomerase E) to the PHD domain, and thus, the biology of the
reciprocal fusion protein AF4-MLL [2]. This “infant version” of the
AF4-MLL fusion protein has been used throughout this study. By contrast, when CRISPR/CAS9 technology was used to
generate balanced chromosomal translocations in target cells,
leukemia development was efficient [7]. This raises again the
general questions about the requirements of the direct and
reciprocal fusion protein and their roles for leukemia onset (for
review see ref. [3]). In one sense, t(4;11) leukemia is quite a peculiar disease
because leukemic cells do not display recurrent secondary
mutations [8, 9]—at least in many infant t(4;11) leukemia cases
[10]. Apart from a few RAS mutations [11, 12] or individual
subclonal mutations, the overall mutation frequency is very low. Thus, this initial translocation event seems to be necessary and
sufficient to cause the onset of acute leukemia. Oncogene www.nature.com/onc The role of reciprocal fusions in MLL-r acute leukemia: studying
the chromosomal translocation t(4;11) Alexander Wilhelm1 and Rolf Marschalek
1✉ © The Author(s) 2021 © The Author(s) 2021 Leukemia patients bearing the t(4;11)(q21;q23) translocations can be divided into two subgroups: those expressing both reciprocal
fusion genes, and those that have only the MLL-AF4 fusion gene. Moreover, a recent study has demonstrated that patients
expressing both fusion genes have a better outcome than patients that are expressing the MLL-AF4 fusion protein alone. All this
may point to a clonal process where the reciprocal fusion gene AF4-MLL could be lost during disease progression, as this loss may
select for a more aggressive type of leukemia. Therefore, we were interested in unraveling the decisive role of the AF4-MLL fusion
protein at an early timepoint of disease development. We designed an experimental model system where the MLL-AF4 fusion
protein was constitutively expressed, while an inducible AF4-MLL fusion gene was induced for only 48 h. Subsequently, we
investigated genome-wide changes by RNA- and ATAC-Seq experiments at distinct timepoints. These analyses revealed that the
expression of AF4-MLL for only 48 h was sufficient to significantly change the genomic landscape (transcription and chromatin)
even on a longer time scale. Thus, we have to conclude that the AF4-MLL fusion protein works through a hit-and-run mechanism,
probably necessary to set up pre-leukemic conditions, but being dispensable for later disease progression. Oncogene (2021) 40:6093–6102; https://doi.org/10.1038/s41388-021-02001-2 Received: 16 March 2021 Revised: 10 August 2021 Accepted: 20 August 2021
Published online: 6 September 2021 Received: 16 March 2021 Revised: 10 August 2021 Accepted: 20 August 2021
Published online: 6 September 2021
1Institute of Pharmaceutical Biology/DCAL, Goethe-University of Frankfurt, Frankfurt/Main, Germany. ✉email: Rolf.Marschalek@em.uni-frankfurt.de 1Institute of Pharmaceutical Biology/DCAL, Goethe-University of Frankfurt, Frankfurt/Main, Germany. ✉email: Rolf.Marschalek@em.un INTRODUCTION However, the
question remains what kind of functions are exerted by both
fusion proteins and whether their actions are required temporarily
or throughout the whole process of leukemia onset. In the past 25 years, researchers have tried to dissect the role of
the MLL-AF4 and AF4-MLL fusion proteins. In most studies, only
the MLL-AF4 fusion gene has been tested in functional assays, but
most studies have failed to demonstrate the oncogenicity of the
MLL-AF4 fusion protein (summarized in ref. [3]). In vivo studies in
mice have also mostly failed, except for two studies, one of which
comes from our own laboratory. Of note, Lin et al. were only able
to recapitulate leukemia development in mice when using a
marinized MLL-Af4 construct in distinct target cells [4]. They also MLL and AF4 wildtype protein complexes have some very basic
functions in mammalian cells. The MLL wildtype protein complex
is known to confer active chromatin marks on target gene A. Wilhelm and R. Marschalek omotors which enable target gene transcription [13–15]. The
F4 complex [16, 17], also termed “super-elongation complex”
eviewed in ref. [18]), is responsible for transcriptional elongation
9]. Both biological processes are crucial for any living cell, and
erefore, pathological functions deriving from t(4;11) fusion
oteins should be easily monitored when investigating changes
in gene transcription. This is important to mention as
aim to mimic leukemia development, rather study the
changes in chromatin and gene transcription in comb
long-time effects. Moreover, we were also interested in finding
explanation for the elimination of the AF4-MLL alle
A. Wilhelm and R. Marschalek
Oncogene (2021) 4 6094 promotors which enable target gene transcription [13–15]. The
AF4 complex [16, 17], also termed “super-elongation complex”
(reviewed in ref. [18]), is responsible for transcriptional elongation
[19]. Both biological processes are crucial for any living cell, and
therefore, pathological functions deriving from t(4;11) fusion
proteins should be easily monitored when investigating changes
in gene transcription. This is important to mention as we
aim to mimic leukemia development, rather study the imm
changes in chromatin and gene transcription in combinati
long-time effects. Moreover, we were also interested in finding a
explanation for the elimination of the AF4-MLL allele in promotors which enable target gene transcription [13–15]. The
AF4 complex [16, 17], also termed “super-elongation complex”
(reviewed in ref. [18]), is responsible for transcriptional elongation
[19]. INTRODUCTION Both biological processes are crucial for any living cell, and
therefore, pathological functions deriving from t(4;11) fusion
proteins should be easily monitored when investigating changes in gene transcription. This is important to mention as we did not
aim to mimic leukemia development, rather study the immediate
changes in chromatin and gene transcription in combination with
long-time effects. Moreover, we were also interested in finding a rational
explanation for the elimination of the AF4-MLL allele in about Cloning of t(4;11) fusion genes and establishment of a t(4;11)
model system g
The signature deriving from the co-expression of MLL-AF4 and
AF4-MLL at day 3 exhibits a total of 264,289 reads. The majority of
reads derived from upregulated pseudogenes (n = 255; 110,862
reads), non-annotated genes (n = 235; 45,116 reads), 13 LincRNA
genes (328 reads), 5 microRNA genes (1831 reads), and 2 SnoRNA
genes (23 reads). The protein-coding genes (n = 107) caused
22,656
reads,
while
mitochondrial
genes
were
also
highly
expressed (n = 11; 83,435 reads). The downregulated gene
signature was composed of 16 pseudogenes, 13 non-annotated
genes, 1 microRNA gene, 2 SnoRNA genes, and 24 protein-coding
genes with a total of 1616 reads. model system
The t(4;11) fusion gene cassettes were cloned into existing
Sleeping Beauty vector systems [21]. This resulted in 3 different
constructs depicted in Fig. 1A: (1) pSBbi::MLL-AF4 (MLL exons
1–10::AF4 exons 4–20), (2) pSBTet::AF4-MLL (AF4 exon 1–3::MLL
exons 12–37) and (3) pSBTet::MLL-AF4 (MLL exons 1–10::AF4 exons
4–20). In order to develop a cell line model system that allowed us
to address the above-mentioned scientific questions, we stably
transfected the pSBbi::MLL-AF4 vector into HEK293T cells (ATCC
CRL-3216™). These cells were then used to isolate the first RNA
samples (d0) and were then transiently transfected with pSBTet::
AF4-MLL (d1). Doxycycline-induction (1 µg/ml for 48 h) was carried
out to achieve a population of cells that expressed both fusion
genes. The expression of the reciprocal AF4-MLL fusion protein
was terminated on day 3 by a medium exchange without
Doxycycline. Cells were grown for an additional 25 days and
aliquots were taken on days 3, 12, and 28, respectively. All this is
summarized in Fig. 1B, where the RT-PCR experiments are shown
for all experimental timepoints (d0 - d28; all three biological
replicates). Strong expression of the AF4-MLL transgene was visible
on day 3, before the shutdown of expression of this transgene. This was independently shown by the decreasing amount of red-
fluorescent cells from d2 to d7 (see Fig. 1C). A Western blot
experiment performed on day 3 demonstrated the co-expression
of both MLL-AF4 and AF4-MLL, respectively (Fig. 1B). Isolated total
RNA from d0–d28 in biological replicates was used to perform the
MACE-Seq experiments. The signature at day 12 was strongly reduced but still comprises
375,996 reads. The signature was derived from 14 pseudogenes
(43.329 reads), 4 non-annotated genes (53 reads), and 11 protein-
coding genes (9.328 reads). MACE analyses revealed again the synergism between the
MLL-AF4 and AF4-MLL fusion proteins From this type of analysis, we conclude that the presence of the
AF4-MLL fusion protein enabled transcription of genes that are
usually shut-down, e.g., pseudogenes, non-annotated genes, etc. This type of arbitrary gene activation can be clearly seen on day 3
and day 28. The Venn diagrams displayed in Fig. 2A (lower panels)
show the overlap between signatures at day 3 and 28 (418 genes),
while the downregulated gene sets appear to be idiosyncratic at
each tested timepoint. The overall MACE data analysis is summarized in Fig. 2A (upper
panel). It summarizes the identified number of gene entries for all
6 cell lines. The last 6 rows display the significant signatures that
were identified (>10 reads, p-value < 0.05 and FC > ±4). The constitutive expression of the MLL-AF4 fusion protein alone
caused a tiny signature of 50 upregulated and 42 downregulated
genes. Co-expression of MLL-AF4 together with the reciprocal
fusion protein AF4-MLL resulted in a highly increased gene set
(634 upregulated, 58 downregulated genes). After turning off the
AF4-MLL fusion protein, this large signature disappeared again on
day 12 (48 upregulated, 50 downregulated genes) but re-
appeared on day 28 (555 upregulated, 53 downregulated genes). This looked like a selection process where a subpopulation of cells
from day 3 was positively selected overtime to maintain the
extended day 3 gene signature. Cloning of t(4;11) fusion genes and establishment of a t(4;11)
model system The vast majority of reads derived
again from mitochondrial genes (n = 19; 323,283 reads). The
downregulated gene signature was composed of 4 pseudogenes,
6 non-annotated genes, 1 LincRNA gene, 2 microRNA genes, and
36 protein-coding genes with a total of 674 reads. The signature at day 28 expanded again and was composed of
214,167 reads. The signature derived from 288 pseudogenes
(93.989 reads), 186 non-annotated genes (31,028 reads), 4
LincRNA genes (80 reads), 4 microRNA genes (1497 reads), 3
SnoRNA genes (51 reads), and 51 protein-coding genes (14.799
reads). Still, the vast majority of reads derived again from
mitochondrial genes (n = 17; 72,853 reads). The downregulated
gene signature was composed of 6 pseudogenes, 7 non-
annotated genes, 1 microRNA gene, and 37 protein-coding genes
with a total of 1186 reads. Oncogene (2021) 40:6093 – 6102 Oncogene (2021) 40:6093 – 6102 A. Wilhelm and R. Marschalek 6095 Fig. 1
Fusions genes and established cell culture model. A All 3 vector constructs are depicted. Only the expressed part of the different
Sleeping Beauty vectors is shown. The MLL-AF4 open reading frame was cloned into the constitutive pSBbi vector, while the reciprocal AF4-
MLL expression construct was inserted into the pSBTet vector backbone. A final vector construct was the MLL-AF4 open reading frame
inserted into the pSBTet-GP backbone; the latter construct was used for a control experiment during the ATAC-Seq experiment. B The
construction of the test cell line is shown. HEK293 cells that constitutively express MLL-AF4 were used to transiently transfect the AF4-MLL
construct at day 1. After a short selection and induction of the AF4-MLL transgene, the expression of AF4-MLL was shut down at day 3, and
cells were grown for an additional 25 days. Samples for RNA or DNA isolation were taken at d0, d3, d12, and d28, respectively. RT-PCR
experiments of all biological replicates are shown. While the MLL-AF4 fusion allele was expressed constantly over the observation period, the
AF4-MLL fusion gene was diminishing. A GAPDH control demonstrates that equal amounts of RNA were used throughout this experiment. Western blot experiments were performed on day 3 for MLL-AF4 and AF4-MLL respectively. C Microscopic pictures to demonstrate the
presence of all vectors as outlined. The SB vector encoding MLL-AF4 expresses constitutively GFP, while the inducible AF4-MLL vector
expresses a dTomato protein. This way, the out-segregation of the AF4-MLL plasmid could be visually traced until day 7. 40% of all patients that appears to worsen the disease outcome,
and what triggers this clonal evolutionary process [20]. (33.058 reads), 4 non-annotated genes (54 reads), 1 LincRNA gene
(16 reads) and 13 protein-coding genes (3.083 reads). The vast
majority of reads are derived from mitochondrial genes (n = 17;
305,728 reads). The downregulated gene signature was composed
of 2 pseudogenes, 1 non-annotated gene, and 38 protein-coding
genes with a total of 2769 reads. Heatmap and volcano plot analyses revealed the high
similarity between the day3 and day 28 signatures 2
Data dissection of the RNA-Seq data and heatmap and volcano plot analysis. A Upper panel: summary of the MACE-Seq data. The
first 6 columns summarize the data obtained by MACE-Seq, the last 6 columns display the filtered information, when applying stringent
criteria (>10 reads, p-value < 0.05 log2 > ± 2) Bottom panel: VENN diagrams displaying the shared up- and downregulated genes between the
different signatures. B Heatmaps were created by using the gene signatures (~700 up-and downregulated genes from all 4 timepoints)
obtained from the cell lines at the four independent time points (d0, d3, d12, and d28) using the ClustVis online tool (biit.cs.ut.ee/clustvis/). C Similarly, gene entries of all protein-coding genes at the four independent time points (d0, d3, d12, and d28) were used to visualize the
significant changes by volcano plots. Gene symbols together with p-values, log2 changes, and −log10 (p-value) data were used to perform the
analyses (VolcaNoseR website, huygens.science.uva.nl). The number of gene entries used for the displayed plots is displayed in the upper left
corner. We used stringent parameters to visualize in red and in blue the most significant changes in gene expression. C
0
100
200
300
4
0
4
d0
-log10(p-value) 4
0
4
d3
4
0
4
d12
log2 4
0
4
d28 d12 d3 log2 log2 Fig. 2
Data dissection of the RNA-Seq data and heatmap and volcano plot analysis. A Upper panel: summary of the MACE-Seq data. The
first 6 columns summarize the data obtained by MACE-Seq, the last 6 columns display the filtered information, when applying stringent
criteria (>10 reads, p-value < 0.05 log2 > ± 2) Bottom panel: VENN diagrams displaying the shared up- and downregulated genes between the
different signatures. B Heatmaps were created by using the gene signatures (~700 up-and downregulated genes from all 4 timepoints)
obtained from the cell lines at the four independent time points (d0, d3, d12, and d28) using the ClustVis online tool (biit.cs.ut.ee/clustvis/). C Similarly, gene entries of all protein-coding genes at the four independent time points (d0, d3, d12, and d28) were used to visualize the
significant changes by volcano plots. Gene symbols together with p-values, log2 changes, and −log10 (p-value) data were used to perform the
analyses (VolcaNoseR website, huygens.science.uva.nl). The number of gene entries used for the displayed plots is displayed in the upper left
corner. Heatmap and volcano plot analyses revealed the high
similarity between the day3 and day 28 signatures 6096 entries
total reads
mean
reads
total up
total
down
de novo up
(<2-fold)
MIR
4x
LINC
4x
PS
4x
sig up 4x
mean reads
sig down 4x
mean reads
total reads
of signature
d0
30,083
5,444,783
180.99
15,327 14,756
2,221
1,933 reads
0 / 0
out of 198
1 / 0
out of 619
15 / 2
out of 3,943
50
6838.82
42
65.94
341,941
d3
31,665
5,656,272
178.62
18,172 13,665
3,803
4,885 reads
5 / 1
out of 205
13 / 0
out of 711
255 / 16
out of 4,176
634
416.86
58
27.88
264,289
d12
29,981
5,805,637
193.64
15,280 14,701
2,119
1,960 reads
0 / 2
out of 207
0 / 1
out of 612
14 / 4
out of 3,884
48
7833.24
50
13.49
375,996
d28
31,536
5,751,501
178.04
17,709 13,824
3,672
3,430 reads
4 / 1
out of 221
4 / 0
out of 689
288 / 6
out of 4,356
555
385.89
53
22.39
214,167
shared up- and downregulated genes
A
d0
d3
d12
d28
17
34
14
12
9
6
9
0
11
0
9
3
1
d0
d3
d12
d28
12
216
133
4
25
6
377
10
2
5
0
2
0 A B
entries
total reads
mean
reads
total up
total
down
de novo up
(<2-fold)
MIR
4x
LINC
4x
PS
4x
sig up 4x
mean reads
sig down 4x
mean reads
total reads
of signature
d0
30,083
5,444,783
180.99
15,327 14,756
2,221
1,933 reads
0 / 0
out of 198
1 / 0
out of 619
15 / 2
out of 3,943
50
6838.82
42
65.94
341,941
d3
31,665
5,656,272
178.62
18,172 13,665
3,803
4,885 reads
5 / 1
out of 205
13 / 0
out of 711
255 / 16
out of 4,176
634
416.86
58
27.88
264,289
d12
29,981
5,805,637
193.64
15,280 14,701
2,119
1,960 reads
0 / 2
out of 207
0 / 1
out of 612
14 / 4
out of 3,884
48
7833.24
50
13.49
375,996
d28
31,536
5,751,501
178.04
17,709 13,824
3,672
3,430 reads
4 / 1
out of 221
4 / 0
out of 689
288 / 6
out of 4,356
555
385.89
53
22.39
214,167
shared up- and downregulated genes
A
d0
d3
d12
d28
17
34
14
12
9
6
9
0
11
0
9
3
1
d0
d3
d12
d28
12
216
133
4
25
6
377
10
2
5
0
2
0 entries
total reads
mean
reads
total up
total
down
de novo up
(<2-fold)
MIR
4x
LINC
4x
PS
4x
sig up 4x
mean reads
sig down 4x
mean reads
total reads
of signature
d0
30,083
5,444,783
180.99
15,327 14,756
2,221
1,933 reads
0 / 0
out of 198
1 / 0
out of 619
15 / 2
out of 3,943
50
6838.82
42
65.94
341,941
d3
31,665
5,656,272
178.62
18,172 13,665
3,803
4,885 reads
5 / 1
out of 205
13 / 0
out of 711
255 / 16
out of 4,176
634
416.86
58
27.88
264,289
d12
29,981
5,805,637
193.64
15,280 14,701
2,119
1,960 reads
0 / 2
out of 207
0 / 1
out of 612
14 / 4
out of 3,884
48
7833.24
50
13.49
375,996
d28
31,536
5,751,501
178.04
17,709 13,824
3,672
3,430 reads
4 / 1
out of 221
4 / 0
out of 689
288 / 6
out of 4,356
555
385.89
53
22.39
214,167
A d28
d3
d12
d0
B
C
200
300
d0
d3
d12
d28
p-value)
entries
total reads
mean
reads
total up
total
down
de novo up
(<2-fold)
MIR
4x
LINC
4x
PS
4x
sig up 4x
mean reads
sig down 4x
mean reads
total reads
of signature
d0
30,083
5,444,783
180.99
15,327 14,756
2,221
1,933 reads
0 / 0
out of 198
1 / 0
out of 619
15 / 2
out of 3,943
50
6838.82
42
65.94
341,941
d3
31,665
5,656,272
178.62
18,172 13,665
3,803
4,885 reads
5 / 1
out of 205
13 / 0
out of 711
255 / 16
out of 4,176
634
416.86
58
27.88
264,289
d12
29,981
5,805,637
193.64
15,280 14,701
2,119
1,960 reads
0 / 2
out of 207
0 / 1
out of 612
14 / 4
out of 3,884
48
7833.24
50
13.49
375,996
d28
31,536
5,751,501
178.04
17,709 13,824
3,672
3,430 reads
4 / 1
out of 221
4 / 0
out of 689
288 / 6
out of 4,356
555
385.89
53
22.39
214,167
shared up- and downregulated genes
A
d0
d3
d12
d28
17
34
14
12
9
6
9
0
11
0
9
3
1
d0
d3
d12
d28
12
216
133
4
25
6
377
10
2
5
0
2
0 shared up- and downregulated genes
d0
d3
d12
d28
17
34
14
12
9
6
9
0
11
0
9
3
1
d0
d3
d12
d28
12
216
133
4
25
6
377
10
2
5
0
2
0 ted genes
d0
d3
d12
d28
17
34
14
12
9
6
9
0
11
0
9
3
1 shared u
d0
d3
d12
d28
12
216
133
4
25
6
377
10
2
5
0
2
0 d3 d3 d0 d28
d3
d12
d0
B B C C
0
100
200
300
4
0
4
d0
4
0
4
d3
4
0
4
d12
4
0
4
d28
log2
-log10(p-value)
Fig. Heatmap and volcano plot analyses revealed the high
similarity between the day3 and day 28 signatures For heatmap analyses, we retrieved only the protein-coding genes
of all signatures. The heatmap analysis is displayed in Fig. 2B, where
we analyzed deregulated target genes from all 4 timepoints. The
combined gene set contained a total of 608 genes that were
retrieved from the up- and down-regulated gene signatures at all
timepoints. From the heatmap analysis, it became clear that cells
expressing both fusion proteins at day 3 were clustering together
on day 28, while the day 0 signature clustered together with day 12. Further inspection of these signatures (last row in Figs. 2A and
S1, upper panel), revealed that the signature caused by MLL-AF4
resulted in 341,941 reads that derived from 15 pseudogenes Oncogene (2021) 40:6093 – 6102 A. Wilhelm and R. Marschalek d28
d3
d12
d0
B
C
0
100
200
300
4
0
4
d0
4
0
4
d3
4
0
4
d12
4
0
4
d28
log2
-log10(p-value)
entries
total reads
mean
reads
total up
total
down
de novo up
(<2-fold)
MIR
4x
LINC
4x
PS
4x
sig up 4x
mean reads
sig down 4x
mean reads
total reads
of signature
d0
30,083
5,444,783
180.99
15,327 14,756
2,221
1,933 reads
0 / 0
out of 198
1 / 0
out of 619
15 / 2
out of 3,943
50
6838.82
42
65.94
341,941
d3
31,665
5,656,272
178.62
18,172 13,665
3,803
4,885 reads
5 / 1
out of 205
13 / 0
out of 711
255 / 16
out of 4,176
634
416.86
58
27.88
264,289
d12
29,981
5,805,637
193.64
15,280 14,701
2,119
1,960 reads
0 / 2
out of 207
0 / 1
out of 612
14 / 4
out of 3,884
48
7833.24
50
13.49
375,996
d28
31,536
5,751,501
178.04
17,709 13,824
3,672
3,430 reads
4 / 1
out of 221
4 / 0
out of 689
288 / 6
out of 4,356
555
385.89
53
22.39
214,167
shared up- and downregulated genes
A
d0
d3
d12
d28
17
34
14
12
9
6
9
0
11
0
9
3
1
d0
d3
d12
d28
12
216
133
4
25
6
377
10
2
5
0
2
0
2
Data dissection of the RNA-Seq data and heatmap and volcano plot analysis. A Upper panel: summary of the MACE-Seq data. Heatmap and volcano plot analyses revealed the high
similarity between the day3 and day 28 signatures t 6 columns summarize the data obtained by MACE-Seq, the last 6 columns display the filtered information, when applying string
eria (>10 reads, p-value < 0.05 log2 > ± 2) Bottom panel: VENN diagrams displaying the shared up- and downregulated genes between
erent signatures. B Heatmaps were created by using the gene signatures (~700 up-and downregulated genes from all 4 timepo
ained from the cell lines at the four independent time points (d0, d3, d12, and d28) using the ClustVis online tool (biit.cs.ut.ee/clustv
imilarly, gene entries of all protein-coding genes at the four independent time points (d0, d3, d12, and d28) were used to visualize
nificant changes by volcano plots. Gene symbols together with p-values, log2 changes, and −log10 (p-value) data were used to perform
lyses (VolcaNoseR website, huygens.science.uva.nl). The number of gene entries used for the displayed plots is displayed in the upper
ner. We used stringent parameters to visualize in red and in blue the most significant changes in gene expression. Involvement of mitochondrial gene activation by t(4;11)
fusion proteins Chromosome usage analysis revealed patterns revealing the
pathomolecular power of the different t(4;11) fusion genes
Another analysis we performed was to fingerprint the deregulated
genes on chromosomes (GUDC module). This is depicted in Fig. S3, where the analysis is shown for all 4 timepoints for the
observed up- and down-regulated gene signatures. The different
timepoints are displayed and the “mean gene usage” is given (e.g.,
the signature of 50 up-regulated genes at day 0 has a mean of
0,11% of all genes from all chromosomes at that timepoint). Deviations from the mean usage are given for each chromosome
(except chromosome Y). Relatively more target genes were
expressed e.g., from chromosomes 18, 5, 11, and 21 at day 0. Chromosome usage analysis revealed patterns revealing the
pathomolecular power of the different t(4;11) fusion genes
Another analysis we performed was to fingerprint the deregulated
genes on chromosomes (GUDC module). This is depicted in Fig. S3, where the analysis is shown for all 4 timepoints for the
observed up- and down-regulated gene signatures. The different
timepoints are displayed and the “mean gene usage” is given (e.g.,
the signature of 50 up-regulated genes at day 0 has a mean of
0,11% of all genes from all chromosomes at that timepoint). Deviations from the mean usage are given for each chromosome
(except chromosome Y). Relatively more target genes were
expressed e.g., from chromosomes 18, 5, 11, and 21 at day 0. The co-expression of AF4-MLL increased the mean gene usage
from 0.11% to 1.96% (~18-fold increase), and some chromosomes
showed a higher target gene number (chromosomes 5, 7, 2, 18, 1,
6), others with lower target gene numbers, (chromosomes 19, 17,
14, 20, 13, 12, 15, 8). Thus, target gene activation or repression is
not a random but rather a selective process. Gene usage dropped
again at d12, and the pattern at day 28 appears to be a further
development of the day 3 pattern, indicating a clonal evolution
that is presumably taking place. In the second ATAC-Seq experiment, MLL-AF4 or AF4-MLL were
only short-time induced and then turned off again (Fig. 4A). This
was compared to our initial setting with a constitutive MLL-AF4
and 48 h expression of AF4-MLL. Short-term effects on chromatin
at d3 were observed for both single-transfected cells, however,
this effect was nearly lost on d28. Involvement of mitochondrial gene activation by t(4;11)
fusion proteins One of the interesting findings was that certain mitochondrial
genes were dramatically overexpressed, a mechanism that has
never been described before for t(4;11) leukemia cells. This is
summarized in Fig. S2, where mitochondrial genes are listed. This figure displays the number of reads from mock cells, as
well
as
from
the
gene
signatures
obtained
at
d0–d28,
respectively. The strong increase of distinct mitochondrial
genes appears to be a property of the MLL-AF4 fusion protein,
but could even be enhanced by the presence of the AF4-MLL
fusion protein (e.g., ribosomal RNA genes). The most affected
mitochondrial genes were ATP6, CO1-3, CYB, ND4, and both
mitochondrial rRNAs (12S and 16S), respectively. Our experi-
mental
efforts
to
measure
differences
in
mitochondrial
respiration in our day 0 to day 28 cells versus mock cells
remained inconclusive (data not shown). Therefore, we con-
cluded that the overexpression of these mitochondrial genes
has a yet unknown, metabolic function that occurs in the
presence of t(4;11) fusion proteins. In order to investigate this assumption in more detail, we
carefully analyzed the MACE- and ATAC-Seq data sets (Fig. 3B). MACE-Seq signatures were subclassified according to gene types
(pseudogenes, non-annotated genes, LincRNA genes, microRNA
genes, SNO genes, and protein-coding genes). In particular, the
number of pseudogenes/non-annotated genes (PG/NA) and
protein-coding genes (PCG) were summarized for the up- and
downregulated gene signatures in the small table on the left of
Fig. 3B. Noteworthy, not every target gene identified in the MACE
experiment was also found in the ATAC-Seq experiment. However,
the concordance was quite high (>90%). At day 0, target genes
(385 out of 459 MACE targets) displayed a pattern that most
activated genes derived from the accessible chromatin fractions,
and vice versa, downregulated target genes (381 out of 384 MACE
targets) were associated with less accessible chromatin. This
changed at day 3, because activated target genes (1592 out of
1747 MACE targets) were more deriving from less accessible
chromatin, while down-regulated genes (479 out of 498 MACE
targets) could be attributed equally to both chromatin fractions. At day 28, the ratio for up-regulated genes (1207 out of 1269
MACE targets) was similar to d3, and downregulated genes (344
out of 355 MACE target genes) were again associated with less
accessible chromatin. This clearly indicated that AF4-MLL allowed
activating target genes even from the less accessible chromatin
fraction. Heatmap and volcano plot analyses revealed the high
similarity between the day3 and day 28 signatures We used stringent parameters to visualize in red and in blue the most significant changes in gene expression. Oncogene (2021) 40:6093 – 6102 A. Wilhelm and R. Marschalek 6097 The resulting chromatin data from the first ATAC-Seq experi-
ment (Fig. S1, lower panel) were quite comparable (see mean
reads/gene entry). All data entries were then filtered to select
target-gene signatures (>2 reads, p-value < 0.05, log2 > ± 1 or ± 2)
and displayed by Circos plots in Fig. 3A, where significant
signatures of MACE- and ATAC-Seq data are compared (log2 =
± 1). In MLL-AF4 expressing cells, the total number of deregulated
genes was lower than the ATAC-Seq changes. Cells co-expressing
both fusion proteins displayed already a much higher number of
deregulated target genes when compared with the ATAC-Seq
data. Similarly, the number of deregulated genes at d28 was again
much higher than the observed changes in the ATAC-Seq
experiment. This argues again that AF4-MLL acts like a “chromatin
opener” as visible from the increasing chromatin accessibility from
d3 to d28. Similarly, we performed Volcano plot analyses with all protein-
coding genes sets that are summarized in Fig. 2C. The total
number of gene entries is indicated for each plot. Of interest, MLL/
KMT2A is only visible at day 3, because of the overexpression of
AF4-MLL that exhibits the poly-adenylated 3′-portion of the MLL
gene. Another interesting finding is the MIF gene that can only be
found to be strongly and significantly overexpressed in cells from
d3–d28 (day 0 FC = 4.0, day 3 FC = 21.5, day 12 FC = 5.3, and day
28 FC = 21.3). High MIF expression (Macrophage Inhibitory Factor)
has been recently linked to worse outcomes and high relapse in
leukemia patients (see discussion). Comparison of MACE and ATAC-Seq data revealed a critical
function of the reciprocal fusion protein AF4-MLL Comparison of MACE and ATAC-Seq data revealed a critical
function of the reciprocal fusion protein AF4-MLL Finally, we performed ATAC-Seq experiments to investigate the
chromatin changes mediated by the expression of the single and
co-expressed t(4;11) fusion proteins. For the purpose of our
studies, we performed two different ATAC-Seq experiments. For
the first experiment, we used an analogous setting to the MACE
experiment. In the second experiment, we also analyzed both
single fusion proteins (MLL-AF4 or AF4-MLL) in inducible vectors
to address their functions in a setting where they were expressed
individually. The transgenes in these single fusion protein
experiments were induced for only 48 h and followed until day
28 to understand their individual impact. Involvement of mitochondrial gene activation by t(4;11)
fusion proteins When compared to cells in
which both fusion proteins were present (constitutive MLL-AF4,
inducible AF4-MLL for 48 h). the changes in chromatin accessibility
were still detectable at d28. The maintained gene set at d28 was
significantly higher than within the single transfected cells (20.2%
versus 6.2% or 4.6%). This supported a “hit & run mechanism”
exerted by AF4-MLL because the observed changes did not
diminish after shutting down the expression of the AF4-MLL
fusion protein. Importantly, these data suggest again the fact that
an initial co-expression of both fusion proteins has a benefit,
however, the AF4-MLL fusion could be lost once the process of
clonal evolution has been initiated. The co-expression of AF4-MLL increased the mean gene usage
from 0.11% to 1.96% (~18-fold increase), and some chromosomes
showed a higher target gene number (chromosomes 5, 7, 2, 18, 1,
6), others with lower target gene numbers, (chromosomes 19, 17,
14, 20, 13, 12, 15, 8). Thus, target gene activation or repression is
not a random but rather a selective process. Gene usage dropped
again at d12, and the pattern at day 28 appears to be a further
development of the day 3 pattern, indicating a clonal evolution
that is presumably taking place. De novo gene activation or the shut-down of gene
transcription reveal an important mechanism of MLL fusion
proteins A Comparison of the
genome-wide MACE- with ATAC-Seq data visualized for day 0, day 3, and day 28. It shows the number of deregulated genes in MACE
experiments, as well as the changes in chromatin accessibility in the ATAC-Seq experiments. B. Left Table: summary of deregulated gene
classes (pseudogenes/non-annotated genes (PG/NA)) versus protein-coding genes (PCG) for up and downregulated genes with a log2 values
of ±1. Right part: Pseudogene/non-annotated genes (pink numbers) were compared to PCG’s (blue numbers) by indicating their percentages
in each circle plot. Numbers in the green and red rectangles show e.g., that the up- or downregulated genes and their origin from accessible
or less accessible chromatin fragments. This type of analysis was performed for all 12 subsections. Fig. 3
Comparing MACE- and ATAC-Seq data of the t(4;11) model system in Circos plots and DAGT module analysis. A Comparison of the
genome-wide MACE- with ATAC-Seq data visualized for day 0, day 3, and day 28. It shows the number of deregulated genes in MACE
experiments, as well as the changes in chromatin accessibility in the ATAC-Seq experiments. B. Left Table: summary of deregulated gene
classes (pseudogenes/non-annotated genes (PG/NA)) versus protein-coding genes (PCG) for up and downregulated genes with a log2 values
of ±1. Right part: Pseudogene/non-annotated genes (pink numbers) were compared to PCG’s (blue numbers) by indicating their percentages
in each circle plot. Numbers in the green and red rectangles show e.g., that the up- or downregulated genes and their origin from accessible
or less accessible chromatin fragments. This type of analysis was performed for all 12 subsections. four different time points. Except for a few genes (numbers in
brackets), these de novo or shut-down gene signatures were not
part of the highly up- or down-regulated gene signatures, as they
are expressed at very low levels (with very few reads per gene). Moreover, these signatures could trace them over time. Vice versa,
we observed similar behavior with genes that were completely
turned off. Genes found to be shut down at day 0 could be traced until day 28, but again at days 3, 12, and 28 additional genes were
shut down over time. This type of analysis was quite important as
it contradicted the results presented in Fig. 2A–C. All these figures
have suggested that the d28 signature derived from the day
3 signature, and was probably positively selected over 25 days
(see above). De novo gene activation or the shut-down of gene
transcription reveal an important mechanism of MLL fusion
proteins Finally, we investigated the obtained MACE-Seq data for de novo
gene expression, as well as for the shut-down of genes in the four
different signatures (DAGE & ST module). As shown in Fig. 4B,
several thousand genes became either activated (2221 de novo
genes at d0; 3803 at d3; 2119 at d12; 3672 at d28) or shut down
(3957 shut-down genes at d0; 3171 at d3; 3964 at d12; 2973 at
d28) in the presence of an individual or both fusion proteins at the Oncogene (2021) 40:6093 – 6102 different time points. Except for a few genes (numbers in
ckets), these de novo or shut-down gene signatures were not
of the highly up- or down-regulated gene signatures, as they
expressed at very low levels (with very few reads per gene). eover, these signatures could trace them over time. Vice versa,
b
d
i
il
b h
i
i h
h
l
l
until day 28, but again at days 3, 12, and 28 additional genes were
shut down over time. This type of analysis was quite important as
it contradicted the results presented in Fig. 2A–C. All these figures
have suggested that the d28 signature derived from the day
3 signature, and was probably positively selected over 25 days
(
b
) H
h
DAGE/ST
d l
l
f d
3
Comparing MACE- and ATAC-Seq data of the t(4;11) model system in Circos plots and DAGT module analysis. A Comparison of the
ome-wide MACE- with ATAC-Seq data visualized for day 0, day 3, and day 28. It shows the number of deregulated genes in MACE
eriments, as well as the changes in chromatin accessibility in the ATAC-Seq experiments. B. Left Table: summary of deregulated gene
ses (pseudogenes/non-annotated genes (PG/NA)) versus protein-coding genes (PCG) for up and downregulated genes with a log2 values
1. Right part: Pseudogene/non-annotated genes (pink numbers) were compared to PCG’s (blue numbers) by indicating their percentages
ach circle plot. Numbers in the green and red rectangles show e.g., that the up- or downregulated genes and their origin from accessible
ess accessible chromatin fragments. This type of analysis was performed for all 12 subsections. A. Wilhelm and R. Marschalek A. Wilhelm and R. Marschalek 6098 Fig. 3
Comparing MACE- and ATAC-Seq data of the t(4;11) model system in Circos plots and DAGT module analysis. De novo gene activation or the shut-down of gene
transcription reveal an important mechanism of MLL fusion
proteins However, the DAGE/ST module analyses of de novo
and shut-down genes tell another story, namely that of ongoing Oncogene (2021) 40:6093 – 6102 A. Wilhelm and R. Marschalek 6099 g. 4
Changes in ATAC-Seq experiment with single fusion proteins and co-expression of both fusion proteins. A Circos plots of ATAC
xperiments with inducible MLL-AF4, inducible AF4-MLL, or the combination of constitutive MLL-AF4 in combination with inducible AF4-
eveal that the combination of constitutive MLL-AF4 and 48 h AF4-MLL demonstrated again the long-term impact on chromatin of
articular combination. Inducible transgenes were expressed only for 48 h and analyzed on day 3 and day 28 (inducible transgenes shut d
or 25 days). B Analysis of de novo genes and shut-down genes by the DAGE/ST module. Several thousand genes were found to be induce
ompletely repressed by the expression of the tested t(4;11) fusion proteins. These signatures obtained at the individual timepoints w
aced during the 4 weeks of the experiment and displayed by different colors (day 0 signature = blue; day 3 signature = green;
2 signature = red; day 28 signature is orange). These analyses revealed a clonal evolutionary process that could very likely have b
iggered by the expression of AF4-MLL on days 1–3. Fig. 4
Changes in ATAC-Seq experiment with single fusion proteins and co-expression of both fusion proteins. A Circos plots of ATAC-Seq
experiments with inducible MLL-AF4, inducible AF4-MLL, or the combination of constitutive MLL-AF4 in combination with inducible AF4-MLL
reveal that the combination of constitutive MLL-AF4 and 48 h AF4-MLL demonstrated again the long-term impact on chromatin of this
particular combination. Inducible transgenes were expressed only for 48 h and analyzed on day 3 and day 28 (inducible transgenes shut down
for 25 days). B Analysis of de novo genes and shut-down genes by the DAGE/ST module. Several thousand genes were found to be induced or
completely repressed by the expression of the tested t(4;11) fusion proteins. These signatures obtained at the individual timepoints were
traced during the 4 weeks of the experiment and displayed by different colors (day 0 signature = blue; day 3 signature = green; day
12 signature = red; day 28 signature is orange). These analyses revealed a clonal evolutionary process that could very likely have been
triggered by the expression of AF4-MLL on days 1–3. Fig. 4
Changes in ATAC-Seq experiment with single fusion proteins and co-expression of both fusion proteins. DISCUSSION
Thi
i This manuscript describes experiments addressing two important
questions concerning the pathology of t(4;11) MLL fusion proteins. First of all, we aimed to understand the pathological relevance of
the direct and reciprocal fusion protein, MLL-AF4 and AF4-MLL, on
a genome-wide level. The second question addressed whether
continued expression of AF4-MLL is required, or whether AF4-MLL
could act by a hit-and-run mechanism. The latter question was
raised by several observations, namely that some patients with t
(4;11) leukemia are diagnosed with an MLL-AF4 fusion gene, but
lack the AF4-MLL allele. Several laboratories have already shown
that the missing AF4-MLL allele could be explained by complex
translocations [22], but there are also cases where the reciprocal
fusion gene is not expressed (either transcriptionally inactive or
deleted allele). A recent publication has also demonstrated that t
(4;11) patients lacking the expression of the AF4-MLL allele have a
worse prognosis [20]. y
p
p
y
p
To verify these initial findings, we also examined the protein-
coding genes in the traceable signatures shown in Fig. S4B. Again,
we extracted the protein-coding genes and analyzed these gene
sets by pathway analyses. The blue signature did not reveal any
known pathway. The green signature at day 3 revealed the
pathways
already
identified,
“G-protein-coupled
signaling”,
“innate cell response” and “humoral immune response”. Of
interest, only the combination of two protein-coding gene
signatures (orange and green) at day 28 revealed the “B cell
proliferation” pathway. In addition, “Calcium homeostasis”, “Chlor-
ide transport” and “PLC-activating G-protein signaling” pathways
became overt. All this indicates that the cells were changing their
behavior and diverging significantly from the original cells at day
0 due to clonal evolution. To address both questions at the same time, we decided to use
a setting where MLL-AF4 is constitutively expressed, while AF4-
MLL expression could be temporarily turned on by using an
inducible vector backbone. Constitutive expressing MLL-AF4 alone resulted in a relatively
small signature (for details see Fig. 2A). When AF4-MLL was co-
expressed for a very short time period (48 h), an immense
upregulation of target genes occurred. The observed gene
signature contained roughly 8-times more protein-coding genes,
but also several hundred pseudogenes and non-annotated genes. Such a strong increase of genes could only be explained by non-
specific chromatin activation that enabled this massive increase of
gene transcription. The heatmap analysis (Fig. De novo gene activation or the shut-down of gene
transcription reveal an important mechanism of MLL fusion
proteins A Circos plots of ATAC-Seq
experiments with inducible MLL-AF4, inducible AF4-MLL, or the combination of constitutive MLL-AF4 in combination with inducible AF4-MLL
reveal that the combination of constitutive MLL-AF4 and 48 h AF4-MLL demonstrated again the long-term impact on chromatin of this
particular combination. Inducible transgenes were expressed only for 48 h and analyzed on day 3 and day 28 (inducible transgenes shut down
for 25 days). B Analysis of de novo genes and shut-down genes by the DAGE/ST module. Several thousand genes were found to be induced or
completely repressed by the expression of the tested t(4;11) fusion proteins. These signatures obtained at the individual timepoints were
traced during the 4 weeks of the experiment and displayed by different colors (day 0 signature = blue; day 3 signature = green; day
12 signature = red; day 28 signature is orange). These analyses revealed a clonal evolutionary process that could very likely have been
triggered by the expression of AF4-MLL on days 1–3. Oncogene (2021) 40:6093 – 6102 A. Wilhelm and R. Marschalek 6100 evolution and selection that started at the timepoint when the
reciprocal fusion protein AF4-MLL was expressed. This is a
completely different view and is not apparent when analyzing
heatmaps, volcano plots, or gene signatures. This ongoing
evolution occurs at the level of lowly expressed genes and
suggests that cells initiated such an evolutionary process as a
result of the presence of the AF4-MLL fusion protein. evolution and selection that started at the timepoint when the
reciprocal fusion protein AF4-MLL was expressed. This is a
completely different view and is not apparent when analyzing
heatmaps, volcano plots, or gene signatures. This ongoing
evolution occurs at the level of lowly expressed genes and
suggests that cells initiated such an evolutionary process as a
result of the presence of the AF4-MLL fusion protein. In conclusion, the expression of AF4-MLL between d1 and d3
induced a gene signature of which a large portion (~80%) was still
present at d28. Of interest, most identified genes that were
upregulated at d3 or d28 represent pseudogenes or non-
annotated genes, indicative of non-specific activation of chroma-
tin which may allow other transcription factors, e.g., the MLL-AF4
fusion protein, to activate novel gene signatures that may
promote the conversion of normal cells into pre-malignant cells. DISCUSSION
Thi
i 2B) clearly clustered
d0 with d12, and d3 with d28, while the volcano plot analysis (Fig. 2C) already indicated certain genes to be expressed from day 3
onward, indicating already here that the expression of AF4-MLL
initiated an ongoing clonal evolution (e.g., the disappearance of
ALOXB12 at day 12, or the maintenance of MIF (d3–d28)). MIF has
been recently identified as a critical target gene that correlated
with a worse outcome in leukemia patients, as it was defined as an
independent prognostic factor important for OS and DSF [23] We did the same analysis for the shut-down genes to determine
which programs were switched off. In Fig. S5A we first analyzed
the shut-down gene signatures in a VENN diagram and analyzed
all intersections and idiosyncratic signatures by pathway analyses. Most of these analyses for intersections and full signatures
revealed
that
“cell
adhesion”,
“cell
migration”
or
“cell-cell
interactions” are lost from these cells. The idiosyncratic program
at day 3 (153 genes) also revealed a relief of “cell fate
commitment”, which by contrast would suggest a “cell de-
differentiation process”. independent prognostic factor important for OS and DSF [23]. A detailed analysis of the ATAC-Seq data is displayed in Fig. S1
(lower panel). At day 0, most gene types (PS, NA, LINC, MIR, SNO,
and PCG) were equally up- and down-regulated. From d3
onwards, the number of all gene types that were now increasingly
linked with accessible chromatin (PG, NA, LINC, and PCG). This
indicated again that the transcriptional patterns were evolving
and become selected over time, due to short-term expression of
the AF4-MLL fusion protein (d1–d3). p
Further analysis of the traceable genetic shut-down program is
summarized in Fig. S5B. Here, it became obvious that the
“biological” or “cell adhesion” as well as “migration” were already
shut down at d0. The signature at d3 also inhibited cell-cell
interactions, while the MLL-AF4 mediated signature at d12 was
focussed on inhibiting “T cell functions”. Combining these
signatures at d3 and d12 supports this finding (blue/green at d3
and d12). The green signature at d28 once more inhibits “T helper
cell functions”. The detailed comparison of MACE- and ATAC-Seq data (Fig. 3A/
B) revealed a clear shift from protein-coding genes to pseudo-
genes/non-annotated genes in the upregulated signatures. This
could well be interpreted as an evolutionary process where many
genes were activated for gene transcription, even from less
accessible chromatin. De novo gene activation or the shut-down of gene
transcription reveal an important mechanism of MLL fusion
proteins To this end, AF4-MLL prepares the ground for an adaptive cell
type that may change according to internal or external signals. p
p
Therefore, we decided to examine these signatures in more
detail. The upregulated protein-coding genes were analyzed in a
VENN diagram (Fig. S4A). The overlap of these protein-coding
genes was 95 genes in all four signatures. All subsignatures
(overlapping and idiosyncratic) were analyzed by gene ontology
enrichment analysis, but only those which are marked in green
resulted in successful output. These were the gene sets with 41,
95, 541, 189, and 315 de novo genes, as well as the full signatures
at day 3 (1195) and day 28 (889). Of interest, gene set 541
contained an “innate response program” (response to bacteria;
antimicrobial response) as well as a “humoral immune response”,
while the full signature of de novo genes at day 28 displayed “B
cell proliferation”. In addition, the major pathways identified on
days 3 and 28 were “G-protein-coupled receptor signaling”,
“Regulation of signaling receptor activity” and “Calcium signaling”,
respectively. This is interesting, as it points to unexpected
signaling pathways. As these signatures were mostly related to
the presence of AF4-MLL, we can assume that the changes made
by this fusion protein are probably important. Outline of our experimental setting and bioinformatic
pipeline: data evaluation and establishment of novel tools p
Based on these data, we pose the hypothesis that the disruption
of the MLL protein between the CXXC domain and the PHD/BD
domain causes a dramatic effect: it results in a direct fusion protein
that is able to strongly enhance target gene transcription, but that
the additional presence of a complementary, reciprocal fusion
protein AF4-MLL enables a “broader use” of the genome, namely
the activation of certain genes within repressed chromatin. Such an
“adaptive genome usage” would be important, as it allows a given
cell to change its cell fate rapidly, depending on triggers from the
outside. These novel features make a pre-tumor cell almost
omnipotent with regard to deregulated gene expression. Over
time and depending on external signals, this will convert a normal
cell into an aberrant cell, and most likely causes the onset of cancer,
combined with strong features of pluripotency. According to our
experience with MLL-r leukemias, this is presumably one of the best
definitions we can make for the most prevalent MLL-r leukemias,
namely those bearing a t(4;11) translocation. p p
As summarized in Fig. S6, our experimental setting was used to perform
MACE- and ATAC-Seq experiments. Differential expression analysis was
performed using the R-Bioconductor DESeq2 library. Raw counts were
normalized by the geometric mean-based method [28]. These data were
used to define a simple algorithm (>10 reads, p-values < 0.05, log2 > ± 2)
that allows the definition of highly significant gene signatures. The
resulting data were used to prepare Circos plots [29] for the visualization of
genome-wide changes in gene transcription or the ATAC-Seq data. In
addition, we used these data sets to generate heatmaps, volcano plots,
and pathway analyses. p
y
y
In addition, we used the FileMaker database program to import all the
DESeq2 data for further analysis and to apply additional algorithms. This
resulted in three additional analytic modules, named GUDC, DAGT, and
DAGE/ST, respectively. The GUDC module analyzes the “gene usage on
different chromosomes”, which results in a kind of “chromosome
fingerprint”. The result of the analysis is displayed for each chromosome
as more (positive) or less (negative) gene expression in comparison to the
mathematical mean expression for each chromosome. The DAGT module
(“differential analysis by gene type”) automatically subclassifies each gene
entry in our signatures to one of the different gene types (pseudogenes,
non-annotated genes, LINC RNAs, MIR RNAs SNO RNAs, mitochondrial
genes, and protein-coding genes. DISCUSSION
Thi
i It seems that genes are tested for any
benefit they may offer to these cells, with genes not supporting a
cellular advantage presumably being shut down. That these genes
are derived from the pool of pseudogenes or non-annotated
genes is not a surprise since those genes may harbor benefits for In summary, the analysis of de novo and shut-down genes over
the four different time points clearly suggests a clonal evolution of
these cells when triggered by the expression of the reciprocal
fusion protein AF4-MLL for 48 h (d1–d3). The induced changes in
the genetic program trigger the cells towards B cell pathways and
inhibit T cell functions, and moreover inhibit migration and cell
adhesion processes, while supporting PLC- and G-protein coupled
signaling pathways, which are important for several chemokine
signaling pathways. Oncogene (2021) 40:6093 – 6102 A. Wilhelm and R. Marschalek 6101 malignant cell growth [24]. Thus, it seems that the cells are
gambling with their genetic coding potential in order to find the
best way to adapt to a new cell fate. In biology, this is equivalent
to “cell plasticity” and is usually a typical sign of stem or
progenitor cells. RNeasy® Mini Kit (Qiagen) and cDNA synthesis was performed using
SuperScript® II (Invitrogen) at days 0, day 3, day 12, and day28. All
isolated RNAs were quality checked (Agilent Bioanalyzer) and final
concentrations were determined. Equal amounts of total RNA were used throughout all
experiments, and all experiments were performed with 3 biological
replicates. All primers used for RT-PCR analyses are as follows: MLL8•3
(5′–CCCAAAACCACTCCTAGTGAG–3′), MLL13•5 (5′–CAGGGTGATAGCTG
TTTCGG–3′),
AF4•3
(5′–GTTGCAATGCAGCAGAAGCC–3′),
and
AF4•5
(5′–ACTGTCACTGTCCTCACTGTCA–3′). With these 4 oligonucleotides,
transcription of all vector-derived transgenes was successfully tested. g
This evolutionary process can not be maintained by the single
fusion proteins alone, as seen in Fig. 4A. Thus, co-expression of
both fusion proteins– even when giving them only a short time of
action—seems to be a necessary step in order to induce changes
in the chromatin that remain over longer time periods. This interpretation is supported by findings within the low
expressed gene signatures (de novo and shut-down genes by using
the DAGE/ST module; Fig. 4B). We could clearly see in the identified
de novo and shut-down gene signatures that the expression of
AF4-MLL induced an evolutionary process. Differential gene expression profiling by MACE-Seq Differential gene expression profiling by MACE Seq
The chimeric genes were either expressed constitutively or induced for
48 h with 1 µg/ml Doxycycline and total RNA were isolated from
transfected cell lines. After testing the correct expression of both
transgenes, differential gene expression (DGE) profiles were obtained by
MACE (Massive Analysis of cDNA Ends)—Seq experiments following the
manufacturer protocol (GenXPro, Frankfurt, Germany). Further details are
given in the Supplementary data file. The MACE data have been deposited
on the GEO server with the Accession number GSE178569. Cell culture and transfections HEK293T cells were grown in DMEM with 10% (v/v) FCS (Capricon
Scientific), 2 mM L-Glutamine (Capricon Scientific), and 1% (v/v) Pen Strep
(GE Healthcare) at 37 °C and 5% CO2. The single-transfected stable cell line
expressing MLL-AF4 (pSBbi::MLL-AF4) in a constitutive fashion was
established by using a low amount (50 ng) of SB transposase vector
SB100X. After 24 h, cells were subjected to Puromycin (1 µg/ml). The cells
were incubated with selection markers for 3–10 days and terminated when
virtually all cells were emitting the expected green color derived from their
corresponding reporter genes (eGFP). The cells were further cultivated for
several weeks before being used for the second round of transient
transfection with an inducible AF4-MLL construct (pSBTet::AF4-MLL). Blasticidin and Doxycycline were administered only for 48 h and then
relived to allow segregation. The transfected cell line continued to express
the respective reporter and selection marker (eGFP and Puromycin), while
the second plasmid became out-segregated due to lack of selection. ATAC-Seq experiments As a side note, strong overexpression of specific mitochondrial
genes (ATP6, CO1-3, CYB, ND4, and both mitochondrial rRNAs (12S
and 16S)) were overexpressed in the presence of t(4;11) fusion
proteins (Fig. S2). We have no rational explanation for this
experimental observation, despite the fact that others have
observed that the overexpression of certain mitochondrial genes
was associated with poor clinical outcomes [25, 26]. Preparation of ATAC samples was performed according to a published
protocol [27]. Further details are given in the Supplementary data file. The
ATAC-Seq data have been deposited on the GEO server with the Accession
number GSE178567 and GSE178568. DISCUSSION
Thi
i Of note, we observed
within the de novo gene fractions a “B-cell specific” gene signature,
while genes responsible for “T-cell activities” were downregulated
at d28 (Figs. S4 and S5). In addition, from day 3 onwards, we saw
gene signatures that resemble innate immune cell activities, and
thus, recalls a “myeloid gene program”. This would fit perfectly the
well-known mixed-lineage phenotype of t(4;11) leukemia cells. 1. Meyer C, Burmeister T, Gröger D, Tsaur G, Fechina L, Renneville A, et al. The MLL
recombinome of acute leukemias in 2017. Leukemia. 2018;32:273–84.
2. Rössler T, Marschalek R. An alternative splice process renders the MLL protein
either into a transcriptional activator or repressor. Pharmazie. 2013;68:601–7.
3. Marschalek R. The reciprocal world of MLL fusions: a personal view. Biochim
Biophys Acta Gene Regul Mech 1. 2020;863:194547.
4. Lin S, Luo RT, Ptasinska A, Kerry J, Assi SA, Wunderlich M, et al. Instructive role of
MLL-fusion proteins revealed by a model of t(4;11) Pro-B acute lymphoblastic
leukemia. Cancer Cell. 2016;30:737–49. Outline of our experimental setting and bioinformatic
pipeline: data evaluation and establishment of novel tools Finally, the DAGE/ST module “differential
analysis of de novo or shut-down gene expression”) uses the DESeq2 data
to identify “de novo induced genes” or “shut-down genes” after t(4;11)
transgene expression. For this purpose, a log2var discriminator (defined as
“Ln(fold change)/Ln2”) was used, because the DESeq2 provides log2 data
even when mock or experimental data displayed zero reads. By using the
log2var discriminator, we were able to quickly identify all “de novo
transcribed genes” or “shut-down genes” and included these critical gene
sets in our analyses. The “signature tracing” (ST) module allows tracing
gene signatures over time. 4. Lin S, Luo RT, Ptasinska A, Kerry J, Assi SA, Wunderlich M, et al. Instructive role of
MLL-fusion proteins revealed by a model of t(4;11) Pro-B acute lymphoblastic
leukemia. Cancer Cell. 2016;30:737–49. ADDITIONAL INFORMATION Supplementary information The online version contains supplementary material
available at https://doi.org/10.1038/s41388-021-02001-2. Supplementary information The online version contains supplementary material
available at https://doi.org/10.1038/s41388-021-02001-2. 18. Luo Z, Lin C, Shilatifard A. The super elongation complex (SEC) family in tran-
scriptional control. Nat Rev Mol Cell Biol. 2012;13:543–7. 19. Cho S, Schroeder S, Ott M. CYCLINg through transcription: posttranslational
modifications
of
P-TEFb
regulate
transcription
elongation. Cell
Cycle. 2010;9:1697–705. Correspondence and requests for materials should be addressed to Rolf Marschalek. Reprints and permission information is available at http://www.nature.com/
reprints Reprints and permission information is available at http://www.nature.com/
reprints 20. Agraz-Doblas A, Bueno C, Bashford-Rogers R, Roy A, Schneider P, Bardini M, et al. Unraveling the cellular origin and clinical prognostic markers of infant B-cell
acute lymphoblastic leukemia using genome-wide analysis. Haematologica. 2019;104:1176–88. Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims
in published maps and institutional affiliations. Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims
in published maps and institutional affiliations. 21. Kowarz E, Löscher D, Marschalek R. Optimized sleeping beauty transposons
rapidly generate stable transgenic cell lines. Biotechnol J. 2015;10:647–53. 22. Kowarz E, Burmeister T, Lo Nigro L, Jansen MW, Delabesse E, Klingebiel T, et al. Complex MLL rearrangements in t(4;11) leukemia patients with absent AF4.MLL
fusion allele. Leukemia. 2007;21:1232–8. 23. Sharaf-Eldein M, Elghannam D, Elderiny W, Abdel-Malak C. Prognostic implication
of MIF gene expression in childhood acute lymphoblastic leukemia. Clin Lab. 2018;64:1429–37. Open Access This article is licensed under a Creative Commons
Attribution 4.0 International License, which permits use, sharing,
adaptation, distribution and reproduction in any medium or format, as long as you give
appropriate credit to the original author(s) and the source, provide a link to the Creative
Commons license, and indicate if changes were made. The images or other third party
material in this article are included in the article’s Creative Commons license, unless
indicated otherwise in a credit line to the material. If material is not included in the
article’s Creative Commons license and your intended use is not permitted by statutory
regulation or exceeds the permitted use, you will need to obtain permission directly
from the copyright holder. To view a copy of this license, visit http://creativecommons. org/licenses/by/4.0/. 24. Wang A, Hai R. Noncoding RNAs serve as the deadliest universal regulators of all
cancers. Cancer Genomics Proteom. 2021;18:43–52. 25. Wu S, Fahmy N, Alachkar H. REFERENCES 1. Meyer C, Burmeister T, Gröger D, Tsaur G, Fechina L, Renneville A, et al. The MLL
recombinome of acute leukemias in 2017. Leukemia. 2018;32:273–84. 2. Rössler T, Marschalek R. An alternative splice process renders the MLL protein
either into a transcriptional activator or repressor. Pharmazie. 2013;68:601–7. 3. Marschalek R. The reciprocal world of MLL fusions: a personal view. Biochim
Biophys Acta Gene Regul Mech 1. 2020;863:194547. RNA extraction, cDNA synthesis, and RT-PCR experiments
The cell line expresses the MLL-AF4 transgene throughout the 4-weeks
observation time. The AF4-MLL transgene was induced by using 1 µg/ml
Doxycycline to the cell culture for 48 h. Total RNA was isolated using 4. Lin S, Luo RT, Ptasinska A, Kerry J, Assi SA, Wunderlich M, et al. Instructive role of
MLL-fusion proteins revealed by a model of t(4;11) Pro-B acute lymphoblastic
leukemia. Cancer Cell. 2016;30:737–49. Oncogene (2021) 40:6093 – 6102 A. Wilhelm and R. Marschalek 6102 28. Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion
for RNA-seq data with DESeq2. Genome Biol. 2014;15:550. 5. Bursen A, Schwabe K, Rüster B, Henschler R, Ruthardt M, Dingermann T, et al. The
AF4-MLL fusion protein is capable of inducing ALL in mice without requirement
of MLL-AF4. Blood. 2010;115:3570–9. 28. Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion
for RNA-seq data with DESeq2. Genome Biol. 2014;15:550. 29. Krzywinski M, Schein J, Birol I, Connors J, Gascoyne R, Horsman D, et al. Circos: an
information aesthetic for comparative genomics. Genome Res. 2009;19:1639–45. 29. Krzywinski M, Schein J, Birol I, Connors J, Gascoyne R, Horsman D, et al. Circos: an
information aesthetic for comparative genomics. Genome Res. 2009;19:1639–45. 6. Wilkinson AC, Ballabio E, Geng H, North P, Tapia M, Kerry J, et al. RUNX1 is a key
target in t(4;11) leukemias that contributes to gene activation through an AF4-
MLL complex interaction. Cell Rep. 2013;3:116–27. 7. Secker KA, Bruns L, Keppeler H, Jeong J, Hentrich T, Schulze-Hentrich JM, et al. Only hematopoietic stem and progenitor cells from cord blood are susceptible to
malignant transformation by MLL-AF4 translocations. Cancers. 2020;12:1487. ACKNOWLEDGEMENTS We thank GenXPro GmbH in Frankfurt, specially Mohammed Alkhatib, Lukas Jost, and
Björn Rotter, for their ongoing help in setting up MACE- and ATAC-Seq experiments
as well as the helpful discussions for the bioinformatic processing of data. We are also
thankful to Arpita Kundu and Silvia Bracharz for their handling help with cell cultures,
and Dr. Constantin Kondak and Prof. Jochen Klein for their help with the
mitochondrial respiration experiments. 8. Bardini M, Spinelli R, Bungaro S, Mangano E, Corral L, Cifola I, et al. DNA copy-
number abnormalities do not occur in infant ALL with t(4;11)/MLL-AF4. Leukemia. 2010;24:169–76. 9. Prelle C, Bursen A, Dingermann T, Marschalek R. Secondary mutations in t(4;11)
leukemia patients. Leukemia. 2013;27:1425–7. 10. Andersson AK, Ma J, Wang J, Chen X, Gedman AL, Dang J, et al. The landscape of
somatic mutations in infant MLL-rearranged acute lymphoblastic leukemias. Nat
Genet. 2015;47:330–7. AUTHOR CONTRIBUTIONS Cloning, sequencing, acquisition of data, analysis, and interpretation of data (e.g.,
statistical analysis, biostatistics, computational analysis): AW. Constructing the
database, visualization of data, writing and reviewing the original draft, of the
manuscript, project administration: RM. 11. Mansur MB, Ford AM, Emerenciano M. The role of RAS mutations in MLL-
rearranged leukaemia: a path to intervention? Biochim Biophys Acta Rev Cancer. 2017;1868:521–6. 12. Chu SH, Song EJ, Chabon JR, Minehart J, Matovina CN, Makofske JL, et al. Inhi-
bition of MEK and ATR is effective in a B-cell acute lymphoblastic leukemia model
driven by Mll-Af4 and activated Ras. Blood Adv. 2018;2:2478–90. FUNDING 13. Yokoyama A, Wang Z, Wysocka J, Sanyal M, Aufiero DJ, Kitabayashi I, et al. Leukemia proto-oncoprotein MLL forms a SET1-like histone methyltransferase
complex
with
menin
to
regulate
Hox
gene
expression. Mol
Cell
Biol. 2004;24:5639–49. This work is funded by DFG grants Ma 1876/12-1 and Ma 1876/13-1, and grant
2018.070.1 and 2018.070.2 from the Wilhelm Sander Foundation. Open Access
funding enabled and organized by Projekt DEAL. 14. Milne TA, Dou Y, Martin ME, Brock HW, Roeder RG, Hess JL. MLL associates
specifically with a subset of transcriptionally active target genes. Proc Natl Acad
Sci USA. 2005;102:14765–70. COMPETING INTERESTS 15. Dou Y, Milne TA, Ruthenburg AJ, Lee S, Lee JW, Verdine GL, et al. Regulation of
MLL1 H3K4 methyltransferase activity by its core components. Nat Struct Mol
Biol. 2006;13:713–9. The authors declare no competing interests. 16. Bitoun E, Oliver PL, Davies KE. The mixed-lineage leukemia fusion partner
AF4 stimulates RNA polymerase II transcriptional elongation and mediates
coordinated chromatin remodeling. Hum Mol Genet. 2007;16:92–106. CONSENT FOR PUBLICATION We have obtained consent to publish this paper from all the participants. 17. Benedikt A, Baltruschat S, Scholz B, Bursen A, Arrey TN, Meyer B, et al. The
leukemogenic AF4-MLL fusion protein causes P-TEFb kinase activation and
altered epigenetic signatures. Leukemia. 2011;25:135–44. ADDITIONAL INFORMATION The mitochondrial transcription machinery genes are
upregulated in acute myeloid leukemia and associated with poor clinical out-
come. Metab Open. 2019;2:100009. 26. Tyagi A, Pramanik R, Bakhshi R, Singh A, Vishnubhatla S, Bakhshi S. Expression of
mitochondrial genes predicts survival in pediatric acute myeloid leukemia. Int J
Hematol. 2019;110:205–12. 27. Corces MR, Trevino AE, Hamilton EG, Greenside PG, Sinnott-Armstrong NA,
Vesuna S, et al. An improved ATAC-seq protocol reduces background and
enables interrogation of frozen tissues. Nat Methods. 2017;14:959–62. © The Author(s) 2021 Oncogene (2021) 40:6093 – 6102
|
https://openalex.org/W2883513488
|
https://link.springer.com/content/pdf/10.1007%2Fs11307-018-1252-5.pdf
|
English
| null |
Pre-therapy Somatostatin Receptor-Based Heterogeneity Predicts Overall Survival in Pancreatic Neuroendocrine Tumor Patients Undergoing Peptide Receptor Radionuclide Therapy
|
Molecular imaging and biology
| 2,018
|
cc-by
| 7,957
|
Correspondence to: Ralph Bundschuh; e-mail: ralph.bundschuh@ukbonn.de Mol Imaging Biol (2019) 21:582Y590
DOI: 10.1007/s11307-018-1252-5
* The Author(s), 2018; corrected publication, July 2018
Published Online: 16 July 2018
RESEARCH ARTICLE
Pre-therapy Somatostatin Receptor-Based
Heterogeneity Predicts Overall Survival
in Pancreatic Neuroendocrine Tumor Patients
Undergoing Peptide Receptor Radionuclide
Therapy
Rudolf A. Werner,1,2 Harun Ilhan,3 Sebastian Lehner,3,4 László Papp,5 Norbert Zsótér,6
Imke Schatka,7 Dirk O. Muegge,2 Mehrbod S. Javadi,1 Takahiro Higuchi,2,8
Andreas K. Buck,2 Peter Bartenstein,3 Frank Bengel,9 Markus Essler,10
Constantin Lapa,2 Ralph A. Bundschuh10
1The Russell H. Morgan Department of Radiology and Radiological Science, Division of Nuclear Medicine and Molecular Imaging, Johns
Hopkins University School of Medicine, Baltimore, MD, USA
2Department of Nuclear Medicine, University Hospital Würzburg, Würzburg, Germany
3Department of Nuclear Medicine, University Hospital, LMU Munich, Munich, Germany
4Ambulatory Healthcare Center Dr. Neumaier & Colleagues, Radiology, Nuclear Medicine, Radiation Therapy, Regensburg, Germany
5Department of Nuclear Medicine, Medical University of Vienna, Vienna, Austria
6Mediso Medical Imaging Systems Ltd., Budapest, Hungary
7Department of Nuclear Medicine, Charité – Universitätsmedizin Berlin, Berlin, Germany
8Department of Bio Medical Imaging, National Cardiovascular and Cerebral Research Center, Suita, Japan
9Department of Nuclear Medicine, Hannover Medical School, Hannover, Germany
10Department of Nuclear Medicine, University Hospital Bonn, Sigmund-Freud-Str. 25, 53127, Bonn, Germany
Abstract
Purpose: Early identification of aggressive disease could improve decision support in pancreatic
neuroendocrine tumor (pNET) patients prior to peptide receptor radionuclide therapy (PRRT).
The prognostic value of intratumoral textural features (TF) determined by baseline somatostatin
receptor (SSTR)-positron emission tomography (PET) before PRRT was analyzed.
Th
i i
l
i
f thi
ti l
i
d d
t
t
ti
O Mol Imaging Biol (2019) 21:582Y590
DOI: 10.1007/s11307-018-1252-5
* The Author(s), 2018; corrected publication, July 2018
Published Online: 16 July 2018 RESEARCH ARTICLE Pre-therapy Somatostatin Receptor-Based
Heterogeneity Predicts Overall Survival
in Pancreatic Neuroendocrine Tumor Patients
Undergoing Peptide Receptor Radionuclide
Therapy Rudolf A. Werner,1,2 Harun Ilhan,3 Sebastian Lehner,3,4 László Papp,5 Norbert Zsótér,6
Imke Schatka,7 Dirk O. Muegge,2 Mehrbod S. Javadi,1 Takahiro Higuchi,2,8
Andreas K. Buck,2 Peter Bartenstein,3 Frank Bengel,9 Markus Essler,10
Constantin Lapa,2 Ralph A. Bundschuh10 1The Russell H. Morgan Department of Radiology and Radiological Science, Division of Nuclear Medicine and Molecular Imaging, Johns
Hopkins University School of Medicine, Baltimore, MD, USA
2Department of Nuclear Medicine, University Hospital Würzburg, Würzburg, Germany
3Department of Nuclear Medicine, University Hospital, LMU Munich, Munich, Germany
4Ambulatory Healthcare Center Dr. Neumaier & Colleagues, Radiology, Nuclear Medicine, Radiation Therapy, Regensburg, Germany
5Department of Nuclear Medicine, Medical University of Vienna, Vienna, Austria
6Mediso Medical Imaging Systems Ltd., Budapest, Hungary
7Department of Nuclear Medicine, Charité – Universitätsmedizin Berlin, Berlin, Germany
8Department of Bio Medical Imaging, National Cardiovascular and Cerebral Research Center, Suita, Japan
9Department of Nuclear Medicine, Hannover Medical School, Hannover, Germany
10Department of Nuclear Medicine, University Hospital Bonn, Sigmund-Freud-Str. 25, 53127, Bonn, Germany The original version of this article was revised due to a retrospective Open
Access order: This article was originally published electronically on the
publisher’s internet portal (currently SpringerLink) on July 16, 2018 without
open access. With the author(s)’ decision to opt for Open Choice the
copyright of the article changed on July 25, 2018 to © The Author(s) 2018
and the article is forthwith distributed under the terms of the Creative
Commons Attribution 4.0 International License (http://creativecommon-
s.org/licenses/by/4.0/), which permits use, duplication, adaptation, distribu-
tion and reproduction in any medium or format, as long as you give
appropriate credit to the original author(s) and the source, provide a link to
the Creative Commons license and indicate if changes were made.
Rudolf A. Werner, Harun Ilhan, Constantin Lapa and Ralph A. Bundschuh
contributed equally to this work.
Electronic supplementary material The online version of this article (https://
doi.org/10.1007/s11307-018-1252-5) contains supplementary material,
which is available to authorized users. Introduction diversity, earlier relapse of pNET patients undergoing radionu-
clide therapy, or mechanisms of tumor escape in dedifferentiated
tumors [14–16]. Decoding a general prognostic phenotype [10],
we hypothesized that intratumoral textural feature (TF) analysis
assessed by a baseline SSTR-PET might address the urgent
clinical need of prognostication in G1/2 pNET patients prior to
PRRT. Patients with potentially poor response to PRRT may be
identified and different therapeutic regimens might be applicable
(e.g., systemic therapies). Therefore, we aimed to elucidate the
prognostic capability of a baseline PET scan in a homogenous
cohort of G1/G2 pNET patients. As an orphan malignancy, the incidence of neuroendocrine
tumors with pancreatic origin (pNET) is continuously rising,
mainly due to technical progress in diagnostic imaging and
improved awareness of treating physicians [1, 2]. Surgical
resection is the only curative approach [3]. In advanced settings,
treatment options include cytotoxic chemotherapy, somatostatin
analogs, or targeted therapies such as tyrosine kinase and mTOR
inhibitors [4–7]. Recently, favorable results have been reported for
unresectable midgut as well as for bronchial NET using peptide
receptor radionuclide therapy (PRRT) with [177Lu]DOTA-D-Phe-
Tyr3-octreotate ([177Lu]DOTATATE) [8, 9]. Materials and Methods Tailored medical treatment mainly focuses on proteomics or
gene sequencing; however, their prognostic ability is rather
limited due to small sample sizes, ongoing tumor development,
and incomplete reflection of the entire tumor burden [10, 11]. Recently, the Delphic Consensus Assessment for Gastroenter-
opancreatic (GEP)-NET disease management reported on the
limitations of chromogranin A (CgA) alterations as well as Ki67
in identification of therapy responders. More precise clinical
decision-making increased demand for real-time multidimen-
sional information regarding tumor behavior [12]. Non-invasive
determination of intratumoral heterogeneity as assessed by
baseline somatostatin receptor (SSTR)-positron emission tomog-
raphy (PET) before PRRT has already proven its prognostic
performance by outperforming conventional PET parameters,
such as mean/maximum standardized uptake values (SUVmean/
max) in a mixed cohort of patients scheduled for endoradiotherapy
[13]. However, in particular for pNET, PRRT efficacy prediction
has not been elucidated yet due to considerable heterogeneous Since our study comprises a retrospective analysis of routinely
acquired data, the local ethic committees waived the need for
further approval. All patients gave written and informed
consent to the procedures as well as all patients provided
informed consent for scientific analysis of the obtained data. Werner R.A. et al.: Heterogeneity in pNET Undergoing PRRT Procedures: Thirty-one patients with G1/G2 pNET were enrolled (G2, n = 23/31). Prior to PRRT with
[177Lu]DOTATATE (mean, 3.6 cycles), baseline SSTR-PET computed tomography was performed. By segmentation of 162 (median per patient, 5) metastases, intratumoral TF were computed. The
impact of conventional PET parameters (SUVmean/max), imaging-based TF, and clinical parameters
(Ki67, CgA) for prediction of both progression-free survival (PFS) and overall survival (OS) after
PRRT were evaluated. Results: Within a median follow-up of 3.7 years, tumor progression was detected in 21 patients
(median, 1.5 years) and 13/31 deceased (median, 1.9 years). In ROC analysis, the TF entropy,
reflecting derangement on a voxel-by-voxel level, demonstrated predictive capability for OS (cutoff =
6.7, AUC = 0.71, p = 0.02). Of note, increasing entropy could predict a longer survival (9 6.7, OS =
2.5 years, 17/31), whereas less voxel-based derangement portended inferior outcome (G 6.7, OS =
1.9 years, 14/31). These findings were supported in a G2 subanalysis (9 6.9, OS = 2.8 years, 9/23
vs. G 6.9, OS = 1.9 years, 14/23). Kaplan–Meier analysis revealed a significant distinction between
high- and low-risk groups using entropy (n = 31, p G 0.05). For those patients below the ROC-derived
threshold, the relative risk of death after PRRT was 2.73 (n = 31, p = 0.04). Ki67 was negatively
associated with PFS (p = 0.002); however, SUVmean/max failed in prognostication (n.s.). Conclusions: In contrast to conventional PET parameters, assessment of intratumoral heteroge-
neity demonstrated superior prognostic performance in pNET patients undergoing PRRT. This novel
PET-based strategy of outcome prediction prior to PRRT might be useful for patient risk stratification. Key words: Pancreatic neuroendocrine tumor, Tumor heterogeneity, [68Ga], [177Lu]-DOTA-
TATE/-DOTATOC, PET/CT, SSTR Abstract Abstract
Purpose: Early identification of aggressive disease could improve decision support in pancreatic
neuroendocrine tumor (pNET) patients prior to peptide receptor radionuclide therapy (PRRT). The prognostic value of intratumoral textural features (TF) determined by baseline somatostatin
receptor (SSTR)-positron emission tomography (PET) before PRRT was analyzed. 583 PET/CT Imaging, Imaging Interpretation PRRT was performed with a mean of 7.2 ± 1.0 GBq (194.6 ±
27 mCi; range, 3.3–8.9 GBq, 89.2–240.5 mCi) per cycle using
[177Lu]DOTATATE. In total, the enrolled subjects underwent
112 treatment cycles (median, 4, range, 1–6; mean 3.6 ± 1.2)
aiming at a standard interval of 3 months on a compassionate use
basis [18, 19]. The majority of cases (21/31, 67.7 %) received at
least four subsequent treatment cycles. PRRT was performed
according to The joint IAEA, EANM, and SNMMI practical
guidance on a compassionate use basis or in accordance with the
Rotterdam protocol as published by Kwekkeboom et al., i.e., at
time point of disease progression [18, 19]. Long-acting and short-
acting release formulations were also discontinued according to
[18]. Imaging including both functional (SSTR-PET) and/or
morphologic imaging (CT) modalities was performed every 3–
6 months after PRRT [18, 19]. As a prerequisite for treatment initiation, all patients had to
demonstrate sufficient uptake in pre-therapeutic SSTR-PET
computed tomography (CT) [18, 19], i.e., lesional uptake
higher than physiological liver uptake [21]. A mean of 132 ±
35.7 MBq (3.6 ± 0.9 mCi; range, 72–185 MBq, 1.9–5 mCi) of
[68Ga]DOTATATE/-TOC (n = 27, [68Ga]DOTATATE and
n = 4, -[68Ga]DOTATOC) was administered intravenously. After 60 min, imaging was performed using the following
scanners: Bonn, Biograph 2 PET/CT (Siemens Medical Sol-
utions, Erlangen, Germany); Wuerzburg, Biograph 64 (Siemens
Medical Solutions, Erlangen, Germany); Munich, Gemini TF
PET/CT (Philips Medical, Eindhoven, Netherlands) or Siemens
Biograph 64 (Siemens Medical Solutions, Erlangen, Germany);
Hannover, Biograph 2 (Siemens Medical Solutions, Erlangen, Table 1. PET/CT Imaging, Imaging Interpretation Detailed patient’s characteristics according to Ki67/grading
Case
Sex
Age (y)
Metastases at time
of baseline SSTR-PET
Prior therapy
Number
of treatment
cycles
Cumulative activity
in GBq (mCi)
Ki67 (%)
Grading
#1
m
79
Liver, LN
Surgery
4
29.7 (802.7)
1
1
#2
f
73
Liver
Surgery
3
21.6 (583.8)
1
1
#3
m
71
Liver, bone, LN
SSA
4
31.4 (848.6)
2
1
#4
f
65
Liver, bone, LN
SSA, CTx
4
29.5 (797.3)
2
1
#5
m
61
Liver, LN
SSA
5
32.5 (878.4)
2
1
#6
m
71
Liver
None
3
19.9 (537.8)
2
1
#7
m
67
Liver
Surgery, SSA, CTx, RTx
2
4.6 (124.3)
2
1
#8
m
62
Liver
SSA
4
29.9 (808.1)
2
1
#9
m
57
Liver, LN
Surgery, SSA
4
27.9 (754.1)
4
2
#10
m
60
Liver
SSA
3
22.5 (608.1)
4
2
#11
m
44
Liver
Surgery, SSA
4
32.4 (875.7)
4
2
#12
f
70
Liver
None
6
40.6 (1097.3)
5
2
#13
m
52
Liver, bone, LN
SSA, CTx
4
29.6 (800)
5
2
#14
m
56
Liver, bone, LN
SSA, CTx
4
31.2 (843.2)
5
2
#15
f
70
Liver, LN, lung
Surgery
4
26 (702.7)
5
2
#16
f
71
Liver, bone, LN
None
4
30.2 (816.2)
5
2
#17
m
63
Liver, bone
Surgery
6
44.6 (1205.4)
5
2
#18
f
39
Liver
SSA, CTx
1
6 (162.2)
5
2
#19
m
62
Liver, LN
Surgery, SSA
4
29.6 (800)
7
2
#20
m
51
Liver
SSA
4
30.1 (813.5)
10
2
#21
f
62
Liver, LN
Surgery, SSA, CTx
4
27.8 (751.4)
10
2
#22
f
60
Liver
None
5
37.3 (1008.1)
10
2
#23
f
50
Liver
SSA
3
22.4 (605.4)
10
2
#24
m
53
Liver
Surgery
4
29.1 (786.5)
12
2
#25
f
41
LN
Surgery, SSA, CTx
2
14.4 (389.2)
15
2
#26
f
53
Liver
None
2
15.8 (427)
15
2
#27
m
66
Liver
SSA
4
15.9 (429.7)
15
2
#28
m
67
Liver, bone
SSA
4
29.7 (802.7)
15
2
#29
f
47
Liver, LN
CTx
4
29.6 (800)
20
2
#30
f
68
Liver
Surgery, SSA
2
15.2 (410.8)
20
2
#31
f
74
Liver, bone, LN
Surgery, CTx
1
8.6 (232.4)
20
2
y years, SSTR-PET somatostatin receptor-positron emission tomography, GBq gigabecquerel, m male, f female, LN lymph node, SSA somatostatin analog, CTx
chemotherapy, RTx radiation therapy, GBq gigabecquerel Table 1. Werner R.A. et al.: Heterogeneity in pNET Undergoing PRRT Werner R.A. et al.: Heterogeneity in pNET Undergoing PRRT Werner R.A. et al.: Heterogeneity in pNET Undergoing PRRT one (25.8 %) were classified as G1 NET and 23/31 (74.2 %) as
G2 NET. In G2 NET, the median Ki67 was 8 % (range, 4–20 %). one (25.8 %) were classified as G1 NET and 23/31 (74.2 %) as
G2 NET. In G2 NET, the median Ki67 was 8 % (range, 4–20 %). rogression-free survival (PFS) was defined according to
Response Evaluation Criteria in Solid Tumors 1.1 (RECIST1.1)
by follow-up examinations starting from the time point of baseline
imaging [18, 20]. For the calculation of overall survival (OS), the
time interval between the baseline SSTR-PET examination and
date of death was analyzed. Analysis of CgA levels before PRRT revealed a range
between 35 and 64.700 μg/l (median, 924 μg/l). Twenty-five
out of thirty-one (80.1 %) patients were pre-treated (somato-
statin analogs, n = 19/31, (61.2 %); surgery, n = 13/31
(41.9 %); chemotherapy, n = 9/31 (29 %); and external beam
radiation, n = 1/31 (3.2 %)). Clinical characteristics of the
patient cohort are given in Table 1. Analysis of CgA levels before PRRT revealed a range
between 35 and 64.700 μg/l (median, 924 μg/l). Twenty-five
out of thirty-one (80.1 %) patients were pre-treated (somato-
statin analogs, n = 19/31, (61.2 %); surgery, n = 13/31
(41.9 %); chemotherapy, n = 9/31 (29 %); and external beam
radiation, n = 1/31 (3.2 %)). Clinical characteristics of the
patient cohort are given in Table 1. PET/CT Imaging, Imaging Interpretation y years, SSTR-PET somatostatin receptor-positron emission tomography, GBq gigabecquerel, m male, f female, LN lymph node, SSA somatostatin analog, CTx
chemotherapy, RTx radiation therapy, GBq gigabecquerel in receptor-positron emission tomography, GBq gigabecquerel, m male, f female, LN lymph node, SSA somatostatin analog, C
therapy, GBq gigabecquerel Patient Population A total of 31 subjects (14/31 females (45.2 %), mean 60 ±
10 years (y), range, 39–79 y) of four university medical centers
with histologically proven pNET were enrolled. The patients
enrolled in the present subanalysis were part of a larger patient
cohort [13]. The study population was restricted to G1/2 pNET,
as G3 tumors normally suffer from rapid disease progression
under PRRT [17]. Ki67 ranged between 1 and 20 % with a
median of 5 % for the entire cohort (n = 31). Eight out of thirty- 584 Results Germany). System spatial resolutions are 4.8 mm for the Gemini
TF, 4.4 mm for the Biograph 64, and 9.3 mm for the Biograph 2
[22–24]. All data was reconstructed using iterative algorithms
implemented by the manufacturer and depending on the routine
protocol of the different medical centers. Scatter and attenuation
correction was performed based on the different transmission
data [13]. To allow for valid pooling of the results between
Siemens and Philips PET/CT scanners, phantom studies based
on the National Electrical Manufacturers Association NU2-2001
standard were conducted in Munich. According to a recent
published study investigating the robustness of TF in GEP-NET
patients using SSTR-PET in a multicentric setting, the following
TF were taken into account [25]: from the gray-level co-
occurrence matrix (Entropy, Homogeneity), from the gray-level
run length matrix (high gray-level run emphasis (HGRE)), and
from the gray-level size zone matrix (intensity variation, high
gray-level zone emphasis (HGZE), zone length non-uniformity
(ZLNU), short-zone high gray-level emphasis (SZHGE), zone
percentage (ZP)). In addition, metabolic tumor volume (MTV)
and total receptor expression (TRE) were assessed. Lesions were
identified by reviewing the SSTR-PET, CT, and fused hybrid
imaging by board-certified nuclear medicine physicians. In case
of multiorgan involvement, a maximum of three lesions per
organ (largest in size and metabolically most active lesion) was
segmented. A manual segmentation method was preferred in
order to exclude adjacent physiological SSTR-avid structures on
PET/CT images [26]. TF analysis was performed by using the
Interview Fusion Workstation (Mediso Medical Imaging Sys-
tems Ltd., Budapest, Hungary) [13]. As previously described, CT
images were available only for localization and not used to guide
delineation of the VOI [27]. Further, conventional PET
parameters (SUVmean/max, metabolic tumor volume (MTV) and
tissue receptor expression (TRE = MTV × SUVmean)) were also
investigated [13]. The radiotracer concentration in the ROIs was
normalized to the injected dose per kilogram of patient’s body
weight to derive the SUVs. For the assessment of TF, 162
volume of interest (median, 5, range, 1–12 per patient) were
manually segmented. In the majority of the cases (22/31, 71 %),
at least four lesions were investigated. Metastases with an MTV
smaller than 10 cm3 were excluded [28]. Almost all subjects suffered from liver metastases (30/31,
96.8 %), less than half of the cohort demonstrated lymph
node metastases (14/31, 45.2 %), 8/31 suffered from bone
lesions (25.8 %), and 1/31 (3.2 %) showed pulmonary
metastases (Table 1). Entire Cohort In ROC analysis of TF, entropy demonstrated a significant
predictive ability for OS (cutoff = 6.7, AUC = 0.71, p = 0.02)
with an accuracy of 71 %. Increasing entropy could predict
longer survival (9 6.7, OS = 2.5 y, 17/31), whereas less
entropy portended inferior outcome (G 6.7, OS = 1.9 y, 14/
31, Table 2, Supplementary Table a: see electronic supple-
mentary material (ESM)). All the investigated conventional
PET parameters (SUVmean/max, MTV, TRE) failed in
response prediction (Supplementary Table b). Subsequent Kaplan–Meier analysis revealed a significant
distinction between high- and low-risk patients for OS using
entropy (p = 0.045) in the whole cohort (PFS, n.s.). In Cox hazard analysis, entropy and intensity variation
demonstrated significance for OS (p G 0.05, respectively). For PFS, none of the investigated conventional PET
parameters (SUVmean/max) or other TF were significant. Regarding clinical parameters, the cumulative administered
dose reached significance for OS (p = 0.04, r = 0.37) and
Ki67 for PFS prediction (p = 0.002, r = −0.54). For those
patients below the ROC-derived threshold for entropy, the
RR of cancer-related death after PRRT was 2.73 (n = 31, CI
1.07–7.01; p = 0.04) (Fig. 1, Supplementary Table c in
ESM). Results During an observation period/follow-up of median
3.7 years, disease progression occurred in 21/31 subjects
(67.7 %) after a median of 1.5 y from the baseline PET scan
(range, 0.8 months–4.5 y). Thirteen out of thirty-one
(41.9 %) patients died from their tumor after a median of
1.9 y (range, 0.8 months–4.6 y). Of those, 11/13 (84.6 %)
belonged to the G2 group (mean Ki67, 11 %). The median
proliferation index Ki67 in the deceased patients was 5 %
(range, 2–20 %). Statistical Analysis Neither Ki67 nor grading demonstrated significant corre-
lation with the independent heterogeneity parameters (e.g.,
entropy/Ki67, r = −0.27, n.s.). Statistical analysis was performed using SPSS Statistics 22 and
MedCalc (Vers. 17.4.4). The cutoff values of each parameter for
the prediction of PFS and OS were identified through receiver
operating characteristic (ROC) analysis using the Youden Index
for maximization of specificity and sensitivity [29]. Kaplan–Meier
analysis (univariate analysis) was performed using thresholds
established by ROC analysis in cases in which ROC showed
statistical significant results. A multivariate Cox hazard analysis
was conducted to determine independent prognostic parameters as
well as relative risks (RR) [26, 30]. Non-parametric log-rank tests
were used to assess the differences in the Kaplan–Meier curves;
statistical significance was considered with a p value G 0.05. PET/CT Imaging, Imaging Interpretation Detailed patient’s characteristics according to Ki67/grading 585 Werner R.A. et al.: Heterogeneity in pNET Undergoing PRRT Subanalysis of G2 NET In ROC analysis of G2 NET, entropy reached significance
for OS prediction with an accuracy of 70 % (ROC, cutoff =
6.9, AUC = 0.72, p = 0.03). Regarding OS prediction, the
findings of the entire cohort were supported in a subanalysis
of G2 NET (9 6.9, OS = 2.8 y, 9/23 vs. G 6.9, OS = 1.9 y, 14/
23, Table 2). Werner R.A. et al.: Heterogeneity in pNET Undergoing PRRT Werner R.A. et al.: Heterogeneity in pNET Undergoing PRRT 586 Table 2. Receiver operating characteristic (ROC) analysis for progression-free (PFS) and overall survival (OS) for the textural feature entropy (independent
according to Cox analysis) Table 2. Receiver operating characteristic (ROC) analysis for progression-free (PFS) and overall survival (OS) for the textural feature entropy (independent
according to Cox analysis)
p value
Cutoff value
Sensitivity (%)
Specificity (%)
Accuracy (%)
PPV (%)
NPV (%)
AUC
≤cutoff
9 cutoff
Entropy—entire cohort (n = 31)
PFS
0.4
6.7
50
63.6
54.8
71.4
41.2
0.47
1.3 years (14/31)
1.9 years (17/31)
OS
0.02*
6.7
69.2
72.2
71
64.3
76.5
0.71
1.9 years (14/31)
2.5 years (17/31)
Entropy—G2 NET (n = 23)
PFS
0.14
6.9
58.8
33.3
52.2
71.4
22.2
0.66
1.3 years (14/23)
1.8 years (9/23)
OS
0.03*
6.9
81.8
58.3
70
64.3
77.8
0.72
1.9 years (14/23)
2.8 years (9/23)
Given for entire cohort (n = 31) and the G2 neuroendocrine tumors (NET, n = 23) subgroup. PFS and OS below the cutoff (≤cut-off) and above to the cutoff
(9 cutoff) with the number of patients for each group is shown in brackets. With increasing entropy, PFS and OS increase
PPV positive predictive value, NPV negative predictive value, AUC area under the curve
*Statistically significant Given for entire cohort (n = 31) and the G2 neuroendocrine tumors (NET, n = 23) subgroup. PFS and OS below the cutoff (≤cut-off) and above to the cutoff
(9 cutoff) with the number of patients for each group is shown in brackets. With increasing entropy, PFS and OS increase
PPV positive predictive value, NPV negative predictive value, AUC area under the curve
*Statistically significant wide range of NET from 2 to 20 %, i.e., that the therapeutic
response of a Blow^ G2 NET scheduled for PRRT might
differ from a G2 NET with an increased Ki67 [16]. In Kaplan–Meier analysis, no statistical significance was
reached in the G2 subgroup (p = 0.072). Subanalysis of G2 NET Results are
displayed for the entire cohort and the G2 group (Table 3);
respective Kaplan–Meier plots for OS are given in Fig. 2. Biopsy carries the potential for tumor under-sampling,
and as a consequence, inaccurate therapeutic decisions can
be made [31]. Hence, as a non-invasive whole-body
molecular tool considering the extent of disease, PET-
based assessment of intratumoral heterogeneity might serve
as a novel diagnostic biomarker reflecting the entire
phenotypical tumor burden. As previously described, the
prognostic value of TF derived by PET has been success-
fully investigated in different tumor types [26, 32–34]. In
our previous trial investigating various disease entities, TF
analysis of a baseline SSTR-PET/CT proved prognostic
value in PRRT candidates [13]. In the present study, we
focused on subjects only suffering from NET of pancreatic
origin. For NET, treatment options have improved in the last
years [35]: The NETTER-1 trial revealed impressive find-
ings using PRRT in midgut NET [8]. Moreover, recent
developments of systemic agents such as everolimus or
axitinib have also demonstrated favorable results. However,
attention should also be paid to cardiac adverse events In Cox analysis, entropy reached significance (p = 0.03) for
OS prediction. In accordance with the findings for the entire
cohort, the RR of cancer-related death after PRRT was 2.89 (CI
0.8–10.44; p = 0.1) for the G2 subgroup (Fig. 1, Supplementary
Table c in ESM). The parameter intensity variation [13]
trended to be significant in Cox analysis (p = 0.05). For both the entire cohort as well as the G2 subgroup,
results for ROC and Cox analyses of investigated parameters
are presented in Supplementary Table b in ESM. Werner R.A. et al.: Heterogeneity in pNET Undergoing PRRT 587 Table 3. Results of Kaplan–Meier analysis for overall survival (OS) for the entire cohort (n = 31) and G2 neuroendocrine tumors (NET, n = 23) for the
textural feature entropy. Asterisk denotes statistically significant
ξ2
p value
HR negative
CI
HR positive
CI
Entire cohort (n = 31)
4.03
0.045*
0.32
0.11–0.97
3.12
1.04–9.38
G2 NET (n = 23)
3.24
0.072
0.27
0.08–0.89
3.60
1.13–12.10
CI confidence interval, HR hazard ratio Table 3. Results of Kaplan–Meier analysis for overall survival (OS) for the entire cohort (n = 31) and G2 neuroendocrine tumors (NET, n = 23) for the
textural feature entropy. Asterisk denotes statistically significant
ξ2
p value
HR negative
CI
HR positive
CI
Entire cohort (n = 31)
4.03
0.045*
0.32
0.11–0.97
3.12
1.04–9.38
G2 NET (n = 23)
3.24
0.072
0.27
0.08–0.89
3.60
1.13–12.10
CI confidence interval, HR hazard ratio CI confidence interval, HR hazard ratio scan in advanced pNET patients treated with [177Lu]DOTA-
TATE was linked to a better outcome after PRRT; however,
[18F]FDG PET is not routinely assessed in treatment
planning [40]. (grade 3/4 hypertension) leading to axitinib withdrawal in
20 % of patients [6, 36]. Hence, novel risk stratification
approaches for this tumor entity might be intensively sought
for: As demonstrated in this study, imaging-based survival
prediction using TF analysis might be helpful to differentiate
between low-risk and high-risk groups. Of note, entropy
reached significance in all three statistical tests (ROC,
Kaplan–Meier, Cox analysis) emphasizing its potential in
response prediction independent of other investigated vari-
ables at least for the entire cohort. However, clinical
implications have to be made with extreme caution, as the
herein presented findings should rather be interpreted as a
Bproof-of-concept^ and further research in larger, more
homogeneous cohorts is definitely warranted. Higher entropy values are related to superior outcome in
our study. A multivariate Cox analysis corroborated these
findings: the RR for cancer-related death for those patients
below the ROC-derived threshold of entropy indicated an
almost threefold increased mortality compared to that for the
low-risk group (Fig. 1). These results are contrary to
findings in [18F]FDG PET studies investigating TF in
pancreatic ductal carcinoma or non-small cell lung cancer
(low entropy associated with longer OS) [41, 42]. Under-
standably, results from [18F]FDG PET in highly metaboli-
cally active tumors cannot be directly transferred to SSTR-
PET; however, these findings emphasize the value of tumor
heterogeneity assessment. Discussion This is the first study to assess intratumoral heterogeneity as
a risk stratification tool for pNET patients scheduled for
PRRT. Entropy, reflecting derangement on a voxel-by-voxel
level, outperformed standard conventional PET parameters
in prognostication. These findings were further corroborated
in a G2 subanalysis. However, this group per se includes a Fig. 1. Relative risk (RR) charts with 95 % confidence interval (CI) using the ROC-derived threshold (Table 2) of entropy for
overall survival. a Entire cohort (n = 31) and b G2 neuroendocrine tumor subgroup (n = 23). When the RR is exactly 1, the risk is
unchanged. For those patients below the ROC-derived threshold of entropy, the RR of cancer-related death after PRRT
increases (indicating worse outcome, applies to both the entire cohort and the G2 subcohort). Asterisk denotes statistically
significant. Fig. 1. Relative risk (RR) charts with 95 % confidence interval (CI) using the ROC-derived threshold (Table 2) of entropy for
overall survival. a Entire cohort (n = 31) and b G2 neuroendocrine tumor subgroup (n = 23). When the RR is exactly 1, the risk is
unchanged. For those patients below the ROC-derived threshold of entropy, the RR of cancer-related death after PRRT
increases (indicating worse outcome, applies to both the entire cohort and the G2 subcohort). Asterisk denotes statistically
significant. Fig. 1. Relative risk (RR) charts with 95 % confidence interval (CI) using the ROC-derived threshold (Table 2) of entropy for
overall survival. a Entire cohort (n = 31) and b G2 neuroendocrine tumor subgroup (n = 23). When the RR is exactly 1, the risk is
unchanged. For those patients below the ROC-derived threshold of entropy, the RR of cancer-related death after PRRT
increases (indicating worse outcome, applies to both the entire cohort and the G2 subcohort). Asterisk denotes statistically
significant. Werner R.A. et al.: Heterogeneity in pNET Undergoing PRRT have recently reported on the
predictive role of asphericity in GEP-NET patients sched-
uled for PRRT: a higher level of asphericity was associated
with poorer outcome. However, compared to the present
study investigating SSTR-PET, heterogeneity parameters
were derived from [111In-DTPA0]octreotide scintigraphy
[44], which has a lower affinity to SSTR2A compared to
its PET counterparts [45]. Moreover, entropy and asphericity
differ in their equations, which also serves as a possible
explanation for the different results: The latter one quanti-
tates the deviation from spherical of the shape of the MTV,
and it takes both the mean surface S and the mean volume V
into account [44, 46]. On the contrary, entropy considers I as
the voxel value in the ROI and P(I) as the probability of the
occurrence of that pixel value [47]. Apart from that, in the
present study, SSTR-PET had been used, while Wetz and
co-workers performed a SPECT approach [44]. Taken
together, the exact association between the herein presented
imaging-derived analysis of tumor lesion texture and the
underlying tumor biology must be further determined in
prospective, longitudinal studies. this reflects a typical clinical situation. Furthermore, in
our cohort, the number of treatment cycles ranged from 1
to 6; however, the majority of the cases (70 %) received
at least four radiopeptide administrations (median
4 cycles). Changes of imaging-derived parameters be-
tween subsequent scans might be also of prognostic value;
however, functional follow-up imaging was not available
in every patient. Moreover, in only 70 % of the cases, at
least four lesions could be manually segmented and
metastases with MTV smaller than 10 cm3 were not
considered. Ki67 is prone to sampling bias as well as
particular inaccuracy regarding the time lag of assessment
and subsequent SSTR-PET scan. Due to the different
affinities to SSTR2A, the use of [68Ga]DOTATATE/-TOC
might have also led to a certain data bias [45]. In
accordance with the reported robustness of certain TF
published in [25], a pre-selection of heterogeneity param-
eters has been performed in the present study. Hence, a
limited number of TF had been investigated, and
therefore, no correction for p values was applied to adjust
for multiple tests, but such procedures could be subject of
future studies [50]. Consequently, as a Bonferroni correc-
tion had not been applied, the herein derived findings
must be interpreted with caution. Werner R.A. et al.: Heterogeneity in pNET Undergoing PRRT g
y
Analyzing pre-therapeutic [68Ga]DOTATOC scans of
liver metastases in pNET patients scheduled for [Y-90/Lu-
177] treatment, a SUVmax threshold of 9 16.4 for achieving
radiologic response was proposed [37]. In our study, a
cohort treated with the less nephrotoxic and more common
Lu-177 was enrolled [38]. However, comparative thresholds
were reached, but the SUV was not significant in our
analysis. Similar to our findings, Gabriel also reported that
the SUV profile of a baseline [68Ga]DOTATOC PET does
not add additional information for response prediction in
PRRT patients [39]. Sansovini et al. have recently proven
that a negative 2-deoxy-2-[18F]fluoro-D-glucose (FDG) PET The value of entropy in patients with esophageal cancer
undergoing RTx has been recently evaluated. Although
responders were associated with greater local heterogeneity
than non-responders, responders presented lower entropy
values [33]. The response of NET tumor tissue to radiation
exposure, however, might vary [43], and the included
patients in the present study were heavily pre-treated with
CTx and RTx (30 %), which could also have a certain
impact on the SSTR fluctuations on the tumor cell surface. Fig. 2. Kaplan–Meier plots and number-at-risk tables for the probability of overall survival. a Entire cohort, n = 31, and b G2
neuroendocrine tumor subgroup, n = 23. Low-risk group (solid lines) was identified by various textural parameters measured on
somatostatin receptor-positron emission tomography/computed tomography (SSTR-PET/CT) before peptide receptor
radionuclide therapy. Cutoff values derived by receiver operating characteristics (ROC) analysis were used (Table 2). Only
entropy was significant in both ROC and Cox analysis; d days. b bilit
f
ll
i
l
E ti
h
t
31
d b G2 Fig. 2. Kaplan–Meier plots and number-at-risk tables for the probability of overall survival. a Entire cohort, n = 31, and b G2
neuroendocrine tumor subgroup, n = 23. Low-risk group (solid lines) was identified by various textural parameters measured on
somatostatin receptor-positron emission tomography/computed tomography (SSTR-PET/CT) before peptide receptor
radionuclide therapy. Cutoff values derived by receiver operating characteristics (ROC) analysis were used (Table 2). Only
entropy was significant in both ROC and Cox analysis; d days. Werner R.A. et al.: Heterogeneity in pNET Undergoing PRRT 588 Moreover, as intratumoral phenotypic heterogeneity is
frequently observed in NET even between synchronous or
metachronous metastases, no attempt was performed to
correlate these histopathological findings with patient out-
come [15]. Wetz et al. Werner R.A. et al.: Heterogeneity in pNET Undergoing PRRT Moreover, a more
homogenous study setting might strengthen our prelimi-
nary findings, in particular by enrolling a larger, prospec-
tive cohort using the same scanners and without variances
in the imaging protocol. In contrast to previous findings, hepatic tumor burden did
not turn out as an independent survival predictor, mainly
since almost the entire cohort of our patients (97 %) suffered
from liver metastases [48]. Of note, Ki67 did not correlate
significantly with the investigated heterogeneity parameters
emphasizing their independence. Correlating the prolifera-
tion index with outcome variables, Ki67 demonstrated its
potential in PFS prediction but failed for OS. This might be
also caused by sampling variability, as needle biopsies are
typically not guided to regions with a higher proliferative
rate [15]. Conflict of Interest RB has a non-commercial research contract with Mediso Medical Imaging
Systems; RB is on the speaker’s bureau for Mediso Medical Imaging
Systems and consultant for Bayer HealthCare. Mediso Medical Imaging
Systems employ NZ; LP was employed by Mediso Medical Imaging
Systems while largest parts of the study were performed. All other authors
declare that they have no conflict of interest. Conclusion As demonstrated in this multicenter study, application of
entropy as obtained by baseline SSTR-PET might be useful
for differentiating high-risk from low-risk groups in pNET
patients scheduled for PRRT. This multicenter analysis has several limitations. First,
only a limited number of patients could be included in
this study, even though pNETs have a low annual
incidence [2]. Additionally, imaging protocols differ
slightly from center to center, including various PET
reconstruction algorithms and different used PET scan-
ners. Moreover, no harmonization between the used
Biograph 2 and 64 PET has been performed. Compared
to previous investigations [37], this might explain why the
SUVmax did not turn out as a significant predictive
parameter and other conventional PET parameters, like
SUVpeak, could be subject of future studies. The herein
stated OS for pNET patients under PRRT is significantly
lower as described previously [16, 49]: However, it
remains a matter of debate if the OS should be defined
from diagnosis and treatment initiation or from baseline
SSTR-PET. Moreover, therapeutic algorithms might also
vary between centers, as the treating nuclear medicine
physician has to adjust treatment planning under current
circumstances (e.g., due to renal impairment); nonetheless, Acknowledgments. We express our gratitude to Michaela Mooz (Department
of Nuclear Medicine, University Hospital Würzburg) for her assistance in
data collection. This project has received funding from the European
Union’s Horizon 2020 research and innovation programme under the Marie
Sklodowska-Curie grant agreement No 701983. References Lancet 387:968–977 g
(
)
21. Kwekkeboom DJ, Krenning EP (2002) Somatostatin receptor imag-
ing. Semin Nucl Med 32:84–91 22. Surti S, Kuhn A, Werner ME, Perkins AE, Kolthammer J, Karp JS
(2007) Performance of Philips Gemini TF PET/CT scanner with
special consideration for its time-of-flight imaging capabilities. J Nucl
Med 48:471–480 23. Marti-Climent JM, Prieto E, Dominguez-Prado I et al (2013)
Contribution of time of flight and point spread function modeling to
the performance characteristics of the PET/CT Biograph mCT
scanner. Rev Esp Med Nucl Imagen Mol 32:13–21 7. Raymond E, Dahan L, Raoul JL, Bang YJ, Borbath I, Lombard-Bohas
C, Valle J, Metrakos P, Smith D, Vinik A, Chen JS, Hörsch D,
Hammel P, Wiedenmann B, van Cutsem E, Patyna S, Lu DR,
Blanckmeister C, Chao R, Ruszniewski P (2011) Sunitinib malate for
the treatment of pancreatic neuroendocrine tumors. N Engl J Med
364:501–513 24. Prieto E, Marti-Climent JM, Arbizu J et al (2010) Evaluation of spatial
resolution of a PET scanner through the simulation and experimental
measurement of the recovery coefficient. Comput Biol Med 40:75–80 8. Strosberg J, El-Haddad G, Wolin E et al (2017) Phase 3 trial of
177Lu-dotatate for midgut neuroendocrine tumors. N Engl J Med
376:125–135 25. Bailly C, Bodet-Milin C, Couespel S, Necib H, Kraeber-Bodéré F,
Ansquer C, Carlier T (2016) Revisiting the robustness of PET-based
textural features in the context of multi-centric trials. PLoS One
11:e0159984 9. Brabander T, van der Zwan WA, Teunissen JJM, Kam BLR, Feelders
RA, de Herder WW, van Eijck CHJ, Franssen GJH, Krenning EP,
Kwekkeboom DJ (2017) Long-term efficacy, survival, and safety of
[177Lu-DOTA(0),Tyr(3)]octreotate in patients with gastroentero-
pancreatic and bronchial neuroendocrine tumors. Clin Cancer Res
23:4617–4624 26. Nakajo M, Jinguji M, Nakabeppu Y, Nakajo M, Higashi R, Fukukura
Y, Sasaki K, Uchikado Y, Natsugoe S, Yoshiura T (2017) Texture
analysis of
18F-FDG PET/CT to predict tumour response and
prognosis of patients with esophageal cancer treated by chemo-
radiotherapy. Eur J Nucl Med Mol Imaging 44:206–214 10. Aerts HJ, Velazquez ER, Leijenaar RT et al (2014) Decoding tumour
phenotype by noninvasive imaging using a quantitative radiomics
approach. Nat Commun 5:4006 27. Li X, Rowe SP, Leal JP, Gorin MA, Allaf ME, Ross AE, Pienta KJ,
Lodge MA, Pomper MG (2017) Semiquantitative parameters in
PSMA-targeted PET imaging with
18F-DCFPyL: variability in
normal-organ uptake. J Nucl Med 58:942–946 11. References 17. Ezziddin S, Opitz M, Attassi M, Biermann K, Sabet A, Guhlke S,
Brockmann H, Willinek W, Wardelmann E, Biersack HJ, Ahmadza-
dehfar H (2011) Impact of the Ki-67 proliferation index on response to
peptide receptor radionuclide therapy. Eur J Nucl Med Mol Imaging
38:459–466 1. Fraenkel M, Faggiano A, Valk GD (2015) Epidemiology of
neuroendocrine tumors. Front Horm Res 44:1–23 2. Lawrence B, Gustafsson BI, Chan A et al (2011) The epidemiology of
gastroenteropancreatic neuroendocrine tumors. Endocrinol Metab Clin
N Am 40:1–18 vii 18. Bodei L, Mueller-Brand J, Baum RP, Pavel ME, Hörsch D, O'Dorisio
MS, O'Dorisio TM, Howe JR, Cremonesi M, Kwekkeboom DJ,
Zaknun JJ (2013) The joint IAEA, EANM, and SNMMI practical
guidance on peptide receptor radionuclide therapy (PRRNT) in
neuroendocrine tumours. Eur J Nucl Med Mol Imaging 40:800–816 3. Cauley CE, Pitt HA, Ziegler KM, Nakeeb A, Schmidt CM, Zyromski
NJ, House MG, Lillemoe KD (2012) Pancreatic enucleation:
improved outcomes compared to resection. J Gastrointest Surg
16:1347–1353 19. Kwekkeboom DJ, de Herder WW, Kam BL, van Eijck CH, van Essen
M, Kooij PP, Feelders RA, van Aken MO, Krenning EP (2008)
Treatment with the radiolabeled somatostatin analog [177Lu-DOTA
0,Tyr3]octreotate: toxicity, efficacy, and survival. J Clin Oncol
26:2124–2130 4. Panzuto F, Di Fonzo M, Iannicelli E et al (2006) Long-term clinical
outcome of somatostatin analogues for treatment of progressive,
metastatic, well-differentiated entero-pancreatic endocrine carcinoma. Ann Oncol 17:461–466 5. Rinke A, Muller HH, Schade-Brittinger C et al (2009) Placebo-
controlled, double-blind, prospective, randomized study on the effect
of octreotide LAR in the control of tumor growth in patients with
metastatic neuroendocrine midgut tumors: a report from the PROMID
Study Group. J Clin Oncol 27:4656–4663 20. Eisenhauer EA, Therasse P, Bogaerts J, Schwartz LH, Sargent D, Ford
R, Dancey J, Arbuck S, Gwyther S, Mooney M, Rubinstein L,
Shankar L, Dodd L, Kaplan R, Lacombe D, Verweij J (2009) New
response evaluation criteria in solid tumours: revised RECIST
guideline (version 1.1). Eur J Cancer 45:228–247 6. Yao JC, Fazio N, Singh S, Buzzoni R, Carnaghi C, Wolin E, Tomasek
J, Raderer M, Lahner H, Voi M, Pacaud LB, Rouyrre N, Sachs C,
Valle JW, Fave GD, van Cutsem E, Tesselaar M, Shimada Y, Oh DY,
Strosberg J, Kulke MH, Pavel ME (2016) Everolimus for the
treatment of advanced, non-functional neuroendocrine tumours of
the lung or gastrointestinal tract (RADIANT-4): a randomised,
placebo-controlled, phase 3 study. Informed Consent 14. Yachida S, Jones S, Bozic I, Antal T, Leary R, Fu B, Kamiyama M,
Hruban RH, Eshleman JR, Nowak MA, Velculescu VE, Kinzler KW,
Vogelstein B, Iacobuzio-Donahue CA (2010) Distant metastasis
occurs late during the genetic evolution of pancreatic cancer. Nature
467:1114–1117 Informed consent was obtained from all individual participants included in
the study. 15. Couvelard A, Deschamps L, Ravaud P, Baron G, Sauvanet A, Hentic
O, Colnot N, Paradis V, Belghiti J, Bedossa P, Ruszniewski P (2009)
Heterogeneity of tumor prognostic markers: a reproducibility study
applied to liver metastases of pancreatic endocrine tumors. Mod
Pathol 22:273–281 Open Access This article is distributed under the terms of the Creative
Commons Attribution 4.0 International License (http://creativecommon-
s.org/licenses/by/4.0/), which permits unrestricted use, distribution, and
reproduction in any medium, provided you give appropriate credit to the
original author(s) and the source, provide a link to the Creative Commons
license, and indicate if changes were made. 16. Ezziddin S, Khalaf F, Vanezi M, Haslerud T, Mayer K, al Zreiqat A,
Willinek W, Biersack HJ, Sabet A (2014) Outcome of peptide
receptor radionuclide therapy with 177Lu-octreotate in advanced grade
1/2 pancreatic neuroendocrine tumours. Eur J Nucl Med Mol Imaging
41:925–933 license, and indicate if changes were made. Statement of Human and Animal Rights All procedures performed in studies involving human participants were in
accordance with the ethical standards of the institutional and/or national
research committee and with the 1964 Helsinki declaration and its later
amendments or comparable ethical standards. For this type of study, formal
consent is not required. This article does not contain any studies with
animals performed by any of the authors. 589 Werner R.A. et al.: Heterogeneity in pNET Undergoing PRRT Werner R.A. et al.: Heterogeneity in pNET Undergoing PRRT FDG uptake predicts survival in patients with pancreatic ductal
adenocarcinoma. Eur J Nucl Med Mol Imaging 43:1461–1468 31. Komaki K, Sano N, Tangoku A (2006) Problems in histological
grading of malignancy and its clinical significance in patients with
operable breast cancer. Breast Cancer 13:249–253 42. Pyka T, Bundschuh RA, Andratschke N, Mayer B, Specht HM, Papp
L, Zsótér N, Essler M (2015) Textural features in pre-treatment [F18]-
FDG-PET/CT are correlated with risk of local recurrence and disease-
specific survival in early stage NSCLC patients receiving primary
stereotactic radiation therapy. Radiat Oncol 10:100 32. 32. Bundschuh RA, Dinges J, Neumann L, Seyfried M, Zsoter N, Papp L,
Rosenberg R, Becker K, Astner ST, Henninger M, Herrmann K,
Ziegler SI, Schwaiger M, Essler M (2014) Textural parameters of
tumor heterogeneity in
18F-FDG PET/CT for therapy response
assessment and prognosis in patients with locally advanced rectal
cancer. J Nucl Med 55:891–897 43. Koukouraki S, Strauss LG, Georgoulias V, Schuhmacher J, Haberkorn
U, Karkavitsas N, Dimitrakopoulou-Strauss A (2006) Evaluation of
the pharmacokinetics of 68Ga-DOTATOC in patients with metastatic
neuroendocrine tumours scheduled for 90Y-DOTATOC therapy. Eur J
Nucl Med Mol Imaging 33:460–466 33. Tixier F, Le Rest CC, Hatt M et al (2011) Intratumor heterogeneity
characterized by textural features on baseline 18F-FDG PET images
predicts response to concomitant radiochemotherapy in esophageal
cancer. J Nucl Med 52:369–378
18 44. Wetz C, Apostolova I, Steffen IG, Hofheinz F, Furth C, Kupitz D, Ruf
J, Venerito M, Klose S, Amthauer H (2017) Predictive value of
asphericity in pretherapeutic [111In]DTPA-octreotide SPECT/CT for
response to peptide receptor radionuclide therapy with [177Lu]DOTA-
TATE. Mol Imaging Biol 19:437–445 34. Hatt M, Visvikis D, Pradier O, Cheze-le Rest C (2011) Baseline 18F-
FDG PET image-derived parameters for therapy response prediction
in oesophageal cancer. Eur J Nucl Med Mol Imaging 38:1595–1606 35. Hicks RJ, Kwekkeboom DJ, Krenning E, et al. (2017) ENETS
consensus guidelines for the standards of care in neuroendocrine
neoplasia: peptide receptor radionuclide therapy with radiolabeled
somatostatin analogues. Neuroendocrinology 105(3):295-309 45. Garcia-Carbonero R, Garcia-Figueiras R, Carmona-Bayonas A et al
(2015) Imaging approaches to assess the therapeutic response of
gastroenteropancreatic neuroendocrine tumors (GEP-NETs): current
perspectives and future trends of an exciting field in development. Cancer Metastasis Rev 34:823–842 36. Werner R.A. et al.: Heterogeneity in pNET Undergoing PRRT Strosberg JR, Cives M, Hwang J, Weber T, Nickerson M, Atreya CE,
Venook A, Kelley RK, Valone T, Morse B, Coppola D, Bergsland EK
(2016) A phase II study of axitinib in advanced neuroendocrine
tumors. Endocr Relat Cancer 23:411–418 46. Apostolova I, Ego K, Steffen IG, Buchert R, Wertzel H, Achenbach
HJ, Riedel S, Schreiber J, Schultz M, Furth C, Derlin T, Amthauer H,
Hofheinz F, Kalinski T (2016) The asphericity of the metabolic
tumour volume in NSCLC: correlation with histopathology and
molecular markers. Eur J Nucl Med Mol Imaging 43:2360–2373 37. Kratochwil C, Stefanova M, Mavriopoulou E, Holland-Letz T,
Dimitrakopoulou-Strauss A, Afshar-Oromieh A, Mier W, Haberkorn
U, Giesel FL (2015) SUV of [68Ga]DOTATOC-PET/CT predicts
response probability of PRRT in neuroendocrine tumors. Mol Imaging
Biol 17:313–318 47. Cook GJ, O'Brien ME, Siddique M et al (2015) Non-small cell lung
cancer treated with erlotinib: heterogeneity of 18F-FDG uptake at
PET-association with treatment response and prognosis. Radiology
276:883–893 38. Bergsma H, Konijnenberg MW, van der Zwan WA, Kam BLR,
Teunissen JJM, Kooij PP, Mauff KAL, Krenning EP, Kwekkeboom
DJ (2016) Nephrotoxicity after PRRT with 177Lu-DOTA-octreotate. Eur J Nucl Med Mol Imaging 43:1802–1811 48. Ezziddin S, Attassi M, Yong-Hing CJ, Ahmadzadehfar H, Willinek
W, Grunwald F, Guhlke S, Biersack HJ, Sabet A (2014) Predictors of
long-term outcome in patients with well-differentiated gastroenter-
opancreatic neuroendocrine tumors after peptide receptor radionuclide
therapy with 177Lu-octreotate. J Nucl Med 55:183–190 39. Gabriel M, Oberauer A, Dobrozemsky G, Decristoforo C, Putzer D,
Kendler D, Uprimny C, Kovacs P, Bale R, Virgolini IJ (2009) 68Ga-
DOTA-Tyr3-octreotide PET for assessing response to somatostatin-
receptor-mediated radionuclide therapy. J Nucl Med 50:1427–1434 49. Sharma N, Naraev BG, Engelman EG, Zimmerman MB, Bushnell DL
Jr, OʼDorisio TM, OʼDorisio MS, Menda Y, Müller-Brand J, Howe
JR, Halfdanarson TR (2017) Peptide receptor radionuclide therapy
outcomes in a North American cohort with metastatic well-
differentiated neuroendocrine tumors. Pancreas 46:151–156 40. Sansovini M, Severi S, Ianniello A, Nicolini S, Fantini L, Mezzenga
E, Ferroni F, Scarpi E, Monti M, Bongiovanni A, Cingarlini S, Grana
CM, Bodei L, Paganelli G (2017) Long-term follow-up and role of
FDG PET in advanced pancreatic neuroendocrine patients treated with
177Lu-D OTATATE. Eur J Nucl Med Mol Imaging 44:490–499 50. Chalkidou A, O'Doherty MJ, Marsden PK (2015) False discovery
rates in PET and CT studies with texture features: a systematic review. PLoS One 10:e0124165 177Lu-D OTATATE. References Marusyk A, Almendro V, Polyak K (2012) Intra-tumour heterogene-
ity: a looking glass for cancer? Nat Rev Cancer 12:323–334 28. Hatt M, Majdoub M, Vallieres M, Tixier F, le Rest CC, Groheux D,
Hindie E, Martineau A, Pradier O, Hustinx R, Perdrisot R, Guillevin
R, el Naqa I, Visvikis D (2015) F-18-FDG PET uptake characteriza-
tion through texture analysis: investigating the complementary nature
of heterogeneity and functional tumor volume in a multi-cancer site
patient cohort. J Nucl Med 56:38–44 12. Oberg K, Krenning E, Sundin A, Bodei L, Kidd M, Tesselaar M,
Ambrosini V, Baum RP, Kulke M, Pavel M, Cwikla J, Drozdov I,
Falconi M, Fazio N, Frilling A, Jensen R, Koopmans K, Korse T,
Kwekkeboom D, Maecke H, Paganelli G, Salazar R, Severi S,
Strosberg J, Prasad V, Scarpa A, Grossman A, Walenkamp A, Cives
M, Virgolini I, Kjaer A, Modlin IM (2016) A Delphic consensus
assessment: imaging and biomarkers in gastroenteropancreatic neuro-
endocrine tumor disease management. Endocr Connect 5:174–187 29. Youden WJ (1950) Index for rating diagnostic tests. Cancer 3:32 35
30. Hachamovitch R, Di Carli MF (2008) Methods and limitations of
assessing new noninvasive tests: part II: outcomes-based validation
and reliability assessment of noninvasive testing. Circulation
117:2793–2801 13. Werner RA, Lapa C, Ilhan H, Higuchi T, Buck AK, Lehner S,
Bartenstein P, Bengel F, Schatka I, Muegge DO, Papp L, Zsótér N,
Große-Ophoff T, Essler M, Bundschuh RA (2017) Survival prediction
in patients undergoing radionuclide therapy based on intratumoral
somatostatin-receptor heterogeneity. Oncotarget 8:7039–7049 590 Werner R.A. et al.: Heterogeneity in pNET Undergoing PRRT Eur J Nucl Med Mol Imaging 44:490– 41. Hyun SH, Kim HS, Choi SH, Choi DW, Lee JK, Lee KH, Park JO,
Lee KH, Kim BT, Choi JY (2016) Intratumoral heterogeneity of 18F-
|
https://openalex.org/W4387743919
|
https://scipost.org/10.21468/SciPostPhysCore.6.4.068/pdf
|
English
| null |
Quasi-characters in $\widehat{su(2)}$ current algebra at fractional levels
|
SciPost physics core
| 2,023
|
cc-by
| 21,799
|
Abstract We study the even characters of Ø
su(2) conformal field theories (CFTs) at admissible
fractional levels obtained from the difference of the highest weight characters in the
unflavoured limit. We show that admissible even character vectors arise only in three
special classes of admissible fractional levels which include the threshold levels, the pos-
itive half-odd integer levels, and the isolated level at -5/4. Among them, we show that
the even characters of the half-odd integer levels map to the difference of characters of
Ø
su(2)4N+4, with N ∈Z>0, although we prove that they do not correspond to rational
CFTs. The isolated level characters maps to characters of two subsectors with ×
so(5)1
and Ø
su(2)1 current algebras. Furthermore, for the Ø
su(2)1 subsector of the isolated level,
we introduce discrete flavour fugacities. The threshold levels saturate the admissibility
bound and their even characters have previously been shown to be proportional to the
unflavoured characters of integrable representations in Ø
su(2)4N CFTs, where N ∈Z>0
and we reaffirm this result. Except at the three classes of fractional levels, we find spe-
cial inadmissible characters called quasi-characters which are nice vector valued modu-
lar functions but with q-series coefficients violating positivity but not integrality. Copyright S. Grover. This work is licensed under the Creative Commons
Attribution 4.0 International License. Published by the SciPost Foundation. Received 24-01-2023
Accepted 15-09-2023
Published 18-10-2023
Check for
updates
doi:10.21468/SciPostPhysCore.6.4.068
Contents
1
Introduction
2
1.1
Summary of results and organisation
4
2
Highest weight characters of Ø
su(2) at admissible levels
5
2.1
Kaˇc-Wakimoto highest weight characters
6
2.2
Even and odd characters
8
2.3
Unflavoured even characters
11
3
Admissible levels with admissible even character vector
12
3.1
Threshold levels
12
3.2
Curious case of the isolated level m = −5/4 (p = 3,u = 4)
14
3.3
Positive half-integer levels
18
1 Copyright S. Grover. This work is licensed under the Creative Commons
Attribution 4.0 International License. Published by the SciPost Foundation. Received 24-01-2023
Accepted 15-09-2023
Published 18-10-2023
Check for
updates
doi:10.21468/SciPostPhysCore.6.4.068 Quasi-characters in Ø
su(2) current algebra at fractional levels Sachin Grover
Harish-Chandra Research Institute, Chhatnag Road, Jhunsi, Allahabad, India-211019
Homi Bhabha National Institute, Training School Complex,
Anushaktinagar, Mumbai, India-400094 ⋆sachingrover@hri.res.in SciPost Phys. Core 6, 068 (2023) 1The notion of admissibility is extended from the notion of highest weight modules of the vertex operator
algebra and the category O as defined in [10] to the category with relaxed-highest weight modules and its spectral
flows. The definitions can be seen in [11–13]. Contents 1
Introduction
2
1.1
Summary of results and organisation
4
2
Highest weight characters of Ø
su(2) at admissible levels
5
2.1
Kaˇc-Wakimoto highest weight characters
6
2.2
Even and odd characters
8
2.3
Unflavoured even characters
11
3
Admissible levels with admissible even character vector
12
3.1
Threshold levels
12
3.2
Curious case of the isolated level m = −5/4 (p = 3,u = 4)
14
3.3
Positive half-integer levels
18 1
Introduction
2
1.1
Summary of results and organisation
4
2
Highest weight characters of Ø
su(2) at admissible levels
5
2.1
Kaˇc-Wakimoto highest weight characters
6
2.2
Even and odd characters
8
2.3
Unflavoured even characters
11
3
Admissible levels with admissible even character vector
12
3.1
Threshold levels
12
3.2
Curious case of the isolated level m = −5/4 (p = 3,u = 4)
14
3.3
Positive half-integer levels
18 1 SciPost Phys. Core 6, 068 (2023) 4
Conclusions and Discussion
22
A
Quasi-characters examples
23
A.1
n = 0, k = ⌈u/p⌉
23
A.2
Multiple quasi-characters
24
B
Asymptotic behaviour of q-series
25
C
Closed subsets without quasi-characters? 26
C.1
First criterion
26
C.2
Second criterion
29
D
Supplementary table
31
References
35 4
Conclusions and Discussion
22
A
Quasi-characters examples
23
A.1
n = 0, k = ⌈u/p⌉
23
A.2
Multiple quasi-characters
24
B
Asymptotic behaviour of q-series
25
C
Closed subsets without quasi-characters? 26
C.1
First criterion
26
C.2
Second criterion
29
D
Supplementary table
31
References
35 35 1
Introduction The integer level WZW models have a finite number of irreducible and unitary
integrable highest-weight representations (primaries) of the current algebra that close under
fusion, with positive semi-definite integer Verlinde fusion coefficients. Notably, the finite num-
ber of characters of the integrable representations compose the modular invariant partition
function with positive semi-definite and finite integer coefficients.2 rational CFTs. The integer level WZW models have a finite number of irreducible and unitary
integrable highest-weight representations (primaries) of the current algebra that close under
fusion, with positive semi-definite integer Verlinde fusion coefficients. Notably, the finite num-
ber of characters of the integrable representations compose the modular invariant partition
function with positive semi-definite and finite integer coefficients.2 The character of Ö
su(2) at level m in spin-j representation3 is a vector valued modular
function which is represented as χj(τ,z) = TrRjqL0−c/24ωJ0
0 ,
(1) (1) with q = e2πiτ, where τ ∈H and the flavour fugacity ω = e2πiz, where ω ∈u(1) corre-
sponds to the Cartan of su(2) denoted by J0
0. The characters (1) do not have a well-defined
unflavoured limit ω →1 for arbitrary rational spins, i.e., spins which are not integers or half-
odd integers. The highest weight characters of the fractional level Ö
su(2) theories are given by
Kaˇc-Wakimoto [2]. Although the highest weight characters are linearly independent, they do
not form a basis of characters for the extended category of admissible modules, which includes
the relaxed category along with the indecomposables as defined in [11,13]. Moreover, the un-
flavoured limit of the highest weight characters with rational spins which are not half-integers,
obtained by setting the flavour fugacities to one is not a well-defined limit due to the presence
of poles as we will witness in the next section. It was first observed in [22] that the difference
of the characters of the highest weight primaries with equal conformal dimensions, which we
will refer to as the even characters, admits a well-defined unflavoured limit or q-expansion. 1
Introduction The conformal field theories (CFTs) with a real fractional level Kaˇc-Moody algebra as the cur-
rent algebra have been of interest for a long time. Recently, the connection of the fractional
level theories with the unitary four-dimensional N = 2 superconformal field theories (SCFTs)
was brought to light [1]. They showed that the twisted-translated Schur sector of the four-
dimensional N = 2 superconformal field theories (SCFTs) when restricted to a 2-dimensional
plane has a current algebra generically at fractional levels. In particular, the Schur index of the
4-dimensional SCFT is the unflavoured vacuum character of a non-unitary 2-dimensional CFT
with a chiral algebra possibly at a fractional level. This result formed our initial motivation
to study the unflavoured characters of the fractional level theories. In this paper, we classify
the Ö
su(2) current algebra CFTs at admissible fractional levels by imposing conformal bootstrap
constraints on their unflavoured even characters. Ö Fractional levels in the context of Ö
su(2) current algebra, here and from now on refer to the
admissible fractional levels defined by Kaˇc and Wakimoto [2], parametrised as m = p/u −2
with p,u ∈Z≥2 and coprime, as admissibility implies p ≥2. At these levels, the highest
weight modular invariant representations are finite-dimensional representations of the mod-
ular group SL(2,Z). Since they form a finite-dimensional representation, the fractional level
theories were a candidate for rational CFTs (RCFTs). However, it was shown in [3] that the
corresponding RCFT Verlinde fusion rules [4] do not generate positive semi-definite fusion
coefficients. Many CFT proposals with Ö
su(2) current algebra at fractional levels were sug-
gested but could not resolve all the issues plaguing the description. See for example [5–9]. The resolution of the puzzle lies in the fact that the 2-dimensional CFT with current algebra
at a fractional level, not only have a finite number of highest weight modules but also an
infinite number of admissible modules,1 obtained by spectral flow automorphisms with an un-
bounded spectrum, and additionally, it has indecomposable modules. Thus, these theories are
logarithmic CFTs (log CFTs) [11–19]. Contrast this situation to the theories at positive integer
levels. At positive integer levels, the Wess-Zumino-Witten (WZW) models can be described as 2 SciPost Phys. Core 6, 068 (2023) rational CFTs. 4Mathur, Mukhi, and Sen (MMS) [23,24] developed a classification program for RCFTs in terms of the number
of Kaˇc-Moody characters, r, of the RCFT without specifying the Kaˇc-Moody algebra. The unflavoured Kaˇc-Moody
characters satisfy an order-r linear differential equation. See [25–29] for some recent applications of the MLDE
approach to RCFT classification, and [30] for a review. Using this approach, most notably, MMS obtained a partial
classification for the two character theories, the so-called MMS series. The two-character unitary theories with
c < 25 were completely classified very recently in [31]. 3Unless otherwise noted, representations in this context would mean irreducible highest weight represent
tions.i p
yi
y
y
[
]
5We will define admissible, inadmissible and quasi-characters in the next section. The notion of an admissib
character should not be confused with Kaˇc-Wakimoto notion of admissible representations. 6However, [26,31,32] showed that quasi-characters have more applications. Namely, certain linear combina-
tions of the quasi-characters can be used to build admissible characters by appropriately choosing a linear combina-
tion of two quasi-characters with the same modular T- transformation and appropriate modular S-transformation,
or by a multiplication of the quasi-character by a suitable power of the modular j-function. 2Note that a finite number of CFT primaries is insufficient to ensure its rationality (for an example see [20]
See [21] for a few definitions of rationality.
3 c
5
e e co
p e e y c assi
ed e y ece
y
[3 ].
5We will define admissible, inadmissible and quasi-characters in the next section. The notion of an admissible
character should not be confused with Kaˇc-Wakimoto notion of admissible representations.
6However, [26,31,32] showed that quasi-characters have more applications. Namely, certain linear combina-
tions of the quasi characters can be used to build admissible characters by appropriately choosing a linear combina 1.1
Summary of results and organisation In our study of the even characters, we show that the even unflavoured characters of Ö
su(2) CFTs
may generate a q-expansion with negative coefficients, i.e., seem to generically admit quasi-
characters. Also in the cases we consider, the quasi-characters can not be made admissible
characters by choosing a suitable linear combination of characters with the same modular T-
transformations. We thus conclude that the even characters are not necessarily equivalent to
modules of a CFT. Interestingly, at certain special admissible levels labelled by (p = 2,u = 2N + 1), or
(p = 2N + 3,u = 2), or (p = 3,u = 4), where N ∈Z≥0, we show that the even characters
do not contain any quasi-characters. The even characters are admissible if they satisfy the
admissibility conditions defined in section 2. To show the positivity and integrality, which are
necessary conditions for the admissibility of the even characters at these special levels, we
rediscover a relation between the characters of integrable representations of Ö
su(2) at level 4N
and the even characters of the threshold sequence (p = 2,u = 2N + 1),8 and discover new
relations for the half-odd integer levels (p = 2N + 1,u = 2), where N ∈Z>0,9 and in another
new relation for the isolated level (p = 3,u = 4) we show that there are two independent
subsets of unflavoured characters: a subset of characters is related to the unflavoured Ò
so(5)
characters at level 1 and another subset to Ö
su(2) characters at level 1. Although the positivity
of the even characters is not manifest at the positive half-odd integer levels (p = 2N +3,u = 2)
from the analytic relations, we have checked the positivity to a sufficiently high order ∼q2000. We use the mappings of the even characters to the unflavoured characters of integrable repre-
sentations and the available modular data to establish a new correspondence between the two
subsectors of Ö
su(2)−5/4 log CFT, i.e., (p = 3,u = 4), with Ö
su(2)1 and Ö
so(5)1 RCFTs, and prove
that the even characters from positive half-odd integer levels do not correspond to any RCFT. The organisation of the paper is as follows. In section 2 we set up the notation and de-
fine the admissible and quasi-characters. 1
Introduction These unflavoured even characters, which are the principal subject of this paper, are poised to
be the unflavoured characters of a tentative RCFT as they satisfy an (untwisted) modular lin-
ear differential equation (MLDE) with the Wronskian l/6 = 0 [22].4 Usually, not all solutions
of an MLDE are characters of an RCFT and a check is needed to weed out the characters which
are not admissible.5 In particular, not all solutions possess a q-expansion with positive inte-
gral coefficients. The procedural checks have been routinely carried out to generate admissible
characters [24–29] in a constructive program to classify the RCFTs, which is still in its infancy
except for the two-character theories. Besides proving a complete classification of all admis-
sible characters for the two-component character vectors, [26] also introduced some useful
applications of a class of inadmissible characters called quasi-characters. The quasi-characters,
although nice vector valued modular functions with integer coefficients, are not admissible
characters of any CFT because their q-expansion does not necessarily have all positive integer
coefficients. As a result, we cannot interpret them as a partition sum of any CFT.6 The equivalence between the modules of the fractional level theory and the characters is
only within the region of convergence of the characters. The highest weight characters of Ö
su(2) 3 SciPost Phys. Core 6, 068 (2023) modules have a pole at ω = 1 (or z = 0), i.e., the radius of convergence does not include the
unflavoured limit [11, 13, 17]. Nevertheless, the even linear combination, which gets rid of
the pole at ω = 1, has a radius of convergence 1 < |ω| < |q|−1 (or |q| < |ω| < 1) for |q| < 1
with a well-defined unflavoured limit. The unflavoured even characters have been widely
used in the literature [1,33–36] in the context of 4-dimensional SCFTs.7 The Schur index of a
4-dimensional SCFT is a modular object which satisfies a (twisted or untwisted) MLDE [33]. The 4-dimensional SCFTs with a Ö
su(2) chiral algebra at admissible fractional levels will have a
Schur index equal to the vacuum character, which satisfies an MLDE with l = 0 and transforms
under modular transformations to the even characters. The modular behaviour of the Schur
index is connected to the Weyl anomaly coefficient of the 4-dimensional SCFT [33]. yi
p
[
]
8The threshold levels saturates the Kaˇc-Wakimoto admissibility bound on the levels.
9i 9Note that (p = 3,u = 2) has a quasi-character first noticed in [17]. 7Note that the unitary 4-dimensional SCFT exists only if certain unitarity bounds on the central charges of th
4-dimensional theory are satisfied (see for example the lecture notes [37]). y
Note that (p = 3,u = 2) has a quasi-character first noticed in [17]. 1.1
Summary of results and organisation Then, we write the closed form expression of the
flavoured characters of the Ö
su(2) algebra, as given by Kaˇc and Wakimoto [2]. We define the
even and odd combinations of the characters [22] to extract unflavoured characters. We also
define the radius of convergence of these characters which is also the radius of convergence
of the characters of the integrable representations. In section 3 we study the even admissible
Ö
su(2) character vectors of the two infinite sequences (p = 2,u = 2N + 1), (p = 2N + 3,u = 2),
where N ∈Z>0 and the isolated level (p = 3,u = 4). We relate the characters to integrable
highest weight characters of Ö
su(2) and Ö
so(5) at positive integer levels through the process of
unitarisation, which is based on choosing the lowest dimension primary as the unique vacuum 4 SciPost Phys. Core 6, 068 (2023) for the unitary theory. See [36,38–40] for some discussion. The unflavoured characters of the
sequence (p = 2N + 1,u = 2) possess certain peculiar properties which rule them out as pos-
sible unflavoured characters of an RCFT. We will also elaborate on these properties in section
3 for the unitary theory. See [36,38–40] for some discussion. The unflavoured characters of the
sequence (p = 2N + 1,u = 2) possess certain peculiar properties which rule them out as pos-
sible unflavoured characters of an RCFT. We will also elaborate on these properties in section We conclude the paper with a summary of our results and speculations in section 4. The ap-
pendices contain some explicit details and results. A few explicit examples of quasi-characters
are given in appendix A with tabulated data on the quasi-characters in the supplementary
table D. In the appendix B we explicitly write the asymptotic behaviour of the q-expansion,
for small q, of the unflavoured characters for arbitrary values of p and u. The location of the
quasi-characters will be immediately clear from the rearranged q expansion. The results are
in complete agreement with our data on explicit q-expansions. In the last appendix C, we will
argue that the modular S-matrix of the theories with quasi-characters intertwines the quasi-
characters with the admissible characters. This leads to two criteria, based on the modular and
conformal properties of the characters, which may rule out the possibility of a closed subsector
without a quasi-character. 10See [42, 43] on integrality of characters which is closely linked to the representation of the modular group
(2,Z). ( ,
)
11The subscript ‘0’ on the characters and modular matrices will be reserved for the vacuum representation. 2
Highest weight characters of Ø
su(2) at admissible levels The flavoured characters of Ö
su(2) chiral algebra capture the full information of the module,
degeneracy for the su(2) spin and conformal dimension. The unflavoured characters, which
give the degeneracy at a grade (L0 eigenspace) in the module are particularly useful for im-
plementing bootstrap constraints. Apart from the examples mentioned in the introduction,
another interesting example is the recent bootstrap classification of the 4-dimensional rank
two SCFTs which has a Schur index satisfying a 4th-order MLDE [41].l The unflavoured characters of an RCFT are holomorphic functions in the upper half plane
τ ∈H except for the finite number of singularities at the cusp points. This condition is called
weak holomorphicity. In particular, they have the following q expansion, χj(τ) = qhj−c/24 X
i∈Z≥0
a(j)
i qi ,
q = e2πiτ ,
(2) (2) where hj is the conformal dimension or L0 eigenvalue of the highest weight state and c is
the central charge of the tentative CFT. The coefficients a(j)
i
need to satisfy certain admissibil-
ity conditions which we impose on the unflavoured characters. The admissibility conditions
are necessarily satisfied by the characters of an RCFT. We define admissible characters as the
unflavoured characters which satisfy the following character admissibility conditions. where hj is the conformal dimension or L0 eigenvalue of the highest weight state and c is
the central charge of the tentative CFT. The coefficients a(j)
i
need to satisfy certain admissibil-
ity conditions which we impose on the unflavoured characters. The admissibility conditions
are necessarily satisfied by the characters of an RCFT. We define admissible characters as the
unflavoured characters which satisfy the following character admissibility conditions. 1. Positivity and Integrality of the coefficients in the q-expansion. The coefficients in the
formal power series in q are interpreted as the degeneracy of states at an energy level
in the corresponding module.10 1. Positivity and Integrality of the coefficients in the q-expansion. The coefficients in the
formal power series in q are interpreted as the degeneracy of states at an energy level
in the corresponding module.10 2. The requirement of a unique vacuum state implies that the unflavoured vacuum charac-
ter has the form,11
( )
c/24(
(0)
(0)
2
)
( ) 2. The requirement of a unique vacuum state implies that the unflavoured vacuum charac-
ter has the form,11 χ0(τ) = q−c/24(1 + a(0)
1 q + a(0)
2 q2 + ···). l
j
13The corresponding S-matrices also need not satisfy the positivity of the fusion coefficients. 2
Highest weight characters of Ø
su(2) at admissible levels (3) (3) 3. The requirement of positive multiplicities of the characters as building blocks of the 3. The requirement of positive multiplicities of the characters as building blocks of the 5 SciPost Phys. Core 6, 068 (2023) physical modular invariant partition function, i.e., physical modular invariant partition function, i.e., Z(τ, ¯τ) =
X
i,j
¯χi(¯τ)Mi jχj(τ),
(4) (4) i,j where Mi j ∈Z≥0. where Mi j ∈Z≥0. The unflavoured characters transform under the modular S-transformations as vector valued
modular functions,12
X The unflavoured characters transform under the modular S-transformations as vector valued
modular functions,12
X χj(−1/τ) =
X
j′
Sj j′χj′(τ). (5) (5) We can only impose positivity of the Verlinde fusion coefficients as the additional requirement
on the admissible characters if the S-matrix in eqn (2.4) is unitary [24]. To put it a little differ-
ently, we generally do not expect the S and T-modular transformation matrices of eqn (2.4),
which act on the unflavoured characters (called the reduced S and T-modular transformation
matrices in [29]), to be compatible with the MTC structure, in particular, corresponding Ver-
linde fusion coefficients are not necessarily positive integers. However, a unitary S-matrix can
be obtained from the reduced S-matrix following standard procedures, which is by construc-
tion compatible with the MTC structure (see [24] for more details). In particular, the unitary
S-matrix produces positive integer Verlinde fusion coefficients, Ni jk =
X
l
SilSjl(S−1)kl
S0l
∈Z≥0 . (6) (6) The vector valued modular functions which do not satisfy the positivity of the coefficients
in the q-expansion are called quasi-characters.13 Quasi-characters routinely appear as solutions
of modular linear differential equations (MLDEs), but the quasi-characters gained popularity
due to their constructive usage in building admissible characters [26,31,32]. The characters
which are not admissible are called inadmissible characters. 12Strictly speaking, these are the ‘reduced’ S-matrices. The reduced S-matrix relates the transformation of the
flavoured characters. We will call these matrices as just S-matrices.i 2.1
Kaˇc-Wakimoto highest weight characters The admissible levels of the Ö
su(2) algebra were defined by Kaˇc and Wakimoto in [2] from the
modular properties of the characters of the admissible highest weight representations. Kaˇc
and Wakimoto showed that the admissible rational levels of the Ö
su(2)m affine Lie algebra can
be parametrised in terms of two positive integers as m = p/u−2 with p,u ∈Z≥2 and coprime. Ö The classification of all positive energy irreducible representations of the Ö
su(2) algebra at an
admissible level was done in [10] which showed that the finite number of irreducible repre-
sentations (objects) from category O are exactly modular invariant. The set of ‘admissible’
representations was extended to the relaxed highest weight modules in [17]. We will restrict
our attention to the highest weight modules in category O since they are related to integrable
modules of integer level current algebra CFTs, as we will see in the remaining paper. The
highest weight representations at level m = p/u −2 are further parametrised in terms of two
integers 1 ≤n + 1 ≤p −1 and 0 ≤k ≤u −1 as 2j(n, k) + 1 = n + 1 −kp
u ,
(7) (7) 6 SciPost Phys. Core 6, 068 (2023) where j(n, k) is the spin of the representation. The conformal dimension of the primary with
spin j(n, k) is h(n, k) = j(j + 1)u
p
= (nu −kp)2
4up
+ (nu −kp)
2p
. (8) (8) The L0 eigenvalue, (8) of the highest weight states remains invariant under the transformation
σ0 : j 7→−1 −j. This can be stated in terms of n, k as The L0 eigenvalue, (8) of the highest weight states remains invariant under the transformation
σ0 : j 7→−1 −j. This can be stated in terms of n, k as ¯n + 1 = p −(n + 1),
(9)
¯k = u −k ,
k ̸= 0,
(10) (9) (9) (10) nder which j(¯n,¯k) = −1 −j(n, k) and h(¯n,¯k) = h(n, k). The closed form expression for the Kaˇc-Wakimoto character formula [2] is given in term
of The closed form expression for the Kaˇc-Wakimoto character formula [2] is given in terms
of of
b±(n, k) = ±u(n + 1) −kp ,
a = pu. (11) b±(n, k) = ±u(n + 1) −kp ,
a = pu. (11) (11) b±(n, k) = ±u(n + 1) −kp ,
a = pu. 2.1
Kaˇc-Wakimoto highest weight characters (1 Notice, σ0 : b+ 7→−b+. The spin j character is written in terms of theta functions as Notice, σ0 : b+ 7→−b+. The spin j character is written in terms of theta functions as χj(n,k)(τ,z) =
Θb+,a(τ,z/u) −Θb−,a(τ,z/u)
Θ1,2(τ,z) −Θ−1,2(τ,z)
,
(12) (12) where where
Θb,a(τ,z) =
X
qar2ωar ,
(13) Θb,a(τ,z) =
X
r=Z+b/2a
qar2ωar ,
(13) (13) where q = e2πiτ, with τ ∈H/SL(2,Z) and the flavour fugacity is ω = e2πiz ∈u(1). We can
understand the radius of convergence of the character (12) by writing the denominator using
the Jacobi triple product identity [17], Θ1,2(τ,ω) −Θ−1,2(τ,ω) = q1/8ω1/2
∞
Y
i=1
(1 −ω−1qi−1)(1 −qi)(1 −ωqi). (14) (14) The denominator has zeroes at ω = qi, where i ∈Z. The denominator can be power series
expanded in z, The denominator has zeroes at ω = qi, where i ∈Z. The denominator can be power series
expanded in z, Θ1,2(τ,z) −Θ−1,2(τ,z) = 2πizη3(q) + O(z3). (15) (15) The unflavoured limit z →0 is not well-defined due to the pole at z = 0. To see this, we need
to expand the character in a power series of z about z = 0. Using the series form of the Θ
function, Θb+,a(τ,z) =
X
s∈Z
qas2+b2
+/4a+sb+ωas+b+/2 ,
(16) (16) in the characters and then expanding it in powers of z gives χj(n,k)(q,z) =
1
2πizη3(q)
qb2
+/4a X
s∈Z
qas2+sb+ −qb2
−/4a X
s∈Z
qas2+sb−
+ O(z)
+
1
uη3(q)
qb2
+/4a X
s∈Z
qas2+sb+(as + b+/2) −qb2
−/4a X
s∈Z
qas2+sb−(as + b−/2)
. ( (17) SciPost Phys. Core 6, 068 (2023) The pole represents the infinite number of states in each grade generated by the action of su(2)
currents which is due to the fractional su(2) weights. For k = 0, b = b+ = −b−the character
takes the form, as z →0, χj(n,0)(q) =
1
η3(q)q(n+1)2/4(m+2) X
s∈Z
qus((n+1)+us(m+2))(n + 1 + 2us(m + 2)). (18) (18) All of the characters labelled by (n,0) are admissible characters. They have half integer
spins and are denoted by Lj. 14Note that we are interested in the regions |q| < |ω| < 1 or 1 < |ω| < |q|−1 where |q| < 1. 2.1
Kaˇc-Wakimoto highest weight characters Their characters do not have singularity in the z →0 limit and in
fact, the characters are analogous to the integrable highest weight characters of Ö
su(2)M where
M ∈N [44], χ(M)
l
(q) = q(l+1)2/4(M+2)
η3(q)
X
n
l + 1 + 2n(M + 2)
qn(l+1+(M+2)n) ,
(19) (19) where we have used the notation χ(M)
l
to denote the character of spin l/2 representation of
Ö
su(2) at level M. The characters have a q-expansion (19) which corresponds directly to the
energy eigenspace degeneracies of the corresponding CFT modules. 2.2
Even and odd characters Core 6, 068 (2023) These expansions are, of course, related by the transformation ω →1/ω. We note that the
expansion (23) can be recognised with positive coefficients. Similarly, the other character with
spin j = −1/4 has the expansions χ−1/4(q,|ω| < 1) = −q−1/12 ω3/4
1 −ω
1 + q
1
ω + 1
+ O(q2)
,
(24)
χ−1/4(q,|ω| > 1) = q−1/12 ω1/4
1 −ω
1 + q (1 + ω) + O(q2)
. (25) (24) (25) From these characters, we can extract an even character, From these characters, we can extract an even character, χ+
−1/4(q,|ω| > 1) = χ−1/4(q,|ω| > 1) −χ−3/4(q,|ω| > 1)
(26)
= q−1/12 ω1/4(1 −ω1/2)
1 −ω
1 −q
1
pω + pω
+ O(q2)
,
(27) (26) (27) and χ+
−1/4(q,|ω| < 1) obtained by the applying the rule ω →1/ω to (26). The unflavoured
limit on the real axis ω →1 of eqn (26) is well-defined, χ+
−1/4(q,ω →1) = q−1/12 1
2
1 −2q + O(q2)
,
(28) (28) defined in the region |q| < 1. We can scale the character by u = 2 to obtain an integral
q-expansion which is the unflavoured even character or simply as even character, whenever
the meaning is evident. In fact, for Ö
su(2)−1/2 we only have a single linearly independent
unflavoured even character since χ+
−1/4(q,|ω| > 1) + χ+
−3/4(q,|ω| < 1) = 0,
χ+
−1/4(q,|ω| < 1) + χ+
−3/4(q,|ω| > 1) = 0. (29) (29) The above calculation illustrates how to regularise the divergent infinite series in eqn (23) to
the unflavoured limit (28). Next, we generalise the above derivation to an arbitrary Kaˇc-Wakimoto admissible level
m = p/u −2. The character of the irreducible highest weight representation with spin j(n, k)
is15 χj(n,k)(q,ω) =
P
s∈Z ωj
qas2+b2
+/4a+sb+ωas/u −qas2+b2
−/4a+sb−ωas/u−(n+1)
q1/8 Q∞
i=1(1 −ω−1qi−1)(1 −qi)(1 −ωqi)
,
(30) (30) with the character in the conjugate representation can be brought to the form with the character in the conjugate representation can be brought to the form χj(¯n,¯k)(q,ω) =
P
s∈Z ω−j−1
qas2+b2
+/4a+sb+ω−as/u −qas2+b2
−/4a+sb−ω−as/u+(n+1)
q1/8 Q∞
i=1(1 −ω−1qi−1)(1 −qi)(1 −ωqi)
. (31) (31) The even character is just the subtraction of the above two characters, (30) and (31). 2.2
Even and odd characters For the characters with k ̸= 0, we need to define even (χ+) and odd (χ−) characters which
are the following linear combinations of the highest weight characters [22], χ±
j(n,k)(q,ω) = χj(n,k)(q,ω) ∓χj(¯n,¯k)(q,ω). (20) (20) The odd characters are related to the Virasoro minimal model characters [22], thus satisfying
admissibility conditions (2), (3) as well as positivity and integrality of Verlinde fusion coeffi-
cients. The odd characters are the characters of the indecomposable admissible modules, and
the indecomposable modules are proportional to the characters of the irreducible Virasoro
modules [13]. We turn our attention to the even characters. The even combination of characters does
not possess any singularity in the ω →1, which we demonstrate below, first for an example
Ö
su(2)−1/2 theory to convince the reader before we move to the general explanation. Example. We begin by the Ö
su(2)−1/2 example parametrised with integers p = 3 and u = 2. Apart from the vacuum representation, we have two more highest weight representations with
spins j(0,1) = −3/4 and j(1,1) = −1/4. The character χj(0,1)=−3/4 obtained by using (12), is χ−3/4(q,ω) = q−1/12ω−3/4
P
s∈Z q6s2−sω3s −
P
s∈Z q6s2+1−5sω3s−1
P
s∈Z q2s2+sω2s −
P
s∈Z q2s2−sω2s−1
. Performing the polynomial division, we have an infinite series expansion, Performing the polynomial division, we have an infinite series expansion, χ−3/4(q,ω) = −q−1/12ω1/4 1 + ω + ω2 + ···
−q11/12ω1/4 1 + 2ω + 2ω2 + ···
+ O(q23/12)
(21 χ−3/4(q,ω) = −q−1/12ω1/4 1 + ω + ω2 + ···
−q11/12ω1/4 1 + 2ω + 2ω2 + ···
+ O(q23/12). (21) The ω series can be resummed in different radii of convergence to obtain a q-series with
ω dependent coefficients,14 The ω series can be resummed in different radii of convergence to obtain a q-series with
ω dependent coefficients,14 χ−3/4(q,|ω| < 1) = −q−1/12 ω1/4
1 −ω
1 + q(1 + ω) + O(q2)
,
(22)
χ−3/4(q,|ω| > 1) = q−1/12 ω3/4
1 −ω
1 + q(1 + 1
ω) + O(q2)
. (23) (22) (23) 14Note that we are interested in the regions |q| < |ω| < 1 or 1 < |ω| < |q|−1 where |q| < 1. 14Note that we are interested in the regions |q| < |ω| < 1 or 1 < |ω| < |q|−1 where |q| < 1. 8 SciPost Phys. 2.2
Even and odd characters χ+
j(¯n,¯k)(q,ω) =
P
s∈Z qas2+b2
+/4a+sb+ ωas/u+j+1 −ω−as/u−j
q1/8(ω −1)(1 −q)(1 −qω)
Q∞
i=2(1 −ω−1qi−1)(1 −qi)(1 −ωqi)
−
P
s∈Z qas2+b2
−/4a+sb− ωas/u+j+1−(n+1) −ω−as/u−j+(n+1)
q1/8(ω −1)(1 −q)(1 −qω)
Q∞
i=2(1 −ω−1qi−1)(1 −qi)(1 −ωqi)
. (32) (32) q
(
)(
q)(
q
)
Q
i=2(
q
)(
q )(
q )
15We identify j(n, k) ≡j, b+(n, k) = b+ , b−(n, k) = b−here. 15We identify j(n, k) ≡j, b+(n, k) = b+ , b−(n, k) = b−here. 15We identify j(n, k) ≡j, b+(n, k) = b+ , b−(n, k) = b−here. 9 9 SciPost Phys. Core 6, 068 (2023) The unflavoured limit ω →1 is non-trivial for the factors The unflavoured limit ω →1 is non-trivial for the factors The unflavoured limit ω →1 is non-trivial for the factors The unflavoured limit ω →1 is non-trivial for the factors ωas/u+j+1 −ω−as/u−j
ω −1
and
ωas/u+j+1−(n+1) −ω−as/u−j+(n+1)
ω −1
(33) (33) of the even character (32). The rest of the character is regular near the point ω = 1 in the
region |q| < 1. We can obtain an infinite series by expanding the denominator in the region
|ω| > 1 (or equivalently |ω| < 1). The limit ω →1 is well-defined for all such characters only
on the real ω line, which yields the unflavoured limit of the even character (32). The rest of the character is regular near the point ω = 1 in the
region |q| < 1. We can obtain an infinite series by expanding the denominator in the region
|ω| > 1 (or equivalently |ω| < 1). The limit ω →1 is well-defined for all such characters only
on the real ω line, which yields the unflavoured limit χ+
j(n,k)(q) =
1
uη3(q)
qb2
+/4a X
s∈Z
qas2+sb+(2as + b+) −qb2
−/4a X
s∈Z
qas2+sb−(2as + b−)
. (34) (34) Note that the even character (34) is precisely the z independent coefficient in the infinite series
(17) multiplied by a factor of 2. Also, note that χ+
j(¯n,¯k)(q,ω) = χj(¯n,¯k)(q,ω) −χj(n,k)(q,ω) = −χ+
j(n,k)(q,ω). (35) (35) Due to this relation, we have a pair of identical characters up to a negative sign in the set of
even combinations of characters. Let us now discuss the modular properties of the even characters. 2.3
Unflavoured even characters In this subsection, we will analyse the q-series expansion of even characters χ+ in the ω →1
limit. Since we have eliminated the pole at ω = q0 = 1, the even characters have a well-
defined q-expansion in the region |q| < 1 in the limit ω →1, with finite integer coefficients
which are also positive for specific cases. Let us now focus on the unflavoured limit of the even
characters. Due to the σ0 symmetry, even characters always appear in pairs, one with b+ < 0 and the
her with b+ > 0. We, therefore, restrict ourselves to b+ < 0 in this subsection.16 The q-expansion of even characters is unique in the well defined limit z →0 and is given
by χ+
j(n,k)(q) = qb2
+/4a−1/8
η3(q)
X
s∈Z
2spqs2a+sb+ + b+
u
X
s∈Z
qs2a+sb+
−qb2
−/4a−1/8
η3(q)
X
s∈Z
2spqs2a+sb−+ b−
u
X
s∈Z
qs2a+sb−
. (41) (41) The exponent of the leading term in the q-series in the τ →i∞limit can be identified as b2
+/4a −1/8 = −c(p,u)/24 + h(n, k),
(42) (42) where c(p,u) is the Virasoro central charge and h(n, k) is the conformal dimension. The set
of even characters needs to be multiplied by u which ensures that the coefficients are integers
and the modular transformations remains invariant, however, as we will see, positivity of the
coefficients is not guaranteed. The terms can now be rearranged in an ascending order of powers of q. There are two
orderings of the expansion based on the values of b−. 16We will not restrict ourselves to this condition later and will choose the even sector as per convenience. 2.2
Even and odd characters While the action of
modular transformation T on the characters is by a left multiplication by a diagonal matrix Tnk,n′ k′ = e2πih(n,k)δnk,n′ k′ ,
(36) (36) the characters transform non-trivially under the modular S-transformation. The S-matrix com
ponents [22] are Snk,n′k′ =
v
t 2
pu(−1)k′(n+1)+k(n′+1)e−iπkk′p/u sin
π(n + 1)(n′ + 1)u
p
. (37) (37) The modular S-transformations of the even characters can then be written in the following
manner, χ+
j(n,k̸=0)
−1
τ, z
τ
=
X
n′,k′̸=0
2S+
nk,n′k′χ+
j(n′,k′)(τ,z) +
X
n′
2Snk,n′0χj(n′,0)(τ,z),
(38) (38) and χj(n,0)
−1
τ, z
τ
=
X
n′,k′̸=0
Sn0,n′k′χ+
j(n′,k′)(τ,z) +
X
n′
Sn0,n′0χj(n′,0)(τ,z),
(39) (39) where the sum runs over the labels (n′, k′) of even characters. The components of the modular
S+-transformation matrix are S+
nk,n′k′ = 1
2
Snk,n′k′ −S¯n¯k,n′k′
. (40) (40) Thus we see that the even characters form a closed modular invariant set under the modular S
and T transformations which we will refer to as the even sector. As mentioned above we will
work with the even characters and in particular we will focus on the unflavoured characters,
i.e., they are functions of the modular parameter τ only. Thus we see that the even characters form a closed modular invariant set under the modular S
and T transformations which we will refer to as the even sector. As mentioned above we will
work with the even characters and in particular we will focus on the unflavoured characters,
i.e., they are functions of the modular parameter τ only. 10 SciPost Phys. Core 6, 068 (2023) 18At the very least, the conformal dimensions and central charge are positive after relabelling. 2.3
Unflavoured even characters That is, for |b−| < a, and b+ < 0 we
have χ+
j(n,k)(τ) =
b+
uη3(q)qb2
+/4a−1/8
1 −b−
b+
qk(n+1) −
X
s∈N
qk(n+1)+s2a+sb−
2usp
b+
+ b−
b+
+
X
s∈N
qs2a+sb+
1 + 2usp
b+
+
X
s∈N
qs2a−sb+
1 −2usp
b+
+
X
s∈N
qk(n+1)+s2a−sb−
2usp
b+
−b−
b+
,
(43 (43) or, for |b−| > a, and b+ < 0 we have or, for |b−| > a, and b+ < 0 we have χ+
j(n,k)(τ) =
b+
uη3(q)qb2
+/4a−1/8
1 +
−2up
b+
−b−
b+
qk(n+1)+a+b−−b−
b+
qk(n+1)
−
X
s−1∈N
qk(n+1)+s2a+sb−
2usp
b+
+ b−
b+
+
X
s∈N
qs2a+sb+
1 + 2usp
b+
+
X
s∈N
qs2a−sb+
1 −2usp
b+
+
X
s∈N
qk(n+1)+s2a−sb−
2usp
b+
−b−
b+
. (44) (44) This rearrangement paves the way for a hierarchical arrangement of the powers of q. This
is shown in eq. (B.1) for the arrangement with |b−| < a. The asymptotic expansion makes
the coefficients of the q-series up to q4a+2b−+k(n+1) appear manifestly as a function of (n, k) 11 SciPost Phys. Core 6, 068 (2023) for a level parametrised as (p,u). To investigate if the character is a quasi-character,17 we
look for a sign flip of the coefficients in the q-expansion. From these q-expansions we have
found that there are quasi-character(s) at every level except the levels (p = 2,u = 2N + 1),
(p = 2N + 3,u = 2), where N ∈Z>0 and (p = 3,u = 4). 3
Admissible levels with admissible even character vector In this section, we will focus on the even characters of the Ö
su(2) algebra at the special frac-
tional levels which only have admissible even characters. If we look at the q-series expansion,
(43) or (44), we see that quasi-characters are ubiquitous except at very special points in the
space of admissible even characters. These special levels correspond to (p = 2,u = 2N + 1),
(p = 2N + 3,u = 2), where N ∈Z>0 and an isolated point at (p = 3,u = 4). We show
the positivity of the even characters by relating them to characters of integrable representa-
tions of a current algebra at an integer level. Thus, we discover two new maps from the even
unflavoured characters of fractional level Ö
su(2) current algebra log CFTs to the characters of
integer level Ö
su(2) and Ö
so(5) current algebra RCFTs. We will also present examples of the two
infinite sequences in the subsections 3.1 and 3.3. We will focus on these special levels in this
section, but before we discuss that, a comment on the appearance of quasi-characters in other
cases is in order. It is obvious from the q-expansions of the character formula at these levels that, ex-
cept at the special levels mentioned above, we always find one quasi-character with labels
(n = 0, k = ⌈(u/p)⌉), where ⌈x⌉denotes the ceiling function of x. It is easy to verify this from
the small q expansion in eq. (B.1). Furthermore, for higher k and n values, one may find ad-
ditional quasi-character(s). While we have studied the q-expansions for a large but finite set
of p and u values, the existence of a quasi-character for the labels (n = 0, k = ⌈(u/p)⌉) guaran-
tees that every Ö
su(2) theory at fractional levels, except for the special ones, have at least one
quasi-character. To get a better picture of this pattern, we present some illustrative examples
in the appendix A. In appendix C we look at the fractional levels with quasi-characters in their
even sector. 17After suitable rescaling by the integer u.
18 3.1
Threshold levels Among the special levels mentioned at the beginning of this section, the set with p = 2 satu-
rates the admissibility bound. For this reason, the Ö
su(2) levels are called the threshold levels. They are also called boundary levels for the same reason. The term boundary levels was first
used in [45]. Since p = 2, n = 0 is the only possible value. Due to the condition gcd(p,u) = 1,
u = 2N + 1, therefore k = 0,··· ,2N which we restrict to 0 ≤k ≤N due to the σ0 transforma-
tions (35). The central charge for these theories is (45) c = −6N. (45) c = −6N. To satisfy the integrality condition, the even characters are scaled by u = 2N + 1. For
simplicity, we will call χ+
j (n, k)(q) from now on to denote the even characters with integer
coefficients, that is, all the even characters are multiplied by u. We need to relabel the highest
weight representations such that the lowest dimensional primary is identified with the vacuum
of the tentative unitary RCFT. This procedure, which we will call unitarisation in anticipation
that the relabelling produces a unitary RCFT,18 chooses the correct vacuum character of the 12 SciPost Phys. Core 6, 068 (2023) form given in eqn. (3). If the unitarisation procedure gives a representation theory of a known
current algebra RCFT then it automatically guarantees that all the characters are admissible. form given in eqn. (3). If the unitarisation procedure gives a representation theory of a known
current algebra RCFT then it automatically guarantees that all the characters are admissible. g
y g
Having set up the procedure and the expectations, we illustrate this for the threshold levels. The lowest dimensional primary in the set of highest weight representations is given by hmin = hN = −N
2
N + 1
2N + 1
. (46) (46) We identify this representation with the vacuum representation of the unitarised theory. The
character of this vacuum representation has its leading coefficient b+(N) = 1, hence it is non-
degenerate. The central charge of unitarised CFT is We identify this representation with the vacuum representation of the unitarised theory. The
character of this vacuum representation has its leading coefficient b+(N) = 1, hence it is non-
degenerate. 3.1
Threshold levels The central charge of unitarised CFT is cU = c −24hmin = 3(4N)
4N + 2 ,
(47) (47) which is the central charge of Ö
su(2)4N CFT. The spectrum of conformal dimensions in this
theory is shifted by hmin, hU
k = (N −k)(N −k + 1)
4N + 2
. (48) (48) The spins are labelled as l/2 = N −k. The spins are labelled as l/2 = N −k. The spins are labelled as l/2 = N −k. The unitarisation procedure for the threshold cases provides a map to the D-type characters
of Ö
su(2)4N theories with N ∈Z>0 [22].ll The unitarisation procedure for the threshold cases provides a map to the D-type characters
of Ö
su(2)4N theories with N ∈Z>0 [22].ll The unflavoured even characters are equal to the unflavoured characters of the D-type
modular invariants of Ö
su(2) theory at level 4N. This can be easily shown by manipulating
the numerators of the characters, the denominator factor is independent of the level and does
not play any role in the manipulation. Instead of using the label k, we choose the spin of the
representation l of the related integrable representation, χ+
j(0,k)(q) = 2(2N + 1)
η3(q)
q(l+1)2/4(4N+2) X
s
qs2(4N+2)+s(l+1)
2s +
l + 1
4N + 2
−qk+s2(4N+2)+s(−4N+l−1)
2s −4N −l + 1
4N + 2
= χ(4N)
l
(q) + χ(4N)
4N−l(q), = χ(4N)
l
(q) + χ(4N)
4N−l(q),
(49) (49) and similarly χj(0,0) = χ(4N)
2N
. (50) (50) This establishes the positivity of the characters of threshold cases as well. The final expression
of the characters can be compared with eq. (19), with the identification 4N = M. The flavoured characters do not match for arbitrary fugacities, ω ∈u(1), but the vacuum
character of the even sector with u = 2N +1 match with the corresponding Ö
su(2)4N characters
with spin j = N at discrete fugacities [36], χ(4N)
N
(q,ω) = χsu(2)
2N
(ω)χ+
0 (q,ω),
ω = eπil/(N+1) ,
(51) (51) where χsu(2)
2N
(ω) =
Pi=N
i=−N ωi is the finite su(2) character in spin N representation, and
0 ≤l ≤2N + 1. When l = 0 the (51) reduces to the unflavoured limit (49). The conse-
quences of (51) for the 4-dimensional SCFT are discussed in [36] which we will not pursue
here. In retrospect, we should have been confident of the threshold cases to produce admis-
sible characters due to the relations in section 5.2 of [33]. We will demonstrate a similar 13 SciPost Phys. Core 6, 068 (2023) Table 1: The primaries in the even sector of Ö
su(2) current algebra at level m = −5/4
are labelled by (n, k). The spins of the primaries are labelled by j(n, k) and conformal
dimensions by h(n, k). 19Note that this is not the actual modular partition function of the Ö
su(2)−4/3 log CFT since there are infinite
mber of modules contributing the same character to the partition function. The spins are labelled as l/2 = N −k. (n, k)
j(n, k)
h(n, k)
(0,0)
0
0
(1,0)
1/2
1
(0,1)
-3/8
-5/16
(1,1)
1/8
3/16
(1,2)
-3/4
-1/4 computation in the next subsection to flavour a subsector of characters at the isolated level
m = −5/4 to the characters of Ö
su(2)1. computation in the next subsection to flavour a subsector of characters at the isolated level
m = −5/4 to the characters of Ö
su(2)1. computation in the next subsection to flavour a subsector of characters at the isolated level
m = −5/4 to the characters of Ö
su(2)1. The modular invariant partition function composed of the even characters is a D-type mod-
ular invariant of the A-D-E classification [46] when written in terms of characters in integrable
representation, as mentioned above. Z(4N) = 2|χ(4N)
2N |2 +
N−1
X
l=0
|χ(4N)
2l
+ χ(4N)
4N−2l|2 . (52) (52) Let us demonstrate the results through an example of Ö
su(2)−4/3 theory, where we have the
A-type modular invariant partition function given by19 Z−4/3 = |χj(0,0)|2 + 1
2|χ+
j(0,1)|2 ,
(53) (53) where, j(n, k) represents the spin of the highest weight state. This partition function does not
have integer coefficients but we can write another modular invariant partition function, where, j(n, k) represents the spin of the highest weight state. This partition function does not
have integer coefficients but we can write another modular invariant partition function, 2Z−4/3 = 2|χj(0,0)|2 + |χ+
j(0,1)|2 ,
(54) (54) which has a q-expansion with integer coefficients. We can use the relations given in eq (49)
to write the partition function as the D-type modular invariant of Ö
su(2)4. The correspondence
also holds at the level of modular S and T matrices and consequently the fusion algebra. 3.2
Curious case of the isolated level m = −5/4 (p = 3, u = 4) (n, k)
hU(n, k)
(0,0)
5/16
(1,0)
21/16
(0,1)
0
(1,1)
1/2
(1,2)
1/16 (n, k)
hU(n, k)
(0,0)
5/16
(1,0)
21/16
(0,1)
0
(1,1)
1/2
(1,2)
1/16 We can construct an A-type modular invariant partition function using only the even charac-
ters, We can construct an A-type modular invariant partition function using only the even charac-
ters,
X We can construct an A-type modular invariant partition function using only the even charac-
ters, Z(τ, ¯τ) =
X
i,j∈Γ
¯χiMi jχj ,
(56) (56) where we have already rescaled the characters by a factor u = 4 so that all coefficients in
the q-expansion are integers. The scaling forces us to choose a different vacuum as we had
seen for the threshold levels. To further the reason, note that the S-matrix is not unitary but
S†MS = M, where M = diag(1, 1, 1
2, 1
2, 1
2). The factors 1/2 imply that we need to rescale
the χn,0 characters by 2 since it is the smallest integer which makes the coefficients integers. Hence the physical modular invariant is Z(τ, ¯τ) = 2|χj(0,0)|2 + 2|χj(1,0)|2 + |χ+
j(0,1)|2 + |χ+
j(1,1)|2 + |χ+
j(1,2)|2 . (57) (57) This implies that the choice of the vacuum for the non-unitary theory is not correct anymore
and we are forced to consider a different vacuum. We can choose the lowest dimensional
primary as the vacuum character. In this example, it corresponds to the choice of character
χ+
j(0,1) to belong to the vacuum module. The resulting unitarised theory has non-negative
conformal dimensions and central charge. The RCFT description is in terms of the unitarised theory, where the primary with labels
(0,1) is identified with the vacuum. Table 2 lists the unitarised conformal dimensions. The
unitarised theory has the central charge cU = c −24hmin = −5 + 24(5/16) = 5/2. (58) (58) The primary field with conformal dimension 1/2 indicates the presence of a fermion. Com-
bining this with the fact that the central charge cU = 5/2 indicates that this theory has a free
field representation in terms of 5 fermions which form a representation of the Ò
so(5)1 current
algebra. 3.2
Curious case of the isolated level m = −5/4 (p = 3, u = 4) In this subsection, we will look at the isolated case of m = −5/4. This is a particularly inter-
esting case, firstly because it is a single entry and not an infinite sequence. The labels (n, k),
spins j(n, k), and the conformal dimensions h(n, k) of the even sector are listed in table 1. The
central charge of this theory is c = −5. Corresponding modular S-matrix can be read off from eq.(37) and eq.(40),
χj(0,0)(−1/τ)
χj(1,0)(−1/τ)
χ+
j(0,1)(−1/τ)
χ+
j(1,1)(−1/τ)
χ+
j(1,2)(−1/τ)
=
1
2
p
2
−1
1
1
−1
1
1
−1
1
−1
−1
2
2
p
2
p
2
0
−2
−2
p
2
p
2
0
2
−2
0
0
2
χj(0,0)(τ)
χj(1,0)(τ)
χ+
j(0,1)(τ)
χ+
j(1,1)(τ)
χ+
j(1,2)(τ)
. (55) (55) 19Note that this is not the actual modular partition function of the Ö
su(2)−4/3 log CFT since there are infinite
number of modules contributing the same character to the partition function. 14 SciPost Phys. Core 6, 068 (2023)
Table 2: The unitarised value of the conformal dimension associated to the primaries
in the even sector of Ö
su(2) current algebra at level m = −5/4 denoted by hU(n, k). (n, k)
hU(n, k)
(0,0)
5/16
(1,0)
21/16
(0,1)
0
(1,1)
1/2
(1,2)
1/16
We can construct an A t pe modular in ariant partition function using onl
the e en charac SciPost Phys. Core 6, 068 (2023) Table 2: The unitarised value of the conformal dimension associated to the primaries
in the even sector of Ö
su(2) current algebra at level m = −5/4 denoted by hU(n, k). Table 2: The unitarised value of the conformal dimension associated to the primaries
in the even sector of Ö
su(2) current algebra at level m = −5/4 denoted by hU(n, k). 3.2
Curious case of the isolated level m = −5/4 (p = 3, u = 4) Note that the conformal dimensions of (n = 0, k = 0) and (n = 1, k = 0) differ by
an integer which may give an impression that this theory can be a logarithmic CFT, however
the sum of the two characters is required to build characters of the unitarised theory. Also,
log CFTs are always non-unitary and there is no unitary counterpart of it. A straightforward
analysis rules out the A-type modular invariant, we can look for a D-type modular invariant
to accommodate this spectrum since we have the characters whose conformal dimensions are
separated by an integer which can be combined to build the non-diagonal partition function. The non-diagonal modular invariant partition function has the form Z(τ, ¯τ) = |χj(0,1)|2 + |χj(0,0) + χj(1,0)|2 + |χj(1,1)|2 . (59) (59) Since the vacuum character corresponds to χ+
j(0,1), which has a unit leading coefficient in its
q-expansion, ensures that the vacuum is unique for this theory. The modular S-transformation 15 SciPost Phys. Core 6, 068 (2023) of the characters is given by of the characters is given by of the characters is given by of the characters is given by
χ+
j(0,1)(−1/τ)
χj(1,0)(−1/τ) + χj(0,0)(−1/τ)
χ+
j(1,1)(−1/τ)
=
1/2
1/
p
2
1/2
1/
p
2
0
−1/
p
2
1/2
−1/
p
2
1/2
χ+
j(0,1)(τ)
χj(1,0)(τ) + χj(0,0)(τ)
χ+
j(1,1)(τ)
. (60)
(60) Note that the above S-matrix is now orthogonal STS = I (and therefore unitary). Since we
have distinct modules in the
Û
so(2r + 1) WZW models corresponding to distinct unflavoured
characters, we can demand positivity of fusion rules as mentioned in section 2. We compute
the Verlinde Fusion rules from the reduced S-matrix (60) using the Verlinde formula [4]. The
fusion rules can be conveniently written in the form of the fusion matrices (Ni)j
k, N(0,1) =
1
0
0
0
1
0
0
0
1
,
N(0,0)+(1,0) =
0
1
0
1
0
1
0
1
0
,
N(1,1) =
0
0
1
0
1
0
1
0
0
,
(61) (61) where the fusion matrices are written in the basis of the vector in (60). The resulting unitary
CFT is the Ò
so(5)1 WZW model which has central charge c = 5/2. 20The decomposition will be manifest below when we write the block-diagonal form of the modular S-matri 3.2
Curious case of the isolated level m = −5/4 (p = 3, u = 4) It can also be described in
terms of 5 copies of the Ising Model. We can identify the primaries in the following manner, (0,1) ≡I,
(0,0) + (1,0) ≡σ ,
(1,1) ≡ε. (62) (0,1) ≡I,
(0,0) + (1,0) ≡σ ,
(1,1) ≡ε. (62) With this identification these primaries have same fusion rules as the Ising model, ε × ε = I,
σ × σ = I + ε,
(63)
σ × ε = σ . (63) We can explicitly show that these characters are identical to the characters of Ò
so(5)1. χj(0,1) = χ ˆω0 = 1
2
θ 5/2
3
+ θ 5/2
4
η5/2
,
(64)
χj(1,1) = χ ˆω1 = 1
2
θ 5/2
3
−θ 5/2
4
η5/2
,
(65)
χj(0,0) + χj(1,0) = χ ˆω2 = 1
p
2
θ2
η
5/2
. (66) (64) (65) (66) Recall, we are looking only at the even characters, it is worth pointing out that the odd
combination of characters for this level gives us the characters of the Ising model. There is an interesting alternative to the choice of the sub-sector, with the choice of the
representation (1,2) as the vacuum representation and taking an orthogonal non-diagonal
combination of (0,0) and (1,0), namely, χj(0,0) −χj(1,0).20 This set forms a modular invari-
ant combination with the central charge c = 1 and corresponds to Ö
su(2)1 WZW model. The
spectrum of conformal dimensions is hU(1,2) = 0 and hU((0,0) −(1,0)) = 1/4. The explicit 16 SciPost Phys. Core 6, 068 (2023) equivalence of the characters is given below.21 equivalence of the characters is given below.21 χ+
j(1,2)(q) = 2q1/12
η3(q)
X
s
q12s2+2s
6s + 1
2
−q12s2−14s+4
6s −7
2
= q1/12
η3(q)
X
s
(1 + 6s)qs+3s2 = χ(1)
0 (q). (67) χ+
j(1,2)(q) = 2q1/12
η3(q)
X
s
q12s2+2s
6s + 1
2
−q12s2−14s+4
6s −7
2
χ+
j(1,2)(q) = 2q1/12
η3(q)
X
s
q12s2+2s
6s + 1
2
−q12s2−14s+4
6s −7
2
q1/12 X
(
6 ) s+3s2
(1)( )
(6 = q1/12
η3(q)
X
s
(1 + 6s)qs+3s2 = χ(1)
0 (q). (67) (67) Similarly, χj(0,0)(q) −χj(1,0)(q) = 2q1/3
η3(q)
X
s
q12s2+4s(6s + 1) −q12s2+8s+1(6s + 2)
= q1/3
η3(q)
X
s
(2 + 6s)q3s2+2s = χ(1)
1 (q). 21The common multiplicative scaling factor u = 4 for the even characters is reduced to u = 2 for this subsector.
The modular-S matrix is in a block diagonal form anyways, so the two subsectors can have two different scaling
factors. 3.2
Curious case of the isolated level m = −5/4 (p = 3, u = 4) (75) (75) Thus, we have two independent subsectors in the even sector of the fractional level m = −5/4. We will explore whether a subsector can be extracted for theories with quasi-character in their
even sector in appendix C. We return to the last class of special fractional levels which occur
at half-odd integer values in the next subsection. Thus, we have two independent subsectors in the even sector of the fractional level m = −5/4. We will explore whether a subsector can be extracted for theories with quasi-character in their
even sector in appendix C. We return to the last class of special fractional levels which occur
at half-odd integer values in the next subsection. 3.2
Curious case of the isolated level m = −5/4 (p = 3, u = 4) (68) (68) This identification guarantees that all coefficients in the q-expansion are positive. These rela-
tions complete the proof of the positivity of the characters in the representation of (p=3,u=4). The modular S-matrix for this two character subsector is This identification guarantees that all coefficients in the q-expansion are positive. These rela-
tions complete the proof of the positivity of the characters in the representation of (p=3,u=4). The modular S-matrix for this two character subsector is S = 1
p
2
1
1
1
−1
,
(69) (69) where the first row corresponds to the vacuum character χ(1)
0 (q). The S-matrix is again unitary
and hence the positivity of Verlinde fusion rules can be imposed on the characters. However,
fusion rules and the modular S and T matrix alone do not characterise the RCFT. It is straight-
forward to flavour the vacuum character of this subsector (akin to the threshold levels [36])
since we have the same number of fugacities on the unitary and non-unitary side. The numer-
ator of the non-vacuum character of the unitary theory is χj(0,0)(q,ω) −χj(1,0)(q,ω) = q1/3 X
s
q12s2+4s(ω3s+1/2 −ω−3s−1/2) −q12s2+8s+1(ω3s+1 −ω−3s−1)
. (70) (70) This will be related to χ(1)
1 (q,ω) due to (68) in the unflavoured limit, χ(1)
1 (q,ω) = 2q1/3 X
s
q12s2+4s(ω6s+1 −ω−6s−1) −q12s2+8s+1(ω6s+2 −ω−6s−2)
. (71) (71) If these two flavoured characters are equal upto a function of ω, then the relation can be
expressed as χ(1)
1 (q,ω) = χsu(2)
2
(ω)
χj(0,0)(q,ω) −χj(1,0)(q,ω)
,
ω = e
2πik
3 . (72) (72) For this to be true, the ratios, For this to be true, the ratios, r1(s,z) = sin(2πz(6s + 1))
sin(πz(6s + 1)) = 2cos(6πzs + πz),
r2(s,z) = sin(2πz(6s + 2))
sin(πz(6s + 2)) = 2cos(6πzs + 2πz),
(73) (73) should be equal and independent of s. At z = k/3 for k ∈Z≥0 the ratios are independent
of s, but not always equal. At only k = 0 and k = 2 are the ratios equal, where the former 21The common multiplicative scaling factor u = 4 for the even characters is reduced to u = 2 for this subsector. The modular-S matrix is in a block diagonal form anyways, so the two subsectors can have two different scaling
factors. 17 SciPost Phys. 3.2
Curious case of the isolated level m = −5/4 (p = 3, u = 4) Core 6, 068 (2023) value yields the relation between the unflavoured characters (68). This proves (72) at discrete
fugacities ω = e
2πik
3 , where k = 0 or k = 2.il As a final remark, we see that the set of independent even unflavoured characters of this
theory decomposes into a couple of sub-sectors. This can be explicitly seen if we write the
S-matrix for the characters. For convenience, let us rewrite the S-matrix of the m = −5/4
theory again, S =
1
2
p
2
−1
1
1
−1
1
1
−1
1
−1
−1
2
2
p
2
p
2
0
−2
−2
p
2
p
2
0
2
−2
0
0
2
. (74) (74) Let us label the characters as χ1 = χj(0,0), χ2 = χj(1,0), χ3 = χ+
j(0,1), χ4 = χ+
j(1,1), and
χ5 = χ+
j(1,2), which form the basis of the above S-matrix. The conformal dimensions of the
primary fields corresponding to χ1 and χ2 differ by an integer. Although the S-matrix (74)
is not manifestly block-diagonal, we can utilise the fact that we can take linear combinations
of χ1 and χ2 to write new characters. A rearrangement of the column vector after taking the
linear combination gives
(χ1 + χ2)(−1/τ)
χ3(−1/τ)
χ4(−1/τ)
(χ1 −χ2)(−1/τ)
χ5(−1/τ)
=
1
2
p
2
0
2
−2
0
0
2
p
2
p
2
0
0
−2
p
2
p
2
0
0
0
0
0
−2
2
0
0
0
2
2
(χ1 + χ2)(τ)
χ3(τ)
χ4(τ)
(χ1 −χ2)(τ)
χ5(τ)
. (75)
(χ1 + χ2)(−1/τ)
χ3(−1/τ)
χ4(−1/τ)
(χ1 −χ2)(−1/τ)
χ5(−1/τ)
=
1
2
p
2
0
2
−2
0
0
2
p
2
p
2
0
0
−2
p
2
p
2
0
0
0
0
0
−2
2
0
0
0
2
2
(χ1 + χ2)(τ)
χ3(τ)
χ4(τ)
(χ1 −χ2)(τ)
χ5(τ)
. 3.3
Positive half-integer levels The unitarised conformal dimensions are obtained by adding hmin to all the conformal
dimensions of the original theory. For k = 0, set This is the central charge of Ö
su(2)4N WZW models from the traditional Sugawara construc-
tion. The unitarised conformal dimensions are obtained by adding hmin to all the conformal
dimensions of the original theory. For k = 0, set hU(n,0) = (n + 1/2)((n + 1/2) + 1)
4N + 2
. (81) (81) We identify the spin l/2 = n + 1/2 with odd l values of the representation of Ö
su(2)4N. The
other set of characters (with k = 1) have dimensions We identify the spin l/2 = n + 1/2 with odd l values of the representation of Ö
su(2)4N. The
other set of characters (with k = 1) have dimensions hU(n,1) = (n −N)(n −N + 1)
4N + 2
. (82) (82) The identification of spins in this case is l/2 = (n −N) with even l. We can choose the set
N ≤n ≤2N −1 so that l is positive.ll The identification of spins in this case is l/2 = (n −N) with even l. We can choose the set
N ≤n ≤2N −1 so that l is positive.ll The k = 0 unflavoured characters match with the unflavoured characters of integrable
highest weight representations of Ö
su(2)4N with spins l/2 = n + 1/2, χj(n,0)(q) =
1
η(q)3 q(2n+2)2/4(4N+2) X
s∈Z
q(4N+2)s2+s(2n+2)(2(4N + 2)s + 2n + 2)
= χ(4N)
2n+1(q) = χ(4N)
l
(q). (83) (83) The characters with k = 1, on the other hand, can be written as the difference of the
characters (χ(4N)
l
−χ(4N)
4N−l) of the integrable highest weight representations of the unitarised
theory with spins l = 2(n −N) and 4N −l = 6N −2n respectively, with n restricted to take
values in N ≤n ≤2N −1 for both cases. We will use the variable l instead of n to write the q-expansion of the character, χ+
j(n,1)(q) = 2q(l+1)2/4(4N+2)
η3(q)
X
s
qs2(4N+2)+s(l+1)(s(4N + 2) + (l + 1)/2)
−qs2(4N+2)−s(l+1)−s(4N+2)+n+1
s(4N + 2) −l + 1 + 4N + 2
2
= χ(4N)
l
(q) −χ(4N)
4N−l(q),
(84) = χ(4N)
l
(q) −χ(4N)
4N−l(q),
(84) (84) where we have used the map σ0 takes l/2 →−l/2 −1 to rearrange the characters with l
an even integer. 3.3
Positive half-integer levels In this subsection, we will look at the other infinite series of theories with admissible characters. These theories are parametrised by (p = 2N+3,u = 2) with N ∈Z>0. The corresponding levels
are half odd integers m = p/u −2 = (2N −1)/2. We have checked that the q-expansions of
all the characters for 5 ≤p ≤29 have positive coefficients (up to ∼q2000). As an aside, the
(p = 3,u = 2) theory has a quasi-character but we will deal with this case separately. For now,
we focus on the unitarisation of the entire (p = 2N + 1,u = 2) sequence with N ∈Z>0. The central charge of these theories is given by c = 3 −
12
2N + 1 . (76) (76) Since we have u = 2, k can take two values 0 and 1. The number of independent even
characters for p = 2N + 1 is 3N. The 2N characters corresponding to k = 0 are equal to
the half-integer spin characters of Ö
su(2)4N theory. The remaining N even characters χ+
j(n,1),
corresponding to k = 1 can be written in terms of the difference of integer spin characters of
the same affine Lie algebra Ö
su(2)4N. The spins j(n, k) are The spins j(n, k) are j(n,0) = 1
2n,
j(n,1) = 1
2(n −N) −1
4 ,
(77) (77) 18 SciPost Phys. Core 6, 068 (2023) with the conformal dimensions, imensions, h(n,0) = n(n + 2)
2(2N + 1) ,
h(n,1) = (n −N)2 + (n −N) −3/4
2(2N + 1)
. (78) (78) The minimum conformal dimension among them is The minimum conformal dimension among them is The minimum conformal dimension among them is hmin = h(N,1) =
−3
8(2N + 1) . (79) (79) Following the usual unitarisation procedure, i.e., identifying the character χ+
j(N,1) with the
character of the identity of the unitary theory gives the unitarised central charge Following the usual unitarisation procedure, i.e., identifying the character χ+
j(N,1) with the
character of the identity of the unitary theory gives the unitarised central charge cU = 3 −
12
2N + 1 +
9
2N + 1 = 3(4N)
4N + 2 . (80) (80) This is the central charge of Ö
su(2)4N WZW models from the traditional Sugawara construc-
tion. p
y
y
p y
23The m = −1/2 case has been studied in detail [11–13,15–18] and was the second example where the log
rithmic nature of Ö
su(2)−1/2 was concretely established. 22It is tempting to think of this expression not as a character but as a Witten Index of a theory with broke
supersymmetry as the alternating signs are reminiscent of the operator (−1)F insertion in the character. Howeve
we do not have a reason to believe supersymmetry is at play here. 3.3
Positive half-integer levels It is not obvious from this expression for the characters that the resulting 19 SciPost Phys. Core 6, 068 (2023) character χ+
j(n,1) has manifestly positive coefficients. In fact, there is a counter-example, at the
level m = −1/2, i.e. N = 1, the character χ+
j(0,1) = −χ+
j(1,1) has the q-expansion22 character χ+
j(n,1) has manifestly positive coefficients. In fact, there is a counter-example, at the
level m = −1/2, i.e. N = 1, the character χ+
j(0,1) = −χ+
j(1,1) has the q-expansion22 χ+
j(1,1)(q) = q−1/12 1 −2q1 + q2 −2q3 + 4q4 −4q5 + 5q6 −6q7 + 9q8 + O(q9)
. (85) χ+
j(1,1)(q) = q−1/12 1 −2q1 + q2 −2q3 + 4q4 −4q5 + 5q6 −6q7 + 9q8 + O(q9)
. (85) (85) The unitarised theory has 3 independent even characters with the vacuum character given by
eq. (85). We have checked for large powers of q, that the quasi-character χ+
j(1,1) alternates
between positive and negative signs which belong to the class D of quasi-characters and the
three character theory in class DAA as defined in section 4.3 of [32].23 i
For the theories with only admissible characters, i.e., p = 2N + 3 (N ≥1), given the fact
that the character χ+
j(n,1) is obtained by subtraction of two characters, it is not guaranteed that
it will have manifestly positive definite coefficients in the q-expansion. Nevertheless, except
in the case of m = −1/2, it is possible to show that the coefficients of χ(4N)
l
are consistently
larger than those for χ(4N)
4N−l for N ≥2 and hence the character χ+
j(n,1) is an admissible character. Although we do not have explicit proof of this assertion, we have checked this up to sufficiently
high orders in the q-expansion. For example, consider the N = 2 case, which has the characters χ+
j(2,1)(q) = q−1/10 1 + 3q + 6q2 + 3q3 + 9q4 + 4q5 + 18q6 + 9q7 + 30q8 + 12q9 + O(q9)
,
χ+
j(3,1)(q) = q1/10 3 + 2q + 6q2 + 3q3 + 12q4 + 6q5 + 24q6 + 9q7 + 39q8 + 18q9 + O(q9)
. j( , )
(86) (86) For higher N the coefficients increase at faster rates. The subtraction of the characters also
means that the character χ(4N)
2N
will never appear in the partition function. 3.3
Positive half-integer levels Let us continue with the example of N = 2. This is an illustrative example. All higher N
values work out essentially in an identical manner. Z = |χj(0,0)|2 + |χj(1,0)|2 + |χj(2,0)|2 + |χj(3,0)|2 + 1
2|χ+
j(2,1)|2 + 1
2|χ+
j(3,1)|2 . (87) (87) The factor of 1/2 in front of a couple of characters is potentially problematic for we need
the partition integers to be positive integers. This can be avoided by rescaling the partition
function by a factor of 2, ˜Z = 2Z = |χ+
j(2,1)|2 + |χ+
j(3,1)|2 + 2|χj(0,0)|2 + 2|χj(1,0)|2 + 2|χj(2,0)|2 + 2|χj(3,0)|2 . (88) (88) At this point, it appears that we have a couple of choices for the vacuum character, namely, it
could be χ+
j(2,1) or χ+
j(3,1), but a glance at eq.(86) would convince us that it is the former one
that is a vacuum character and not the latter as predicted from eq. (79). Since the characters are written as subtraction of two integrable highest weight characters,
the partition function when expanded in terms of the integrable characters have negative
coefficients. The presence of these negative signs does not allow the description of the partition
function as an ordinary CFT. ˜Z = 2Z = |χ(8)
0
−χ(8)
8 |2 + |χ(8)
2
−χ(8)
6 |2 + 2|χ(8)
1 |2 + 2|χ(8)
3 |2 + 2|χ(8)
5 |2 + 2|χ(8)
7 |2 . (89) (89) owever, partition functions with negative coefficients are quite common in defect CFT, but the
efect partition functions have different modular properties. In this case, we have a partition 22It is tempting to think of this expression not as a character but as a Witten Index of a theory with broken
supersymmetry as the alternating signs are reminiscent of the operator (−1)F insertion in the character. However,
we do not have a reason to believe supersymmetry is at play here. 20 SciPost Phys. Core 6, 068 (2023) function which is modular invariant under SL(2,Z), and as a result, it is not a partition function
of a defect CFT. 3.3
Positive half-integer levels This effectively rules out the possibility of the half-integer level Ö
su(2) highest
weight states being related to defect CFT.i Although, we find an interesting correspondence between the characters at the fractional
level m = (2N −3)/2 and the character at the level M = 4N of the Ö
su(2) CFT, the choice of the
vacuum in the unitary theory poses a new puzzle. In general, for the tentative unitary RCFT
with the unflavoured characters as the even unflavoured characters at half-odd integer levels,
we find that the RCFT has a unique vacuum but the modular invariant partition function has
negative multiplicities when the character is expressed in terms of the Ö
su(2)4N characters, and
the naive Verlinde fusion coefficients are not well-defined. To understand this better, let us look at the modular transformation of the characters χ+
j(n,1)
in the ‘unitary’ theory, χ+
j(n,1)(−1/τ) = 2
X
n′∈Γ
S+
n1,n′1χ+
j(n′,1)(τ) +
X
n′
Sn1,n′0χj(n′,0)(τ)
,
(90) (90) where the modular S-matrix element Sn1,n′0 is given by eq. (37) and S+
n1,n′1 = −i(−1)n+n′+N
p
2N + 1
sin
π(n′ + 1)
cos
π
2N + 1(n′ + 1)(2n −2N + 1)
= 0. (91) (91) Notice, S+
n1,n′1 vanishes for all n, n′, and the vacuum character of the unitarised theory belongs
to this sector. It therefore follows that the S-matrix elements involving the vacuum character
χ+
j(N,1) and any character belonging to the χ+
j(n,1) set vanishes. Since the S-matrix is not unitary,
we can not straight away impose Verlinde fusion rules. However, using standard methods
[24], we can write the unitary S-matrix which will also have SN1,N1 = 0 since the vacuum
is unique. In particular, since some matrix elements in the vacuum row of the modular S-
transformation vanishes, the corresponding fusion coefficients are not well-defined. This leads
to the conclusion that the tentative theory will not be an RCFT. For completeness, we also write the modular S-transformation of χ+
j(n,1) where we will
relabel n = l
2 + N and n′ = l′−1
2
so that we can relate them to the integrable representations
of Ö
su(2) at level 4N, χ+
j(l/2+N,1)(−1/τ) =
2
p
2N + 1
X
l′∈odd
sin
π
4N + 2(l + 1)(l′ + 1)
χj(l′/2−1/2,0)(τ). 4
Conclusions and Discussion In this paper, we complete the picture of the even characters of the highest weight repre-
sentations of the Ö
su(2) log CFTs at fractional levels in the well-defined unflavoured limit by
implementing bootstrap constraints on the even characters. We found admissible character
vectors at only three special classes of fractional levels. These special fractional levels are clas-
sified into two infinite sequences, the threshold level sequence with (p = 2,u = 2N + 1) and
the sequence with positive half-integer levels with (p = 2N + 3,u = 2) where N ∈Z>0, and
the isolated level (p = 3,u = 4). Our classification of the Ö
su(2) admissible levels includes the
two new mappings for the isolated level and the half-odd integer levels. The isolated level
is the most curious case as its characters with correct fusion rules can be decomposed into
two independent modular invariant sectors, one sector corresponds to Ö
so(5)1 and the other
to Ö
su(2)1. The reason for such a decomposition is not clear to us. We have expanded the
unflavoured correspondence for the 2-character subsector with Ö
su(2)1 algebra to a flavoured
correspondence for the non-vacuum character of the unitary theory at two discrete values of
fugacity. We find the even characters of the half-odd integer levels to be related to the differ-
ence of characters of the Ö
su(2)4N+4 algebra. Although not manifest, we have checked their
positivity to a reasonably large power of q-expansion. However, when Ö
su(2)4N+4 integrable
representation characters are used as the building blocks of the partition function, it has neg-
ative multiplicities. The negative multiplicities in a modular invariant partition function do
not correspond to a physical partition function. Moreover, since the Verlinde fusion rules are
ill-defined, the characters do not represent any RCFT as they are incompatible with an MTC
structure. For the threshold cases, we have shown the positivity and integrality of the q-series
coefficients by writing these characters as sums of integrable highest weight characters of the
Ö
su(2)4N WZW models, reaffirming an old observation [22]. The correspondence between the
unflavoured even characters at the threshold levels and the unflavoured characters of Ö
su(2)4N
was already extended to the flavoured vacuum character of the non-unitary theory to the spin
N character of the unitary theory at discrete flavour fugacities in [36]. 3.3
Positive half-integer levels (92) (92) Since these characters are written in terms of subtraction of two characters of Ö
su(2) at level
M = 4N theory, we write the modular S-transformation as χ(4N)
l
−χ(4N)
4N−l(−1/τ) =
2
p
2N + 1
X
l′∈odd
sin
π
4N + 2(l′ + 1)(l + 1)
χ(4N)
l′
(τ). (93) (93) In summary, we find that the unitary RCFT does not have correct fusion coefficients due to
vanishing matrix elements involving the vacuum character. It may be interesting to explore
the possibility of expanding the Hilbert space of the CFT in such a way that the modular S-
matrix has non-vanishing vacuum row and column entries. There is an alternate way of salvaging these theories by looking for a suitable subsector of
the theory which forms a closed set under the fusion algebra and admits a modular invariant
partition function comprising only the characters in the subsector. While this seems a worth-
while exercise for the theories containing the quasi-characters, it is easy to see that there is no 21 SciPost Phys. Core 6, 068 (2023) such suitable subsector in any theory with u = 2. The argument is as follows. The primaries
with the characters χ(4N)
2l
−χ(4N)
4N−2l have integer spins. One may be tempted to think that the
integer spin subsector can very well have closed fusion algebra, but for the u = 2 case, we
have seen that the modular S-transformation eq. (93) relates the integer spin sector to the
half odd-integer spin sector. On the other hand, the sector with characters χ(4N)
2l+1 have half
odd-integer spins, which by themselves do not close under the Ö
su(2)4N fusion algebra. A little
generalisation of this argument can show that no mixing of half odd-integer and integer spin
subsectors can have a closed fusion algebra. 4
Conclusions and Discussion However, note that the
modular data alone is not enough to completely classify the tentative RCFT. In other words,
although we have introduced discrete flavour fugacity for the Ö
su(2) sectors, we have ignored
the flavour fugacity while classifying the RCFT. From another point of view, our results regarding the even sector of the Kaˇc-Wakimoto
characters at the admissible levels of Ö
su(2) current algebra classify the fractional levels into
two sets: with and without quasi-characters, i.e., in many cases the q-expansion is a vector
valued modular function with nice modular properties which do not satisfy the positivity of
the q-series coefficients. Apart from the special fractional levels mentioned above, we find
quasi-characters at every other level. Note that the modules of fractional level theories are
equivalent to formal power series in q and ω, and within its radius of convergence to the ana- 22 SciPost Phys. Core 6, 068 (2023) lytically continued characters of Kaˇc-Wakimoto. In this paper, the difference of the analytically
continued characters, i.e, the even characters were expanded in the region 1 < |ω| < |q|−1 for
|q| < 1. Equivalently they could be expanded in the region |q| < |ω| < 1 for |q| < 1 due to the
linear dependence, for example (29). Both of the expansions have a well-defined ω →1 limit
with an integral q-series. Surely, there are other regions of expansions of the Kaˇc-Wakimoto
character where we will not encounter quasi-characters but those characters are not realised
by an RCFT. In particular when the expansions are equivalent to the modules of the fractional
level theories. The even characters with the naive radius of convergence have been used in
earlier works, for example, [33, 35, 36], however, it remains to be seen if we can draw any
conclusions between the unitary and non-unitary theories while expanding in other regions of
convergence. The fact that the even sector of only a few special fractional level theories admits admissible
unflavoured character vectors points to some intricate relations between the integrable and
fractional level Ö
su(2) theories. It would also be interesting to explore the consequences of our
results on the corresponding 4-dimensional SCFT. In light of [33,47,48], we expect non-trivial
constraints on which non-unitary Ö
su(2) chiral algebras might be relevant, particularly in the
presence of non-local operators such as lines or surface defects. We hope to shed some light
on the correspondence soon. Acknowledgments We are immensely thankful to Dileep P. Jatkar for valuable suggestions, discussions, and guid-
ance while preparing the first draft. We benefited greatly from illuminating comments and
suggestions by Sujay Ashok, Subramanya Hegde, Sunil Mukhi and K.P. Yogendran. We are
grateful to the anonymous referees who commented on a previous version and the incredibly
constructive reviews from the referee at SciPost, which guided the current version of the paper. 4
Conclusions and Discussion We hasten to point out that the existence of quasi-characters is not necessarily a lost cause. As shown earlier for the two [26] and three [32] component character vectors, admissible
characters can be obtained from the quasi-characters either by adding two quasi-characters
with proper coefficients or by multiplication of an appropriate power of the modular invariant
j-function. It will be interesting to answer whether the quasi-characters found in the even
sector of fractional level theories are related to admissible characters of some RCFT. A
Quasi-characters examples In this appendix, we list examples of quasi-characters to shed light on the types of quasi-
characters we encountered. All the characters have only a finite number of positive or negative
coefficients respectively. This implies the sign flips only once in the q-expansion. We encountered only one example of a quasi-character with infinitely many positive or
negative signs, for Ö
su(2)−1/2 as discussed in section 3. p = 4 sequence: • u = 3, m = −2/3, c = −3/2,
χ+
(2,2)(q) = q−5/48(1 −4q −12q2 −41q3 −103q4 −249q5 −518q6 −O(q7)),
(A.4)
• u = 5, m = −6/5, c = −9/2,
χ+
(2,3)(q) = q−1/80(3 + 9q + 14q2 + 27q3 + 36q4 + 38q5 −117q6 −O(q7)),
(A.5)
• u = 7, m = −10/7, c = −15/2,
χ+
(2,5)(q) = q−13/112(1 + 3q −6q2 −23q3 −84q4 −222q5 −544q6 −O(q7)). (A.6) χ+
(2,5)(q) = q−13/112(1 + 3q −6q2 −23q3 −84q4 −222q5 −544q6 −O(q7)). (A.6) For additional examples, we refer the reader to the table D where we have provided data for
p = 3,4,··· ,10. In most of these examples, we see that the sign flip occurs at fairly small
powers in the q-expansion. However, we have checked all these cases up to ∼q100 to see if
there is another sign flip. We generically find that there is no instance except the (p = 3,u = 2)
case corresponding to m = −1/2 mentioned above. In the table D among other things, we
have listed the first coefficient of the q-series including its sign as well as the sign of the 90th
coefficient of the q-series to indicate that the sign flip in all those cases up to that order happens
only once. As can be seen from these two data points that there is a single sign flip for all cases
except the (p = 3,u = 2) case. p = 3 sequence: • u = 5, m = −7/5, c = −7,
χ+
(1,3)(q) = q−13/120(1 + 3q −2q2 −11q3 −48q4 −134q5 −321q6 −O(q7)),
(A.1)
• u = 7, m = −11/7, c = −11,
χ+
(1,4)(q) = q−13/168(2 + 6q + 18q2 + 28q3 + 54q4 + 72q5 −··· −82q8 + O(q9)), (A.2)
• u = 8, m = −13/8, c = −13,
χ+
(1 5)(q) = q−11/96(1 + 3q + 9q2 + 5q3 −45q4 −153q5 −219q8 −O(q7)). (A.3) u = 8, m = −13/8, c = −13, χ+
(1,5)(q) = q−11/96(1 + 3q + 9q2 + 5q3 −45q4 −153q5 −219q8 −O(q7)). (A.3) (A.3) A.1
n = 0, k = ⌈u/p⌉ We find that we always get a quasi-character for (n = 0, k = ⌈u/p⌉) labels except at the special
admissible levels described in section 3. These characters typically do not contain integer
entries but have a common denominator. They all have integer coefficients after they are
multiplied by u. 23 SciPost Phys. Core 6, 068 (2023) Let us now consider a few explicit examples. We will list them using the parameter u for
p = 3,4. A.2
Multiple quasi-characters As mentioned earlier, most of the fractional level cases except the special fractional levels
mentioned in sec. 3 possess at least one quasi-character. But, as can be seen from the table
D that there are several cases where we have multiple quasi-characters. We will look at some
examples of this type here. For the level m = −10/7, we encounter a second quasi-character with b+ = −5, which
takes us beyond the criteria of finding a quasi-character at n = 0, k = ⌈u/p⌉. χ+
(2,4)(q) = q−11/112(5 + 15q + 45q2 + 91q3 + 198q4 + ··· −51908q21 + O(q22)). (A.7) (A.7) In this example we encounter two quasi-characters in the character vector. As we progress to
larger values of u and p we encounter more quasi-characters for the set of characters. 24 SciPost Phys. Core 6, 068 (2023) • p = 6, u = 5, m = −4/5, c = −2. • p = 6, u = 5, m = −4/5, c = −2. • p = 6, u = 5, m = −4/5, c = −2. The first one is the quasi-character satisfies n = 0 and k = ⌈u/p⌉, The first one is the quasi-character satisfies n = 0 and k = ⌈u/p⌉,
χ+
(4,4)(q) = q−7/60(1 −8q −24q2 −77q3 −191q4 −453q5 −967q6 −O(q7)). (A.8) The first one is the quasi-character satisfies n = 0 and k = ⌈u/p⌉, The first one is the quasi-character satisfies n = 0 and k = ⌈u/p⌉, χ+
(4,4)(q) = q−7/60(1 −8q −24q2 −77q3 −191q4 −453q5 −967q6 −O(q7)). (A.8) The other two quasi-characters are, The other two quasi-characters are, χ+
(4,3)(q) = q17/60(7 + 21q + 46q2 + 103q3 + 204q4 + ··· −30886q20 −O(q7)),
(A.9)
χ+
(3,3)(q) = q−11/120(2 + 6q + 18q2 + 44q3 + 80q4 + ··· −240q11 −O(q7)). (A.10) χ+
(4,3)(q) = q17/60(7 + 21q + 46q2 + 103q3 + 204q4 + ··· −30886q20 −O(q7)),
(A.9) χ+
(3,3)(q) = q−11/120(2 + 6q + 18q2 + 44q3 + 80q4 + ··· −240q11 −O(q7)). (A.10) In the supplementary table D, we list about the first 20 characters independent up to the
relation (35) found for each value of p using the q-expansion 41 up to q100. A.2
Multiple quasi-characters In the table, a0
is the leading order coefficient of the character, and as is the term when the sign flips, i.e., the
character χ(q) = qh−c/24(−a0 −a1q −··· −as−1qs−1 + asqs + ···). 24We will use the choice b+ < 0. B
Asymptotic behaviour of q-series To derive the conditions for the positivity of coefficients, we arrange the q-expansion in such
a way that different q-series have different leading behaviour and have no overlapping expo-
nents. We will write the expansion for the characters with k ̸= 0 only since the characters with
k = 0 are well defined as discussed in section 2.24 χj(n,k)(τ) = b+
2uqb2
+/4a−1/8
k(n+1)−1
X
l=0
clql +
a+b−−1
X
l=0
ck(n+1)+l −b−
b+
cl
qk(n+1)+l
+
b+−k(n+1)−b−−1
X
l=0
ck(n+1)+a+b−+l −b−
b+
ca+b−+l −
2up
b+
+ b−
b+
cl
qk(n+1)+a+b−+l
+
−2b+−1
X
l=0
ca+b++l −b−
b+
ca+b+−k(n+1)+l −
2up
b+
+ b−
b+
cb+−k(n+1)−b−+l)
+
2up
b+
+ 1
cl
qa+b++l
+
−b−+k(n+1)+b+−1
X
l=0
ca−b++l −b−
b+
ca−b+−k(n+1)+l −
2up
b+
+ b−
b+
c−b+−b−−k(n+1)+l
+
2up
b+
+ 1
c−2b++l +
−2up
b+
+ 1
cl
qa−b++l
+
(2+1)(a+b−)−1
X
l=0
ck(n+1)+a−b−+l −b−
b+
ca−b−+l −
2up
b+
+ b−
b+
c−2b−+l
+
2up
b+
+ 1
ck(n+1)−sb−−b++l +
−2up
b+
+ 1
ck(n+1)−b−+b++l
+
2up
b+
−b−
b+
cl
qk(n+1)+a−b−+l + Σ
. (B.1) (B.1) 25 SciPost Phys. Core 6, 068 (2023) The cl is the coefficient of the series 1 + 3q + 9q2 + ··· =
P
i ciqi obtained from expanding the
denominator of the character as a q-series on the numerator. Finally, Σ is the correction to the
character at order q4a+2b−+k(n+1), i.e., due to s = 2 terms in the sum (41). Typically, eq. (B.1)
is a good approximation even at higher orders if we let the sum over l run over larger values
in the last summation. The cl is the coefficient of the series 1 + 3q + 9q2 + ··· =
P
i ciqi obtained from expanding the
denominator of the character as a q-series on the numerator. Finally, Σ is the correction to the
character at order q4a+2b−+k(n+1), i.e., due to s = 2 terms in the sum (41). Typically, eq. (B.1)
is a good approximation even at higher orders if we let the sum over l run over larger values
in the last summation. C
Closed subsets without quasi-characters? In this section, we report our preliminary investigations into the problem of finding sub-sectors
which are devoid of quasi-characters. So far we have been looking at theories at those frac-
tional levels, which only have admissible characters. However, as we have seen, there are
theories at other fractional levels which possess at least one quasi-character. It is easy to see
that these theories have a modular invariant partition function with the inclusion of the quasi-
character, but the presence of the quasi-character in the modular invariant makes difficult an
interpretation of the partition function in terms of state counting of a physical system.i To remedy this problem, we try to find subsectors of these theories. The subsectors are
an admissible subset of characters closed under the fusion algebra equipped with a modular
invariant partition function. It is a priori not guaranteed that theories with such sub-sectors
without quasi-characters exist at all. The reason for this suspicion stems from the observation
that the modular S-matrix for these theories typically does not decompose into a manifestly
block-diagonal form so that quasi-characters can be isolated. It is straightforward to show
that if the S-transformation matrix is in a block-diagonal form, it is a sufficient and necessary
condition for the existence of a closed sub-sector. We have seen a glimpse of a couple of sub-sectors when we studied the case of m = −5/4. Note that the linear combination of the characters provided a new basis, in terms of which
we could block-diagonalise the S-matrix. In general a block-diagonal S-matrix is not easy to
achieve. It is even more difficult when the eigenvalues of the modular T-transformation are
non-degenerate. In the remaining part of this section we set up two simple criteria. The admissible fractional
level theories with quasi-characters which satisfy the criteria do not have a subsector without
the quasi-character. 25Here S(nqkq),(nk) denotes the modular S-matrix for the entire even sector and not just k = 0 sector. 27The characters which transform in the same way under the modular T-transform should also transform ap-
propriately under the modular S-transformations. Thus, an arbitrary linear combination of the characters is not
possible. C.1
First criterion We can therefore derive conditions on n and n′ irrespective
of the value of k, k′.i The second equality follows from the first by using the fact that p and u are co-prime. This con-
straint is independent of k and k′. We can therefore derive conditions on n and n′ irrespective
of the value of k, k′. Our first criterion comes from the case when p is a prime number. Since p cannot be
factored, eq. (C.2) implies p is either n+1 or n′ +1. However, values of n and n′ are bounded,
0 ≤n, n′ ≤p −2. Therefore, eq. (C.2) is not satisfied for any prime number p. This criterion
will be sharpened momentarily when we consider the second condition on the S-matrix, but
before we do that, let us mention that if p is not a prime number then it is possible to obtain
solutions to eq. (C.2). The non-prime p can be factorised in at least one way as p = p1p2, where p1 and p2 are
(non-unique) integer factors of p at this step. As can be easily seen, the equation eq. (C.2)
can have multiple solutions in general. We easily see that if we choose p1 and p2 such that
p = p1p2 and (p1+p2) is minimised, we have max(r) = (p−p1)(p−p2)/p. It can also be easily
seen that for every value of 1 ≤r ≤max(r) there exists at least one solution to eq. (C.2). Another place where zeros of S-matrix can occur comes from the condition Snk,n′ k′ =S¯n¯k,n′ k′. ′ Another place where zeros of S-matrix can occur comes from the condition Snk,n′ k′ =S¯n¯k,n′ k′. This implies a condition on k and k′ and this condition is independent of the values n
and n′, This implies a condition on k and k′ and this condition is independent of the values n
and n′, (C.3) 2kk′ = (2s −1)u,
s ∈N. (C.3) Since the left-hand side of eq. (C.3) is an even integer, it implies that unless u is an even
integer, there are no solutions k, k′ of this condition. We then reach an interesting conclusion,
combining the results for this eq. (C.3) and the first condition eq. 26For a simpler criterion, we can only look at the (nq = 0, kq = ⌈u/p⌉) row of the S-matrix, but since there can
be more quasi-characters, we will look at a more general constraint. C.1
First criterion One simple way a quasi-character with labels (nq, kq) would decouple from a subsector is when
S(nqkq),(nk) = 0,∀(n, k) ̸= (nq, kq).25 This implies that the quasi-character forms a meromorphic
(single-character) CFT imposing strict bootstrap constraints on the conformal dimension of the
quasi-character and the central charge of the theory. Namely, c −24hq = 0 mod(8). We per-
formed a simple check for the quasi-character with labels (0,⌈u/p⌉) for (3≤p≤30,3≤u ≤30)
and did not find any quasi-character of this type which forms a single character CFT. We will
therefore consider a weaker criterion of imposing the vanishing of the S-matrix element condi-
tion only on the sub-sector labelled by (n, k) within all allowed values of (n, k). This will imply
that the quasi-character can have non-vanishing S-matrix entries with the characters outside
the chosen sub-sector. On the other extreme, if there is no entry of the modular S-matrix which vanishes and all
characters have non-degenerate eigenvalues of the modular T-transformation then it is guar-
anteed that the S-matrix can not be reduced to a block-diagonal form. This implies that a 26 SciPost Phys. Core 6, 068 (2023) subsector is not possible in such a theory. This is the first criterion we propose in this subsec-
tion.ii subsector is not possible in such a theory. This is the first criterion we propose in this subsec-
tion.ii To explore whether we can find such theories with no subsectors, we need to find the level
(p,u) which do not have any vanishing entry of the modular S-matrix for the even sector. Since
the even characters have two classes, k = 0 and k ̸= 0, we have two conditions corresponding
to the S-matrix elements for the two classes. We will evaluate the constraints separately as
they are independent. We first set Snk,n′ k′ = 0,
(C.1) (C.1) to find out what constraints this condition imposes on (n, k), (n′, k′) and (p,u).26 The con-
straint coming from eq. (C.1) is u(n + 1)(n′ + 1)
p
= r′ ,
r′ ∈Z+ ,
(n + 1)(n′ + 1) = rp ,
r ∈Z+. (C.2) (C.2) The second equality follows from the first by using the fact that p and u are co-prime. This con-
straint is independent of k and k′. C.1
First criterion (C.2), that the S-matrix has
no zeros for the level parametrised by (p ∈prime,u ∈odd) which translates to the statement
that there are no sub-sectors for these levels, up to the existence of characters with degener-
ate eigenvalues of the modular T-transformation, which by suitable addition of rows of the
S-matrix could, in principle make a sub-sector possible.27 We tabulate a few special cases
where the dimensions of the characters belonging to the same theory are separated by an in-
teger in table 3. A closer look at the table makes it clear that the criteria works well only for
(p ∈prime,u ∈prime). Since there is no sub-sector, more so without a quasi-character, the Since the left-hand side of eq. (C.3) is an even integer, it implies that unless u is an even
integer, there are no solutions k, k′ of this condition. We then reach an interesting conclusion,
combining the results for this eq. (C.3) and the first condition eq. (C.2), that the S-matrix has
no zeros for the level parametrised by (p ∈prime,u ∈odd) which translates to the statement
that there are no sub-sectors for these levels, up to the existence of characters with degener-
ate eigenvalues of the modular T-transformation, which by suitable addition of rows of the
S-matrix could, in principle make a sub-sector possible.27 We tabulate a few special cases
where the dimensions of the characters belonging to the same theory are separated by an in-
teger in table 3. A closer look at the table makes it clear that the criteria works well only for
(p ∈prime,u ∈prime). Since there is no sub-sector, more so without a quasi-character, the 27 SciPost Phys. Core 6, 068 (2023) Table 3: Theories with (p ∈3 ≤prime ≤29,u ∈3 ≤odd ≤29) which have charac-
ters of degenerate eigenvalue of the modular T-transformation. p
u
3
25
5
9,21,27
7
9,15,25,27
11
9, 15, 21, 25, 27
13
9, 15, 21, 25, 27
17
9, 15, 21, 25, 27
19
9, 15, 21, 25, 27
23
9,15,21, 25, 27
29
9,15, 21,25, 27
th
i
ith (
∈
i
∈
i
)
b
li
i
t d Thi i th
fi
d
i
f th fi t theories with (p ∈prime,u ∈prime) can be eliminated. This is the refined version of the first
criterion.i We can continue to find solutions to the condition eq. C.1
First criterion (C.3) for u = 2w, w ∈Z+. This
condition can be subdivided into whether w is prime or not. If w is prime, eq. (C.3) can only be
satisfied when either k = w or k′ = w as k, k′ ≤u/2 = w. For k = w, we have ⌈w/2⌉= (w+1)/2
number of k′ as solutions of this condition. Whereas if w ̸∈prime we can write w = w1w2
as a (non-unique) factorisation of w where w1 and w2 are integer factors of w at this step. Again, we have the set of solutions due to k = w, with s up to the same value max(s), given by
max(s) = ⌈w/2⌉. However, we will have additional solutions, as compared with w ∈prime, due
to different factorisations of w. For the quasi-character labelled by (n = 0, k = ⌈u/p⌉) we see
that the first condition eq. (C.2) does not give any solutions due to the bound 0 ≤n′ ≤p −2. To locate a subsector with labels (n′, k′), which decouples the quasi-character, they should
satisfy 2⌈u/p⌉k′
s = (2s −1)u,s ∈N,
∀n′ . (C.4) (C.4) Thus, the solutions to the above condition, in addition to the conformal data, which may allow
for suitable linear combinations of the characters and cause the vanishing of some entries of
the modular S-matrix, should be enough to provide a necessary condition for the existence of
a sub-sector. We hope to provide more details elsewhere. Let us look at an example where
both parameters p, and u are prime numbers, e.g., (p = 3,u = 5). This theory satisfies the
criterion. This theory has 6 characters in the even sector corresponding to the set Γ tabulated
below. The S-transformation matrix is given below in the ordered basis of the table 4. Note Thus, the solutions to the above condition, in addition to the conformal data, which may allow
for suitable linear combinations of the characters and cause the vanishing of some entries of
the modular S-matrix, should be enough to provide a necessary condition for the existence of
a sub-sector. We hope to provide more details elsewhere. Let us look at an example where
both parameters p, and u are prime numbers, e.g., (p = 3,u = 5). This theory satisfies the
criterion. This theory has 6 characters in the even sector corresponding to the set Γ tabulated
below. C.2
Second criterion The (p = 3,u = 5) example also illustrates a simple feature. For the original non-unitary
theory, the vacuum is degenerate with the degeneracy u = 5 and a multiplicity in the number of
primaries equal to 2. Since the lowest dimension primary has |b+(0,2)| = 1, it can be identified
with the vacuum because it is non-degenerate. In this example, the lowest dimension primary
also happens to be a quasi-character of the type (n = 0, k = ⌈5/3⌉). Since the vacuum is the
identity element of the fusion algebra, the simple identification of the quasi-character with the
vacuum character of the unitary theory prohibits any closed fusion subalgebra for the theory
without the quasi-character, provided we do not change the vacuum identification. A natural
option is to look for any other vacuum candidate for a sub-sector. But all other characters are
degenerate once we have scaled the characters by u to guarantee integrality and preserving
the form of the modular S-transformation. As we had seen for the (p = 3,u = 4) example,
once we have identified a subsector, we could also rescale the characters of the subsector
without affecting the form of the modular S-transformations. Thus, the candidate vacuum
must preserve integrality with leading coefficient unity after rescaling by the common factor
and be a part of a sub-sector without the quasi-character. In general, this criterion is hard to
satisfy for the subsector, although not outright impossible. Before we state the criterion for a generic level let us consider another example, that of
(p = 4,u = 3). The (p = 4,u = 3) theory has in total 9 primaries and 6 independent even
characters. The shaded entry in the table 5 below belongs to (n, k) = (0,1) which is a quasi-
character and happens to be the lowest dimension primary. From the conformal data it looks
like it has the same fate as that of (p = 3,u = 5), but Awata-Yamada (AY) fusion rules allows
a closed sub-sector with the fusion algebra, (0,0) × (2,0) = (2,0),
(2,0) × (2,0) = (0,0). (C.6) (C.6) These primaries have spins 0 and 1 respectively, hence from the su(2) point of view, their
closure under fusion is expected. For this sub-sector, the degeneracy of the vacuum character
from the q-expansion can be factored out by a rescaling. 28The Awata-Yamada fusion algebra [8] gives the fusion coefficients either 0 or 1 for all fractional level Ö
su(2)
theories. C.1
First criterion The S-transformation matrix is given below in the ordered basis of the table 4. Note Table 4: Conformal data for the even sector of (p = 3,u = 5). The shaded entry
corresponds to the only quasi-character in the even sector. Table 4: Conformal data for the even sector of (p = 3,u = 5). The shaded entry
corresponds to the only quasi-character in the even sector. (n, k)
j(n, k)
h(n, k)
hU(n, k)
(0,0)
0
0
2/5
(1,0)
1/2
5/4
33/20
(0,1)
-3/10
-7/20
1/20
(1,1)
1/5
2/5
4/5
(0,2)
-3/5
-2/5
0
(1,2)
-1/10
-3/20
1/4 28 SciPost Phys. Core 6, 068 (2023) that all entries of the S-transformation are non-zero and the conformal dimensions are not
integer separated to allow any simplification. that all entries of the S-transformation are non-zero and the conformal dimensions are not
integer separated to allow any simplification. S =
1
p
10
−1
−1
1
1
−1
−1
−1
1
−1
1
−1
1
2
−2
2cos(2π/5)
−2cos(2π/5)
−2cos(π/5)
2cos(π/5)
2
2
−2cos(2π/5)
−2cos(2π/5)
−2cos(π/5)
−2cos(π/5)
−2
−2
−2cos(π/5)
−2cos(π/5)
−2cos(2π/5)
−2cos(2π/5)
−2
2
2cos(π/5)
−2cos(π/5)
−2cos(2π/5)
2cos(2π/5)
. (C.5) S =
1
p
10
−1
−1
1
1
−1
−1
−1
1
−1
1
−1
1
2
−2
2cos(2π/5)
−2cos(2π/5)
−2cos(π/5)
2cos(π/5)
2
2
−2cos(2π/5)
−2cos(2π/5)
−2cos(π/5)
−2cos(π/5)
−2
−2
−2cos(π/5)
−2cos(π/5)
−2cos(2π/5)
−2cos(2π/5)
−2
2
2cos(π/5)
−2cos(π/5)
−2cos(2π/5)
2cos(2π/5)
. (C.5) (C.5) This S-matrix satisfies S†MS = M where M = diag(2,2,1,1,1,1). We want to identify a
possible sub-sector without the quasi-character, but a diagonal modular invariant for a possible
subsector theory does not exist as the S-matrix can not be written as a block-diagonal matrix,
nor does it allow an Awata-Yamada fusion subalgebra.28 This S-matrix satisfies S†MS = M where M = diag(2,2,1,1,1,1). We want to identify a
possible sub-sector without the quasi-character, but a diagonal modular invariant for a possible
subsector theory does not exist as the S-matrix can not be written as a block-diagonal matrix,
nor does it allow an Awata-Yamada fusion subalgebra.28 30Note that the sectors (n, k = 0) and (n, k ̸= 0) generally have different degeneracies and we do not expect a
mixing. Again, the theories with degenerate eigenvalues of modular T-transformation have more options, but we
restrict ourselves to the non-degenerate cases. C.2
Second criterion However, their conformal data does
not agree with the general two primary criteria [4]. We also find that the corresponding 2 × 2 29 SciPost Phys. Core 6, 068 (2023) Table 5: Conformal data of (p = 4,u = 3). The shaded entry corresponds to the only
quasi-character in the even sector. (n, k)
j(n, k)
h(n, k)
hU(n, k)
(0,0)
0
0
1/6
(1,0)
1/2
9/16
35/48
(2,0)
1
3/2
5/3
(0,1)
-2/3
-1/6
0
(1,1)
-1/6
-5/48
1/16
(2,1)
1/3
1/3
1/2
(0,2)
-4/3
1/3
1/2
(1,2)
-5/6
-5/48
1/16
(2,2)
-1/3
-1/6
0
S
f
i
i
i
i
l
d h
fl
d
i i
f
i
d
f h S-transformation matrix is singular and the unflavoured partition function composed of the
two elements does not exist.29 transformation matrix is singular and the unflavoured partition function composed of the
wo elements does not exist.29 It seems like at an arbitrary level, if the primary with the lowest conformal dimension has
a quasi-character, then the corresponding unitarised theory has no closed fusion subalgebra. It seems like at an arbitrary level, if the primary with the lowest conformal dimension has
a quasi-character, then the corresponding unitarised theory has no closed fusion subalgebra. As discussed above, the natural way out is to identify some other primary (not the lowest
dimension primary) as the vacuum of a subsector theory which excludes the quasi-character. Note that |b+| = 1 occurs only once each [22] in the sets (n, k = 0) and (n, k ̸= 0) ∈Γ for the
lowest dimension primary. Therefore, there are no other modules with non-degenerate grade
zero subspace to begin with, unless there is a common factor in the q-expansion of a set of
characters, forming a subsector, which can be factored out as we have seen for the example
(p = 3,u = 4). A closer look at the q-expansion given in (43) or (44) implies that the factor
b+ can be factored out while preserving the integrality of the candidate vacuum character if
the following two conditions are simultaneously satisfied, b−
b+
∈Z,
and,
2up
b+
∈Z. (C.7) (C.7) Apart from the known solutions with |b+| = 1, an easy solution is the character with the labels
(n = 1, k = 0) with b+ = −b−= 2u, which is a candidate vacuum for a subsector theory. 29Although there exists another AY sub-sector which excludes only the (1, k) primaries, this subsector is irrele-
vant because the subsector includes the quasi-character. C.2
Second criterion Another solution is the character with the common factor |b+| = 2 and |b−| = 2r with r ∈N
, among others. If we can locate a vacuum character from the above criteria, it should have
a degeneracy which can be factored out from all the characters in the subsector, that is, it
should have the lowest factor which is a divisor of all other degeneracies such that its leading
coefficient is unity, preserving integrality of every character in the subsector and the form of
modular S-transformation. A great choice to fulfil these criteria is the (n, k = 0) sector as it looks plausible from
the point of view of su(2) representation theory (and Awata-Yamada fusion algebra). This
subsector has degeneracy u which can be factored out to identify a vacuum and although it
does not include a quasi-character, its S-matrix elements with the quasi-character given by
labels (0,⌈u/p⌉) are non-vanishing (C.2). We have expectedly not found an evidence of the
existence of a k = 0 subsector.30 Hence if a subsector without the quasi-character exists, it 30 SciPost Phys. Core 6, 068 (2023) must be in the (n, k ̸= 0) subsector which must include a candidate vacuum. must be in the (n, k ̸= 0) subsector which must include a candidate vacuum. We have explored a few possibilities of salvaging theories where the minimum of the con-
formal dimensions is equal to the conformal dimension of the primary associated with the
quasi-character. We now propose that the characters associated with these theories do not
allow an RCFT description. This proposal rules out a large number of theories. A subset of
theories which can be ruled out in this fashion can be located by equating the value of (n, k)
for which the conformal dimension is the minimum, with the labels of the quasi-character
(n = 0, k = ⌈u
p⌉). We have found the set of levels (p,u) = (p, N p −1),
N ∈N,
p ≥3,
(C.8) (C.8) which satisfies the proposed condition. It is worth reiterating that this criterion is based on
the assumption that a subsector excluding the quasi-character, which is the vacuum of the
unitarised theory, and including a candidate vacuum is in general not possible and the eigen-
values of T-transformation are non-degenerate. We hope to provide a refined criterion and
more details elsewhere. D
Supplementary table Table 6: Quasi-characters. Table 6: Quasi-characters. Table 6: Quasi-characters. S. No. p
u
m
(n, k)
a0
as
sign(a90)
1
3
2
−1/2
(0,1)
−1
a1 = 2
−1
2
3
5
−7/5
(0,2)
−1
a2 = 2
+1
3
3
7
−11/7
(0,3)
−2
a8 = 82
+1
4
3
8
−13/8
(0,3)
−1
a5 = 45
+1
5
3
10
−17/10
(0,4)
−2
a11 = 240
+1
6
3
11
−19/11
(0,4)
−1
a11 = 77
+1
7
3
11
−19/11
(0,5)
−4
a34 = 2003808
+1
8
3
13
−23/13
(0,5)
−2
a15 = 2758
+1
9
3
13
−23/13
(0,6)
−5
a56 = 940473641
+1
10
3
14
−25/14
(0,5)
−1
a10 = 492
+1
11
3
14
−25/14
(0,6)
−4
a37 = 5016840
+1
12
3
16
−29/16
(0,6)
−2
a19 = 5286
+1
13
3
16
−29/16
(0,7)
−5
a57 = 981308226
+1
14
3
17
−31/17
(0,6)
−1
a13 = 1806
+1
15
3
17
−31/17
(0,7)
−4
a43 = 63653866
+1
16
3
19
−35/19
(0,7)
−2
a24 = 110282
+1
17
3
19
−35/19
(0,8)
−5
a63 = 6440955585
+1
18
3
20
−37/20
(0,7)
−1
a16 = 3417
+1
19
3
20
−37/20
(0,8)
−4
a50 = 676392192
+1
20
3
22
−41/22
(0,8)
−2
a29 = 570762
+1
21
3
22
−41/22
(0,9)
−5
a71 = 116158108500
+1
1
4
3
−2/3
(0,1)
−1
a1 = 4
+1 31 SciPost Phys. Core 6, 068 (2023) Table 6: Quasi-characters. S. No. D
Supplementary table Core 6, 068 (2023) Table 6: Quasi-characters. S. No. D
Supplementary table p
u
m
(n, k)
a0
as
sign(a90)
2
4
5
−6/5
(0,2)
−3
a7 = 117
+1
3
4
7
−10/7
(0,2)
−1
a2 = 6
+1
4
4
7
−10/7
(0,3)
−5
a21 = 51908
+1
5
4
9
−14/9
(0,3)
−3
a9 = 204
+1
6
4
9
−14/9
(0,4)
−7
a45 = 27281095
+1
7
4
11
−18/11
(0,3)
−1
a3 = 1
+1
8
4
11
−18/11
(0,4)
−5
a23 = 74916
+1
9
4
11
−18/11
(0,5)
−9
a82 = 919103509269
+1
10
4
13
−22/13
(0,4)
−3
a13 = 3264
+1
11
4
13
−22/13
(0,5)
−7
a45 = 190474799
+1
12
4
15
−26/15
(0,4)
−1
a6 = 58
+1
13
4
15
−26/15
(0,5)
−5
a28 = 856005
+1
14
4
15
−26/15
(0,6)
−9
a75 = 453611549547
+1
15
4
17
−30/17
(0,5)
−3
a17 = 18470
+1
16
4
17
−30/17
(0,6)
−7
a50 = 371458008
+1
17
4
19
−34/19
(0,5)
−1
a8 = 48
+1
18
4
19
−34/19
(0,6)
−5
a34 = 10392333
+1
19
4
19
−34/19
(0,7)
−9
a80 = 1111163267691
+1
20
4
21
−38/21
(0,6)
−3
a21 = 37557
+1
21
4
21
−38/21
(0,7)
−7
a57 = 3315025959
+1
1
5
3
−1/3
(0,1)
−2
a1 = 2
+1
2
5
4
−3/4
(0,1)
−1
a1 = 6
+1
3
5
6
−7/6
(0,2)
−4
a7 = 12
+1
4
5
7
−9/7
(0,2)
−3
a5 = 50
+1
5
5
7
−9/7
(0,3)
−8
a38 = 14688216
+1
6
5
7
−9/7
(1,3)
−1
a14 = 1278
+1
7
5
8
−11/8
(0,2)
−2
a3 = 10
+1
8
5
8
−11/8
(0,3)
−7
a26 = 395427
+1
9
5
9
−13/9
(0,2)
−1
a2 = 10
+1
10
5
9
−13/9
(0,3)
−6
a19 = 41868
+1
11
5
9
−13/9
(0,4)
−11
a81 = 753671188875
+1
12
5
9
−13/9
(1,4)
−2
a35 = 5057358
+1
13
5
11
−17/11
(0,3)
−4
a10 = 594
+1
14
5
11
−17/11
(0,4)
−9
a44 = 17115861
+1
15
5
11
−17/11
(1,5)
−3
a67 = 12235696960
+1
16
5
12
−19/12
(0,3)
−3
a7 = 90
+1
17
5
12
−19/12
(0,4)
−8
a35 = 8460776
+1
18
5
12
−19/12
(1,5)
−1
a31 = 1386570
+1
19
5
13
−21/13
(0,3)
−2
a5 = 36
+1 32 SciPost Phys. D
Supplementary table Core 6, 068 (2023) Table 6: Quasi-characters. S. No. D
Supplementary table p
u
m
(n, k)
a0
as
sign(a90)
20
5
13
−21/13
(0,4)
−7
a28 = 1381824
+1
21
5
13
−21/13
(0,5)
−12
a83 = 1008775829062
+1
1
6
5
−4/5
(0,1)
−1
a1 = 8
+1
2
6
5
−4/5
(0,2)
−7
a20 = 30886
+1
3
6
5
−4/5
(1,2)
−2
a11 = 240
+1
4
6
7
−8/7
(0,2)
−5
a8 = 149
+1
5
6
7
−8/7
(0,3)
−11
a59 = 1640260669
+1
6
6
7
−8/7
(1,3)
−4
a37 = 5016840
+1
7
6
11
−16/11
(0,2)
−1
a2 = 14
+1
8
6
11
−16/11
(0,3)
−7
a18 = 32466
+1
9
6
11
−16/11
(0,4)
−13
a71 = 121595449667
+1
10
6
11
−16/11
(1,4)
−2
a29 = 570762
+1
11
6
13
−20/13
(0,3)
−5
a10 = 99
+1
12
6
13
−20/13
(0,4)
−11
a46 = 106991913
+1
13
6
13
−20/13
(1,5)
−4
a65 = 15300126568
+1
14
6
17
−28/17
(0,3)
−1
a3 = 13
+1
15
6
17
−28/17
(0,4)
−7
a21 = 77790
+1
16
6
17
−28/17
(0,5)
−13
a66 = 42209473672
+1
17
6
17
−28/17
(1,6)
−2
a53 = 1214988222
+1
18
6
19
−32/19
(0,4)
−5
a14 = 4305
+1
19
6
19
−32/19
(0,5)
−11
a48 = 246563262
+1
20
6
23
−40/23
(0,4)
−1
a5 = 33
+1
21
6
23
−40/23
(0,5)
−7
a26 = 817329
+1
22
6
23
−40/23
(0,6)
−13
a69 = 53152478501
+1
23
6
23
−40/23
(1,8)
−2
a81 = 99284868952
+1
1
7
3
1/3
(0,1)
−4
a3 = 2
+1
2
7
3
1/3
(1,1)
−1
a2 = 4
+1
3
7
4
−1/4
(0,1)
−3
a1 = 2
+1
4
7
5
−3/5
(0,1)
−2
a1 = 6
+1
5
7
5
−3/5
(0,2)
−9
a29 = 660901
+1
6
7
5
−3/5
(1,2)
−4
a20 = 1764
+1
7
7
6
−5/6
(0,1)
−1
a1 = 10
+1
8
7
6
−5/6
(0,2)
−8
a18 = 2288
+1
9
7
6
−5/6
(1,2)
−2
a10 = 466
+1
10
7
8
−9/8
(0,2)
−6
a8 = 2
+1
11
7
8
−9/8
(0,3)
−13
a60 = 7895499881
+1
12
7
8
−9/8
(1,3)
−5
a39 = 26356888
+1
13
7
9
−11/9
(0,2)
−5
a6 = 68
+1
14
7
9
−11/9
(0,3)
−12
a45 = 117148224
+1 33 SciPost Phys. D
Supplementary table p
u
m
(n, k)
a0
as
sign(a90)
15
7
9
−11/9
(1,3)
−3
a24 = 280605
+1
16
7
9
−11/9
(2,4)
−1
a43 = 12703782
+1
17
7
10
−13/10
(0,2)
−4
a4 = 12
+1
18
7
10
−13/10
(0,3)
−11
a35 = 12260214
+1
19
7
10
−13/10
(1,3)
−1
a12 = 1193
+1
20
7
10
−13/10
(1,4)
−8
a2 = 1297242310936
+1
1
8
3
2/3
(0,1)
−5
a5 = 21
+1
2
8
3
2/3
(1,1)
−2
a4 = 24
+1
3
8
5
−2/5
(0,1)
−3
a1 = 4
+1
4
8
5
−2/5
(0,2)
−11
a40 = 31317813
+1
5
8
5
−2/5
(1,2)
−6
a31 = 168402
+1
6
8
5
−2/5
(2,2)
−1
a13 = 90
+1
7
8
7
−6/7
(0,1)
−1
a1 = 12
+1
8
8
7
−6/7
(0,2)
−9
a17 = 425
+1
9
8
7
−6/7
(1,2)
−2
a9 = 282
+1
10
8
7
−6/7
(1,3)
−10
a94 = 13594626504306 +1
11
8
7
−6/7
(2,3)
−3
a51 = 394666608
+1
12
8
9
−10/9
(0,2)
−7
a9 = 372
+1
13
8
9
−10/9
(0,3)
−15
a62 = 3795042033
+1
14
8
9
−10/9
(1,3)
−6
a41 = 29850606
+1
15
8
11
−14/11
(0,2)
−5
a5 = 54
+1
16
8
11
−14/11
(0,3)
−13
a38 = 13804479
+1
17
8
11
−14/11
(1,3)
−2
a16 = 6996
+1
18
8
13
−18/13
(0,2)
−3
a2 = 2
+1
19
8
13
−18/13
(0,3)
−11
a25 = 500376
+1
20
8
13
−18/13
(0,4)
−19
a95 = 33887464629093 +1
21
8
13
−18/13
(1,4)
−6
a55 = 898776774
+1
22
8
13
−18/13
(2,5)
−1
a56 = 876492702
+1
1
9
4
1/4
(0,1)
−5
a3 = 7
+1
2
9
4
1/4
(1,1)
−1
a2 = 8
+1
3
9
5
−1/5
(0,1)
−4
a1 = 2
+1
4
9
5
−1/5
(0,2)
−13
a52 = 517821576
+1
5
9
5
−1/5
(1,2)
−8
a44 = 159785964
+1
6
9
5
−1/5
(2,2)
−3
a27 = 344835
+1
7
9
7
−5/7
(0,1)
−2
a1 = 10
+1
8
9
7
−5/7
(0,2)
−11
a23 = 14469
+1
9
9
7
−5/7
(1,2)
−4
a15 = 4844
+1
10
9
7
−5/7
(2,3)
−6
a88 = 2597931389088
+1
11
9
8
−7/8
(0,1)
−1
a1 = 14
+1 34 SciPost Phys. D
Supplementary table Core 6, 068 (2023) Table 6: Quasi-characters. S. No. p
u
m
(n, k)
a0
as
sign(a90)
12
9
8
−7/8
(0,2)
−10
a17 = 16286
+1
13
9
8
−7/8
(1,2)
−2
a8 = 114
+1
14
9
8
−7/8
(1,3)
−11
a86 = 2892946308046
+1
15
9
8
−7/8
(2,3)
−3
a45 = 32005560
+1
16
9
10
−11/10
(0,2)
−8
a9 = 180
+1
17
9
10
−11/10
(0,3)
−17
a64 = 27582873039
+1
18
9
10
−11/10
(1,3)
−7
a43 = 83907339
+1
19
9
11
−13/11
(0,2)
−7
a7 = 129
+1
20
9
11
−13/11
(0,3)
−16
a51 = 948409950
+1
21
9
11
−13/11
(1,3)
−5
a30 = 2409958
+1
1
10
3
4/3
(0,1)
−7
a9 = 177
+1
2
10
3
4/3
(1,1)
−4
a7 = 12
+1
3
10
3
4/3
(2,1)
−1
a4 = 6
+1
4
10
7
−4/7
(0,1)
−3
a1 = 8
+1
5
10
7
−4/7
(0,2)
−13
a30 = 734574
+1
6
10
7
−4/7
(1,2)
−6
a22 = 123386
+1
7
10
7
−4/7
(3,3)
−2
a60 = 5880640588
+1
8
10
9
−8/9
(0,1)
−1
a1 = 16
+1
9
10
9
−8/9
(0,2)
−11
a16 = 4005
+1
10
10
9
−8/9
(1,2)
−2
a8 = 318
+1
11
10
9
−8/9
(1,3)
−12
a83 = 2147938578036
+1
12
10
9
−8/9
(2,3)
−3
a42 = 46836198
+1
13
10
11
−12/11
(0,2)
−9
a10 = 1350
+1
14
10
11
−12/11
(0,3)
−19
a65 = 9147426057
+1
15
10
11
−12/11
(1,3)
−8
a45 = 330185968
+1
16
10
13
−16/13
(0,2)
−7
a6 = 136
+1
17
10
13
−16/13
(0,3)
−17
a44 = 183378033
+1
18
10
13
−16/13
(1,3)
−4
a22 = 50592
+1
19
10
13
−16/13
(2,4)
−1
a36 = 7004477
+1
20
10
17
−24/17
(0,2)
−3
a2 = 10
+1
21
10
17
−24/17
(0,3)
−13
a22 = 179790
+1
22
10
17
−24/17
(0,4)
−23
a85 = 2870243892555
+1 [1] C. Beem, M. Lemos, P. Liendo, W. Peelaers, L. Rastelli and B. C. van Rees, Infi-
nite chiral symmetry in four dimensions, Commun. Math. Phys. 336, 1359 (2015),
doi:10.1007/s00220-014-2272-x. References [1] C. Beem, M. Lemos, P. Liendo, W. Peelaers, L. Rastelli and B. C. van Rees, Infi-
nite chiral symmetry in four dimensions, Commun. Math. Phys. 336, 1359 (2015),
doi:10.1007/s00220-014-2272-x. 35 SciPost Phys. Core 6, 068 (2023) [2] V. G. Kac
and
M. Wakimoto,
Modular
invariant
representations
of
infinite-
dimensional Lie algebras and superalgebras, Proc. Natl. Acad. Sci. 85, 4956 (1988),
doi:10.1073/pnas.85.14.4956. [3] I. G. Koh and P. Sobra, Fusion rules and (sub-)modular invariant partition functions in non-
unitary theories, Phys. Lett. B 215, 723 (1988), doi:10.1016/0370-2693(88)90049-4. [4] E. Verlinde, Fusion rules and modular transformations in 2D conformal field theory, Nucl. Phys. B 300, 360 (1988), doi:10.1016/0550-3213(88)90603-7. [5] D. Bernard and G. Felder, Fock representations and BRST cohomology in SL(2) current
algebra, Commun. Math. Phys. 127, 145 (1990), doi:10.1007/BF02096498. [6] P. Mathieu and M. A. Walton,
Fractional-level Kaˇc-Moody algebras and nonuni-
tary
coset
conformal
theories,
Prog. Theor. Phys. Suppl. 102,
229
(1990),
doi:10.1143/PTPS.102.229. [7] V. S. Dotsenko, Solving the SU(2) conformal field theory using the Wakimoto free field
representation, Nucl. Phys. B 358, 547 (1991), doi:10.1016/0550-3213(91)90424-V. [8] H. Awata and Y. Yamada, Fusion rules for the fractional level Õ
sl(2) algebra, Mod. Phys. Lett. A 07, 1185 (1992), doi:10.1142/S0217732392003645. [9] B. Feigin. F. Malikov, Modular functor and representation theory of c
sl2 at a rational level,
(arXiv preprint) doi:10.48550/arXiv.q-alg/9511011. [10] D. Adamovic and A. Milas, Vertex operator algebras associated to modular invariant rep-
resentations for A(1)
1 , Math. Res. Lett. 2, 563 (1995), doi:10.4310/MRL.1995.V2.N5.A4. [11] T. Creutzig and D. Ridout, Modular data and Verlinde formulae for fractional level WZW
models I, Nucl. Phys. B 865, 83 (2012), doi:10.1016/j.nuclphysb.2012.07.018. [12] T. Creutzig and D. Ridout, Logarithmic conformal field theory: Beyond an introduction, J. Phys. A: Math. Theor. 46, 494006 (2013), doi:10.1088/1751-8113/46/49/494006. [13] T. Creutzig and D. Ridout, Modular data and Verlinde formulae for fractional level WZW
models II, Nucl. Phys. B 875, 423 (2013), doi:10.1016/j.nuclphysb.2013.07.008. [14] M. R. Gaberdiel, Fusion rules and logarithmic representations of a WZW model at fractional
level, Nucl. Phys. B 618, 407 (2001), doi:10.1016/S0550-3213(01)00490-4. [15] F. Lesage, P. Mathieu, J. Rasmussen and H. Saleur, The bsl(2)−1/2 WZW model and the βγ
system, Nucl. Phys. B 647, 363 (2002), doi:10.1016/S0550-3213(02)00905-7. [16] F. Lesage, P. Mathieu, J. Rasmussen and H. Saleur, Logarithmic lift of the model, Nucl. Phys. B 686, 313 (2004), doi:10.1016/j.nuclphysb.2004.02.039. [17] D. Ridout,
bsl(2)−1/2:
A
case
study,
Nucl. Phys. B
814,
485
(2009),
doi:10.1016/j.nuclphysb.2009.01.008. References [18] D. Ridout, Fusion in fractional level -theories with, Nucl. Phys. B 848, 216 (2011),
doi:10.1016/j.nuclphysb.2011.02.015. [19] D. Ridout and S. Wood, The Verlinde formula in logarithmic CFT, J. Phys.: Conf. Ser. 597,
012065 (2015), doi:10.1088/1742-6596/597/1/012065. 36 SciPost Phys. Core 6, 068 (2023) [20] M. R. Gaberdie and H. G. Kausch, Indecomposable fusion products, Nucl. Phys. B 477, 293
(1996), doi:10.1016/0550-3213(96)00364-1. [21] M. R. Gaberdiel, An introduction to conformal field theory, Rep. Prog. Phys. 63, 607
(2000), doi:10.1088/0034-4885/63/4/203. [22] S. Mukhi and S. Panda, Fractional-level current algebras and the classification of characters,
Nucl. Phys. B 338, 263 (1990), doi:10.1016/0550-3213(90)90632-N. [23] S. D. Mathur, S. Mukhi and A. Sen, On the classification of rational conformal field theories,
Phys. Lett. B 213, 303 (1988), doi:10.1016/0370-2693(88)91765-0. [24] S. D. Mathur, S. Mukhi and A. Sen, Reconstruction of conformal field theories from
modular geometry on the torus, Nucl. Phys. B 318, 483 (1989), doi:10.1016/0550-
3213(89)90615-9. [25] H. R. Hampapura and S. Mukhi, On 2d conformal field theories with two characters, J. High Energy Phys. 01, 005 (2016), doi:10.1007/JHEP01(2016)005. [26] A. R. Chandra and S. Mukhi, Towards a classification of two-character rational conformal
field theories, J. High Energy Phys. 04, 153 (2019), doi:10.1007/JHEP04(2019)153. [27] A. Das, C. N. Gowdigere and J. Santara, Wronskian indices and rational conformal field
theories, J. High Energy Phys. 04, 294 (2021), doi:10.1007/JHEP04(2021)294. [28] A. Das, C. N. Gowdigere and J. Santara, Classifying three-character RCFTs with
Wronskian
index
equalling
0
or
2,
J. High
Energy
Phys. 11,
195
(2021),
doi:10.1007/JHEP11(2021)195. [29] J. Kaidi, Y.-H. Lin and J. Parra-Martinez, Holomorphic modular bootstrap revisited, J. High
Energy Phys. 12, 151 (2021), doi:10.1007/JHEP12(2021)151. [30] S. Mukhi, Classification of RCFT from holomorphic modular bootstrap: A Status report,
(arXiv preprint) doi:10.48550/arXiv.1910.02973. [31] S. Mukhi and B. C. Rayhaun, Classification of unitary RCFTs with two primaries and central
charge less than 25, Commun. Math. Phys. 401, 1899 (2023), doi:10.1007/s00220-023-
04681-1. [32] S. Mukhi, R. Poddar and P. Singh, Rational CFT with three characters: The quasi-character
approach, J. High Energy Phys. 05, 003 (2020), doi:10.1007/JHEP05(2020)003. [33] C. Beem and L. Rastelli, Vertex operator algebras, Higgs branches, and modular differential
equations, J. High Energy Phys. 08, 114 (2018), doi:10.1007/JHEP08(2018)114. [34] M. Buican and T. Nishinaka, On the superconformal index of Argyres-Douglas theories, J. Phys. A: Math. Theor. 49, 015401 (2015), doi:10.1088/1751-8113/49/1/015401. [35] M. Buican
and
Z. Laczko,
Nonunitary
lagrangians
and
unitary
non-
Lagrangian
conformal
field
theories,
Phys. Rev. Lett. References 120,
081601
(2018),
doi:10.1103/PhysRevLett.120.081601. [36] M. Buican and Z. Laczko, Rationalizing CFTs and anyonic imprints on Higgs branches, J. High Energy Phys. 03, 025 (2019), doi:10.1007/JHEP03(2019)025. [37] M. Lemos, Lectures on chiral algebras of N ⩾2 superconformal field theories, (arXiv
preprint) doi:10.48550/arXiv.2006.13892. 37 SciPost Phys. Core 6, 068 (2023) [38] S. Guruswamy and A. W. W. Ludwig, Relating c < 0 and c > 0 conformal field theories,
Nucl. Phys. B 519, 661 (1998), doi:10.1016/S0550-3213(98)00059-5. [39] T. Gannon, Comments on nonunitary conformal field theories, Nucl. Phys. B 670, 335
(2003), doi:10.1016/j.nuclphysb.2003.07.030. [40] J. A. Harvey, Y. Hu and Y. Wu, Galois symmetry induced by Hecke relations in rational
conformal field theory and associated modular tensor categories, J. Phys. A: Math. Theor. 53, 334003 (2020), doi:10.1088/1751-8121/ab8e03. [41] J. Kaidi, M. Martone, L. Rastelli and M. Weaver, Needles in a haystack. An algorithmic
approach to the classification of 4d N = 2 SCFTs, J. High Energy Phys. 03, 210 (2022),
doi:10.1007/JHEP03(2022)210. [42] S. D. Mathur and A. Sen, Group theoretic classification of rational conformal field the-
ories with algebraic characters, Nucl. Phys. B 327, 725 (1989), doi:10.1016/0550-
3213(89)90312-X. [43] J. Kaidi and E. Perlmutter, Discreteness and integrality in conformal field theory, J. High
Energy Phys. 02, 064 (2021), doi:10.1007/JHEP02(2021)064. [44] P. Di Francesco, P. Mathieu and D. Sénéchal, Conformal field theory, Springer, New York,
USA, ISBN 9781461274759 (1997), doi:10.1007/978-1-4612-2256-9. [45] V. Kac, S.-S. Roan and M. Wakimoto, Quantum reduction for affine superalgebras, Com-
mun. Math. Phys. 241, 307 (2003), doi:10.1007/s00220-003-0926-1. [46] A. Cappelli, C. Itzykson and J. B. Zuber, The A-D-E classification of minimal and A(1)
1 confor-
mal invariant theories, Commun. Math. Phys. 113, 1 (1987), doi:10.1007/BF01221394. [47] C. Córdova, D. Gaiotto and S.-H. Shao, Infrared computations of defect Schur indices, J. High Energy Phys. 11, 106 (2016), doi:10.1007/JHEP11(2016)106. [48] C. Córdova, D. Gaiotto and S.-H. Shao, Surface defects and chiral algebras, J. High Energy
Phys. 05, 140 (2017), doi:10.1007/JHEP05(2017)140. 38
|
https://openalex.org/W1973621084
|
https://figshare.com/ndownloader/files/468181
|
unk
| null |
Systematic Identification of Cellular Signals Reactivating Kaposi Sarcoma–Associated Herpesvirus
|
PLOS pathogens
| 2,007
|
cc-by
| 57
|
RTA amount
1
10
100
1000
10000
Relative luciferase intensity
-
Figure S1
1ng 5ng 25ng 125ng RTA amount
1
10
100
1000
10000
Relative luciferase intensity
-
Figure S1
1ng 5ng 25ng 125ng Figure S1 Figure S1 Figure S1 Figure S1 RTA amount
1
10
100
1000
10000
Relative luciferase intensity
-
1ng 5ng 25ng 125ng
|
https://openalex.org/W4247005078
|
https://www.qeios.com/read/SL36LP/pdf
|
English
| null |
Actual Accrual
|
Definitions
| 2,020
|
cc-by
| 65
|
Qeios · Definition, February 7, 2020 Open Peer Review on Qeios Actual Accrual National Cancer Institute National Cancer Institute Qeios ID: SL36LP · https://doi.org/10.32388/SL36LP Source National Cancer Institute. Actual Accrual. NCI Thesaurus. Code C71471. A number of subjects that have been enrolled for the participation in a study at a given
specified point in time (including all ever-enrolled subjects). Qeios ID: SL36LP · https://doi.org/10.32388/SL36LP 1/1
|
https://openalex.org/W2060847056
|
https://zenodo.org/records/1229241/files/article.pdf
|
English
| null |
Fatal illness associated with a new hantavirus in Louisiana
|
Journal of medical virology
| 1,995
|
public-domain
| 4,212
|
Accepted for publication February 27, 1995.
Address reprint requests to Dr. Stuart Nichol, SPBIDVRDI
NCID (MS G14), Centers for Disease Control and Prevention,
1600 Clifton Road, Atlanta, GA 30333. Fatal Illness Associated With a New Hantavirus
in Louisiana Ali S. Khan, Christina F. Spiropoulou, Sergey Morzunov, Sherif R. Zaki, Melvin A. Kohn,
Soheir R. Nawas, Louise McFarland, and Stuart T. Nichol Division of Viral and Rickettsia1 Diseases (A.S.K., C.F.S., S.M., S.R.Z., S.T.N), National Centers for Infectious
Diseases, Centers for Disease Control and Prevention, Atlanta, Georgia; E.A. Conway Memorial Hospital (S.R.N.),
Monroe, Louisiana; Louisiana Department of Health and Hospitals (M.A.K., L.M.), New Orleans, Louisiana Division of Viral and Rickettsia1 Diseases (A.S.K., C.F.S., S.M., S.R.Z., S.T.N), National Centers for Infectious
Diseases, Centers for Disease Control and Prevention, Atlanta, Georgia; E.A. Conway Memorial Hospital (S.R.N.),
Monroe, Louisiana; Louisiana Department of Health and Hospitals (M.A.K., L.M.), New Orleans, Louisiana A fatal case of hantaviral illness occurred in Lou-
isiana, outside of the range of P. maniculatus, the
rodent reservoir for Sin Nombre virus. Hantavi-
rus RNA and antigens were detected in patient
autopsy tissues, and nucleotide sequence analy-
sis of amplified polymerase chain reaction (PCR)
products identified a newly recognized unique
hantavirus, provisionally named Bayou virus. Prominent features of the clinical illness are
com pat i b le with ha ntavi rus pu I mona ry sy n-
drome (HPS), but several features such as renal
insufficiency and intraalveolar hemorrhage are
more compatible with hemorrhagic fever with
renal syndrome (HFRS), a disease associated
with Eurasian hantaviruses. polymerase chain reaction (PCR) and immunohis-
tochemical staining data confirming infection with a
hantavirus. In addition, nucleotide sequence analysis
indicates that this virus, provisionally named Bayou
virus, represents a newly recognized unique hantavirus
[Morzunov et al., 19951 distinct from Sin Nombre virus
and the recently discovered Black Creek Canal virus,
detected in cotton rats (Sigmodon hispidus) from Flor-
ida [CDC, 1994a; Rollin et al., 1995133. 0 1995 WILEY-LISS, INC. *This article is a US Government
work and, as such, is in the public domain in the United States of
America. CASE REPORT A 58-year-old Louisiana highway worker was admit-
ted to the regional public health hospital on June 16,
1993, for evaluation of dizziness, fever, vomiting, and
weakness after approximately a two to three day his-
tory consisting of fever, chills, night sweats, headache,
nausea, vomiting, and bilateral lower extremity weak-
ness. On admission, his temperature was 101.6"F, res-
piratory rate of 18, heart rate of 92, and blood pressure
of 170/100 mm/Hg. His physical exam and chest radio-
graph were unremarkable on admission; his urine
analysis showed trace protein. On the third day of hos-
pitalization, he was noted to have scant concentrated
urine despite aggressive administration of fluids (at
least 3.7 liters), and underwent computerized head to-
mography with intravenous contrast for evaluation of
his ongoing headache and mental obtundation. He also
had developed progressive shortness of breath and
chest radiograph evidence of bilateral diffuse intersti-
tial infiltrates necessitating intubation. Later that day,
he became hypotensive and required aggressive pressor
support. By the fourth day of hospitalization, he was
noted to have acute oliguric renal insufficiency, hypoal-
buminemia, and disseminated intravascular coagula-
tion. On the fifth day of hospitalization, he developed
irreversible asystole and expired despite aggressive
medical management for his suspected adult respira-
tory distress syndrome (ARDS) and presumptive septic Journal of Medical Virology 46:281-286 (1995) Journal of Medical Virology 46:281-286 (1995) KEY WORDS: KEY WORDS: bunyaviridae, hantavirus pul-
monary
syndrome,
hantavi-
ruses, nephropathy, Sin Nom-
bre virus, HFRS, Bayou virus INTRODUCTION There was no history of nephrotoxic analgesic or
antibiotic administration either prior to or during hos-
pitalization. Significant laboratory values are depicted
in Table I. All pre- and postmortem bacterial, fungal,
and mycobacterial stains and cultures of sputum and
blood were sterile. A complete clinical description was
rapidly presented to notify clinicians of the increasing
geographic spread of HPS [Steier and Clay, 19931. g
g
p
p
[
y,
Pertinent gross autopsy findings included: bilateral
atelectasis of both lungs, left greater than right, almost
generalized; bilateral severe pleural effusions; moder-
ate peritoneal effusions; mild pericardial effusions; se-
vere cerebral edema; generalized and severe subcuta-
neous edema, and cardiomegaly with left ventricular
hypertrophy. Histopathologic examination of the lung
revealed an interstitial pneumonitis with congestion
and inter- and intra-alveolar mononuclear and neutro-
philic leukocytic infiltrates. Focal areas of atelectasis,
hemorrhage, as well as type I1 pneumocyte hyperplasia
were observed. Extensive amounts of edema fluid and
fibrin as well as large numbers of hemosiderin-laden
macrophages were also seen within the alveoli. Both
kidneys showed changes consistent with early renal
tubular necrosis; the pancreas was histologically nor-
mal. Fresh frozen and formalin-fixed tissues were sent
to CDC for hantavirus diagnostic testing. No serum
sample was available. METHODS
RNA Extraction, Amplification and Nucleotide
Sequence Analysis Ribonucleic acid (RNA) was extracted from lung and
brain autopsy tissues as described previously [Nichol et
al., 19931. Briefly, approximately 100 mg of tissue was
homogenized manually in 500 p1 of acid guanidine thio-
cyanate solution, followed by the phenol-chloroform ex-
traction and purification with an RNA matrix (RNaid
kit, BiolOl, La Jolla, CAI. The extracted RNA was then
assayed for the presence of hantavirus RNA using a
nested reverse-transcriptase polymerase chain amplifi-
cation technique as described previously [Nichol et al.,
1993; Spiropoulou et al., 19941. This assay amplifies a
185 nucleotide region of the G2 encoding region of the
virus M segment. This area was chosen for analysis as
it has been extensively analyzed for a number of hanta-
viruses, and phylogenetic trees generated based on
data from this region have been highly predictive of the
overal phylogeny of hantaviruses [Nichol et al., 1993;
Spiropoulou et al., 1994; Rollin et al., 1995a,b]. The
nucleotide sequence of a 139 bp region of the resulting
product DNA (185 bp in length) was analyzed directly
using the dyedeoxy cycle sequencing technique (Ap-
plied Biosystems, Foster City, CA). INTRODUCTION Recently, hantavirus pulmonary syndrome (HPS)
was described in the southwestern United States
[Nichol et al., 1993; Duchin et al., 1994; Zaki et al.,
19951. This syndrome is characterized by a febrile pro-
drome and severe noncardiogenic pulmonary edema
simulating adult respiratory distress syndrome. HPS is
caused by a novel hantavirus (Sin Nombre virus) and
like other known hantaviruses, is associated with a
specific primary rodent vector [McKee et al., 19911, in
this case Peromyscus maniculatus, the deer mouse
[Nichol et al., 1993; Childs et al., 1994; Elliot et al.,
1994; Hjelle et al., 19941. Shortly after the description
of this clinical entity, additional cases were reported
throughout the United States. As expected, the vast
majority of cases occurred within the known geographic
range of P. maniculatus [CDC, 1994133. We discuss the
salient clinical features of a fatal case of this clinical
syndrome in an individual residing outside the geo-
graphical distribution of P. maniculatus and report 282 Khan et al. TABLE I. Selected Laboratory and Clinical Results*
Normal values
16-Jun
18-Jun
19Jun
20-Jun
WBC (103/cu mmj
7.8-11.8
5.3
28.1
29.0
-neutrophils (%)
50-70
58
56
78
-bands (%)
2-6
10
9
1
-lymphocytes (%)
20-44
27
28
16
-monocytes (%)
2-9
7
6
5
Hemoglobin (g/dl)
16.0-18.0
15.8
14.1
9
Hematocrit (% j
47-52
44.8
43.2
25.7
Platelets (io3/cu mm)
130-400
194
28
23
Amylase (UILj
34-122
66
638
625
Albumin (g/dl)
3.5-4.8
4.4
1.7
LDH (U/L)
94-172
211
BUN (mg/dl)
7-22
13
36
40
53
Creatinine (mg/dl)
0.5-1.7
1.1
0.6
5.6
7.4
Bicarbonate (mmol/l)
23-33
31
14
14
16
PT (secj
11-16
13.5
29.8
29.2
PTT (sec)
25-38
32.8
>240
>240
Fibrinogen (mg/dl)
150-350
144
Fibrin split products
>4,000
(ndml)
*WBC, white blood count; LDH, lactic dehydrogenase; BUN, blood urea nitrogen; F'T, prothrombin time; PTT,
partial thromboplastin time. TABLE I. Selected Laboratory and Clinical Results* dehydrogenase; BUN, blood urea nitrogen; F'T, prothrombin time; PTT,
partial thromboplastin time. He resided in a brick house in the residential area of a
semirural town which faced an open field but no ro-
dents were seen within the house. The family owned a
horse at a local farm but no other household pets. Se-
rum samples from all thirty-three rodents captured in
these environs during the initial case investigation
were seronegative for hantaviral antibodies (T.G. Ksiazek, personal communication). No other family
member was ill during this six week time period. shock. Phylogenetic Analysis related to Black Creek Canal and Sin Nombre viruses
(76% and 71% identity, respectively), the only other
hantaviruses recovered from rodent species indigenous
to North America and associated with HPS (Fig. 1). Phylogenetic analysis of nucleotide sequence differ-
ences between previously characterized hantaviruses
and PCR fragments generated from the hantavirus de-
tected in the Louisiana case lung and brain tissues
were analyzed by the maximum parsimony method us-
ing PAUP software [Swofford, 19911. Analysis was car-
ried out by a branch and bound search using a weight-
ing of transversions over transitions of 4:l. This
weighting was based on prior MacClade software anal-
ysis [Maddison and Maddison, 19921 of trees generated
from nucleotide sequence differences detected among
numerous Sin Nombre virus variants [Spiropoulou et
al., 19941. Such weighting is predicted to improve the
effectiveness of the maximum parsimony method for
estimation of the correct phylogeny [Hillis et al., 19941. Bootstrap confidence limits were calculated by 1,000
heuristic search repetitions of the analysis. Previously
published hantavirus sequences used in the analysis
include: Hantaan strains 76-118 [Genbank M14627,
Schmaljohn et al., 1987; X61034, Yo0 and Kang, 19871,
Lee [Genbank D00377, Schmaljohn et al., 19881, and
HV-114 [Genbank L08753; Xiao et al., 19931, Seoul
strains 80-39 [Genbank S47716, Antic et al., 19921,
SR-11 [Genbank M34882, Arikawa et al., 19901, and
Thai [Genbank L08756, Xiao et al., 19941, Puumala
strains Sotkamo [Genbank X61034, Vapalahti et al.,
19921, Cg18-20 [Genbank M29979, Giebel et al., 19891,
Berkel [Pilaski et al., 19941 and 9013 [Rollin et al.,
1995al; Prospect Hill strain PHV-1 [Genbank X55129,
Parrington et al., 19911; Sin Nombre viruses NM H10
and CC107 [Nichol et al., 1993; Spiropoulou et al., 1994;
Dexin et al., 19951; and Black Creek Canal virus [Rollin
et al., 1995bl. Immunohistochemistry Immunohistochemical assays for virus antigens were
performed as described previously using hantavirus-
specific monoclonal and polyclonal antibodies [Zaki et
al., 19951. Newly Recognized Hantavirus in Louisiana Newly Recognized Hantavirus in Louisiana 283 DISCUSSION The results of the genetic analysis of a fragment of
the virus G2 coding region indicates that the virus de-
tected in the tissues of this patient is a newly recog-
nized highly pathogenic hantavirus most closely re-
lated to Sin Nombre (SN) and Black Creek Canal (BCC)
viruses found in P. maniculatus and S. hispidus, respec-
tively [Child et al., 1994; Rollin et al., 1995bl. Subse-
quent phylogenetic analysis of the nucleotide se-
quences of the complete genome S and M segments of
this and other hantaviruses revealed a similar tree to-
pology to that reported here and suggested naming of
this distinct virus from Louisiana as Bayou virus [Mor-
zunov et al., 19951. As P. maniculatus is absent from
the southeastern United States [Baker, 19681, rodent
trapping and serological testing are currently under-
way to determine the identity of the rodent reservoir,
and to determine the geographic distribution of Bayou
virus. The strong immunostaining of this patient's tis-
sues observed with a broadly reactive hantavirus-spe-
cific monoclonal antibody but not with high-titered
polyclonal sera from SN virus-infected P. maniculatus
is consistent with the genetic findings. It is interesting
that phylogenetic analysis places the hantaviruses as-
sociated with predominantly HPS-like illness, i.e., SN,
BCC, and Bayou virus, in a distinct genetic clade. Immunohistochemical Analysis Viral antigens were detected in several organs in-
cluding the lungs, spleen, and kidneys. The strongest
immunostaining was seen using a cross-reactive mono-
clonal antibody (GB04-BF07) raised against Puumala
virus nucleocapsid protein. This antibody had previ-
ously been shown to cross-react with a number of han-
taviruses including Sin Nombre virus [Ruo et al., 1991;
Zaki et al., 19951. Immunostaining was characteristi-
cally punctate and observed primarily within microvas-
cular endothelial cells, in a pattern similar to that seen
in previously reported cases of HPS LZaki et al., 19951. Lesser amounts of antigens were detected using a high-
titered polyclonal sera from SN virus infected P. manic-
ulatus. INTRODUCTION p
For the six week period prior to his illness onset
(usual extreme of hantavirus incubation periods) he
had no travel outside northeast Louisiana where he
was employed as a bridge and road maintenance
worker. His duties routinely exposed him to rodents
and included: repairing and replacing culverts; clean-
ing drainage pipes; and cleaning drifts and beaver
dams from under bridges. On the weekend he operated
a large mechanical plower on two local farms where
rodent sightings were common in the fields and barns. RESULTS
PCR and Nucleotide Sequence Analysis Specific nested PCR DNA products of the expected
size (185 bp) were obtained using total RNA extracted
from tissues from the lung and brain of the patient
(data not shown). No bands were obtained using RNA
from control autopsy materials or RNA controls. The
amplified DNA from the tissues were extracted from
the agarose gel, and analyzed with an automated ther-
mocycle sequencing technique with the same primers
used for PCR product synthesis. The DNA bands were
found to contain hantavirus-like sequences but the nu-
cleotide sequence differed from that of any of the known
hantaviruses by at least 24%. Phylogenetic analysis of
the nucleotide sequence differences by the maximum
parsimony method indicated that the hantavirus asso-
ciated with this fatal HPS case in Louisiana was novel,
representing a distinct lineage, and was most closely ,
y
,
g
Many of the clinical illness and laboratory findings of
this patient are similar to those reported for SN virus-
induced HPS cases in the Southwest [Duchin et al.,
1994; Shefer et al., 1994; Zaki et al., 19951. These in-
clude the febrile prodrome, the rapid onset of severe
pulmonary disease, hemoconcentration, leukocytosis
with a left shift, severe progressive thrombocytopenia,
hypoalbuminemia, metabolic acidosis, and elevated PT
and PTT. The elevated amylase has not been previously
reported, although abdominal pain is a not uncommon
feature of the SN virus-induced HPS clinical prodrome. The acute and profound renal insufficiency is most
atypical compared to other reported HPS cases [Duchin 24
K
27
55
, 23 fHtnLee
93
l6
SN CC107
BCC
49
55
100
,
39 731
98k Htn7611
110 HtnHvl14
+I
Seo8039
I I r
23
Thai
Mr
65
8
Pu u Be rkel
Puu9013
65
PuuCg1820
23
PuuSotkamo
I
59
PH
. Genetic relationship of virus detected in the Louisiana HPS
previously characterized hantaviruses. Nucleotide sequence
ces among the 139 bp of PCR fragements of the M segment of
s detected in the Louisiana HPS case and previously charac-
hantaviruses were analyzed by the weighted maximum parsi-
ethod (seeMaterials and Methods) A single most-parsimonius
tree was obtained. Horizontal lengths are proportional
step differences (indicated above lines). Vertical dist
graphic representation only. Bootstrap conficence lim
lated by 1,000 heuristic search repetitions, and limits in
are indicated in bold italics at appropriate branch points 284 Khan et al. Fig. 1. Genetic relationship of virus detected in the Louisiana HPS
case to previously characterized hantaviruses. REFERENCES Rollin PE, Ksiazek TG, Elliott LH, Ravkov EV, Martin ML, Morzunov
S, Livingstone W, Monroe M, Glass G, Ruo S, Khan AS, Childs JE,
Nichol ST, Peters CJ (1995b): Isolation of Black Creek Canal virus,
a new hantavirus from Szgmodon Hispidus in Florida. Journal of
Medical Virology 46:3539. Antic D, Kang CY, Spik K, Schmaljohn C, Vapalahti 0, Vaheri A
(1992): Comparison of the deduced gene products of the L, M and S
genome segments of hantaviruses. Virus Research 24:3546. Arikawa J , Lapenotiere HF, Iacono-Connors L, Wang M, Schmaljohn
CS (1990): Coding properties of the S and M genome segment of
SR-11 virus: Comparison to other causative agents of hemorrhagic
fever with renal syndrome. Virology 176:114-125. Ruo SL, Sanchez A, Elliot LH, Brammer LS, McCormick JB, Fisher-
Hoch SP (1991): Monoclonal antibodies to three strains of hantavi-
ruses: Hantaan, R22, Puumala. Archives of Virology 119:l-11. p
g
fever with renal syndrome. Virology 176:114-125. Schmaljohn CS, Schmaljohn AL, Dalrymple JM (1987): Hantaan virus
M RNA Coding strategy, nucleotide sequence, and gene order. Virology 157:31-39. Baker RH (1968): Habitats and distribution. In King JA (ed): “Biology
of Peromyscus (Rodentia).” Lawrence, Kansas: Special Publica-
tions of the American Society of Mammalogists, pp 98-126. Schmaljohn CS, Arikawa J, Hasty SE, Rasmussen L, Lee HW, LPyung
WL, Dalrymple JM (1988): Conservation of antigenic properties
and sequences encoding the envelope proteins of prototype Han-
taan virus and two virus isolates from Korean Haemorrhagic fever
patients. Journal of General Virology 69:1949-1955. CDC (1994a): Newly identified hantavirus-Florida, 1994. Morbidity
and Mortality Weekly Report 43:99-105. CDC (199413): Hantavirus Pulmonary syndrom+Virginia,
1993. Morbidity and Mortality Weekly Report 43876477. Childs JE, Ksiazek TG, Spiropoulou CF, Krebs JW, Morzunov S, Mau-
pin GO, Gage KL, Rollin R, Sarisky J, Enscore R, Peters CJ, Nichol
ST (1994): Serologic and genetic identification of Peromyscus man-
iculatus as the primary rodent reservoir for a new hantavirus in
the southwestern United States. Journal of Infectious Diseases
169:1271-1280. Sheedy JA, Froeb HF, Batson HA, Conley CC, Murphy JP, Hunter RB,
Cugell DW, Giles RB, Bershadsky SC, Vester JW, Yoe RH (1954):
The Clinical course of epidemic hemorrhagic fever. American
Journal of Medicine 16:619-628. Shefer Am, Tapper0 JW, Bresee JS, Peters CJ, Ascher MS, Zaki SR,
Jackson RJ, Werner SB, Rollin PE, Ksiazek TG, Nichol ST, Bert-
man J, Parker S, Failing RM (1994): Hantavirus pulmonary syn-
drome in California: Report of two cases and investigation. RESULTS
PCR and Nucleotide Sequence Analysis It is of
note that renal insufficiency was also a prominent fea-
ture of the nonfatal case of HPS-like illness in Miami,
Florida due to Black Creek Canal virus [CDC, 1994a;
Rollin et al., 1995bI. The phylogenetic analysis places
these two viruses from the southeastern US. together
in another distinct genetic clade, suggesting that this
genetic group of hantaviruses may be associated with
clinical illness somewhat intermediate between HPS
and HFRS, prominent respiratory illness associated
with significant renal insufficiency. However, identifi-
cation and analysis of a greater number of patients
infected with these viruses will be needed to clarify the
full spectrum of clinical disease. ropathia epidemica virus strain Haellnaes B1 molecular cloning
and nucleotide sequence analysis. Virology 172:498-505. q
y
gy
Hillis DM, Huelsenbeck JP, Cunningham CW (1994): Application and
accuracy of molecular phylogenies. Science 264:671-677. Hjelle B, Jenison S, Torrez-Martinez N, Yamada T, Nolte K, Zumwalt
R, MacInnes K, Myers G (1994): A novel hantavirus associated
with an outbreak of fatal respiratory disease in the southwestern
United States: Evolutionary relationships to known hantaviruses. Journal of Virology 68:592-596. Li D, Schmaljohn AL, Anderson K, Schmaljohn CS (1995): Complete
nucleotide sequences of the M and S segments of two hantavirus
isolates from California: Evidence of reassortment in nature
among viruses related to hantavirus pulmonary syndrome. Virol-
ogy 206:963-977. Maddison WP, Maddison DR (1992): “MacClade: Analysis of Phylog-
eny and Character Evolution.” Sunderland, MA: Sinauer. McKee KT Jr, LeDuc JW, Peters CJ (1991): Hantaviruses. In Belshe
RB (ed): “Textbook of Human Virology.” 2nd ed. St. Louis: Mosby
Year Book, pp 615-632. Morzunov SP, Feldmann H, Spiropoulou CF, Semenova VA, Rollin
PE, Ksiazek TG, Peters CJ, Nichol ST (1995): A newly recognized
virus associated with a fatal case of hantavirus pulmonary syn-
drome in Louisiana. Journal of Virology, 691980-1983. Nichol ST,
Spiropoulou CF, Morzunov S, Rollin PE, Ksiazek TG, Feld-
mann H, Sanchez A, Zaki S, Childs J , Peters CJ (1993): Genetic
identification of a hantavirus associated with an outbreak of acute
respiratory illness. Science 262:914-917. Parrington MA, Lee PW, Kang CY (1991): Molecular characterization
of the Prospect Hill virus M RNA segment: Comparison with the M
RNA segments of other hantaviruses. Journal of General Virology
72:1845-1854. ACKNOWLEDGMENTS We are indebted to C.J. Peters for reviewing the
manuscript, Marty Monroe for technical assistance,
and John O’Connor for editorial review. Rollin PE, Bowen MD, Saluzzo JF, Guerard S, Flechaire A, Coudrier
D, Sureau P, Peters CJ, Nichol ST (1995a): Isolation of a Puumala
virus from a human case of Nephropathia Epidemica in France. American Journal of Tropical Medicine and Hygiene (in press). RESULTS
PCR and Nucleotide Sequence Analysis Pilaski J , Feldmann H, Morzunov S, Rollin RE, Ruo SL, Lauer B,
Peters CJ, Nichol ST (1994): Genetic identification of a new Puu-
mala virus strain causing severe hemorrhagic fever with renal
syndrome in Germany. Journal of Infectious Diseases 170:1456-
1462. RESULTS
PCR and Nucleotide Sequence Analysis Nucleotide sequence
differences among the 139 bp of PCR fragements of the M segment of
the virus detected in the Louisiana HPS case and previously charac-
terized hantaviruses were analyzed by the weighted maximum parsi-
mony method (see Materials and Methods). A single most-parsimonius tree was obtained. Horizontal lengths are proportional to nucleotide
step differences (indicated above lines). Vertical distances are for
graphic representation only. Bootstrap conficence limits wre calcu-
lated by 1,000 heuristic search repetitions, and limits in excess of 50%
are indicated in bold italics at appropriate branch points. Newly Recognized Hantavirus in Louisiana 285 et al., 1994; Zaki et al., 19951. Although the renal dis-
ease in this individual may be attributable to the ad-
ministration of intravenous contrast, a history of scant
dark urine preceded the contrast study. Moreover, ele-
ments of the clinical prodrome, laboratory abnormali-
ties, central nervous system manifestations, micro-
scopic findings and the observed renal insufficiency are
typical of hemorrhagic fever with renal syndrome
(HFRS), the clinical entity caused by Eurasian hanta-
viruses [Sheedy et al., 1954; Cohen et al., 19831. It is of
note that renal insufficiency was also a prominent fea-
ture of the nonfatal case of HPS-like illness in Miami,
Florida due to Black Creek Canal virus [CDC, 1994a;
Rollin et al., 1995bI. The phylogenetic analysis places
these two viruses from the southeastern US. together
in another distinct genetic clade, suggesting that this
genetic group of hantaviruses may be associated with
clinical illness somewhat intermediate between HPS
and HFRS, prominent respiratory illness associated
with significant renal insufficiency. However, identifi-
cation and analysis of a greater number of patients
infected with these viruses will be needed to clarify the
full spectrum of clinical disease. ropathia epidemica virus strain Haellnaes B1 molecular cloning
and nucleotide sequence analysis. Virology 172:498-505. et al., 1994; Zaki et al., 19951. Although the renal dis-
ease in this individual may be attributable to the ad-
ministration of intravenous contrast, a history of scant
dark urine preceded the contrast study. Moreover, ele-
ments of the clinical prodrome, laboratory abnormali-
ties, central nervous system manifestations, micro-
scopic findings and the observed renal insufficiency are
typical of hemorrhagic fever with renal syndrome
(HFRS), the clinical entity caused by Eurasian hanta-
viruses [Sheedy et al., 1954; Cohen et al., 19831. Zaki SR, Greer PW, Coffeld LM, Nolte KB, Foucar K, Feddersen RM,
Zumwalt R, Miller GL, Rollin PE, Ksiazek TG, Nichol ST, Mahy
BWJ, Peters CJ (1995): Hantavirus pulmonary syndrome: Patho-
genesis of an emerging infectious disease. American Journal of
PathO1ogy 146:552-579. Yo0 D, Kang CY (1987): Nucleotide sequence of the M segment of the
genomic RNA of Hantaan virus 76-118. Nucleic Acids Research
15:6299-6299. gy
Xiao Sy, hduc JW, Chu YK, Schmaljohn CS (1994): Phylogenetic
analyses of virus isolates in the genus Hantavirus, family Bunya-
viridae. Virology 198:205-217. REFERENCES Clinical
Infectious Disease 19:1105-1109. Cohen MS, Hsi-En K, Chien-Ching C, Hsin-Chen G (1983): CNS man-
ifestations of epidemic hemorrhagic fever. Archives of Internal
Medicine 143:2070-2072. Duchin JS, Koster FT, Peters CJ, Simpson GL, Tempest B, Zaki SR,
Ksiazek TG, Rollin PE, Nichol ST, Umland ET, Moolenaar RL,
Reef SE, Nolte KB, Gallaher MM, Butler JC, Breiman RF, the
Hantavirus Study Group (1994): Hantavirus pulmonary syn-
drome: A clinical description of 17 patients with a newly recog-
nized disease. New England Journal of Medicine 330:949-955. Spiropoulou CF, Morzunov S, Feldmann H, Sanchez A, Peters CJ,
Nichol ST (1994): Genome structure and variability of a virus
causing hantavirus pulmonary syndrome. Virology 200:715-723. Steier KJ, Clay R (1993): Hantavirus pulmonary syndrome (HPS):
Report of first case in Louisiana. Journal of American Osteopath
Association 93:1286-1289. Elliott LH, Ksiazek TG, Rollin PE, Spiropoulou CF, Morzunov S, Mon-
roe M, Goldsmith CS, Humphrey CD, Zaki SR, Krebs JW, Maupin
G, Gage K, Childs JE, Nichol ST,
Peters CJ (1994): Isolation of the
causative agent of hantavirus pulmonary syndrome. American
Journal of Tropical Medicine and Hygiene 51: 102-108. Swofford DL (1991): PAUP Phylogenetic Analysis Using Parsimony,
Version 3.0s. Computer program. Illinois Natural History Survey,
Champaign, Illinois. Vapalahti 0, Kallio-Kokko H, Salonen EM, Brummer-Korvenkontion
M, Vaheri A (1992): Cloning and sequencing of Puumala virus
Sotkamo strain S and M RNA segments: Evidence for strain varia- Giebel LB, Stohwasser R, Zoeller L, Bautz EK, Darai G (1989): Deter-
mination of the coding capacity of the M genome segment of Neph- 286 Khan et al. Khan et al. Khan et al. tion in hantaviruses and expression of the nucleocapsid protein. Journal of General Virology 732329438, gy
Xiao SY, Liang M, Schmaljohn CS (1993): Molecular and antigenic
characterization of HV114, a hantavirus isolated from a patient
with haemorrhagic fever with renal syndrome in China. Journal of
General Virology 74:1657-1659. gy
Xiao Sy, hduc JW, Chu YK, Schmaljohn CS (1994): Phylogenetic
analyses of virus isolates in the genus Hantavirus, family Bunya-
viridae. Virology 198:205-217.
|
https://openalex.org/W2736020490
|
https://ageconsearch.umn.edu/record/265162/files/P9-p106-116.pdf?subformat=pdfa
|
English
| null |
Growing Seed Yams in the Air: the Agronomic Performance of Two Aeroponics Systems Developed in Ghana
|
Sustainable agriculture research
| 2,017
|
cc-by
| 4,933
|
Abstract Aeroponics has been perceived as a technology crammed innovation, far out of reach of the ordinary farmer. Apart from its continuous dependency on electrical power, the technology comes with very sophisticated inputs
such as solenoid valves, timers, misters, CO2 tanks, and air and water pumps. The main objective of this study
was to evaluate the option of using gravity-fed and pressurised aeroponics system for propagating seed yams
from vine cuttings. The study was setup at the CSIR±Crops Research Institute in collaboration with the
Agricultural Engineering Department of the Kwame Nkrumah University of Science and Technology. The basic
advantage of the gravity-fed system is its non-dependency on electrical power, pumps or timers and its ability for
continuous production. The two systems were set-up using conventional materials and equipment available on
the local market. The treatments were arranged in a split-split-plot design with the two aeroponic units as the
main plot, four nutrient concentration levels (C1 - , C2 - , C3 - and C4 -) as the sub plots and vine cuttings from
three Dioscorea rotundata varieties (Dente, Pona and Mankrong Pona) as sub-sub plots respectively. Results
showed there were significant differences (P<0.05) in minituber weight and days to emergence of planted
miniubers. The agronomic response of the two systems in producing mini-tubers was suggestive of the fact that
both systems could be used to effectively produce mini-tubers. Keywords: nutrient concentrations, vine cuttings, propagation, gravity-fed, power-dependent aeroponics,
soil-less cultivation Received: August 19, 2016 Accepted: November 11, 2016 Online Published: July 11, 2017
doi:10.5539/sar.v6n3p106 URL: https://doi.org/10.5539/sar.v6n3p106 Received: August 19, 2016 Accepted: November 11, 2016 Online Published: July 11, 2017
doi:10.5539/sar.v6n3p106 URL: https://doi.org/10.5539/sar.v6n3p106 Give to AgEcon Search
The World’s Largest Open Access Agricultural & Applied Economics Digital Library
This document is discoverable and free to researchers across the
globe due to the work of AgEcon Search.
Help ensure our sustainability.
AgEcon Search
http://ageconsearch.umn.edu
aesearch@umn.edu
Papers downloaded from AgEcon Search may be used for non-commercial purposes and personal study only.
No other use, including posting to another Internet site, is permitted without permission from the copyright
owner (not AgEcon Search) or as allowed under the provisions of Fair Use U S Copyright Act Title 17 U S C AgEcon Search
http://ageconsearch.umn.edu
aesearch@umn.edu Papers downloaded from AgEcon Search may be used for non-commercial purposes and personal study only. No other use, including posting to another Internet site, is permitted without permission from the copyright
owner (not AgEcon Search), or as allowed under the provisions of Fair Use, U.S. Copyright Act, Title 17 U.S.C. Sustainable Agriculture Research; Vol. 6, No. 3; 2017
ISSN 1927-050X E-ISSN 1927-0518
Published by Canadian Center of Science and Education Patricia Oteng-Darko1, Nicholas Kyei-Baffour2, Emmanuel Otoo1 & Wilson Agyei Agyare2
1CSIR-Crops Research Institute, Ghana
2Department of Agricultural Engineering, Kwame Nkrumah University of Science and Technology, Ghana
Correspondence: Patricia Oteng-Darko, CSIR-Crops Research Institute, Ghana. E-mail:
potengdarko@gmail.com Patricia Oteng-Darko1, Nicholas Kyei-Baffour2, Emmanuel Otoo1 & Wilson Agyei Agyare2
1CSIR-Crops Research Institute, Ghana
2Department of Agricultural Engineering, Kwame Nkrumah University of Science and Technology, Ghana
Correspondence: Patricia Oteng-Darko, CSIR-Crops Research Institute, Ghana. E-mail:
potengdarko@gmail.com Patricia Oteng-Darko1, Nicholas Kyei-Baffour2, Emmanuel Otoo1 & Wilson Agyei Agyare2
1CSIR-Crops Research Institute, Ghana
2Department of Agricultural Engineering, Kwame Nkrumah University of Science and Technology, Ghana
Correspondence: Patricia Oteng-Darko, CSIR-Crops Research Institute, Ghana. E-mail:
potengdarko@gmail.com Received: August 19, 2016 Accepted: November 11, 2016 Online Published: July 11, 2017
doi:10.5539/sar.v6n3p106 URL: https://doi.org/10.5539/sar.v6n3p106 1. Introduction Plants require light, water, nutrients, oxygen and carbon dioxide for photosynthesis to grow and thrive. Soil can
be a supplier of nutrients, but is not necessarily in and of itself nutrients - hence the effectiveness of hydroponic
and aeroponics. However, water in itself is becoming more and more scarcer as a commodity and as global
population increases, the concern over water and soil quality also continues to grow. New technologies for
growing foods that are not overly dependent on soils and water are becoming not only a distinct advantage, but a
necessity. The aeroponics and hydroponics technologies have been demonstrated in several ways to be
significantly more water- and energy-efficient means for food production. It is therefore in this regard that the
technology is being adapted for use in this research to propagate seed yam. Aeroponically generated seed yams can improve the seed multiplication ratio of yams and thus make available
more seed yams on the market. It can also reduce disease incidence of seed yams which results in yield losses. Aeroponics, if successfully used in the propagation of seed yams, can significantly increase the incomes of
farmers, improve access to quality seed yams all year round (by making it more accessible and affordable to
commercial growers and small scale farmers) and reduce the production costs of yams. This would improve
IDUPHUV¶OLYHOLKRRGDQGDOVRHQKDQFHIRRGVHFXULW\LQWKHFRXQWU\ In aeroponics, plants are grown in an air or mist environment without engaging soils or any soil aggregate or soil
medium (Arunkumar & Manikand, 2011). Aeroponics gives room for easy access to plant roots since it is not
planted in any aggregate media (Pagliarulo & Hayden, 2002). The growth chamber and fertigation system
employed in aeroponics give room for complete regulation of the root zone setting, including temperature,
humidity, pH, nutrient concentration, mist application frequency and duration. Plants grown using aeroponics 106 Sustainable Agriculture Research Vol. 6, No. 3; 2017 http://sar.ccsenet.org often show signs of accelerated growth and early maturity (Mirza, Younus, Hoyano, & Currie, 1998). These
abilities have made the technology a popular research tool for studying root growth and nutrient uptake (Barak,
Smith, Krueger & Peterson, 1998). often show signs of accelerated growth and early maturity (Mirza, Younus, Hoyano, & Currie, 1998). These
abilities have made the technology a popular research tool for studying root growth and nutrient uptake (Barak,
Smith, Krueger & Peterson, 1998). 1. Introduction Since its introduction into the science arena, aeroponics has offered researchers a non-invasive means to examine
plant roots during development (Mbiyu et al., 2013). It also allows researchers a large number and wide range of
experimental parameters to use in their work (Stoner, 1983). The ability to precisely control the root zone
moisture levels and the amount of water delivered makes aeroponics ideally suited for the study of water stress
and irrigation/fertigation related research. The aeroponic technology has also been successfully used for crops
that are vegatatively propagated, the most recent being the successful application of the technology in the
propagation of yams (Oteng-Darko, Otoo, Kyei-Baffour & Agyare, Unpublished; Maroya, Balogun, & Asiedu,
2014). In further advancement, Oteng-Darko, Kyei-Baffour, Otoo and Agyare (Unpublished) developed the
gravity-fed aeroponic option alongside the pressurised system for seed yam production. This paper presents the
findings and enhancements made to the technology and the successes achieved in its application for seed yam
generation. 2. Method Two aeroponic systems were designed and set up as has been described by Oteng-Darko, Kyei-Baffour, Otoo
and Agyare (Unpublished) at the CSIR-Crops Research Institute, Kumasi, Ghana. Two agronomic evaluations
ZHUH GRQH VXEVHTXHQWO\ WR GHWHUPLQH WKH V\VWHP¶V DELOLW\ WR SURGXFH VHHG \DPV 7KH Dgronomic evaluation
involved two phases: evaluating the two aeroponic systems for its ability to produce mini-tubers and evaluating
the mini-tubers produced for their ability to be used in propagating seed yams. In the first agronomic evaluation,
one and two node cuttings of three yam varieties were planted on the aeroponic units and fertigated with four
different nutrient concentrations. The experimental design was a split-split plot design whereby the aeroponic
units were the main plot, nutrient concentrations the sub plots and yam varieties, the sub-sub plot. In the second agronomic evaluation, three experiments were carried out, all set up in a split-split plot design with
the main plot subjected to mini-tubers harvested from the two aeroponic units, the sub plots to mini-tubers from
the various nutrient concentrations (C1, C2, C3 and C4) and the sub-sub plots subjected to mini-tubers from the
three yam varieties used. The first experiment was subjected to a treatment in which dormant mini-tubers were
planted in pots at a screenhouse, one day after harvesting. The second experiment, non-dormant mini-tubers
were planted directly in the field. In the third experiment, non-dormant mini-tubers were nursed using sawdust in
a screenhouse and transplanted two weeks after emergence. Data was collected on days to rooting, days to tuber initiation, days to emergence (mini-tubers), yield and yield
components. Data collected was analysed using Genstats 9.0 and Mstat 5.4 statistical package. Mean separation
was done using the Fishers unprotected least significant difference. Results were judged significant at p < 0.05. 3. Results Planting with one node cuttings showed significant differences (p<0.05) between Aeroponic units and variety;
and nutrient concentration and variety (Figures 1 and 2). Significant differences (p<0.05) also existed between
the two aeroponics systems and the various nutrient concentrations. Figure 1. Number of roots for one-node cuttings
2
4
6
8
10
12
14
16
18
20
22
24
26
C1
C2
C3
C4
C1
C2
C3
C4
Power-Dependent
Power-Independent
Days to rooting, DAP
Nutrient concentration/Aeroponic units
Dente
Mankrong
Pona 107 Vol. 6, No. 3; 2017 Vol. 6, No. 3; 2017 Sustainable Agriculture Research http://sar.ccsenet.org Figure 2. Number of roots for two-node cuttings
2
4
6
8
10
12
14
16
18
C1
C2
C3
C4
C1
C2
C3
C4
Power-dependant
Power-independent
Days to rooting, DAP
Nutrient concentration/Aeroponic unit
Dente
Mankrong Pona
Pona Figure 2. Number of roots for two-node cuttings Significant difference (p<0.05) existed between the main treatments (pressurised ± PD and gravity-fed ± PI
aeroponics systems) for both the one and two node cuttings. Varieties also showed highly significant differences
at p<0.01 (Figure 3) in their response to the number of roots at root initiation for both the one and two-node
cuttings. The number of roots observed was also significant for the one-node cuttings under the
two-way-aeroponics unit x variety- interaction. Figure 3. Number of mini-tubers of varieties under both power dependent and gravity-fed system
0
0.5
1
1.5
2
2.5
3
3.5
NMATI 2WAP
4WAP NMATI 2WAP
4WAP NMTATI 2WAP
4WAP
Pona
Dente
Mankrong
Pona
Number of mini-tubers
Weeks after planting/Variety
PD C1
PD C2
PD C3
PD C4
P1 C1
P1 C2
P1 C3 Weeks after planting/Variety Figure 3. Number of mini-tubers of varieties under both power dependent and gravity-fed system No significant differences were seen in the number of mini-tubers for any of the treatment interaction at two
weeks after planting (Figure 3) for the one node planting. However, with the two-node cuttings, significant
differences were seen in the main treatment (aeroponic units), sub plot (nutrient concentration) and sub-sub plot
(variety). Significant differences also existed between the two way interactions between aeroponic units x variety
and nutrient concentration x variety for the two node cutting but not the one node cuttings. At four and six weeks
after planting, significant differences (p<0.05) were seen in the nutrient concentration treatments (Figure 3) for 108 Sustainable Agriculture Research Vol. 6, No. 3. Results 3; 2017 http://sar.ccsenet.org both the one and two node cuttings The grand mean for the total number of mini-tubers harvested per plant (from both the first and second harvest)
was 2.38. The aeroponics systems had means of 2.89 and 1.89 for the power dependent and power independent
systems respectively (Figure 4). There were no significant differences in the three-way interaction between
aeroponics system, nutrient concentration and variety. Significant differences were seen in the various nutrient
concentrations used. There was a highly significant difference (p < 0.01) between the varieties. Figure 4. Number of mini-tubers harvested per plant
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
C1
C2
C3
C4
C1
C2
C3
C4
Power-dependent
Power-independent
Total number of mini-tuber
harvested
Aeroponic units
Dente
Mankro
Pona Figure 4. Number of mini-tubers harvested per plant In planting with non-dormant seeds in the field, the mean emergence for both the power dependent and
gravity-fed aeroponics systems was 5.36 days after planting, showing no significant differences between any of
the interaction under this treatment (Figure 5). Figure 5. Emergence of nursed seeds in the screenhouse
0
1
2
3
4
5
6
7
C1
C2
C3
C4
C1
C2
C3
C4
Power Dependent
Power Independent
Emergence, days after planting
Aeroponic units
Dente
Mankrong
Pona Mankrong
Pona Figure 5. Emergence of nursed seeds in the screenhouse Mean emergence for directly planted dormant mini-tubers derived from the power-dependent and
power-independent aeroponics systems were 60.56 and 59.86 days after planting (Figure 6). The mean
emergence for C1, C2, C3 and C4 were 60.17, 58.00, 60.39 and 62.28 DAP respectively. Varietal means were
81.25, 56.08 and 43.29 for Dente, Mankrong Pona and Pona respectively. 109 Vol. 6, No. 3; 2017 Sustainable Agriculture Research http://sar.ccsenet.org Figure 6. Emergence of directly planted dormant mini-tubers
0
10
20
30
40
50
60
70
80
90
100
C1
C2
C3
C4
C1
C2
C3
C4
PD
P1
Emergence, Days after planting
Aeroponic units
Dente
Mankrong Pona
Pona Figure 6. Emergence of directly planted dormant mini-tubers The grand mean for emergence of planted non-dormant mini-tubers was 5.36 days after planting (Figure 7). Mean emergence for both the power-dependent and power-independent aeroponics systems was 5.36 days after
planting. There were no significant differences between any interactions under this treatment. Figure 7. 3. Results Emergence of directly planted non-dormant mini-tubers
0
1
2
3
4
5
6
7
C1
C2
C3
C4
C1
C2
C3
C4
Power Dependent
Power Independent
Emergence, days after planting
Aeroponic units
Dente
Mankrong Pona
Pona Figure 7. Emergence of directly planted non-dormant mini-tubers Significant differences were seen in the mini-tuber weight of the various varieties produced using the various
nutrient concentrations (Figure 8). 110 Vol. 6, No. 3; 2017 Sustainable Agriculture Research http://sar.ccsenet.org Figure 8. Weight of harvested mini-tubers
0
2
4
6
8
10
12
14
C1
C2
C3
C4
C1
C2
C3
C4
Power-dependent
Power-independent
Weight of minituber harvested, g
Aeroponic units
Dente
Mankrong Pona
Pona Figure 8. Weight of harvested mini-tubers The mean number of seed yam tubers produced was significant (p<0.01) for all varieties propagated using C3
(Figure 9). Even though the nutrient effects were not so significant in any of the previous discussions, it stands to
be argued that the nutrient concentration used in propagating the mini-tubers, whether by the power-dependent or
power-independent aeroponic systems has significant impact on the final yield of the second generation seed. Figure 9. Chart showing number of seed tubers per plant
0
0.5
1
1.5
2
2.5
3
3.5
C1
C2
C3
C4
C1
C2
C3
C4
Power-Dependent
Power-Independent
Number of tubers/plant
Aeroponic units/nutrient concentration
Dente
Mankrong Pona
Pona Figure 9. Chart showing number of seed tubers per plant Using the two-node cuttings, a mean of 130 vines were cut per explant for transplanting onto the aeroponic units. The maximum mean number of mini-tubers harvested for dente was attained using nutrient concentration C2
whereas the maximum number of mini-tubers for pona and mankrong pona were attained using nutrient
concentration C3. Propagation using the power-dependent aeroponic system gave a mean multiplication ratio of 404 mini-tubers
per explant (Table 1). Using the power-dependent system, Mankrong Pona had the highest multiplication of 477
mini-tubers/explant followed by Pona and Dente with 390 and 347 mini-tubers/explant respectively. Propagation using the power-independent aeroponic system again showed Mankrong Pona having the highest
multiplication ratio of 347 mini-tubers/explant followed by Pona and Dente with 302 and 173
mini-tubers/explant respectively (Table 1). The mean multiplication ratio of the power-independent system was 111 Vol. 6, No. 3; 2017 Sustainable Agriculture Research http://sar.ccsenet.org p
p
Table 1. 3. Results Mean mini-tuber yields and multiplication ratio of the yam varieties under the two aeroponics systems
Aeroponic unit
Vine cuttings per plant
Mean yield per cutting
Multiplication ratio/explant
Dente
Mankrong Pona
Pona
Dente
Mankrong Pona
Pona
Mean
Power-Dependent
130
2.67
3.67
3.00
347.10
477.10
390.00
404.70
Power-Independent
130
1.33
2.67
2.33
172.90
347.10
302.90
274.30 ni-tuber yields and multiplication ratio of the yam varieties under the two aeroponics systems Table 1. Mean mini-tuber yields and multiplication ratio of the yam varieties under the two aeropon Significant differences were seen in the multiplication ratio of vines planted on the two aeroponic systems used
(Figure 10). Mean seed yam multiplication ratio of the power-dependent aeroponic system for all the varieties
was 1035 mini-tubers/explant. Mankrong Pona had the highest multiplication ratio of 1393 seed yams/explant
followed by Dente and Pona with 895 and 819 seed yams/explant respectively. The highest multiplication ratio using the power-independent aeroponic system was 774 seed yams /explant for
Mankrong Pona. Pona and Dente had 642 and 347 seed yams/explant respectively. The mean seed yam
propagation ratio using the power-independent system was 587 seed yams/explant (Figure 10). Figure 10. Number of seed yam generated per explant
0
200
400
600
800
1000
1200
1400
1600
Dente
Mankrong
Pona
Pona
Seed yam/explant
Variety
Power-Dependent
Power-Independent Power-Dependent
Power-Independent Figure 10. Number of seed yam generated per explant 4. Discussion In yams, mini-tuber production is affected by genotype, as has been seen by Powell, Brown and Caligari (1989)
for potato. This research confirms the assertion that genotypes differ widely in their capacity to produce
mini-tubers, some being more prolific than others (Venkatasalam et al., 2011; Sharma, Venkatasalam & Kumar,
2013). The prolific nature of these genotypes established a positive correlation between days to rooting and days
to mini-tuber initiation (Figure 11). 112 Sustainable Agriculture Research Vol. 6, No. 3; 2017 http://sar.ccsenet.org Figure 11. Correlation between days to rooting and days to mini-tuber initiation
5
10
15
20
5
10
15
20
25
Days to rooting
Days to mini-tuber initiation Figure 11. Correlation between days to rooting and days to mini-tuber initiation Even though no correlation was observed for days to root initiation and number of roots at root initiation, the
same was not the case for days to mini-tuber initiation and number of mini-tubers at mini-tuber initiation. A
negative correlation (R2 = 0.196) was observed between the number of days to mini-tuber formation and the
number of mini-tubers at tuber initiation establishing that early tuberisation does not have any effect on the
number of mini-tubers as shown in Figure 12. This can also be attributed to the genotypic differences between
the yam varieties used. Figure 12. Correlation between days to mini-tuber initiation and number of mini-tubers at tuber initiation
1
1.5
2
2.5
3
5
10
15
20
25
Number of mini-tubers at tuber initiation
Days to mini-tuber initiation Number of mini-tubers at tuber initiation Figure 12. Correlation between days to mini-tuber initiation and number of mini-tubers at tuber initiation 12. Correlation between days to mini-tuber initiation and number of mini-tubers at tuber initiation As has been reported by Soffer and Burger (1988), aeroponics optimizes root aeration resulting in more yields
than classical hydroponics. A positive correlation (R2 = 0.1274), though not very strong was observed between
number of roots and number of mini-tubers. From Figure 13 varieties with the most prolific rooting system also
yielded more mini-tubers, thus confirming the report by Soffer and Burger (1988). As has been reported by Soffer and Burger (1988), aeroponics optimizes root aeration resulting in more yields
than classical hydroponics. A positive correlation (R2 = 0.1274), though not very strong was observed between
number of roots and number of mini-tubers. 4. Discussion From Figure 13 varieties with the most prolific rooting system also
yielded more mini-tubers, thus confirming the report by Soffer and Burger (1988). 113 Sustainable Agriculture Research Vol. 6, No. 3; 2017 http://sar.ccsenet.org Figure 13. Correlation between number of roots and number of mini-tubers at six weeks after planting
10
20
30
40
50
1
2
3
4
Number of roots at 6 weeks after planting
Number of mini-tuber at 6 weeks after planting gure 13. Correlation between number of roots and number of mini-tubers at six weeks after planting There was also a positive correlation (R2 = 0.344) between days to rooting and days to new leaf (Figure 14). Early rooted varieties also had early new leaves. Although the early rooting may have played some role in vine
cuttings expressing new leaf/leaves, one remains free to speculate that the quantitative genotypic reflect the
physiological differences in the cultivars used. Thus, the correlations between the number of days to root
initiation and new leaf/leaves formation may depend on the physiological conditions of the cultivars used and
not necessarily on the nutrient concentration used. There was also a positive correlation (R2 = 0.344) between days to rooting and days to new leaf (Figure 14). Early rooted varieties also had early new leaves. Although the early rooting may have played some role in vine
cuttings expressing new leaf/leaves, one remains free to speculate that the quantitative genotypic reflect the
physiological differences in the cultivars used. Thus, the correlations between the number of days to root
initiation and new leaf/leaves formation may depend on the physiological conditions of the cultivars used and
not necessarily on the nutrient concentration used. Figure 14. Correlation between days to root initiation and number of days to new leaf
20
40
60
80
100
5
10
15
20
Number of days to new leaf/leaves
Days to root initiation Number of days to new leaf/leaves Figure 14. Correlation between days to root initiation and number of days to new leaf The negative correlation (R2 = 0.28) between days to mini-tuber initiation and number of mini-tubers at six
weeks after planting (Figure 15), though weak, is suggestive of the fact that initiating early tuberisation does not
affect final yield or number of mini-tubers that would be produced per plant. This is because many factors have
been reported to affect tuber formation (Kempen, 2012: Menzel, 1980: Sattelmacher & Marschner, 1978). 4. Discussion Maroya, Balogun & Asiedu (2014) reported
using a planting density of 400 and 100 cm2/plant which resulted in mini-tuber every three to five months
whereas this research used a planting density of 36 cm2/plant and reports mini-tuber yields from 4 months
onwards and subsequently every two weeks. In evaluating mini-tuber harvested, plants propagated from mini-tubers nursed in the screenhouse had
rudimentary leaves before transfer to the experimental field 21 days after emergence. According to Haverkort,
Van De Waart & Marinus (1991) and Lommen (1999), glasshouse raised transplants from very early cultivars
sometimes show a poor performance after transplanting into the field. This has been attributed to the fact that
immediately after transplanting, a major part of the daily dry matter production is invested in tuber growth
(Lommen, 1999). This high degree of partitioning to tubers leads to a limited growth of the haulm and thereby
limits the biomass production and final tuber yield (Lommen, 1999). Nutrient concentration did not have any significance on the number of mini-tubers harvested. However,
significant differences were seen in harvested mini-tuber weights and eventual number and weight of seed yams
harvested using the various varieties produced under the different nutrient concentrations. It thus stands to be
argued that the nutrient concentration used in propagating the mini-tubers, whether by the power-dependent or
power-independent aeroponic systems has significant impact on the final yield of the second generation seed. The two aeroponics systems were developed to enhance and optimize the rapid multiplication of seed yams. The
potential benefits herein discussed such as rapid rooting and tuberisation and high multiplication ratios gives
these systems the potential to improve seed yam production in the country. Acknowledgments 7KHDXWKRUV¶ZRXOGOLNHWRDFNQRZOHGJHWKH:HVW$IULFDQ$JULFXOWXUDO3URGXFWLYLW\3URJUDPPH *KDQD DQGWKH
BMGF Africa Yam Project for financially supporting this work. Barak, P., Smith, J. D., Krueger, R., & Peterson, A. (1998). Measurement of short-term nutrient uptake rates in
cranberry by aeroponics. Plant Cell Environ, 19, 237-242.
https:/doi.org/10.1111/j.1365-3040.1996.tb00246.x 4. Discussion According to Kempen (2012), even the bacteria living in the root zone are reported to have an influence;
however, nitrogen levels, temperature and light have the greatest effect. Reports show that short days and cool The negative correlation (R2 = 0.28) between days to mini-tuber initiation and number of mini-tubers at six
weeks after planting (Figure 15), though weak, is suggestive of the fact that initiating early tuberisation does not
affect final yield or number of mini-tubers that would be produced per plant. This is because many factors have
been reported to affect tuber formation (Kempen, 2012: Menzel, 1980: Sattelmacher & Marschner, 1978). According to Kempen (2012), even the bacteria living in the root zone are reported to have an influence;
however, nitrogen levels, temperature and light have the greatest effect. Reports show that short days and cool 114 Sustainable Agriculture Research Vol. 6, No. 3; 2017 http://sar.ccsenet.org night temperatures also promote tuberization whereas long days, high night temperatures, and high nitrogen
fertilisation inhibit or delay the process (Menzel, 1980, Sattelmacher & Marschner, 1978). This research did not
go further to corroborate these findings. Figure 15. Correlation between days to mini-tuber initiation and number of mini-tubers at six weeks after planting
Th
l
i
d
i
b
d i i
f h
k
f
i
i i i
i i
b
d
i
Th
1
2
3
4
5
10
15
20
25
Number of mini-tuber at six weeks after planting
Days to mini-tuber initiation Number of mini-tuber at six weeks after planting Figure 15. Correlation between days to mini-tuber initiation and number of mini-tubers at six weeks after planting The planting density, number and timing of harvests are key factors in optimizing mini-tuber production. The
planting density used resulted in optimized resource use efficiency. Maroya, Balogun & Asiedu (2014) reported
using a planting density of 400 and 100 cm2/plant which resulted in mini-tuber every three to five months
whereas this research used a planting density of 36 cm2/plant and reports mini-tuber yields from 4 months
onwards and subsequently every two weeks. The planting density, number and timing of harvests are key factors in optimizing mini-tuber production. The
planting density used resulted in optimized resource use efficiency. Haverkort, A. J., Van De Waart, M., & Marinus, J. (1991). Field performance of potato micro-tubers as Arunkumar, M., & Manikand, R. (2011). Aeroponics. Accessed online at
http://www.slideboom.com/presentations/263081-/AEROPONICS on 08/02/14 on material. Potato Res, 34, 353-64 https:/doi.org/10.1007/BF02360509 propagation material. Potato Res, 34, 353-64 https:/doi.org/10.1007/BF02360509 Kempen E. (2007). The influence of light intensity, light quality and root zone temperature on potato minituber
production. Thesis presented in partial of the requirements for the degree of Master of Agricultural Science
at the University of Stellenbosch. Lommen, W. J. M., & Struik, P. C. (1992b). Potato Production of potato mini-tubers by repeated harvesting:
Plant productivity and initiation, growth and re-absorption of tubers. Netherlands Journal of Agricultural
Science, 40, 342-359. Maroya, N., Balogun, M., & Asiedu, R. 2014. Seed Yam Production in Aeroponics System: A Novel Technology. YIIFSWA Working Paper No 2. Institute International of Tropical Agriculture, Ibadan, Nigeria; 9 pages. ©
IITA-YIIFSWA; 2014. ISBN 978-978- 8444-37-4. Menzel, C. M. (1980). Tuberization in potato at high temperatures: Responses to gibberellin and growth
inhibitors. Ann. Bot., 46, 259-265. https://doi.org/10.1093/oxfordjournals.aob.a085916 Mbiyu, M. W., Muthoni, J., Kabira, J., Elmar, G., Muchira, C., Pwaipwai, P., Ngaruiya, J., Otieno, S., & Onditi, J. (2013). Use of aeroponics technique for potato (Solanum tuberosum) mini-tubers production in Kenya. Accessed online at http://www.internationalscholarsjournals.org on 17/12/2013 Mirza, M., Younus, M., Hoyano, Y., & Currie, R. (1998). Greenhouse production of Echinacea and other
medicinal plants. Paper presented at Opportunities and Profits II: Special Crops into the 21stCentury. Nov
1-3, 1998, Edmonton, AB, Canada Oteng-Darko P., Otoo E., Kyei-Baffour N., & Agyare W. A. (Unpublished). Growth and yield response of three
aeroponically propagated yam varieties under four different nutrient concentrations. Abstract presented at
the World Congress on Root and Tuber Crops, Nanning, Guangxi, China, Jan 18ϋ22, 2016. Oteng-Darko P., N. Kyei-Baffour, E. Otoo E., & Agyare W.A (Unpublished). Design and Evaluation of Two
Simple Aeroponic Systems for Seed Yam Propagation. Under review, African Crop Science Journal Pagliarulo, C. L., & Hayden, A. L. (2002). Potential for greenhouse aeroponic cultivation of medicinal root crops. Proc. Amer. Plasticult. Soc. Conf., San Diego, Cali. Accessed online at Pagliarulo, C. L., & Hayden, A. L. (2002). Potential for greenhouse aeroponic cultivation of medicinal root crops. Proc. Amer. Plasticult. Soc. Conf., San Diego, Cali. Accessed online at
http://aerofarms.com/wordpress/wpcontent/files_mf/1265411630GrowingMedicinalCropswithAeroponics.p
df aerofarms.com/wordpress/wpcontent/files_mf/1265411630GrowingMedicinalCropswithAeroponics.p Powell, W., Brown, J., & Caligari, P. D. S. 1989. Variability in response of potato cultivars to micro-propagation
II. Subsequent field performance. Ann Applied Biology 115: 123-28 Sattelmacher, B. & Marschner, H. (1978). Relation between nitrogen nutrition, cytokinin activity and
tuberization in Solanum tuberosum. Physiol. Plant, 44, 65-68. https://doi.org/10.1111/j.1399-3054.1978.tb01614.x Sharma, A. K., Venkatasalam. E. P. & Kumar, V. on material. Potato Res, 34, 353-64 https:/doi.org/10.1007/BF02360509 (2013). Potato mini-tuber production during main and off crop
seasons in high hills of north-western Himalaya. Potato J 40(1): 29-37 Soffer, H., & Burger, D. W. (1988). Effects of dissolved oxygen concentration in aeroponics on the formation
and growth of adventitious roots. J. Am. Soc. Hortic. Sci., 113, 218-221 toner, R. J. (1983). Aeroponics versus Bed and Hydroponic Propagation. Florists Rev. 1(173), 4477. Venkatasalam, E. P., Latawa, J., Sharma, S., Sharma, S., Sharma, A. K., Sharma, S., Patial, R., & Singh, S. (2011). In vitro and in vivo performance of potato cultivars for different seed production systems. Potato J.,
38(2), 149-54 This is an open-access article distributed under the terms and conditions of the Creative Commons Attribution
license (http://creativecommons.org/licenses/by/3.0/). Copyright for this article is retained by the author(s), with first publication rights granted to the journal. References Barak, P., Smith, J. D., Krueger, R., & Peterson, A. (1998). Measurement of short-term nutrient uptake rates in
cranberry by aeroponics. Plant Cell Environ, 19, 237-242. https:/doi.org/10.1111/j.1365-3040.1996.tb00246.x Haverkort, A. J., Van De Waart, M., & Marinus, J. (1991). Field performance of potato micro-tubers as 115 Sustainable Agriculture Research Vol. 6, No. 3; 2017 http://sar.ccsenet.org Copyrights py g
Copyright for this article is retained by the author(s), with first publication rights granted to the journal This is an open-access article distributed under the terms and conditions of the Creative Commons Attribution
license (http://creativecommons.org/licenses/by/3.0/). 116
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.